diff --git a/.agents/skills/dsh-code-review/SKILL.md b/.agents/skills/dsh-code-review/SKILL.md
index 3502e4454a..2b562cc321 100644
--- a/.agents/skills/dsh-code-review/SKILL.md
+++ b/.agents/skills/dsh-code-review/SKILL.md
@@ -19,19 +19,26 @@ description: Use when reviewing a pull request in the deepseek-harness repo —
 
 ## Blocking requirements
 
-1. **Docs match the code.** Config, defaults, errors, wire fields, events, and public behavior update the package README and JSDoc in the same diff. Comments state non-obvious contracts; flag implementation narration, test walkthroughs, review history, and duplicated rationale for deletion or a link to their one home.
-2. **Core type docs match.** Changes to spine or seam vocabulary update the appropriate [core-data-structures](../../../docs/core-data-structures/core.md) page and any `type-equiv` entry. Internal types need no catalog entry.
-3. **Registrations clean up.** A new registry contribution has a test that disposes its owner and observes removal.
-4. **Required gates pass.** Trust the [current readiness sequence](../../../AGENTS.md#run-the-ci-gates-locally-before-marking-a-pr-ready) and `pnpm run check:pre-push` for their enforced inventory; review the semantic gaps they cannot detect.
+1. **New prose receives semantic review.** Use [dsh-prose-standard](../dsh-prose-standard/SKILL.md) to critically review every added or changed Markdown passage, JSDoc, comment, prompt, description, diagnostic, and visible string. Verify required coverage, accuracy, placement, and editorial quality against the owning code or behavior; automated checks do not establish those properties.
+2. **Docs match the code.** Config, defaults, errors, wire fields, events, and public behavior update the package README and JSDoc in the same diff. Comments state non-obvious contracts; flag implementation narration, test walkthroughs, review history, and duplicated rationale for deletion or a link to their one home.
+3. **Core type docs match.** Changes to spine or seam vocabulary update the appropriate [core-data-structures](../../../docs/core-data-structures/core.md) page and any `type-equiv` entry. Internal types need no catalog entry.
+4. **Registrations clean up.** A new registry contribution has a test that disposes its owner and observes removal.
+5. **Required gates pass.** Trust the [current readiness sequence](../../../AGENTS.md#run-the-ci-gates-locally-before-marking-a-pr-ready) and `pnpm run check:pre-push` for their enforced inventory; review the semantic gaps they cannot detect.
 
 ## Manual checks
 
 - **Intent and seam contracts:** trace both sides of every changed interface. Confirm the implementation matches the PR and any RFC, including errors, cancellation, ownership, and disposal.
-- **Lifecycle and concurrency:** for async setup, callbacks, processes, or teardown, apply [defensive-patterns.md](../../../docs/defensive-patterns.md). Check races before publication, cancellation during awaits, independent error reporting, callback containment, and quiescent disposal.
+- **Lifecycle and concurrency:** for async setup, callbacks, processes, or teardown, apply [defensive-patterns.md](../../../docs/defensive-patterns.md). Check races before publication, cancellation during awaits, independent error reporting, callback containment, ownership before reentry, complete detach cleanup, and quiescent disposal.
 - **Capability shape:** a swappable capability follows the interface / implementation / consumer split. Consumers depend on the interface, not a backend.
+- **Scope, ownership, and necessity:** tie each abstraction, state machine, option, defensive copy, and compatibility path to a current contract or production consumer. Challenge unrelated features, speculative generality, and behavior placed outside its owning plugin or service.
 - **Configuration:** deployment-varying timeouts, caps, models, URLs, paths, and retry counts are validated `Config` fields, not literals or `DEFAULT_*` constants.
+- **Enforcement boundaries:** hidden schema fields, filtered prompts, facades, wrappers, and listener ordering are not authoritative enforcement when direct or alternate callers can bypass them. Exercise denial paths at the boundary that actually executes the operation.
+- **Borrowed and derived state:** determine whether retained caller-owned values are borrowed or snapshotted by contract; do not demand copies at typed same-process seams. Materialize mutable values that cross queues, model/tool JSON, durable logs or files, workers, processes, or wire boundaries. Commit notifications and derived state only at the documented success boundary, and trace caches, prompts, UI echoes, replay, and query views to one authoritative source.
+- **Bounds cover the final operation:** verify byte, token, item, and time limits at the boundary that owns the complete emitted or retained result, including wrappers and metadata. Probe tiny limits, exact thresholds, oversized single chunks, and multibyte text for byte limits.
 - **Real entry path:** tests exercise the shipped Loader, bin, worker, ACP bridge, or subprocess where relevant. A hand-mounted plugin does not catch Loader export-shape failures; a function plugin must named-export its namespace and have no default export.
 - **Test strength:** assertions fail on the intended regression and verify external state, logs, events, or disposal rather than restating the implementation or trusting an agent's report. Coverage is necessary but not evidence that the scenario is correct.
+- **Changed checks have a negative control:** a new automated check, or a changed acceptance path in one, has a deliberately invalid case that reaches the real top-level runner and fails for the intended rule; a green happy path does not prove the check is wired.
+- **Implemented RFCs match shipped reality:** when a PR implements a proposed RFC, move and rewrite it as present-tense shipped state in the same diff, then verify paths, names, and mechanisms against the implementation.
 - **Transcript changes:** editor-visible or model-visible changes update snapshots or explain why no snapshot applies. Review golden diffs as behavior changes, not formatting noise.
 - **Bilingual changes:** compare meaning and terminology on both sides; a green pairing hash does not prove translation quality.
 
diff --git a/.agents/skills/dsh-translate-docs/SKILL.md b/.agents/skills/dsh-translate-docs/SKILL.md
index 553cf6f9d1..d935c5c4b7 100644
--- a/.agents/skills/dsh-translate-docs/SKILL.md
+++ b/.agents/skills/dsh-translate-docs/SKILL.md
@@ -14,7 +14,7 @@ These are authoritative; read them at the source so this skill never drifts out
 - **[docs/i18n/README.md](../../../docs/i18n/README.md)** — the pairing contract: the three-file pair (`foo.md`, `foo.zh.md`, `foo.i18n.yaml`), the consistency record's both-side blob hashes, the language-switcher lines, scope/exclusions, and the rollout manifest.
 - **[docs/i18n/translation-rules.md](../../../docs/i18n/translation-rules.md)** — how to translate: faithfulness, structure preservation, terminology discipline, typography (MUST/SHOULD levels).
 - **[docs/i18n/terminology.md](../../../docs/i18n/terminology.md)** — the terminology table, binding in both directions. Load it BEFORE translating, not when a term feels uncertain; the terms you don't notice are the ones that drift.
-- **[docs/i18n/translation-prompt.md](../../../docs/i18n/translation-prompt.md)** — the automated pipeline's machine-consumed template. Agents using this skill do not render it; keep rules shared with this skill synchronized.
+- **[docs/i18n/translation-prompt.md](../../../docs/i18n/translation-prompt.md)** — the automated pipeline's machine-consumed template. Agents using this skill do not render it; the renderer injects `translation-rules.md` so rules have only one home.
 - **[dsh-prose-standard](../dsh-prose-standard/SKILL.md)** — required prose coverage and editorial judgment. Apply it to both sides without adding or dropping source propositions.
 
 ## Find the work
@@ -42,8 +42,9 @@ Do not process every file the same way:
 - **Pass 1 — write, don't transpose.** Read a semantic unit, then restate it as a native technical author in the nearest [style sample's](../../../docs/i18n/style-samples.md) register. Preserve the required frame without forcing sentence-by-sentence correspondence.
 - **Pass 2 — verify against the source, clause by clause.** Fidelity is checked here, not written in: confirm nothing was added or dropped, every term follows the table, and each code span survived verbatim. Fix by rewriting the sentence natively, not by patching words into it.
 - Write only the final text to the file, never drafts or notes.
-- Every term in [terminology.md](../../../docs/i18n/terminology.md) renders exactly as specified, in both directions, including first-occurrence annotations. A term the table misses: translate only with a citable precedent from a major Chinese OSS/vendor doc; otherwise keep the English and add it to the PR's 「待定术语」 list with your suggested rendering. Never invent a rendering inline — that decision belongs to a human and then to the table.
+- Every term in [terminology.md](../../../docs/i18n/terminology.md) renders exactly as specified. For a Chinese target, use the Chinese and first-occurrence columns; an unlisted term needs a citable Chinese OSS/vendor precedent or stays English under 「待定术语」. For an English target, use the English column and an established English technical term; preserve an ambiguous source term with a short gloss and list it as pending. Never invent a rendering inline.
 - Code blocks are byte-identical across the pair, comments included. Relative links keep their `.md` targets; only the switcher line links `.zh.md`.
+- The pairing gate checks heading depths, fenced blocks, table row and column counts, list kinds, ordered-list starts, list item counts, and link targets. In Pass 2, manually verify list and table order, noncanonical list numbering, inline code, emphasis, meaning, terminology, and tone.
 
 ## Finish the pair
 
@@ -51,9 +52,9 @@ Do not process every file the same way:
 2. Record consistency: `pnpm run verify-translation-pairing --write` recomputes and records both sides' full blob hashes in `foo.i18n.yaml`. The yaml diff in your PR is the reviewable statement "I confirmed these two say the same thing" — only run it after you actually have.
 3. New batch landed? Add the `.md` paths to `required` in [scripts/translation-pairing.manifest.json](../../../scripts/translation-pairing.manifest.json) so the gate ratchets forward.
 
-## Verify — the gate, not your eyes
+## Verify the mechanical and human halves
 
-Run `pnpm run verify-translation-pairing`, then the rest of the Markdown gates (`pnpm run verify-md-wrap && pnpm run verify-md-links`, or full `pnpm run doc-sync` before the PR). Fix what they report; do not hand-check what they cover. What they can NOT check — whether the two sides truly say the same thing, terminology judgment calls, tone — is exactly what the PR reviewer will read for, so keep the PR reviewable: state which pairs are new vs minimally updated, and list 「待定术语」 prominently.
+Run `pnpm run verify-translation-pairing`, then the rest of the Markdown gates (`pnpm run verify-md-wrap && pnpm run verify-md-links`, or full `pnpm run doc-sync` before the PR). Fix what they report and manually verify the obligations listed in Pass 2 that the gates do not encode. Keep the PR reviewable: state which pairs are new versus minimally updated and list 「待定术语」 prominently.
 
 ## How to respond to translation review
 
diff --git a/.github/workflows/e2e.yml b/.github/workflows/e2e.yml
index a2015696f4..dd3965f5cd 100644
--- a/.github/workflows/e2e.yml
+++ b/.github/workflows/e2e.yml
@@ -110,9 +110,14 @@ jobs:
           fi
           echo "DEEPSEEK_API_KEY present."
 
+      # The e2e suites boot the example bins in `lib` mode (DSH_EXAMPLE_MODE=lib):
+      # the built artifact under plain Node, resolving plugins through real package
+      # exports — the shape a real consumer runs. That requires a prior build.
+      - name: Build (lib for the e2e example bins)
+        run: pnpm run build
+
       # Real-API end-to-end tests only. The keyless gates (lint/typecheck/
-      # coverage/snapshot/build/etc.) already run in ci.yml on every push/PR;
-      # no need to repeat them or build first (tests run unbuilt via tsx).
+      # coverage/snapshot/etc.) already run in ci.yml on every push/PR.
       # DEEPSEEK_BASE_URL is pinned to the external API; the secret is scoped to
       # this step (and preflight) only — never exposed to checkout/setup/install.
       - name: E2E tests (real DeepSeek API)
@@ -120,4 +125,5 @@ jobs:
           DEEPSEEK_API_KEY: ${{ secrets.DEEPSEEK_API_KEY_EXTERNAL }}
           DEEPSEEK_BASE_URL: https://api.deepseek.com
           DSH_E2E_MAX_WORKERS: 14
+          DSH_EXAMPLE_MODE: lib
         run: pnpm run test:e2e
diff --git a/.gitignore b/.gitignore
index 90709e81c0..bb700f23e5 100644
--- a/.gitignore
+++ b/.gitignore
@@ -11,6 +11,7 @@ examples/*/*.jsonl
 examples/*/.sessions/
 coverage/
 .doc-typecheck-*/
+.node-next-types-*/
 .humanize/
 tmp/
 .claude/commands/
diff --git a/AGENTS.md b/AGENTS.md
index 23d7c0263b..d53a26ec08 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -10,8 +10,9 @@ DeepSeek Harness SDK is a plugin-based agent harness on vendored Cordis: **every
 
 ```
 vendor/      Vendored Cordis source — manifest + sync procedure in vendor/README.md
-packages/    Harness packages at packages/<group>/<pkg>/, all named @deepseek-ai/dsh-<pkg>
+packages/    @deepseek-ai/dsh-<pkg> workspaces at packages/<group>/<pkg>/
   core/        product API spine: session, system-prompt, tools, agent, agent-loop
+  prompt/      workspace instructions
   llm/         LLM seam + the DeepSeek adapters (hand-rolled + pi-ai design twin)
   bash/        bash executor seam + local impl + model-facing bash tools
   fs/          filesystem seam + local impl + policy gate + read/write/edit tools
@@ -34,6 +35,7 @@ python/      Python SDK and bundled runtime (see python/README.md)
 examples/    Runnable cordis.yml leaves over packages/examples bundles (see examples/AGENTS.md)
 docs/        architecture, generated catalogs, RFCs, postmortems, cookbook (see docs/AGENTS.md)
 scripts/     repo gates and generators
+website/     VitePress projection of selected bilingual docs/ sources
 ```
 
 Package groups: [packages/README.md](packages/README.md).
@@ -53,12 +55,17 @@ pnpm run duplication    # cross-file TypeScript clone detection
 pnpm run build          # tsc emits lib/types, tsdown bundles runtime
 pnpm run hygiene        # knip + publint + workspace constraints + NodeNext consumer check
 pnpm run doc-sync       # all documentation gates; see the doc-sync script in package.json
+pnpm run website:build  # VitePress build (doubles as the site's dead-link check)
 pnpm run demo:echo      # mock-model REPL, no key needed
 pnpm run demo:repl      # real REPL coding agent (needs DEEPSEEK_API_KEY)
 pnpm run demo:cordis    # self-referential demo: the agent modifies its own runtime (needs key)
 pnpm run demo:acp       # ACP server agent (needs DEEPSEEK_API_KEY)
 ```
 
+### Host sandbox failures
+
+When required `gh`, `pnpm`, build, test, or generator commands fail because the agent sandbox blocks credentials, network, IPC, file watching, or nested `sandbox-exec`, retry unchanged with the narrowest host escalation before diagnosing authentication or project failure. Require sandbox evidence; never bypass genuine test failures or the product sandbox under test.
+
 ### Run the CI gates locally before marking a PR ready
 
 Run narrow checks during implementation and this CI-equivalent sequence before marking a PR ready. Fresh worktrees need `pnpm run build` before publint and NodeNext inspect `lib/`:
@@ -71,6 +78,7 @@ pnpm run duplication
 pnpm run test:coverage
 pnpm run test:snapshot
 pnpm run doc-sync
+pnpm run website:build
 pnpm run verify-module-graph
 pnpm run build
 pnpm run hygiene
@@ -91,7 +99,7 @@ Real-API tests and demos read `DEEPSEEK_API_KEY`, optional `DEEPSEEK_BASE_URL`,
 ## Conventions
 
 - Every npm package is `@deepseek-ai/dsh-<name>`; vendored packages keep upstream names and are `private: true`. `cordis` is a peerDependency (+ dev) of every harness package.
-- ESM everywhere (`"type": "module"`). Cross-package imports use package names, never relative paths; in-package relative imports use explicit `.ts` extensions. Dev/test/demo run unbuilt via tsx + the root tsconfig `paths` map; builds are for outside consumers only.
+- ESM everywhere (`"type": "module"`). Cross-package imports use package names; in-package relative imports include `.ts`. CI subprocesses that boot examples or Cordis configs run built `lib/` under plain Node; only explicit source-path regressions use tsx ([testing policy](docs/testing.md#test-subprocess-launch-modes)).
 - **Registrations are effects**: every contribution goes through `ctx.effect()` / `ctx.on()`; a registry's `register()` returns the disposer.
 - **Typed events use declaration merging** and merge-extensible maps. Event JSDoc needs `@mode` and payload `@param`; scoped keys absent from payloads need `@dshScopeScan unsupported`. Public service methods document parameters and non-void returns.
 - **Switch on discriminant tags.** Closed unions end in `assertNever`; merge-extensible unions fall through a documented default.
diff --git a/README.i18n.yaml b/README.i18n.yaml
index 37a95ccdac..790812344d 100644
--- a/README.i18n.yaml
+++ b/README.i18n.yaml
@@ -3,4 +3,4 @@
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
 README.md: 53dd3896eb15800125673e7c44f7de02daca9376
-README.zh.md: 2119dabf0ae2d2e16e274e44fda7cbbaec146dc9
+README.zh.md: ab826f62658248249ec18c57b35c0065c0f909d1
diff --git a/README.zh.md b/README.zh.md
index 2119dabf0a..ab826f6265 100644
--- a/README.zh.md
+++ b/README.zh.md
@@ -2,7 +2,7 @@
 
 [English](README.md) | 中文
 
-**DeepSeek Harness SDK** 是用于构建 agent harness 的 SDK，采取基于插件的设计。
+**DeepSeek Harness SDK** 是用于构建 agent harness（智能体框架）的 SDK，采取基于插件的设计。
 
 ## 开发
 
diff --git a/docs/agent-lifecycle.md b/docs/agent-lifecycle.md
index fa2c4bd0fb..c753f59992 100644
--- a/docs/agent-lifecycle.md
+++ b/docs/agent-lifecycle.md
@@ -34,10 +34,19 @@ sequenceDiagram
   Session-->>SDK: <code>session/event</code> <code>assistant/chunk</code>*
   Driver->>Hooks: <code>agent/step-result</code> waterfall
   Driver->>Session: <code>assistant/message</code>
-  Driver->>Session: <code>tool/call</code>
-  Driver->>Tools: execute through pre and post waterfalls
-  Tools-->>Session: tool-owned events when applicable
-  Driver->>Session: <code>tool/result</code> and <code>step/end</code>
+  Driver->>Tools: classify pending call by executionMode
+  loop barriers and bounded rolling pool, reclassify before start
+    opt call starts
+      Driver->>Session: <code>tool/call</code>
+      Driver->>Tools: ordered pre, concurrent execute
+      Tools-->>Session: tool-owned events when applicable
+    end
+    opt next model-order result ready
+      Driver->>Tools: ordered post
+      Driver->>Session: <code>tool/result</code>
+    end
+  end
+  Driver->>Session: <code>step/end</code>
   Driver->>Hooks: <code>agent/turn-continuation</code> waterfall
   Driver->>Hooks: <code>agent/turn-stop</code> serial terminal checkpoint
   Driver->>Session: <code>turn/end</code>
@@ -45,6 +54,8 @@ sequenceDiagram
   Driver-->>SDK: <code>agent/status</code> idle
 ```
 
+The `assistant/message` edge records every successful provider call, including content-less and `max-tokens` finishes. Empty content stays out of derived history while the durable anchor retains usage and exact chunk provenance, including an explicit empty source set.
+
 SDK users that need replayable transcript data should consume `session/event`; `agent/*` is the live coordination surface for queue/status, prompt interception, request shaping, steering, continuation, and errors.
 
 Maintenance mode: curated Mermaid sequence; exact event signatures live in the generated Cordis catalog.
diff --git a/docs/architecture.md b/docs/architecture.md
index 4049049e25..d4abf10283 100644
--- a/docs/architecture.md
+++ b/docs/architecture.md
@@ -24,6 +24,7 @@ A harness is one [Cordis](cordis-primer.md) context. Packages contribute service
 | ctx key | Package family | Role |
 |---|---|---|
 | `ctx.llm` | [`llm/`](../packages/llm/README.md) | adapter registry and streaming model calls |
+| `ctx.tokenMeter` | [`llm/token-meter`](../packages/llm/token-meter/README.md) | singleton replay-aware request/surface pressure |
 | `ctx.bash` | [`bash/`](../packages/bash/README.md) | foreground/background command execution |
 | `ctx.sandbox` | [`sandbox/`](../packages/sandbox/README.md) | same-world process confinement (argv wrapping, per-call policy) |
 | `ctx.codeRuntime` | [`code-runtime/`](../packages/code-runtime/README.md) | model-written program execution |
@@ -32,9 +33,10 @@ A harness is one [Cordis](cordis-primer.md) context. Packages contribute service
 | `ctx.web` | [`web/`](../packages/web/README.md) | search/fetch provider registries |
 | `ctx.compact` | [`compact/`](../packages/compact/README.md) | session-log compaction |
 | `ctx.subagents` | [`subagent/`](../packages/subagent/README.md) | named delegation providers |
+| `ctx.tasks` | [`tasks/`](../packages/tasks/README.md) | background task registry + generic `task_*` control tools |
 | `ctx.workflows` | [`workflow/`](../packages/workflow/README.md) | script-driven multi-agent orchestration |
 | `ctx.sessionPersistence` | [`session-persistence/`](../packages/session-persistence/README.md) | durable storage for session logs |
-| `ctx.sessionQuery` | [`session-query/`](../packages/session-query/README.md) | live-preferred logical-corpus and exact-event reads |
+| `ctx.sessionQuery` | [`session-query/`](../packages/session-query/README.md) | live-preferred logical-corpus exact reads and relationship traces |
 
 ## Event
 
@@ -42,9 +44,9 @@ Events form the service extension API; see the exhaustive [events catalog](cordi
 
 ### Event Domains
 
-- **Session events** are durable, replayable facts. Turn and step boundaries, user input, assistant output, tool calls, tool results, steering, compaction records, and tool-owned durable facts append to the session log and flow through `session/event`.
-- **Agent events** carry the live `Agent` handle for status, diagnostics, prompt admission, call-config shaping, result validation, and continuation policy.
-- **Capability events** belong to the seam that owns the action. `tools/*`, `llm/*`, `system-prompt/*`, `fs/*`, and `subagent/*` let policy and adapters attach without importing the loop.
+- **Session events** are durable, replayable facts: boundaries, messages, tool activity, steering, compaction, and tool-owned records append to the log and flow through `session/event`.
+- **Agent events** carry the live `Agent` handle for status, diagnostics, prompt admission, request shaping, result validation, and continuation policy.
+- **Capability events** belong to their owning seam; `tools/*`, `llm/*`, `system-prompt/*`, `fs/*`, and `subagent/*` attach policy and adapters without importing the loop.
 
 ### Interception Semantics
 
@@ -52,7 +54,7 @@ Waterfall events behave like around-middleware: a listener delegates by calling
 
 ## Default Loop Lifecycle
 
-The shipped loop drains work, assembles requests, streams model answers, executes tools, applies continuation policy, and checkpoints state. Every pause is a service call or event available to plugins.
+The shipped loop drains work from prompt through checkpoint. Every pause is a service call or event available to plugins.
 
 A **session** is one agent's append-only event log. A **turn** drains one queued batch and runs until the model stops asking for tools and no plugin requests continuation. A **step** is one model request plus the tool executions caused by that response. In the flow below ([sequence companion](agent-lifecycle.md)), quoted names are durable session events and event names are extension points.
 
@@ -80,12 +82,13 @@ forever:
       agent/request (config only) -> log request/header -> llm/stream (frozen)
         'assistant/chunk'
       agent/step-result
-      'assistant/message'
-      each tool call:
-        'tool/call'
-        tools/pre-execute -> monotonic guards -> tools/execute -> tools/post-execute -> tools/result
-        'tool/result'
-      append post-tool context and steering
+      'assistant/message' (transformed content or empty success anchor after step-result rejection)
+      schedule tool calls by ctx.tools.executionMode:
+        exclusive -> one-call barrier
+        parallel -> rolling pool, <= maxParallelToolCalls in flight; reclassify before start
+        each start -> 'tool/call' -> ordered tools/pre-execute -> concurrent tools/execute
+        each model-order result -> ordered tools/post-execute -> 'tool/result'
+      append accepted tool-batch context after all recorded results, then steering
       'step/end'
       agent/turn-continuation
       agent/turn-stop (terminal policy)
@@ -94,23 +97,23 @@ forever:
     checkpoint persistence and notify idle/running status
 ```
 
-The loop renders one prompt assembly per step. Plugins contribute ordered sections, tool schemas, and `{{name}}` variables; unknown or valueless references fail the turn instead of shipping a hole. `dsh-system-prompt` owns the harness identity and default deployment persona; an agent-scoped persona may shadow the default. The loop supplies `model` and `cwd`. See the [prompt-ownership RFC](rfc/implemented/architecture/2026-07-05-prompt-variables-and-tool-guidance-ownership.md).
+Each step assembles ordered prompt sections, tool schemas, and `{{name}}` variables; unknown or valueless references fail the turn. `dsh-system-prompt` owns the harness identity and default persona, which an agent scope may shadow. The loop supplies `model` and `cwd` ([prompt-ownership RFC](rfc/implemented/architecture/2026-07-05-prompt-variables-and-tool-guidance-ownership.md)).
 
-Post-tool context lands after all tool results so tool-call/result adjacency stays stable. Steering drains between steps; ordinary leftover steering after a turn is re-queued as input. A terminal `agent/turn-stop` is the explicit exception: it runs after ordinary continuation and steering folding, then remains authoritative through turn close and flush so steering from those later listeners is discarded rather than becoming another step or turn; ordinary queued prompts are preserved.
+Context accepted during tool execution—including async `agent.inject()` notices and post-tool `additionalContexts`—waits for settlement, then follows every recorded result. Successful batches preserve call/result adjacency; interrupted batches drain that context before turn closure. Steering drains between steps; ordinary leftover steering becomes queued input. Terminal `agent/turn-stop` runs after continuation and steering folding, stays authoritative through turn close and flush, and discards later steering while preserving queued prompts.
 
 ### Failure Boundaries
 
-The turn is the containment boundary. A throwing listener, adapter error finish, or failed step ends the current turn with an error reason and reports live diagnostics through `agent/error`; it does not kill the driver loop. `cancel()` clears queued and steering work, aborts the active model/tool boundary when possible, and records the appropriate turn end. Disposal stops the loop, awaits quiescence, unregisters the agent, and lets service disposers drain.
+The turn is the containment boundary. A throwing listener, adapter error finish, or failed step ends it with an error reason and reports `agent/error` without killing the driver. `cancel()` clears queued and steering work, aborts the active model/tool boundary when possible, and records the turn end. Disposal stops the loop, awaits quiescence, unregisters the agent, and drains service disposers.
 
-Every session event is turn-enclosed. Reloading a crashed session preserves the interrupted tail and closes it with a synthetic `interrupted` turn end. A failure after the durable turn has closed reports through `agent/error` only because no safe in-turn position remains. A turn ends with one `TurnEndReason` (`completed`, `aborted`, `error`, `disposed`, `max-tokens`, `rejected`, or `interrupted`); per-variant semantics are in [session.md § TurnEndReasonMap](core-data-structures/session.md#why-a-turn-ended-turnendreasonmap).
+Every session event is turn-enclosed. Reloading preserves an interrupted tail and closes it with a synthetic `interrupted` turn end. Failures after durable turn close report only through `agent/error` because no safe in-turn position remains. Each turn has one `TurnEndReason`; [TurnEndReasonMap](core-data-structures/session.md#why-a-turn-ended-turnendreasonmap) owns the variants.
 
 ### Agent Handles
 
-`ctx.agents` owns live agents and returns an `AgentHandle { agent, dispose() }`. `Agent` is the API other plugins drive: `send()` queues work, `steer()` injects mid-turn content, `inject()` appends context and opens a one-shot injection turn when idle, `cancel()` is the public stop primitive, and `whenIdle()` observes quiescence. The caller fiber and concrete factory provider structurally co-own programmatic lifecycles; a consumer handle is the only non-structural teardown capability, and every owner reaches the same awaited disposer.
+`ctx.agents` owns live agents and returns `AgentHandle { agent, dispose() }`. Plugins drive `Agent` through `send()`, `steer()`, `inject()`, `cancel()`, and `whenIdle()`. The caller fiber and factory provider structurally co-own programmatic lifecycles; the consumer handle is the only other teardown capability. All owners await one disposer.
 
 ### Agent Scope
 
-Every live agent owns a scoped `agent.ctx`. Its registrations shadow same-named globals, receive only that agent's dispatches, and unwind with the agent. `CreateAgentOptions.setup(agentCtx)` composes the scope before publication. The [semantic-gates RFC](rfc/implemented/process/2026-07-14-typescript-program-backed-semantic-gates.md) defines typed resolvers that derive carrier checks from merged `Events` signatures and `scopeTarget`, eliminating the handwritten event table. See the [agent-scope RFC](rfc/implemented/architecture/2026-07-08-agent-scope-contexts.md); subagent composition controls are documented [separately](rfc/implemented/feature/2026-07-12-subagent-persona-tool-filter-and-depth.md).
+Every live agent owns a scoped `agent.ctx`. Its registrations shadow globals, receive only that agent's dispatches, and unwind with it; async effects such as background-task cleanup are awaited. `CreateAgentOptions.setup(agentCtx)` composes the scope before publication. Typed resolvers derive carrier checks from merged `Events` signatures and `scopeTarget` ([semantic-gates RFC](rfc/implemented/process/2026-07-14-typescript-program-backed-semantic-gates.md)). See the [agent-scope RFC](rfc/implemented/architecture/2026-07-08-agent-scope-contexts.md) and [subagent composition controls](rfc/implemented/feature/2026-07-12-subagent-persona-tool-filter-and-depth.md).
 
 ## State
 
@@ -124,9 +127,9 @@ Durability is a plugin concern. Persistence backends buffer synchronous `session
 
 ### Model Content
 
-Messages are arrays of typed content blocks (`text`, `reasoning`, `tool-call`, `tool-result`). The union derives from the merge-extensible `ContentBlockMap`; the same pattern types `MessageSource`, `FinishReason`, `TurnTrigger`, and `TurnEndReason`. New block types are coordinated across adapters, UI bridges, compaction pricing, and persistence, so block types remain a repo-wide contract.
+Messages contain typed blocks (`text`, `reasoning`, `tool-call`, `tool-result`) derived from merge-extensible `ContentBlockMap`; the same pattern types `MessageSource`, `FinishReason`, `TurnTrigger`, and `TurnEndReason`. New block types coordinate adapters, UI bridges, compaction pricing, token metering, and persistence as one repo-wide contract; replay measurement types live in [token-meter.md](core-data-structures/token-meter.md).
 
-Streaming is a raw chunk protocol (`block-start` through `finish`) with `BlockAssembler` as the shared chunk-to-block assembler. The loop logs raw chunks while assembling them for dispatch. `LlmAdapter` is the provider seam: subclass, implement `stream()`, and register with `ctx.llm.registerAdapter(models, adapter)`. StreamChunk conventions live in [llm-streaming.md](core-data-structures/llm-streaming.md).
+Streaming uses raw chunks (`block-start` through `finish`) and `BlockAssembler`. The loop logs and assembles chunks, storing provider/model provenance plus replay state. An `LlmAdapter` implements `stream()`, registers provider routes, and may expose selector metadata; it resolves and validates model ids. Replay state reaches targets only when both routes map to one adapter instance, which owns validation and conversion. The contract lives in [llm-streaming.md](core-data-structures/llm-streaming.md).
 
 ## Extension And Composition
 
@@ -134,11 +137,13 @@ Streaming is a raw chunk protocol (`block-start` through `finish`) with `BlockAs
 
 A swappable capability usually splits into **interface / implementation / consumer**: the interface owns its `ctx` key and events, an implementation registers a backend, and a consumer exposes model behavior through tools or prompts. Bash is the reference; the [capability graph](capability-seams.md) shows every family.
 
-Some seams bend the template deliberately. LLM keeps interface and consumer vocabulary together because adapters are the implementations. Filesystem adds policy gates around provider primitives. Web is one service with search and fetch provider registries, so provider swaps do not rename model tools. Skills and subagents use named provider registries; local skills scan project/user roots, and other providers can add embedded or remote catalogs without registry/tool changes. Subagents spawn fresh, fork from the parent's completed-turn prefix, or use ACP children ([subagent.md](core-data-structures/subagent.md)).
+Some seams bend the template deliberately: LLM combines interface and consumer because adapters implement it; filesystem wraps provider primitives with policy; web keeps search/fetch provider registries behind one service; skills and subagents use named providers. Subagents spawn fresh, fork a completed-turn prefix, or use ACP children ([subagent.md](core-data-structures/subagent.md)).
+
+`dsh-workspace-context` composes baselines on `agent/session-prefix` and appends `ctx.fs`-discovered nested changes on `tools/post-execute`; its [RFC](rfc/implemented/feature/2026-06-24-workspace-context.md) records isolation. `dsh-paths` owns shared paths.
 
 ### Bundles And Apps
 
-`dsh-agent-spine-demo` is the default composition bundle: one plugin loading the shared spine ([README](../packages/examples/agent-spine-demo/README.md)). App packages compose it with a front door and boot `bin`: `dsh-stdio-demo` for terminal REPL, and `dsh-acp-demo` for ACP over JSON-RPC stdio with no stdout logger ([ui/](../packages/ui/README.md)). `dsh-jsonrpc-agent` instead boots an external `cordis.yml`; the Python SDK injects the package default only when no explicit config channel is set and drives `dsh-jsonrpc` over line-delimited stdio JSON-RPC ([Python SDK](../python/README.md)). A deployment is a thin `cordis.yml` leaf: swappable backends, one app entry, and optional product tools ([examples/](../examples/AGENTS.md), [runnable wirings](cookbook/extension-cookbook.md#runnable-wirings), [graph atlas](graph-atlas.md)).
+`dsh-agent-spine-demo` bundles the default spine ([README](../packages/examples/agent-spine-demo/README.md)). `dsh-stdio-demo` adds a terminal front door; `dsh-acp-demo` adds stdout-pure ACP over JSON-RPC ([ui/](../packages/ui/README.md)). `dsh-jsonrpc-agent` boots external `cordis.yml`; the Python SDK supplies its default only without an explicit config channel and drives `dsh-jsonrpc` over line-delimited JSON-RPC ([Python SDK](../python/README.md)). Deployments remain thin leaves with swappable backends and optional product tools ([examples/](../examples/AGENTS.md), [runnable wirings](cookbook/extension-cookbook.md#runnable-wirings), [graph atlas](graph-atlas.md)).
 
 ### Where New Behavior Goes
 
@@ -149,6 +154,7 @@ New behavior should attach to a documented extension point; changing the shipped
 | Add a model provider | register an adapter on `ctx.llm` |
 | Add a model-facing capability | register a tool on `ctx.tools`; schemas flow into prompt assembly |
 | Add command execution | implement and register a `ctx.bash` backend |
+| Add a long-running/background capability | register the work on `ctx.tasks`; the generic `task_*` tools collect/stop it |
 | Add filesystem access or policy | implement a `ctx.fs` provider or listen on `fs/*` policy events |
 | Confine spawned processes | a `ctx.sandbox` backend; consumers wrap their argv before spawning |
 | Intercept prompts, requests, tool use, or continuation | listen on the relevant `agent/*` or `tools/*` waterfall; use serial `agent/turn-stop` for a monotonic terminal stop |
diff --git a/docs/capability-seams.md b/docs/capability-seams.md
index af22ba4c15..1aafbb2817 100644
--- a/docs/capability-seams.md
+++ b/docs/capability-seams.md
@@ -14,6 +14,8 @@ flowchart LR
   pkg_llm_replay["llm-replay"]
   pkg_agent_loop["agent-loop"]
   pkg_compact_basic["compact-basic"]
+  pkg_token_meter["token-meter"]
+  svc_tokenMeter["ctx.tokenMeter<br/>Replay token measurement"]
   pkg_session["session"]
   svc_sessions["ctx.sessions<br/>In-memory session store"]
   pkg_agent["agent"]
@@ -24,8 +26,11 @@ flowchart LR
   svc_sessionPersistence["ctx.sessionPersistence<br/>Durable session persistence seam"]
   pkg_session_persistence_jsonl["session-persistence-jsonl"]
   pkg_session_persistence_sqlite["session-persistence-sqlite"]
+  pkg_tool_bash["tool-bash"]
+  pkg_hooks_claude["hooks-claude"]
+  pkg_hooks_codex["hooks-codex"]
   pkg_acp["acp"]
-  svc_sessionQuery["ctx.sessionQuery<br/>Exact session-history reads"]
+  svc_sessionQuery["ctx.sessionQuery<br/>Exact session-history reads and traces"]
   pkg_system_prompt["system-prompt"]
   svc_systemPrompt["ctx.systemPrompt<br/>System prompt assembly registry"]
   pkg_tools["tools"]
@@ -33,7 +38,6 @@ flowchart LR
   pkg_tool_web["tool-web"]
   svc_tools["ctx.tools<br/>Tool registry and guarded execution pipeline"]
   pkg_tool_ask_user["tool-ask-user"]
-  pkg_tool_bash["tool-bash"]
   pkg_tool_cordis["tool-cordis"]
   pkg_tool_skill["tool-skill"]
   pkg_tool_subagent["tool-subagent"]
@@ -51,8 +55,7 @@ flowchart LR
   svc_bash["ctx.bash<br/>Bash executor seam"]
   pkg_bash_local["bash-local"]
   pkg_bash_sandbox["bash-sandbox"]
-  pkg_hooks_claude["hooks-claude"]
-  pkg_hooks_codex["hooks-codex"]
+  svc_bashEnv["ctx.bashEnv<br/>Managed bash environment registry"]
   pkg_sandbox["sandbox"]
   svc_sandbox["ctx.sandbox<br/>Process-sandbox seam"]
   pkg_sandbox_local["sandbox-local"]
@@ -75,12 +78,19 @@ flowchart LR
   pkg_subagent_fork["subagent-fork"]
   pkg_subagent_acp["subagent-acp"]
   pkg_subagent_mock["subagent-mock"]
+  pkg_tasks["tasks"]
+  svc_tasks["ctx.tasks<br/>Background task registry"]
+  pkg_tool_tasks["tool-tasks"]
   pkg_web["web"]
   svc_web["ctx.web<br/>Web access provider registry"]
   pkg_web_search_exa["web-search-exa"]
   pkg_web_search_perplexity["web-search-perplexity"]
   pkg_web_search_deepseek["web-search-deepseek"]
   pkg_web_fetch_local["web-fetch-local"]
+  pkg_spill["spill"]
+  svc_spillStore["ctx.spillStore<br/>Spill storage seam"]
+  pkg_spill_local["spill-local"]
+  pkg_spill_policy["spill-policy"]
   pkg_workflow["workflow"]
   svc_workflows["ctx.workflows<br/>Workflow script engine"]
   pkg_workflow_workerthread["workflow-workerthread"]
@@ -113,6 +123,8 @@ flowchart LR
   pkg_session_query --> svc_sessionQuery
   pkg_skill --> svc_skills
   pkg_skill_local --> svc_skills
+  pkg_spill --> svc_spillStore
+  pkg_spill_local --> svc_spillStore
   pkg_stdio_demo --> svc_userInteraction
   pkg_subagent --> svc_subagents
   pkg_subagent_acp --> svc_subagents
@@ -120,6 +132,9 @@ flowchart LR
   pkg_subagent_mock --> svc_subagents
   pkg_subagent_spawn --> svc_subagents
   pkg_system_prompt --> svc_systemPrompt
+  pkg_tasks --> svc_tasks
+  pkg_token_meter --> svc_tokenMeter
+  pkg_tool_bash --> svc_bashEnv
   pkg_tools --> svc_tools
   pkg_user_interaction --> svc_userInteraction
   pkg_web --> svc_web
@@ -149,7 +164,10 @@ flowchart LR
   svc_sandbox --> pkg_bash_sandbox
   svc_sessionPersistence --> pkg_acp
   svc_sessionPersistence --> pkg_agent_loop
+  svc_sessionPersistence --> pkg_hooks_claude
+  svc_sessionPersistence --> pkg_hooks_codex
   svc_sessionPersistence --> pkg_session_query
+  svc_sessionPersistence --> pkg_tool_bash
   svc_sessions --> pkg_agent
   svc_sessions --> pkg_agent_loop
   svc_sessions --> pkg_invariants
@@ -157,11 +175,16 @@ flowchart LR
   svc_sessions --> pkg_session_query
   svc_sessions --> pkg_subagent_inprocess
   svc_skills --> pkg_tool_skill
+  svc_spillStore --> pkg_spill_policy
   svc_subagents --> pkg_tool_subagent
   svc_systemPrompt --> pkg_agent_loop
   svc_systemPrompt --> pkg_tool_fs
   svc_systemPrompt --> pkg_tool_web
   svc_systemPrompt --> pkg_tools
+  svc_tasks --> pkg_tool_bash
+  svc_tasks --> pkg_tool_subagent
+  svc_tasks --> pkg_tool_tasks
+  svc_tokenMeter --> pkg_compact_basic
   svc_tools --> pkg_acp
   svc_tools --> pkg_agent_loop
   svc_tools --> pkg_tool_ask_user
@@ -183,9 +206,10 @@ flowchart LR
 | ctx key | Role | Owner | Implementations | Direct consumers | Companion plugins | Note |
 | --- | --- | --- | --- | --- | --- | --- |
 | `ctx.llm` | `seam` | [`llm`](../packages/llm/llm) | [`llm-deepseek`](../packages/llm/llm-deepseek), [`llm-pi-ai`](../packages/llm/llm-pi-ai), [`llm-replay`](../packages/support/llm-replay) | [`agent-loop`](../packages/core/agent-loop), [`compact-basic`](../packages/compact/compact-basic) | - | Adapters register provider implementations; the loop and compaction call the provider-neutral stream service. |
+| `ctx.tokenMeter` | `core` | [`token-meter`](../packages/llm/token-meter) | - | [`compact-basic`](../packages/compact/compact-basic) | - | Owns isolated per-session replay folds; pressure consumers share immutable revisioned measurements. |
 | `ctx.sessions` | `core` | [`session`](../packages/core/session) | - | [`agent-loop`](../packages/core/agent-loop), [`agent`](../packages/core/agent), [`session-persistence`](../packages/session-persistence/session-persistence), [`session-query`](../packages/session-query/session-query), [`subagent-inprocess`](../packages/subagent/subagent-inprocess), [`invariants`](../packages/support/invariants) | - | Owns append-only Session instances and emits the durable session event feed. |
-| `ctx.sessionPersistence` | `seam` | [`session-persistence`](../packages/session-persistence/session-persistence) | [`session-persistence-jsonl`](../packages/session-persistence/session-persistence-jsonl), [`session-persistence-sqlite`](../packages/session-persistence/session-persistence-sqlite) | [`agent-loop`](../packages/core/agent-loop), [`acp`](../packages/ui/acp), [`session-query`](../packages/session-query/session-query) | - | Backends persist the same SessionEvent vocabulary; apps choose a backend at composition time. |
-| `ctx.sessionQuery` | `seam` | [`session-query`](../packages/session-query/session-query) | - | - | - | Resolves live and optional persisted logs into one logical corpus for exact reads. |
+| `ctx.sessionPersistence` | `seam` | [`session-persistence`](../packages/session-persistence/session-persistence) | [`session-persistence-jsonl`](../packages/session-persistence/session-persistence-jsonl), [`session-persistence-sqlite`](../packages/session-persistence/session-persistence-sqlite) | [`agent-loop`](../packages/core/agent-loop), [`tool-bash`](../packages/bash/tool-bash), [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex), [`acp`](../packages/ui/acp), [`session-query`](../packages/session-query/session-query) | - | Backends persist the same SessionEvent vocabulary; apps choose a backend at composition time. |
+| `ctx.sessionQuery` | `seam` | [`session-query`](../packages/session-query/session-query) | - | - | - | Resolves live and optional persisted logs into one logical corpus for exact reads and relationship traces. |
 | `ctx.systemPrompt` | `core` | [`system-prompt`](../packages/core/system-prompt) | - | [`agent-loop`](../packages/core/agent-loop), [`tools`](../packages/core/tools), [`tool-fs`](../packages/fs/tool-fs), [`tool-web`](../packages/web/tool-web) | - | Collects prompt sections and model-facing tool schemas for each step. |
 | `ctx.tools` | `core` | [`tools`](../packages/core/tools) | - | [`agent-loop`](../packages/core/agent-loop), [`tool-ask-user`](../packages/ui/tool-ask-user), [`tool-bash`](../packages/bash/tool-bash), [`tool-cordis`](../packages/cordis/tool-cordis), [`tool-fs`](../packages/fs/tool-fs), [`tool-skill`](../packages/skill/tool-skill), [`tool-subagent`](../packages/subagent/tool-subagent), [`tool-todo`](../packages/todo/tool-todo), [`tool-web`](../packages/web/tool-web), [`acp`](../packages/ui/acp) | - | Registers capabilities, owns Code Mode transport, and routes calls through pre-policy, monotonic guards, around dispatch, post-policy, and final-result observation. |
 | `ctx.userInteraction` | `seam` | [`user-interaction`](../packages/ui/user-interaction) | [`stdio-demo`](../packages/examples/stdio-demo), [`acp`](../packages/ui/acp) | [`tool-ask-user`](../packages/ui/tool-ask-user), [`stdio-demo`](../packages/examples/stdio-demo), [`acp`](../packages/ui/acp) | - | UI front doors provide the active human-answer provider; tool-ask-user pauses a tool call on the provider-neutral ask() promise. |
@@ -193,6 +217,7 @@ flowchart LR
 | `ctx.agents` | `core` | [`agent`](../packages/core/agent) | - | [`agent-loop`](../packages/core/agent-loop), [`acp`](../packages/ui/acp), [`subagent-inprocess`](../packages/subagent/subagent-inprocess), [`stdio-demo`](../packages/examples/stdio-demo), [`invariants`](../packages/support/invariants) | - | Owns live Agent handles and the create/resume factory seam. |
 | `ctx.agentLoop` | `bundle` | [`agent-loop`](../packages/core/agent-loop) | - | [`agent-spine-demo`](../packages/examples/agent-spine-demo) | - | The one concrete loop plugin; extension packages depend on dsh-agent events and services, not on this package. |
 | `ctx.bash` | `seam` | [`bash`](../packages/bash/bash) | [`bash-local`](../packages/bash/bash-local), [`bash-sandbox`](../packages/bash/bash-sandbox) | [`tool-bash`](../packages/bash/tool-bash), [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex) | - | The model-facing bash tools and hook bridges consume this seam; sandboxed or remote executors replace bash-local without touching them. |
+| `ctx.bashEnv` | `core` | [`tool-bash`](../packages/bash/tool-bash) | - | - | - | Plugins declare effect-scoped DSH_* facts; tool-bash collects one trusted snapshot per execution and the executor rebuilds the namespace. |
 | `ctx.sandbox` | `seam` | [`sandbox`](../packages/sandbox/sandbox) | [`sandbox-local`](../packages/sandbox/sandbox-local) | [`bash-sandbox`](../packages/bash/bash-sandbox) | - | Consumers hand over the exact argv they are about to spawn; same-world backends wrap it under a per-call policy and report enforcement. |
 | `ctx.approval` | `seam` | `approval` | [`acp`](../packages/ui/acp) | [`tools`](../packages/core/tools), [`tool-bash`](../packages/bash/tool-bash) | - | One-shot permission decisions dispatched over the `approval/request` waterfall; answerers are listeners (the ACP bridge for its own agents), absence fails closed to `unavailable`. |
 | `ctx.permission` | `core` | [`permission`](../packages/ui/permission) | - | [`acp`](../packages/ui/acp) | - | User-facing preset table (`workspace-write`/`danger-full-access`) bundling the sandbox-mode and approval-policy knobs; a switch writes one `permission/preset` event through to both knob events. |
@@ -200,7 +225,9 @@ flowchart LR
 | `ctx.fs` | `seam` | [`fs`](../packages/fs/fs) | [`fs-local`](../packages/fs/fs-local) | [`tool-fs`](../packages/fs/tool-fs) | [`fs-policy`](../packages/fs/fs-policy) | tool-fs executes read/write/edit through ctx.fs; fs-policy contributes observed-state checks through the fs/* event gate. |
 | `ctx.compact` | `seam` | [`compact`](../packages/compact/compact) | [`compact-basic`](../packages/compact/compact-basic) | [`compact-basic`](../packages/compact/compact-basic) | - | The basic backend currently consumes the pre-step event directly; a model-facing compact tool remains deferred. |
 | `ctx.subagents` | `seam` | [`subagent`](../packages/subagent/subagent) | [`subagent-spawn`](../packages/subagent/subagent-spawn), [`subagent-fork`](../packages/subagent/subagent-fork), [`subagent-acp`](../packages/subagent/subagent-acp), [`subagent-mock`](../packages/support/subagent-mock) | [`tool-subagent`](../packages/subagent/tool-subagent) | - | Providers implement transports; tool-subagent exposes one configured provider as a model-facing tool name. |
+| `ctx.tasks` | `core` | [`tasks`](../packages/tasks/tasks) | - | [`tool-bash`](../packages/bash/tool-bash), [`tool-subagent`](../packages/subagent/tool-subagent), [`tool-tasks`](../packages/tasks/tool-tasks) | - | Producers (tool-bash background commands, tool-subagent background delegations) register running work; tool-tasks is the model-facing control surface that reads, lists, and kills it. |
 | `ctx.web` | `seam` | [`web`](../packages/web/web) | [`web-search-exa`](../packages/web/web-search-exa), [`web-search-perplexity`](../packages/web/web-search-perplexity), [`web-search-deepseek`](../packages/web/web-search-deepseek), [`web-fetch-local`](../packages/web/web-fetch-local) | [`tool-web`](../packages/web/tool-web) | - | Search and fetch providers register into one ctx.web seam; tool-web owns the stable model-facing names. |
+| `ctx.spillStore` | `seam` | [`spill`](../packages/spill/spill) | [`spill-local`](../packages/spill/spill-local) | [`spill-policy`](../packages/spill/spill-policy) | - | The backend saves oversized tool text and returns a model-facing locator plus retrieval hint; spill-policy is the tools/post-execute consumer that decides when to spill. |
 | `ctx.workflows` | `seam` | [`workflow`](../packages/workflow/workflow) | [`workflow-workerthread`](../packages/workflow/workflow-workerthread) | [`tool-workflow`](../packages/workflow/tool-workflow) | - | One engine per context (bash shape, no named-provider registry); the worker-thread engine fans agent() calls out through ctx.subagents. |
 
 Maintenance mode: hybrid: services are discovered from Cordis declarations; interface/implementation/consumer roles are classified in `scripts/gen-doc-graphs.ts` with a completeness guard.
diff --git a/docs/config-catalog.md b/docs/config-catalog.md
index af8ef0832c..4cac932a6f 100644
--- a/docs/config-catalog.md
+++ b/docs/config-catalog.md
@@ -11,11 +11,13 @@ A `Requires:` line lists the service keys the plugin `inject`s: its `cordis.yml`
 
 ## `@deepseek-ai/dsh-acp`
 
-Requires: `agents` · `sessions` · `sessionPersistence` · `tools` · `userInteraction`
+Requires: `agents` · `sessions` · `sessionPersistence` · `tools` · `userInteraction` · `llm` · `systemPrompt`
 
 ```ts config-catalog
 /** Plugin config: the agent template ACP sessions are created from. */
 export interface AcpConfig {
+  /** Provider route for created agents. */
+  provider?: string
   /** Model name for created agents (must have a registered adapter). */
   model?: string
   /** Runtime-only transport override for tests; production uses stdio. */
@@ -25,13 +27,13 @@ export interface AcpConfig {
 
 Depends on: `Stream` (`@agentclientprotocol/sdk`)
 
-Source: [`packages/ui/acp/src/index.ts:203`](../packages/ui/acp/src/index.ts)
+Source: [`packages/ui/acp/src/index.ts:208`](../packages/ui/acp/src/index.ts)
 
 ## `@deepseek-ai/dsh-acp-demo`
 
 ```ts config-catalog
 /**
- * App config: the swappable per-deployment values. `model` configures the
+ * App config: the swappable per-deployment values. `provider` and `model` configure the
  * agent template the ACP bridge creates each session's agent from (NOT a
  * pre-created agent — ACP creates agents at `session/new`); `persona` is the
  * deployment persona (forwarded to the system-prompt plugin); `toolOrder` is
@@ -40,32 +42,49 @@ Source: [`packages/ui/acp/src/index.ts:203`](../packages/ui/acp/src/index.ts)
  * through agent-spine-demo); `persistenceRoot` is the JSONL backend's directory.
  */
 export interface Config {
+  /** Provider route for ACP-created agents. */
+  provider: string
   /** Model name for ACP-created agents (must have a registered adapter). */
   model: string
+  /** Bundled agent-loop concurrency cap; `1` is serial and omission uses its default. */
+  maxParallelToolCalls?: number
   /** Deployment persona (the system-prompt plugin's `persona` config). */
   persona?: string
   /** Explicit model-facing tool order (the system-prompt plugin's `toolOrder` config; see dsh-system-prompt). */
   toolOrder?: string[]
   /** Tool-registry config — its presentation `mode` (forwarded through agent-spine-demo; see dsh-tools). */
   tools?: ToolsConfig
+  /** DeepSeek Harness home directory exposed to bash and used for local skill discovery. */
+  dshHome?: string
   /** Directory the JSONL session backend writes under. Defaults to `./.sessions`. */
   persistenceRoot?: string
+  /** Controls automatic AGENTS.md/CLAUDE.md loading; configure a byte budget or set `false`. */
+  workspaceContext: agentCore.Config['workspaceContext']
   /** Skill registry, local-provider, and model-facing consumer config forwarded to agent-spine-demo. */
   skills?: agentCore.SkillConfig
+  /** Model-facing bash tool config forwarded through agent-core. */
+  toolBash?: NonNullable<agentCore.Config['toolBash']>
+  /** Generic background-task control-tool config forwarded through agent-core. */
+  toolTasks?: NonNullable<agentCore.Config['toolTasks']>
 }
 ```
 
 Depends on: [`agentCore`](../packages/examples/agent-spine-demo/src/index.ts) · [`ToolsConfig`](#deepseek-aidsh-tools)
 
-Source: [`packages/examples/acp-demo/src/index.ts:31`](../packages/examples/acp-demo/src/index.ts)
+Source: [`packages/examples/acp-demo/src/index.ts:32`](../packages/examples/acp-demo/src/index.ts)
 
 ## `@deepseek-ai/dsh-agent-loop`
 
 Requires: `agents` · `sessions` · `llm` · `tools` · `systemPrompt`
 
 ```ts config-catalog
-/** Plugin configuration for declarative startup agents. */
+/** Agent-loop plugin configuration. */
 export interface Config {
+  /**
+   * Maximum parallel-safe calls in flight per agent step. `1` is serial;
+   * omission defaults to {@link DEFAULT_MAX_PARALLEL_TOOL_CALLS}.
+   */
+  maxParallelToolCalls?: number
   /** Agents created or resumed at plugin startup. */
   agents: (AgentOptions & {
     /** Registry identity for the live agent. */
@@ -80,31 +99,47 @@ export interface Config {
 
 Depends on: [`AgentId`](../packages/core/agent/src/index.ts) · [`AgentOptions`](../packages/core/agent/src/index.ts) · [`SessionId`](../packages/core/session/src/index.ts)
 
-Source: [`packages/core/agent-loop/src/index.ts:322`](../packages/core/agent-loop/src/index.ts)
+Source: [`packages/core/agent-loop/src/index.ts:334`](../packages/core/agent-loop/src/index.ts)
 
 ## `@deepseek-ai/dsh-agent-spine-demo`
 
 ```ts config-catalog
 /**
- * Bundle config: each field forwarded verbatim to the child that owns it — `agents` to the
- * agent loop (an app that pre-creates no agents, like the ACP bridge, omits it),
- * `persona` and `toolOrder` to the system-prompt plugin (the deployment's persona section and
- * the explicit model-facing tool order), the `tools` object to the tool registry (its
- * presentation `mode`), and `skills` to the skill registry/local provider/tool consumer.
- * The schema intersects the owners' schemas, which supply defaults for every
- * optional input and keep validation from drifting.
+ * Bundle config: each field forwarded verbatim to the child that owns it —
+ * `agents` to the agent loop (an app that pre-creates no agents, like the ACP
+ * bridge, simply omits it), `persona` and `toolOrder` to the system-prompt
+ * plugin (the deployment's persona section and the explicit model-facing tool
+ * order), the `tools` object to the tool registry (its presentation `mode`),
+ * `dshHome` to bash environment and local skill discovery, `skills` to the
+ * skill registry/local provider/tool consumer, `workspaceContext` to the
+ * workspace-context loader, and `toolBash`/`toolTasks` to the model-facing tool
+ * plugins this bundle owns. Owner schemas supply defaults for optional input;
+ * workspace context instead requires an explicit byte budget or `false` because
+ * it changes model-visible input. Producer opt-in stays producer-local:
+ * `toolBash` configures bash only; independently composed producers keep their
+ * own config.
  */
 export interface Config {
   /** The agent-loop `agents` list (see dsh-agent-loop's `Config`). */
   agents?: AgentLoopConfig['agents']
+  /** Agent-loop concurrency cap; `1` is serial. */
+  maxParallelToolCalls?: AgentLoopConfig['maxParallelToolCalls']
   /** The deployment persona (see dsh-system-prompt's `Config`). */
   persona?: SystemPromptConfig['persona']
   /** The explicit model-facing tool order (see dsh-system-prompt's `Config`). */
   toolOrder?: SystemPromptConfig['toolOrder']
   /** The tool registry's config — its presentation `mode` (see dsh-tools' `Config`). */
   tools?: ToolsConfig
+  /** DeepSeek Harness home directory shared by shell context and local skill discovery. */
+  dshHome?: string
+  /** Workspace-context loader controls with an explicit byte budget; set `false` for hermetic prompts. */
+  workspaceContext: workspaceContext.Config | false
   /** Skill registry, local provider, and model-facing consumer config. */
   skills?: SkillConfig
+  /** Model-facing bash tool config, including this producer's background opt-in. */
+  toolBash?: toolBash.Config
+  /** Generic background-task control-tool wait bounds. */
+  toolTasks?: toolTasks.Config
 }
 
 /** Skill bundle config forwarded to the registry, local provider, and model-facing consumer. */
@@ -118,9 +153,9 @@ export interface SkillConfig {
 }
 ```
 
-Depends on: [`AgentLoopConfig`](#deepseek-aidsh-agent-loop) · [`SkillLocal`](../packages/skill/skill-local/src/index.ts) · [`SkillRegistryConfig`](#deepseek-aidsh-skill) · [`SystemPromptConfig`](#deepseek-aidsh-system-prompt) · [`ToolsConfig`](#deepseek-aidsh-tools) · [`toolSkill`](../packages/skill/tool-skill/src/index.ts)
+Depends on: [`AgentLoopConfig`](#deepseek-aidsh-agent-loop) · [`SkillLocal`](../packages/skill/skill-local/src/index.ts) · [`SkillRegistryConfig`](#deepseek-aidsh-skill) · [`SystemPromptConfig`](#deepseek-aidsh-system-prompt) · [`toolBash`](../packages/bash/tool-bash/src/index.ts) · [`ToolsConfig`](#deepseek-aidsh-tools) · [`toolSkill`](../packages/skill/tool-skill/src/index.ts) · [`toolTasks`](../packages/tasks/tool-tasks/src/index.ts) · [`workspaceContext`](../packages/context/workspace-context/src/index.ts)
 
-Source: [`packages/examples/agent-spine-demo/src/index.ts:46`](../packages/examples/agent-spine-demo/src/index.ts)
+Source: [`packages/examples/agent-spine-demo/src/index.ts:57`](../packages/examples/agent-spine-demo/src/index.ts)
 
 ## `@deepseek-ai/dsh-bash-local`
 
@@ -140,7 +175,7 @@ export interface Config {
 }
 ```
 
-Source: [`packages/bash/bash-local/src/index.ts:18`](../packages/bash/bash-local/src/index.ts)
+Source: [`packages/bash/bash-local/src/index.ts:17`](../packages/bash/bash-local/src/index.ts)
 
 ## `@deepseek-ai/dsh-bash-sandbox`
 
@@ -151,7 +186,7 @@ Requires: `sandbox`
  * Plugin config: the local executor's knobs plus the sandbox policy. All
  * optional — `static Config` supplies the defaults (`mode: 'read-only'` is the
  * fail-safe default; an example that wants a workspace-writable agent opts in
- * explicitly). The runner choice is NOT configured here: which platform
+ * explicitly). The runner choice is not configured here: which platform
  * backend confines the command is the `ctx.sandbox` provider's config.
  */
 export interface Config extends LocalConfig {
@@ -167,7 +202,7 @@ export interface Config extends LocalConfig {
 
 Depends on: [`LocalConfig`](#deepseek-aidsh-bash-local) · [`SandboxMode`](core-data-structures/sandbox.md)
 
-Source: [`packages/bash/bash-sandbox/src/index.ts:26`](../packages/bash/bash-sandbox/src/index.ts)
+Source: [`packages/bash/bash-sandbox/src/index.ts:27`](../packages/bash/bash-sandbox/src/index.ts)
 
 ## `@deepseek-ai/dsh-code-runtime-worker`
 
@@ -207,44 +242,29 @@ Source: [`packages/code-runtime/code-runtime-worker/src/index.ts:21`](../package
 
 ## `@deepseek-ai/dsh-compact-basic`
 
-Requires: `llm`
+Requires: `llm` · `tokenMeter`
 
 ```ts config-catalog
-/**
- * Backend configuration. Every knob is REQUIRED except `auto` and
- * `charsPerToken`: there is no concrete data yet to justify default
- * thresholds/budgets, so a consumer must state each value explicitly rather
- * than inherit a guessed default. `auto` alone defaults to `true`
- * (auto-compaction is the intended posture), and `charsPerToken` defaults to
- * the English-text heuristic its estimator was calibrated on.
- */
+/** Basic compaction configuration; every common field has a deployment default. */
 export interface BasicCompactConfig {
-  /** Context window size in tokens. */
-  contextWindow: number
-  /** Compact when estimated token usage exceeds this fraction of context window. */
-  thresholdRatio: number
-  /** Number of tokens of recent context to retain during compaction. */
-  retainTokens: number
-  /** Model to use for summarization (`''` — uses the agent's model). */
-  summarizationModel: string
-  /** Provider generation cap for the summarization call. */
-  maxTokens: number
-  /** Extra compaction attempts when the first compacted surface is still over threshold. */
-  compactionRetries: number
-  /** Enable automatic compaction on the `agent/pre-step` seam (default true). */
+  /** Compact at this fraction of the token meter's context window. Defaults to `0.8`. */
+  thresholdRatio?: number
+  /** Recent surface tokens retained verbatim. Defaults to `floor(contextWindow * 0.16)`. */
+  retainTokens?: number
+  /** Summary provider; `''` resolves the latest routed pair, then the agent pair. Defaults to `''`. */
+  summarizationProvider?: string
+  /** Summary model; `''` resolves the latest routed pair, then the agent pair. Defaults to `''`. */
+  summarizationModel?: string
+  /** Provider generation cap for summarization. Defaults to `8192`. */
+  maxTokens?: number
+  /** Extra attempts after the first compaction when pressure remains above threshold. Defaults to `1`. */
+  compactionRetries?: number
+  /** Enable the automatic `agent/pre-step` pressure listener. Defaults to `true`. */
   auto?: boolean
-  /**
-   * Text density for the token estimator: estimated tokens = chars /
-   * `charsPerToken`. Defaults to 4 (typical English text). A CJK-heavy
-   * deployment should set ~1-2 — CJK runs at roughly 1-2 chars per token, so
-   * the default UNDERestimates several-fold and compaction fires far too late.
-   * May be fractional.
-   */
-  charsPerToken?: number
 }
 ```
 
-Source: [`packages/compact/compact-basic/src/types.ts:20`](../packages/compact/compact-basic/src/types.ts)
+Source: [`packages/compact/compact-basic/src/types.ts:8`](../packages/compact/compact-basic/src/types.ts)
 
 ## `@deepseek-ai/dsh-fs-local`
 
@@ -256,7 +276,7 @@ export interface Config {
 }
 ```
 
-Source: [`packages/fs/fs-local/src/index.ts:35`](../packages/fs/fs-local/src/index.ts)
+Source: [`packages/fs/fs-local/src/index.ts:38`](../packages/fs/fs-local/src/index.ts)
 
 ## `@deepseek-ai/dsh-hooks-claude`
 
@@ -292,7 +312,7 @@ export interface Config {
 }
 ```
 
-Source: [`packages/hooks/hooks-claude/src/index.ts:43`](../packages/hooks/hooks-claude/src/index.ts)
+Source: [`packages/hooks/hooks-claude/src/index.ts:44`](../packages/hooks/hooks-claude/src/index.ts)
 
 ## `@deepseek-ai/dsh-hooks-codex`
 
@@ -317,7 +337,7 @@ export interface Config {
 }
 ```
 
-Source: [`packages/hooks/hooks-codex/src/index.ts:41`](../packages/hooks/hooks-codex/src/index.ts)
+Source: [`packages/hooks/hooks-codex/src/index.ts:42`](../packages/hooks/hooks-codex/src/index.ts)
 
 ## `@deepseek-ai/dsh-jsonrpc`
 
@@ -355,47 +375,70 @@ export interface Config {
   apiKey?: string
   /** Endpoint base; falls back to $DEEPSEEK_BASE_URL, then the public API. */
   baseURL?: string
-  /** Model names to register (sent verbatim on the wire). */
-  models?: string[]
   /** Thinking-mode default for every request (provider default: enabled). */
   thinking?: 'enabled' | 'disabled'
   /** Thinking effort (only meaningful with thinking enabled). */
   reasoningEffort?: 'high' | 'max'
+  /** Advisory models shown by discovery consumers; defaults to V4 Flash and V4 Pro. */
+  models?: DeepSeekCatalogModel[]
+}
+
+/** One optional model entry advertised by the hand-written adapter. */
+export interface DeepSeekCatalogModel {
+  /** Wire model id accepted by the configured endpoint. */
+  id: string
+  /** Selector label; defaults to {@link id}. */
+  name?: string
+  /** Optional selector detail for deployments with similar model variants. */
+  description?: string
 }
 ```
 
-Source: [`packages/llm/llm-deepseek/src/index.ts:30`](../packages/llm/llm-deepseek/src/index.ts)
+Source: [`packages/llm/llm-deepseek/src/index.ts:36`](../packages/llm/llm-deepseek/src/index.ts)
 
 ## `@deepseek-ai/dsh-llm-pi-ai`
 
 Requires: `llm`
 
 ```ts config-catalog
-/**
- * Plugin config, validated by the same-named schemastery schema. Every field
- * is optional in yml: credentials/endpoint fall back to the environment (a
- * missing API key fails plugin load, not the first call).
- */
+/** Plugin configuration: the non-empty provider profiles this instance owns. */
 export interface Config {
-  /** API key; falls back to $DEEPSEEK_API_KEY. Required one way or the other. */
-  apiKey?: string
-  /** Endpoint base; falls back to $DEEPSEEK_BASE_URL, then the public API. */
-  baseURL?: string
-  /** Model names to register (sent verbatim on the wire). */
-  models?: string[]
-  /**
-   * Thinking level for every request: 'off' disables thinking mode; 'high'
-   * and 'xhigh' (wire 'max') set the effort. Omitted = provider default
-   * (thinking enabled), matching llm-deepseek's omission semantics.
-   */
-  reasoning?: PiAiReasoning
+  /** Non-empty set of pi-ai provider routes this adapter instance owns. */
+  providers: PiAiProviderProfile[]
 }
 
-/** Reasoning levels surfaced by this adapter (DeepSeek wire: high|max). */
-export type PiAiReasoning = 'off' | 'high' | 'xhigh'
+/** Configuration for one pi-ai provider route. */
+export interface PiAiProviderProfile {
+  /** pi-ai provider catalog name and Harness route key. */
+  provider: string
+  /** Provider credential; when absent pi-ai uses its provider-native ambient discovery. */
+  apiKey?: string
+  /** Override the selected catalog model's endpoint without changing its protocol metadata. */
+  baseURL?: string
+  /** Provider request headers; Harness attribution wins reserved names. */
+  headers?: Record<string, string>
+  /** Provider-neutral pi-ai reasoning level. */
+  reasoning?: ThinkingLevel
+  /** Token budgets used by reasoning providers that support them. */
+  thinkingBudgets?: ThinkingBudgets
+  /** Prompt-cache retention preference. */
+  cacheRetention?: CacheRetention
+  /** Streaming transport preference. */
+  transport?: Transport
+  /** HTTP/provider SDK timeout in milliseconds. */
+  timeoutMs?: number
+  /** WebSocket connection timeout in milliseconds. */
+  websocketConnectTimeoutMs?: number
+  /** Provider SDK retry count. */
+  maxRetries?: number
+  /** Maximum provider-requested retry delay in milliseconds. */
+  maxRetryDelayMs?: number
+}
 ```
 
-Source: [`packages/llm/llm-pi-ai/src/index.ts:37`](../packages/llm/llm-pi-ai/src/index.ts)
+Depends on: `CacheRetention` (`@earendil-works/pi-ai`) · `ThinkingBudgets` (`@earendil-works/pi-ai`) · `ThinkingLevel` (`@earendil-works/pi-ai`) · `Transport` (`@earendil-works/pi-ai`)
+
+Source: [`packages/llm/llm-pi-ai/src/config.ts:40`](../packages/llm/llm-pi-ai/src/config.ts)
 
 ## `@deepseek-ai/dsh-llm-replay`
 
@@ -414,10 +457,32 @@ export interface Config {
    * a nested-agent scenario; absent/empty for a single-session scenario.
    */
   childFiles?: string[]
+  /** Optional replay-only provider catalog; absent or empty selects catch-all waterfall replay. */
+  providers?: ReplayProviderConfig[]
+}
+
+/** One provider route exposed by the replay adapter. */
+export interface ReplayProviderConfig {
+  /** Provider route used for replay requests. */
+  id: string
+  /** Selector label; defaults to {@link id}. */
+  name?: string
+  /** Advisory models exposed to clients such as ACP editors. */
+  models?: ReplayModelConfig[]
+}
+
+/** One model exposed by a replay-only provider catalog. */
+export interface ReplayModelConfig {
+  /** Model id used for replay requests. */
+  id: string
+  /** Selector label; defaults to {@link id}. */
+  name?: string
+  /** Optional selector description. */
+  description?: string
 }
 ```
 
-Source: [`packages/support/llm-replay/src/index.ts:306`](../packages/support/llm-replay/src/index.ts)
+Source: [`packages/support/llm-replay/src/index.ts:375`](../packages/support/llm-replay/src/index.ts)
 
 ## `@deepseek-ai/dsh-mcp-client`
 
@@ -560,7 +625,7 @@ export interface Config {
 }
 ```
 
-Source: [`packages/sandbox/sandbox-local/src/index.ts:20`](../packages/sandbox/sandbox-local/src/index.ts)
+Source: [`packages/sandbox/sandbox-local/src/index.ts:19`](../packages/sandbox/sandbox-local/src/index.ts)
 
 ## `@deepseek-ai/dsh-session-persistence-jsonl`
 
@@ -578,7 +643,7 @@ export interface Config {
 }
 ```
 
-Source: [`packages/session-persistence/session-persistence-jsonl/src/index.ts:23`](../packages/session-persistence/session-persistence-jsonl/src/index.ts)
+Source: [`packages/session-persistence/session-persistence-jsonl/src/index.ts:24`](../packages/session-persistence/session-persistence-jsonl/src/index.ts)
 
 ## `@deepseek-ai/dsh-session-persistence-sqlite`
 
@@ -613,14 +678,14 @@ export interface Config {
 export type JournalMode = 'wal' | 'delete' | 'truncate' | 'persist'
 ```
 
-Source: [`packages/session-persistence/session-persistence-sqlite/src/index.ts:38`](../packages/session-persistence/session-persistence-sqlite/src/index.ts)
+Source: [`packages/session-persistence/session-persistence-sqlite/src/index.ts:39`](../packages/session-persistence/session-persistence-sqlite/src/index.ts)
 
 ## `@deepseek-ai/dsh-session-query`
 
 Requires: `sessions`
 
 ```ts config-catalog
-/** Configuration for exact session-query reads. */
+/** Configuration for exact session-query reads and traces. */
 export interface Config {
   /** Maximum accepted raw read context on either side. Defaults to 50. */
   readWindowMax?: number
@@ -657,7 +722,41 @@ export interface Config {
 }
 ```
 
-Source: [`packages/skill/skill-local/src/index.ts:39`](../packages/skill/skill-local/src/index.ts)
+Source: [`packages/skill/skill-local/src/index.ts:40`](../packages/skill/skill-local/src/index.ts)
+
+## `@deepseek-ai/dsh-spill-local`
+
+```ts config-catalog
+/** Plugin config (all optional — `static Config` supplies the defaults). */
+export interface Config {
+  /**
+   * Root directory for spill files. Omitted uses a lazily-created private
+   * (0700) per-process directory under the OS temp dir — the safe default for
+   * a local deployment. Set it to keep spill files under a known location.
+   */
+  root?: string
+}
+```
+
+Source: [`packages/spill/spill-local/src/index.ts:22`](../packages/spill/spill-local/src/index.ts)
+
+## `@deepseek-ai/dsh-spill-policy`
+
+Requires: `tools`
+
+```ts config-catalog
+/** Plugin config. */
+export interface Config {
+  /**
+   * The model-facing context cap for a plain-text tool result, in UTF-8 bytes.
+   * Omitted disables the policy entirely (no-op). When set, a result larger than
+   * this is spilled and replaced with a preview derived from this same budget.
+   */
+  maxInlineBytes?: number
+}
+```
+
+Source: [`packages/spill/spill-policy/src/index.ts:45`](../packages/spill/spill-policy/src/index.ts)
 
 ## `@deepseek-ai/dsh-stdio`
 
@@ -680,7 +779,7 @@ Source: [`packages/ui/stdio/src/index.ts:30`](../packages/ui/stdio/src/index.ts)
 ```ts config-catalog
 /**
  * App config: the swappable per-demo values, each routed to where the app wires
- * it. `model`/`resumeSessionId` configure the pre-created `main` agent (through
+ * it. `provider`/`model`/`resumeSessionId` configure the pre-created `main` agent (through
  * {@link @deepseek-ai/dsh-agent-spine-demo}'s forwarded `agents` list); `persona` is
  * the deployment persona (forwarded to the system-prompt plugin); `toolOrder`
  * is the explicit model-facing tool order (forwarded to the system-prompt plugin);
@@ -689,32 +788,44 @@ Source: [`packages/ui/stdio/src/index.ts:30`](../packages/ui/stdio/src/index.ts)
  * `welcome` is the UI banner.
  */
 export interface Config {
+  /** Provider route for the `main` agent. */
+  provider: string
   /** Model name for the `main` agent (must have a registered adapter). */
   model: string
+  /** Bundled agent-loop concurrency cap; `1` is serial and omission uses its default. */
+  maxParallelToolCalls?: number
   /** Deployment persona (the system-prompt plugin's `persona` config). */
   persona?: string
   /** Explicit model-facing tool order (the system-prompt plugin's `toolOrder` config; see dsh-system-prompt). */
   toolOrder?: string[]
   /** Tool-registry config — its presentation `mode` (forwarded through agent-spine-demo; see dsh-tools). */
   tools?: ToolsConfig
+  /** DeepSeek Harness home directory exposed to bash and used for local skill discovery. */
+  dshHome?: string
   /** Directory the JSONL session backend writes under. Defaults to `./.sessions`. */
   persistenceRoot?: string
   /** stdin-chat banner printed once on start. Defaults to `'ready.'`. */
   welcome?: string
   /** Skill registry, local-provider, and model-facing consumer config forwarded to agent-spine-demo. */
   skills?: agentCore.SkillConfig
+  /** Model-facing bash tool config forwarded through agent-core. */
+  toolBash?: NonNullable<agentCore.Config['toolBash']>
+  /** Generic background-task control-tool config forwarded through agent-core. */
+  toolTasks?: NonNullable<agentCore.Config['toolTasks']>
   /**
    * If set, the `main` agent RESUMES this persisted session id instead of
    * starting fresh. Sourced from an env var in the leaf `cordis.yml`
    * (`resumeSessionId: !!js process.env.RESUME_SESSION_ID`).
    */
   resumeSessionId?: string
+  /** Controls automatic AGENTS.md/CLAUDE.md loading; configure a byte budget or set `false`. */
+  workspaceContext: agentCore.Config['workspaceContext']
 }
 ```
 
 Depends on: [`agentCore`](../packages/examples/agent-spine-demo/src/index.ts) · [`ToolsConfig`](#deepseek-aidsh-tools)
 
-Source: [`packages/examples/stdio-demo/src/index.ts:36`](../packages/examples/stdio-demo/src/index.ts)
+Source: [`packages/examples/stdio-demo/src/index.ts:37`](../packages/examples/stdio-demo/src/index.ts)
 
 ## `@deepseek-ai/dsh-subagent-acp`
 
@@ -849,19 +960,47 @@ Source: [`packages/core/system-prompt/src/index.ts:143`](../packages/core/system
 
 ## `@deepseek-ai/dsh-time-context`
 
-Requires: `systemPrompt`
+Requires: `agents`
 
 ```ts config-catalog
-/** Request-time clock formatting and refresh policy. Invalid values fail plugin load. */
+/** Request-preparation clock formatting and append scheduling. Invalid values fail plugin load. */
 export interface Config {
   /** IANA time zone used for the rendered timestamp. Omit to resolve the Node process's system zone at plugin load. */
   timeZone?: string
-  /** Maximum age of a reading within one turn, in milliseconds (default 60,000; `0` refreshes every step). */
+  /** Minimum milliseconds between durable injections in one session. Omit or set to 0 to inject on every eligible pre-step attempt. */
   refreshIntervalMs?: number
 }
 ```
 
-Source: [`packages/context/time-context/src/index.ts:22`](../packages/context/time-context/src/index.ts)
+Source: [`packages/context/time-context/src/index.ts:20`](../packages/context/time-context/src/index.ts)
+
+## `@deepseek-ai/dsh-token-meter`
+
+```ts config-catalog
+/** Token-meter plugin configuration. */
+export interface TokenMeterConfig {
+  /** Service-wide context-window capacity in tokens. Defaults to `128000`. */
+  contextWindow?: number
+}
+```
+
+Source: [`packages/llm/token-meter/src/types.ts:10`](../packages/llm/token-meter/src/types.ts)
+
+## `@deepseek-ai/dsh-tool-bash`
+
+Requires: `tools` · `bash` · `systemPrompt`
+
+```ts config-catalog
+/** Configuration for the bash tool and its managed child environment. */
+export interface Config {
+  /** Expose `run_in_background` (default true); disabled calls are also rejected. */
+  enableRunInBackground?: boolean
+  /** DeepSeek Harness home directory exposed as `DSH_HOME`; defaults to `$DSH_HOME` or `~/.dsh`. */
+  dshHome?: string
+}
+```
+
+Source: [`packages/bash/tool-bash/src/index.ts:39`](../packages/bash/tool-bash/src/index.ts)
 
 ## `@deepseek-ai/dsh-tool-cordis`
 
@@ -901,6 +1040,28 @@ export interface Config {
 
 Source: [`packages/fs/tool-fs/src/index.ts:22`](../packages/fs/tool-fs/src/index.ts)
 
+## `@deepseek-ai/dsh-tool-fs-search`
+
+Requires: `tools` · `systemPrompt` · `bash`
+
+```ts config-catalog
+/** Plugin config (all optional — `Config` supplies the defaults). */
+export interface Config {
+  /** Max paths one `glob` call retains inline; later paths go to the formatted spill file. */
+  globMaxResults?: number
+  /** Max flat matches one `grep` call retains inline; later matches go to the formatted spill file. */
+  grepMaxMatches?: number
+  /** Max bytes retained for one matched-line preview (the cut preserves UTF-8 boundaries). */
+  grepMaxLineBytes?: number
+  /** Max complete raw `rg` stdout bytes a search will parse; larger raw output fails with `SEARCH_RAW_OUTPUT_OVERFLOW`. */
+  rawOutputMaxBytes?: number
+  /** Cooperative tool-call timeout budget (ms) on both tools, enforced by `@deepseek-ai/dsh-timeout-policy` through `exec.signal`. */
+  timeoutMs?: number
+}
+```
+
+Source: [`packages/fs/tool-fs-search/src/index.ts:59`](../packages/fs/tool-fs-search/src/index.ts)
+
 ## `@deepseek-ai/dsh-tool-skill`
 
 Requires: `tools` · `skills`
@@ -925,34 +1086,29 @@ export interface Config {
   /** The `ctx.subagents` provider name to start runs on (e.g. `spawn`, `acp`). */
   provider: string
   /**
-   * The model-facing tool name to register (default `subagent`). To expose more
-   * than one transport, load this plugin once per provider — each load MUST set
-   * a distinct `toolName` (the tool registry rejects a duplicate name), e.g.
-   * `{ provider: 'spawn', toolName: 'subagent' }` and
-   * `{ provider: 'acp', toolName: 'subagent_acp' }`.
+   * Model-facing tool name (default `subagent`). Each loaded instance must use
+   * a distinct name.
    */
   toolName?: string
   /**
-   * Default per-child agent options (model) applied to every spawned child.
-   * Omitted fields fall back to the child loop's own defaults.
+   * Expose `run_in_background` (default true). Disabled instances omit the
+   * parameter and reject forced background calls.
+   */
+  enableRunInBackground?: boolean
+  /**
+   * Agent options applied to every child; omitted fields use child-loop defaults.
    */
   agentOptions?: AgentOptions
   /**
-   * Per-child persona applied to every child this tool spawns: a scoped
-   * `deployment:persona` section shadowing the deployment's persona for the
-   * child alone. Requires the bound provider's `persona` capability
-   * (in-process backends support it; a request against one that doesn't is
-   * rejected at start). Omitted ⇒ the child renders the deployment persona.
+   * Per-child persona that shadows `deployment:persona`. Requires the
+   * provider's `persona` capability; omission preserves the deployment persona.
    */
   persona?: string
   /**
-   * Tool scoping applied to every child this tool spawns (see
-   * `SubagentStartRequest.toolFilter`): the named global tools vanish from
-   * the child's prompt AND refuse to execute. Requires the provider's
-   * `toolFilter` capability. Unknown names fail the spawn loudly. Note the
-   * child otherwise sees every global tool — including this delegation tool
-   * itself; `deny`-listing it (or setting `maxDepth`) is how a deployment
-   * bounds recursion.
+   * Tool filter applied to every child. Filtered tools disappear from its
+   * prompt and reject execution. Requires the provider's `toolFilter`
+   * capability; unknown names fail startup. Children otherwise see this tool,
+   * so deny it or set `maxDepth` to bound recursion.
    */
   toolFilter?: {
     /** Global tool names the child keeps; everything else is removed. */
@@ -961,12 +1117,8 @@ export interface Config {
     deny?: string[]
   }
   /**
-   * Recursion cap applied to every child this tool spawns (see
-   * `SubagentStartRequest.maxDepth`): a spawn whose child would sit deeper
-   * than this in the delegation tree is rejected. Requires the provider's
-   * `depthLimit` capability. Must be a non-negative safe integer and is
-   * validated when the plugin loads. Omitted ⇒ unbounded (bound it in
-   * deployments that expose this tool to children).
+   * Maximum child depth. Requires the provider's `depthLimit` capability and a
+   * non-negative safe integer. Omission is unbounded.
    */
   maxDepth?: number
 }
@@ -974,7 +1126,23 @@ export interface Config {
 
 Depends on: [`AgentOptions`](../packages/core/agent/src/index.ts)
 
-Source: [`packages/subagent/tool-subagent/src/index.ts:24`](../packages/subagent/tool-subagent/src/index.ts)
+Source: [`packages/subagent/tool-subagent/src/index.ts:23`](../packages/subagent/tool-subagent/src/index.ts)
+
+## `@deepseek-ai/dsh-tool-tasks`
+
+Requires: `tools` · `tasks` · `systemPrompt`
+
+```ts config-catalog
+/** Configures bounded `task_output` waits. */
+export interface Config {
+  /** Wait duration applied when `task_output` sets `wait` without `timeout_ms` (default 30s). */
+  waitTimeoutMs?: number
+  /** Hard cap on any single wait; a larger model-supplied `timeout_ms` is clamped down to it (default 10min). */
+  maxWaitTimeoutMs?: number
+}
+```
+
+Source: [`packages/tasks/tool-tasks/src/index.ts:21`](../packages/tasks/tool-tasks/src/index.ts)
 
 ## `@deepseek-ai/dsh-tool-web`
 
@@ -1034,7 +1202,7 @@ export interface Config {
 export type ToolPresentationMode = 'native' | 'code' | 'both'
 ```
 
-Source: [`packages/core/tools/src/index.ts:307`](../packages/core/tools/src/index.ts)
+Source: [`packages/core/tools/src/index.ts:382`](../packages/core/tools/src/index.ts)
 
 ## `@deepseek-ai/dsh-user-approval`
 
@@ -1108,7 +1276,7 @@ export interface Config {
 }
 ```
 
-Source: [`packages/web/web-fetch-local/src/index.ts:36`](../packages/web/web-fetch-local/src/index.ts)
+Source: [`packages/web/web-fetch-local/src/index.ts:34`](../packages/web/web-fetch-local/src/index.ts)
 
 ## `@deepseek-ai/dsh-web-search-deepseek`
 
@@ -1132,7 +1300,7 @@ export interface Config {
 }
 ```
 
-Source: [`packages/web/web-search-deepseek/src/index.ts:40`](../packages/web/web-search-deepseek/src/index.ts)
+Source: [`packages/web/web-search-deepseek/src/index.ts:38`](../packages/web/web-search-deepseek/src/index.ts)
 
 ## `@deepseek-ai/dsh-web-search-exa`
 
@@ -1154,7 +1322,7 @@ export interface Config {
 }
 ```
 
-Source: [`packages/web/web-search-exa/src/index.ts:39`](../packages/web/web-search-exa/src/index.ts)
+Source: [`packages/web/web-search-exa/src/index.ts:37`](../packages/web/web-search-exa/src/index.ts)
 
 ## `@deepseek-ai/dsh-web-search-perplexity`
 
@@ -1176,7 +1344,7 @@ export interface Config {
 }
 ```
 
-Source: [`packages/web/web-search-perplexity/src/index.ts:33`](../packages/web/web-search-perplexity/src/index.ts)
+Source: [`packages/web/web-search-perplexity/src/index.ts:31`](../packages/web/web-search-perplexity/src/index.ts)
 
 ## `@deepseek-ai/dsh-workflow-workerthread`
 
@@ -1206,6 +1374,26 @@ export interface Config {
 
 Source: [`packages/workflow/workflow-workerthread/src/index.ts:32`](../packages/workflow/workflow-workerthread/src/index.ts)
 
+## `@deepseek-ai/dsh-workspace-context`
+
+```ts config-catalog
+/** User-facing workspace instruction loader configuration. */
+export interface Config {
+  /** Harness home containing the fixed user-global `AGENTS.md`; defaults to `$DSH_HOME` or `~/.dsh`. */
+  dshHome?: string
+  /** Directory entries that identify the project root while walking upward from the session cwd. */
+  projectRootMarkers?: string[]
+  /** UTF-8 byte cap for one rendered baseline or dynamic batch; non-positive or non-finite disables loading. */
+  maxBytes: number
+  /** Maximum UTF-8 bytes read from one instruction file; larger files are ignored. */
+  maxSourceBytes?: number
+  /** Ordered same-directory project candidates; the first existing regular file wins in each scope. */
+  instructionFileCandidates?: string[]
+}
+```
+
+Source: [`packages/context/workspace-context/src/config.ts:16`](../packages/context/workspace-context/src/config.ts)
+
 ## Loadable plugins with no config
 
 These load from a `cordis.yml` entry with no `config:` block; they declare no config surface.
@@ -1216,9 +1404,9 @@ These load from a `cordis.yml` entry with no `config:` block; they declare no co
 - `@deepseek-ai/dsh-llm` ([`packages/llm/llm/src/index.ts`](../packages/llm/llm/src/index.ts))
 - `@deepseek-ai/dsh-session` ([`packages/core/session/src/index.ts`](../packages/core/session/src/index.ts))
 - `@deepseek-ai/dsh-subagent` ([`packages/subagent/subagent/src/index.ts`](../packages/subagent/subagent/src/index.ts))
+- `@deepseek-ai/dsh-tasks` ([`packages/tasks/tasks/src/index.ts`](../packages/tasks/tasks/src/index.ts))
 - `@deepseek-ai/dsh-timeout-policy` — requires `tools` ([`packages/timeout/timeout-policy/src/index.ts`](../packages/timeout/timeout-policy/src/index.ts))
 - `@deepseek-ai/dsh-tool-ask-user` — requires `tools` · `userInteraction` ([`packages/ui/tool-ask-user/src/index.ts`](../packages/ui/tool-ask-user/src/index.ts))
-- `@deepseek-ai/dsh-tool-bash` — requires `tools` · `bash` · `systemPrompt` ([`packages/bash/tool-bash/src/index.ts`](../packages/bash/tool-bash/src/index.ts))
 - `@deepseek-ai/dsh-tool-todo` — requires `tools` ([`packages/todo/tool-todo/src/index.ts`](../packages/todo/tool-todo/src/index.ts))
 - `@deepseek-ai/dsh-user-interaction` ([`packages/ui/user-interaction/src/index.ts`](../packages/ui/user-interaction/src/index.ts))
 
@@ -1232,19 +1420,27 @@ Abstract service classes — a deployment loads a concrete implementation packag
 - `@deepseek-ai/dsh-fs` — abstract `FileSystem` ([`packages/fs/fs/src/index.ts`](../packages/fs/fs/src/index.ts))
 - `@deepseek-ai/dsh-sandbox` — abstract `SandboxProvider` ([`packages/sandbox/sandbox/src/index.ts`](../packages/sandbox/sandbox/src/index.ts))
 - `@deepseek-ai/dsh-session-persistence` — abstract `SessionPersistence` ([`packages/session-persistence/session-persistence/src/index.ts`](../packages/session-persistence/session-persistence/src/index.ts))
+- `@deepseek-ai/dsh-spill` — abstract `SpillStore` ([`packages/spill/spill/src/index.ts`](../packages/spill/spill/src/index.ts))
 - `@deepseek-ai/dsh-workflow` — abstract `WorkflowService` ([`packages/workflow/workflow/src/index.ts`](../packages/workflow/workflow/src/index.ts))
 
 ## Library packages (no plugin entry)
 
 Imported as libraries by other packages; a `cordis.yml` cannot load them.
 
+- `@deepseek-ai/create-sdk` ([`packages/sdk/create-sdk/src/index.ts`](../packages/sdk/create-sdk/src/index.ts))
 - `@deepseek-ai/dsh-acp-snapshot` ([`packages/support/acp-snapshot/src/index.ts`](../packages/support/acp-snapshot/src/index.ts))
+- `@deepseek-ai/dsh-agent-loop-testkit` ([`packages/support/agent-loop-testkit/src/index.ts`](../packages/support/agent-loop-testkit/src/index.ts))
 - `@deepseek-ai/dsh-app-boot` ([`packages/ui/app-boot/src/index.ts`](../packages/ui/app-boot/src/index.ts))
 - `@deepseek-ai/dsh-brand` ([`packages/util/brand/src/index.ts`](../packages/util/brand/src/index.ts))
+- `@deepseek-ai/dsh-helper` ([`packages/sdk/helper/src/index.ts`](../packages/sdk/helper/src/index.ts))
+- `@deepseek-ai/dsh-home` ([`packages/util/home/src/index.ts`](../packages/util/home/src/index.ts))
 - `@deepseek-ai/dsh-hook-protocol` ([`packages/hooks/hook-protocol/src/index.ts`](../packages/hooks/hook-protocol/src/index.ts))
 - `@deepseek-ai/dsh-jsonrpc-demo` ([`packages/examples/jsonrpc-demo/src/index.ts`](../packages/examples/jsonrpc-demo/src/index.ts))
 - `@deepseek-ai/dsh-loader-smoke` ([`packages/support/loader-smoke/src/index.ts`](../packages/support/loader-smoke/src/index.ts))
+- `@deepseek-ai/dsh-paths` ([`packages/util/paths/src/index.ts`](../packages/util/paths/src/index.ts))
+- `@deepseek-ai/dsh-retention` ([`packages/util/retention/src/index.ts`](../packages/util/retention/src/index.ts))
 - `@deepseek-ai/dsh-scope` ([`packages/core/scope/src/index.ts`](../packages/core/scope/src/index.ts))
+- `@deepseek-ai/dsh-scripts` ([`packages/sdk/scripts/src/index.ts`](../packages/sdk/scripts/src/index.ts))
 - `@deepseek-ai/dsh-subagent-inprocess` ([`packages/subagent/subagent-inprocess/src/index.ts`](../packages/subagent/subagent-inprocess/src/index.ts))
 - `@deepseek-ai/dsh-subagent-subprocess` ([`packages/subagent/subagent-subprocess/src/index.ts`](../packages/subagent/subagent-subprocess/src/index.ts))
 - `@deepseek-ai/dsh-timeout` ([`packages/util/timeout/src/index.ts`](../packages/util/timeout/src/index.ts))
diff --git a/docs/cookbook/adding-a-tool.i18n.yaml b/docs/cookbook/adding-a-tool.i18n.yaml
index 7ff1ffff7a..6ec964c335 100644
--- a/docs/cookbook/adding-a-tool.i18n.yaml
+++ b/docs/cookbook/adding-a-tool.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-adding-a-tool.md: 4920c98894326fb8eab3b3d5df298baf1da33c1d
-adding-a-tool.zh.md: 3caf1e62f15f2f15103836b4d3f22be20ba02385
+adding-a-tool.md: da214702939e01fedf3d0d69be7560bbafe0696a
+adding-a-tool.zh.md: b216d18b1593cd7e6074685bd39684f1b9694eac
diff --git a/docs/cookbook/adding-a-tool.md b/docs/cookbook/adding-a-tool.md
index 4920c98894..da21470293 100644
--- a/docs/cookbook/adding-a-tool.md
+++ b/docs/cookbook/adding-a-tool.md
@@ -45,9 +45,9 @@ Registration is effect-based: disposing the plugin fiber unregisters the tool (w
 
 ## Long-running work
 
-Follow tool-bash's background pattern: a `run_in_background` flag returns a task id immediately; companion tools poll incrementally and kill; completion notices arrive via `agent.inject()`. Bound buffers and spill full output to disk so nothing is silently lost.
+Gate `run_in_background` with producer config, reject a pre-aborted call, then register through `ctx.tasks.start({ kind, label, owner: exec.agent, run })`. The runtime validates ownership and control-surface availability before `run()` starts work, then supplies the id, session fence, generic control tools, notices, and owner cleanup.
 
-> TODO: each tool reimplements this background pattern by hand today. At some point we need a generic long-running-tool layer that handles task ids, incremental polling, kill, and completion notices uniformly.
+The producer supplies synchronous `cancel`, non-rejecting `done` that settles after resource cleanup, and optional consuming `readOutput` with bounded-output formatting. Once the id is returned, use a task-owned cancellation signal rather than `exec.signal`. See the [background task runtime RFC](../rfc/implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md) and `dsh-tool-bash` for a stream producer.
 
 ## Execution policy and observation
 
diff --git a/docs/cookbook/adding-a-tool.zh.md b/docs/cookbook/adding-a-tool.zh.md
index 3caf1e62f1..b216d18b15 100644
--- a/docs/cookbook/adding-a-tool.zh.md
+++ b/docs/cookbook/adding-a-tool.zh.md
@@ -45,9 +45,9 @@ export function apply(ctx: Context) {
 
 ## 长时间运行的工作
 
-遵循 tool-bash 的后台模式：`run_in_background` 标志立即返回一个 task id；配套工具增量轮询和终止；完成通知通过 `agent.inject()` 到达。限定缓冲区大小，将完整输出溢写到磁盘，避免静默丢失。
+通过 producer 配置控制 `run_in_background`，拒绝已预先中止的调用，然后使用 `ctx.tasks.start({ kind, label, owner: exec.agent, run })` 注册任务。运行时会在 `run()` 启动工作前校验 owner 和控制面是否可用，随后提供 id、会话围栏、通用控制工具、通知和 owner cleanup。
 
-> TODO: 目前每个工具都手动重新实现这套后台模式。未来需要一个通用的长时间运行工具层，统一处理 task id、增量轮询、终止和完成通知。
+producer 提供同步的 `cancel`、在资源清理后 settle 且不 reject 的 `done`，以及可选的消费式 `readOutput`（负责有界输出的格式化）。返回 id 后，应使用 task 自有的取消信号，而不是 `exec.signal`。流式 producer 的示例和完整契约见[后台 task 运行时 RFC](../rfc/implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md)与 `dsh-tool-bash`。
 
 ## 执行策略与观测
 
diff --git a/docs/cookbook/adding-an-llm-adapter.i18n.yaml b/docs/cookbook/adding-an-llm-adapter.i18n.yaml
index 37f934f3e6..497ae08c32 100644
--- a/docs/cookbook/adding-an-llm-adapter.i18n.yaml
+++ b/docs/cookbook/adding-an-llm-adapter.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-adding-an-llm-adapter.md: 70306ccf119f523dd812a859eef1e8628383bb48
-adding-an-llm-adapter.zh.md: e6d151adfc793a829a876031d5d7280ba078c8e9
+adding-an-llm-adapter.md: f20442b8c2ce823452a3ea13409f202185d12d04
+adding-an-llm-adapter.zh.md: 2864dd1e18742c7449e24f22504a5a38976ab450
diff --git a/docs/cookbook/adding-an-llm-adapter.md b/docs/cookbook/adding-an-llm-adapter.md
index 70306ccf11..f20442b8c2 100644
--- a/docs/cookbook/adding-an-llm-adapter.md
+++ b/docs/cookbook/adding-an-llm-adapter.md
@@ -16,11 +16,11 @@ export const inject = ['llm']
 export const Config: z<Config> = z.object({ apiKey: z.string(), … })
 
 export function apply(ctx: Context, config: Config) {
-  ctx.llm.registerAdapter(['model-a', 'model-b'], new MyAdapter(…))
+  ctx.llm.registerAdapter(['my-provider'], new MyAdapter(…))
 }
 ```
 
-Registration is effect-based (HMR-safe); one adapter per model name — duplicates throw. Secrets are cordis-native: schemastery Config with env fallbacks, fed from cordis.yml via `!!js process.env.MY_KEY`. Never read ad-hoc key files in code.
+Registration is effect-based (HMR-safe); one adapter per provider route — duplicates throw, and multi-route registration is all-or-nothing. `options.provider` selects the adapter and `options.model` is the provider model id, so a dynamic catalog adapter can serve new models without lifecycle reconfiguration. Secrets are cordis-native: schemastery Config with env fallbacks, fed from cordis.yml via `!!js process.env.MY_KEY`. Never read ad-hoc key files in code.
 
 ## Protocol obligations (the contract two implementations verified)
 
@@ -30,6 +30,7 @@ Registration is effect-based (HMR-safe); one adapter per model name — duplicat
 - Errors have exactly two sanctioned paths: THROW from `stream()` (transport and protocol failures — use `LlmError` with a stable code), or end the stream with `finish {kind: 'error' | 'aborted'}` (provider in-band failures). Consumers handle both; pick per failure class and document it.
 - Honor `options.signal` (pass it to fetch / your SDK).
 - A `GenerateOptions` field your provider cannot honor (e.g. a `stop` list on a provider without stop sequences): throw `LlmError(..., 'UNSUPPORTED')` rather than silently dropping it.
+- If the provider requires response ids, signatures, or other native metadata on follow-up calls, emit the minimal lossless-JSON projection as `finish.replayState`. Validate it when rebuilding history. `LlmService` passes it only when the historical provider route and target provider route are currently owned by the exact same adapter instance; your adapter decides whether same-model, cross-model, or cross-provider restoration is legal. Never infer native replay from provider/model names alone when state is absent.
 
 Provider-specific request knobs (thinking modes, effort levels) belong in the ADAPTER's Config, not in `GenerateOptions` — the core vocabulary stays provider-neutral.
 
@@ -41,5 +42,5 @@ Split the adapter into testable stages (llm-deepseek's layout): wire types (`typ
 
 - **Unit: mock the provider, not the harness.** A scripted `node:http` server speaking the provider's wire format covers happy paths, every error status, malformed payloads, premature closes, and aborts — no network, and it drives the 100% per-file coverage gate. Works for SDK-backed adapters too (point the SDK's baseURL at the mock).
 - **Hostile framing tests.** Split stream payloads at arbitrary byte positions (including mid-UTF-8) — real networks do.
-- **E2E: `tests/*.e2e.ts`** under `pnpm run test:e2e`, gated with `describe.skipIf(!process.env.MY_KEY)` so CI (no secrets) stays green. Cover each model × each provider mode you map (thinking on/off, effort levels), a tool-call round trip INCLUDING the follow-up turn with results in history, and loose assertions only (substring/structure, bounded maxTokens — real models are nondeterministic).
+- **E2E: `tests/*.e2e.ts`** under `pnpm run test:e2e`, gated with `describe.skipIf(!process.env.MY_KEY)` so CI (no secrets) stays green. Cover representative model/provider/API families and every provider mode you map, a tool-call round trip INCLUDING the follow-up turn with results in history, and loose assertions only (substring/structure, bounded maxTokens — real models are nondeterministic).
 - Register the e2e file pattern in `knip.json` (per-workspace `entry` override) or knip flags it unused.
diff --git a/docs/cookbook/adding-an-llm-adapter.zh.md b/docs/cookbook/adding-an-llm-adapter.zh.md
index e6d151adfc..2864dd1e18 100644
--- a/docs/cookbook/adding-an-llm-adapter.zh.md
+++ b/docs/cookbook/adding-an-llm-adapter.zh.md
@@ -16,11 +16,11 @@ export const inject = ['llm']
 export const Config: z<Config> = z.object({ apiKey: z.string(), … })
 
 export function apply(ctx: Context, config: Config) {
-  ctx.llm.registerAdapter(['model-a', 'model-b'], new MyAdapter(…))
+  ctx.llm.registerAdapter(['my-provider'], new MyAdapter(…))
 }
 ```
 
-注册基于副作用（HMR 安全）；每个模型名称对应一个适配器，重复注册会抛出异常。密钥采用 Cordis 原生方式管理：schemastery Config 带环境变量回退，通过 cordis.yml 的 `!!js process.env.MY_KEY` 注入。代码中禁止临时读取密钥文件。
+注册基于副作用（HMR 安全）；每个提供方路由仅对应一个适配器，重复注册会抛出异常，多路由注册要么全部成功，要么全部失败。`options.provider` 用于选择适配器，`options.model` 是提供方模型 ID，因此动态模型目录适配器无需重新配置生命周期即可提供新模型。密钥采用 Cordis 原生方式管理：schemastery Config 带环境变量回退，通过 cordis.yml 的 `!!js process.env.MY_KEY` 注入。代码中禁止临时读取密钥文件。
 
 ## 协议义务（两个实现共同验证的契约）
 
@@ -30,6 +30,7 @@ export function apply(ctx: Context, config: Config) {
 - 错误有且仅有两条合法路径：从 `stream()` **抛出**（传输与协议故障——使用带稳定 code 的 `LlmError`），或以 `finish {kind: 'error' | 'aborted'}` 结束流（提供方带内故障）。消费方两者都处理；按故障类别选择路径并加以文档化。
 - 遵守 `options.signal`（将其传递给 fetch 或你的 SDK）。
 - 如果 `GenerateOptions` 中某个字段你的提供方无法支持（例如提供方不支持 stop sequences 时收到 `stop` 列表）：抛出 `LlmError(..., 'UNSUPPORTED')`，而非静默丢弃。
+- 如果提供方在后续调用中需要响应 ID、签名或其他原生元数据，请将其最小无损 JSON 投影作为 `finish.replayState` 发出。重建历史时验证该状态。只有历史提供方路由和目标提供方路由当前由完全相同的适配器实例拥有时，`LlmService` 才会传递该状态；由适配器决定同模型、跨模型或跨提供方恢复是否合法。状态缺失时，切勿仅根据提供方/模型名称推断原生回放。
 
 提供方特有的请求旋钮（thinking 模式、effort 级别）放在**适配器**的 Config 中，而非 `GenerateOptions` 中——核心词汇保持提供方无关。
 
@@ -41,5 +42,5 @@ export function apply(ctx: Context, config: Config) {
 
 - **单元测试：mock 提供方，而非 harness。** 用脚本化的 `node:http` 服务器模拟提供方的协议格式，覆盖正常路径、所有错误状态码、畸形载荷、连接提前关闭和中止——无需网络，且能满足 100% 逐文件覆盖率门禁。对基于 SDK 的适配器同样适用（将 SDK 的 baseURL 指向 mock 服务器）。
 - **恶意分帧测试。** 在任意字节位置（包括 UTF-8 字符中间）切割流载荷——真实网络环境正是如此。
-- **E2E：`tests/*.e2e.ts`**，通过 `pnpm run test:e2e` 运行，以 `describe.skipIf(!process.env.MY_KEY)` 守卫，确保无密钥的 CI 保持绿色。覆盖你映射的每个模型 × 每种提供方模式（thinking 开/关、effort 级别）、一次包含后续轮次（历史中带工具结果）的工具调用往返，以及仅做宽松断言（子串/结构匹配、有界的 maxTokens——真实模型是非确定性的）。
+- **E2E：`tests/*.e2e.ts`**，通过 `pnpm run test:e2e` 运行，以 `describe.skipIf(!process.env.MY_KEY)` 守卫，确保无密钥的 CI 保持绿色。覆盖具有代表性的模型/提供方/API 系列以及你映射的每种提供方模式、一次包含后续轮次（历史中带工具结果）的工具调用往返，以及仅做宽松断言（子串/结构匹配、有界的 maxTokens——真实模型是非确定性的）。
 - 在 `knip.json` 中注册 e2e 文件模式（per-workspace `entry` 覆盖），否则 knip 会将其标记为未使用。
diff --git a/docs/cookbook/maintaining-dsh-code-review.md b/docs/cookbook/maintaining-dsh-code-review.md
new file mode 100644
index 0000000000..5a6c2df120
--- /dev/null
+++ b/docs/cookbook/maintaining-dsh-code-review.md
@@ -0,0 +1,62 @@
+# Maintaining the dsh-code-review skill
+
+The [`dsh-code-review`](../../.agents/skills/dsh-code-review/SKILL.md) skill is kept current by a single designated operator running a private periodic maintenance tool. This cookbook is the entry point for that operator — and for anyone taking over the role — and for repo contributors who want to understand why skill updates arrive as small periodic PRs rather than one-off audits. The workflow itself is specified in the [human-review skill-maintenance RFC](../rfc/proposed/process/2026-07-13-human-review-skill-maintenance.md).
+
+## What the maintainer receives
+
+Run the private tool daily with a two-UTC-day overlap; until the proposed scheduler has completed its acceptance run, the operator invokes the wrapper manually at the same cadence. A manual weekly recovery run uses a seven-day window. The workflow:
+
+1. It selects PRs merged in the chosen window (default two UTC days for the daily cadence, seven for weekly) whose merge commit is reachable from `origin/master`. PRs whose merge commit is not reachable (stacked branches whose parent was squashed) or that exceed a 250-commit acquisition cap are logged to `skipped-pulls.json` and skipped rather than aborting the run.
+2. It collects pre-merge human review feedback with commit anchors (inline comments and review submissions), then compares feedback-time and final landed PR patches. It does not acquire PR conversation comments because current GitHub state cannot give them a force-push-safe feedback-time baseline, and it excludes target-branch-only changes from adoption evidence.
+3. Two independently configured reviewer adapters classify provenance and adoption, then classify agreed-adopted items against the current skill.
+4. The primary adapter drafts a complete revised `SKILL.md`; both adapters review the same diff; blocking findings loop until both approve.
+5. `pnpm run doc-sync` and `pnpm run lint` run against the candidate before the tool declares success.
+
+Each run stores its artifacts on the operator's machine. The saved diff, candidate `SKILL.md`, and promotion manifest land under `~/dsh-code-review-outputs/` named by timestamp. The manifest records the source master commit and skill blob, source feedback IDs and URLs, landed evidence ranges, adapter verdicts, and gate results; raw per-adapter I/O stays in a private temp directory whose path is written to the notification and to the daily log under `~/Library/Logs/dsh-code-review-maintainer/`. The maintenance worktree itself is restored clean after every run so the operator is never tempted to edit the maintenance copy in place.
+
+## What the operator does with a candidate diff
+
+When a run produces a candidate, a macOS notification arrives with a `dsh-code-review-promote <timestamp>` hint.
+
+1. **Read the diff on its own merits.** Do not defer to "the reviewers approved" — the maintainer contract is that the operator is the final judgment. Look for checklist bloat, historical prose, unsupported extrapolation from a single incident, and duplicated coverage with existing skill or authoritative-doc content.
+
+   ```sh
+   ls ~/dsh-code-review-outputs/                         # every candidate ever produced
+   less ~/dsh-code-review-outputs/2026-07-16T02-00-00Z.diff
+   less ~/dsh-code-review-outputs/2026-07-16T02-00-00Z.SKILL.md
+   less ~/dsh-code-review-outputs/2026-07-16T02-00-00Z.manifest.json
+   ```
+
+2. **Cross-check against the run artifacts.** The promotion manifest maps each proposed rule to source feedback and landed evidence; detailed per-adapter I/O, consensus, and adopted evidence live under the run's private temp directory (path shown in the log). Spot-check at least one candidate: does the linked human comment actually support the added rule? Does the linked PR actually adopt it?
+
+3. **Decide one of three:**
+   - **Discard.** Delete the saved candidate. The tool re-considers the same feedback on the next run under whatever the current skill then says.
+
+     ```sh
+     rm ~/dsh-code-review-outputs/2026-07-16T02-00-00Z.{diff,SKILL.md,manifest.json}
+     ```
+   - **Batch.** Keep the candidate aside if the update is small and could combine with a future one. The source-skill check still applies; rerun the analysis or manually rebase and re-review the diff if `master` changes first.
+   - **Promote.** From a clean `master` checkout of the repo, run the promote helper. It refreshes `master`, verifies that the current skill matches the recorded source blob, applies the saved diff, and opens a draft PR whose body carries the manifest's provenance summary. It stops on skill drift rather than overwriting newer guidance; the operator still reviews the PR on GitHub and either merges it or closes it.
+
+     ```sh
+     cd ~/path/to/deepseek-harness   # clean master
+     dsh-code-review-promote 2026-07-16T02-00-00Z
+     ```
+
+4. **Do not commit adapter output verbatim.** Small edits during promotion — tightening wording, removing an example that only makes sense with the source PR's context, folding a rule into an existing one — are expected and preserve the "reviewer judgment" the workflow depends on. Amend the branch before merging.
+
+## When a run produces no candidate
+
+That is the common case after every nonempty classification stage has produced at least one valid adapter result. The tool records "no candidate" in its daily log, sends no notification (to avoid alert fatigue), and moves on. Days without a skill update are the workflow behaving correctly, not a stall.
+
+## Interruptions and handoff
+
+The mechanism lives on one machine. Interruptions the operator handles as they arise:
+
+- **Daily run missed.** The two-day overlap window catches one skipped day automatically; longer gaps recover by running the wrapper manually with `DSH_CODE_REVIEW_SINCE=<Nd>`. Overlapping windows are idempotent: guidance already in the current skill is classified `covered` and does not re-enter as a candidate.
+- **Adapter provider outage.** The tool refuses to run when the two reviewer commands resolve to byte-identical executables. A single batch whose adapter response fails schema or id validation is failed closed at the batch level (every item in the batch marked unclear) and the run continues; the raw output is preserved for debugging. If either adapter produces no valid result for any nonempty batch in an operation, the run fails, writes a failure record, and notifies the operator; it never collapses a total-provider outage into "no candidate."
+- **Handoff to another maintainer.** Open a follow-up RFC that supersedes the current one: either move the mechanism into the repository or record the new operator's private setup. Do not silently transfer the tool — the "single-maintainer bus factor" in the RFC's Risks section is the reason the handoff needs a documented decision.
+
+## Where the operator's private setup lives
+
+The tool source, reviewer adapters, provider credentials, and scheduler are the operator's private infrastructure and are outside this repository by design (see the RFC's "Where the mechanism lives" section). This cookbook and the RFC describe **what the workflow guarantees**; **how** those guarantees are implemented is a private-infrastructure concern. If you are the new operator, the RFC's `## Proposal` sections are the specification you build against.
diff --git a/docs/cordis-catalog/events.md b/docs/cordis-catalog/events.md
index ebd7b75bc0..6fbffc1798 100644
--- a/docs/cordis-catalog/events.md
+++ b/docs/cordis-catalog/events.md
@@ -23,7 +23,7 @@ A fully configured agent and live session were published. Setup is composition-o
 
 Types: [Agent](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:139`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:154`](../../packages/core/agent/src/types.ts)
 
 ### `agent/disposed` — emit
 
@@ -35,7 +35,7 @@ An agent left the registry; AgentLoop emits this after driver quiescence but bef
 
 Types: [Agent](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:148`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:163`](../../packages/core/agent/src/types.ts)
 
 ### `agent/error` — emit
 
@@ -47,7 +47,7 @@ A step or turn errored. The loop reports a failure here (plus the logger) even w
 
 Types: [Agent](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:283`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:298`](../../packages/core/agent/src/types.ts)
 
 ### `agent/pre-step` — serial
 
@@ -59,7 +59,7 @@ Awaited serial checkpoint for session-surface mutation after prompt assembly and
 
 Types: [Agent](../core-data-structures/core.md) · [Message](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:202`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:217`](../../packages/core/agent/src/types.ts)
 
 ### `agent/prompt-submit` — waterfall
 
@@ -71,7 +71,7 @@ Allow, rewrite, or block one drained prompt before it becomes a user message. Ca
 
 Types: [Agent](../core-data-structures/core.md) · [ContentBlock](../core-data-structures/core.md) · [MessageSource](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:212`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:227`](../../packages/core/agent/src/types.ts)
 
 ### `agent/queued` — emit
 
@@ -83,7 +83,7 @@ Detached, frozen content entered the agent's inbox. Source defaults have already
 
 Types: [Agent](../core-data-structures/core.md) · [ContentBlock](../core-data-structures/core.md) · [MessageSource](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:167`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:182`](../../packages/core/agent/src/types.ts)
 
 ### `agent/request` — waterfall
 
@@ -95,7 +95,7 @@ Replace the frozen call configuration. Model-visible content must use logged cha
 
 Types: [Agent](../core-data-structures/core.md) · [LlmCallConfig](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:224`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:239`](../../packages/core/agent/src/types.ts)
 
 ### `agent/session-prefix` — waterfall
 
@@ -107,7 +107,7 @@ Compose request-only messages placed before derived history. The frozen result i
 
 Types: [Agent](../core-data-structures/core.md) · [Message](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:239`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:254`](../../packages/core/agent/src/types.ts)
 
 ### `agent/session-start` — emit
 
@@ -119,7 +119,7 @@ The session lifecycle began, once before the first turn. Use `agent.inject()` to
 
 Types: [Agent](../core-data-structures/core.md) · [SessionStartSource](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:180`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:195`](../../packages/core/agent/src/types.ts)
 
 ### `agent/status` — emit
 
@@ -131,7 +131,7 @@ Agent status changed (`idle` ⇄ `running`, or → `disposed`). `send()` does no
 
 Types: [Agent](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:157`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:172`](../../packages/core/agent/src/types.ts)
 
 ### `agent/step-result` — waterfall
 
@@ -143,7 +143,7 @@ Waterfall: post-process the assembled assistant Message before tool dispatch (va
 
 Types: [Agent](../core-data-structures/core.md) · [Message](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:250`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:265`](../../packages/core/agent/src/types.ts)
 
 ### `agent/turn-continuation` — waterfall
 
@@ -155,7 +155,7 @@ Override whether the turn continues. The default continues after tool calls or s
 
 Types: [Agent](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:260`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:275`](../../packages/core/agent/src/types.ts)
 
 ### `agent/turn-stop` — serial
 
@@ -167,7 +167,7 @@ Monotonic terminal-stop checkpoint after continuation and steering are folded; a
 
 Types: [Agent](../core-data-structures/core.md)
 
-Source: [`packages/core/agent/src/types.ts:270`](../../packages/core/agent/src/types.ts)
+Source: [`packages/core/agent/src/types.ts:285`](../../packages/core/agent/src/types.ts)
 
 ## `approval/*`
 
@@ -195,7 +195,7 @@ Single-slot decision for the next FileSystem.editText. Calling `next()` yields a
 
 Types: [FsTarget](../core-data-structures/filesystem.md) · [FsVersion](../core-data-structures/filesystem.md)
 
-Source: [`packages/fs/fs/src/index.ts:59`](../../packages/fs/fs/src/index.ts)
+Source: [`packages/fs/fs/src/index.ts:61`](../../packages/fs/fs/src/index.ts)
 
 ### `fs/observed` — emit
 
@@ -207,7 +207,7 @@ Record a successful observation. Listeners must be synchronous recorders: throws
 
 Types: [FsTarget](../core-data-structures/filesystem.md) · [FsVersion](../core-data-structures/filesystem.md)
 
-Source: [`packages/fs/fs/src/index.ts:68`](../../packages/fs/fs/src/index.ts)
+Source: [`packages/fs/fs/src/index.ts:70`](../../packages/fs/fs/src/index.ts)
 
 ### `fs/write-intent` — waterfall
 
@@ -219,7 +219,7 @@ Single-slot decision for the next FileSystem.writeText. Calling `next()` yields
 
 Types: [FsTarget](../core-data-structures/filesystem.md) · [FsWriteIntent](../core-data-structures/filesystem.md)
 
-Source: [`packages/fs/fs/src/index.ts:51`](../../packages/fs/fs/src/index.ts)
+Source: [`packages/fs/fs/src/index.ts:53`](../../packages/fs/fs/src/index.ts)
 
 ## `llm/*`
 
@@ -233,7 +233,7 @@ Waterfall around every streaming model call (retry, replay, routing). Bound to t
 
 Types: [GenerateOptions](../core-data-structures/core.md) · [StreamChunk](../core-data-structures/llm-streaming.md)
 
-Source: [`packages/llm/llm/src/index.ts:39`](../../packages/llm/llm/src/index.ts)
+Source: [`packages/llm/llm/src/index.ts:40`](../../packages/llm/llm/src/index.ts)
 
 ## `session/*`
 
@@ -245,7 +245,7 @@ Creation announcement during session publication. A synchronous throw vetoes and
 'session/created'(this: Scoped<Session>, session: Session): void
 ```
 
-Source: [`packages/core/session/src/index.ts:47`](../../packages/core/session/src/index.ts)
+Source: [`packages/core/session/src/index.ts:46`](../../packages/core/session/src/index.ts)
 
 ### `session/disposed` — emit
 
@@ -255,7 +255,7 @@ Emitted once when an announced session leaves the store, including publication r
 'session/disposed'(this: Scoped<Session>, session: Session): void
 ```
 
-Source: [`packages/core/session/src/index.ts:57`](../../packages/core/session/src/index.ts)
+Source: [`packages/core/session/src/index.ts:56`](../../packages/core/session/src/index.ts)
 
 ### `session/event` — emit
 
@@ -267,7 +267,7 @@ Post-commit, fire-and-forget append feed. The listener snapshot resolves before
 
 Types: [SessionEvent](../core-data-structures/core.md)
 
-Source: [`packages/core/session/src/index.ts:69`](../../packages/core/session/src/index.ts)
+Source: [`packages/core/session/src/index.ts:68`](../../packages/core/session/src/index.ts)
 
 ### `session/flush` — parallel
 
@@ -277,7 +277,7 @@ Awaited parallel durability checkpoint: every listener runs and the caller await
 'session/flush'(this: Scoped<Session>, session: Session): Promise<void> | void
 ```
 
-Source: [`packages/core/session/src/index.ts:79`](../../packages/core/session/src/index.ts)
+Source: [`packages/core/session/src/index.ts:78`](../../packages/core/session/src/index.ts)
 
 ## `subagent/*`
 
@@ -289,7 +289,7 @@ A ready child settled. Scope-filtered dispatch uses the same delegating parent c
 'subagent/end'(this: Scoped<SubagentService>, info: SubagentRunEndInfo): void
 ```
 
-Source: [`packages/subagent/subagent/src/index.ts:92`](../../packages/subagent/subagent/src/index.ts)
+Source: [`packages/subagent/subagent/src/index.ts:108`](../../packages/subagent/subagent/src/index.ts)
 
 ### `subagent/provider-added` — emit
 
@@ -299,7 +299,7 @@ A provider became resolvable in the registry.
 'subagent/provider-added'(provider: SubagentProvider): void
 ```
 
-Source: [`packages/subagent/subagent/src/index.ts:66`](../../packages/subagent/subagent/src/index.ts)
+Source: [`packages/subagent/subagent/src/index.ts:82`](../../packages/subagent/subagent/src/index.ts)
 
 ### `subagent/provider-removed` — emit
 
@@ -309,7 +309,7 @@ A provider left the registry. Accepted runs remain holder-owned.
 'subagent/provider-removed'(name: string): void
 ```
 
-Source: [`packages/subagent/subagent/src/index.ts:72`](../../packages/subagent/subagent/src/index.ts)
+Source: [`packages/subagent/subagent/src/index.ts:88`](../../packages/subagent/subagent/src/index.ts)
 
 ### `subagent/start` — emit
 
@@ -319,7 +319,7 @@ A provider established a ready child. For in-process providers, `ctx.agents.get(
 'subagent/start'(this: Scoped<SubagentService>, info: SubagentRunInfo): void
 ```
 
-Source: [`packages/subagent/subagent/src/index.ts:83`](../../packages/subagent/subagent/src/index.ts)
+Source: [`packages/subagent/subagent/src/index.ts:99`](../../packages/subagent/subagent/src/index.ts)
 
 ## `system-prompt/*`
 
@@ -469,14 +469,14 @@ Source: [`packages/workflow/workflow/src/index.ts:45`](../../packages/workflow/w
 
 The framework events every plugin also sees, beyond the harness vocabulary above. This is pinned vendor source ([vendoring policy](../../vendor/README.md)); it is summarized here so the page is a complete picture of the event bus, without elevating framework internals to the harness tier's prominence.
 
-- `internal/plugin` — A plugin fiber was created. ([`vendor/cordis/src/events.ts:197`](../../vendor/cordis/src/events.ts))
-- `internal/status` — A fiber changed lifecycle state. ([`vendor/cordis/src/events.ts:198`](../../vendor/cordis/src/events.ts))
-- `internal/service` — Interception hook for a service binding (no core producer). ([`vendor/cordis/src/events.ts:199`](../../vendor/cordis/src/events.ts))
-- `internal/update` — Waterfall: a fiber config update is being applied. ([`vendor/cordis/src/events.ts:200`](../../vendor/cordis/src/events.ts))
-- `internal/get` — Waterfall: a service is being read from the store. ([`vendor/cordis/src/events.ts:201`](../../vendor/cordis/src/events.ts))
-- `internal/set` — Waterfall: a service is being written to the store. ([`vendor/cordis/src/events.ts:202`](../../vendor/cordis/src/events.ts))
-- `internal/listener` — A listener was registered. ([`vendor/cordis/src/events.ts:203`](../../vendor/cordis/src/events.ts))
-- `internal/dispatch` — An event is being dispatched to listeners. ([`vendor/cordis/src/events.ts:204`](../../vendor/cordis/src/events.ts))
+- `internal/plugin` — A plugin fiber was created. ([`vendor/cordis/src/events.ts:328`](../../vendor/cordis/src/events.ts))
+- `internal/status` — A fiber changed lifecycle state. ([`vendor/cordis/src/events.ts:330`](../../vendor/cordis/src/events.ts))
+- `internal/service` — Interception hook for a service binding (no core producer). ([`vendor/cordis/src/events.ts:332`](../../vendor/cordis/src/events.ts))
+- `internal/update` — Waterfall: a fiber config update is being applied. ([`vendor/cordis/src/events.ts:334`](../../vendor/cordis/src/events.ts))
+- `internal/get` — Waterfall: a service is being read from the store. ([`vendor/cordis/src/events.ts:336`](../../vendor/cordis/src/events.ts))
+- `internal/set` — Waterfall: a service is being written to the store. ([`vendor/cordis/src/events.ts:338`](../../vendor/cordis/src/events.ts))
+- `internal/listener` — A listener was registered. ([`vendor/cordis/src/events.ts:340`](../../vendor/cordis/src/events.ts))
+- `internal/dispatch` — An event is being dispatched to listeners. ([`vendor/cordis/src/events.ts:342`](../../vendor/cordis/src/events.ts))
 - `hmr/change` — A watched source file changed on disk. ([`vendor/hmr/src/index.ts:20`](../../vendor/hmr/src/index.ts))
 - `hmr/reload` — Plugins are being reloaded after a change. ([`vendor/hmr/src/index.ts:21`](../../vendor/hmr/src/index.ts))
 - `exit` — The process is exiting on a signal. ([`vendor/loader/src/index.ts:23`](../../vendor/loader/src/index.ts))
diff --git a/docs/cordis-catalog/services.md b/docs/cordis-catalog/services.md
index 83888c02e1..b858cb6260 100644
--- a/docs/cordis-catalog/services.md
+++ b/docs/cordis-catalog/services.md
@@ -19,7 +19,7 @@ async createAgent(ownerCtx: Context, options: CreateAgentOptions): Promise<Agent
 async resume(ownerCtx: Context, options: ResumeAgentOptions): Promise<AgentHandle>
 ```
 
-Source: [`packages/core/agent-loop/src/index.ts:335`](../../packages/core/agent-loop/src/index.ts)
+Source: [`packages/core/agent-loop/src/index.ts:352`](../../packages/core/agent-loop/src/index.ts)
 
 ## `ctx.agents` — `AgentRegistry`
 
@@ -54,23 +54,38 @@ Source: [`packages/ui/user-approval/src/index.ts:229`](../../packages/ui/user-ap
 
 ## `ctx.bash` — `BashExecutor` (abstract seam)
 
-Registers one `ctx.bash` implementation. Runtime command failures resolve as BashRunResult; only infrastructure failures reject. Background starts return immediately without a timeout, report completion exactly once while live, and remain cancellable by signal or kill. Output reads are incremental and flag lost buffered data; disposal kills and awaits all tasks.
+Abstract bash execution service. Subclass, implement the abstract methods, and load the subclass as a plugin — it registers as `ctx.bash` (one implementation per context; loading a second throws, which is cordis' standard duplicate-service behavior).
+
+Implementations must honor these semantics:
+
+- run rejects only for infrastructure failures. Nonzero exits, timeout kills, and abort kills resolve with a BashRunResult.
+- start returns immediately; no timeout applies to background processes. `done` settles at process close and never rejects; spawn failures settle as `killed` with the error on stderr.
+- BashProcess.readOutput is incremental: consecutive reads never repeat output. Lossy reads report truncation and available spill files.
+- Disposal kills all running background processes and awaits their exit.
 
 ```ts cordis-catalog
 abstract resolve(request: BashExecRequest): BashExecSpec
 abstract run(spec: BashExecSpec): Promise<BashRunResult>
-abstract start(spec: BashExecSpec): BashTask
-abstract get(id: BashTaskId): BashTask | undefined
-abstract ownerOf(id: BashTaskId): OwnerToken | undefined
-abstract list(): BashTask[]
-abstract readOutput(id: BashTaskId): BashTaskRead
-abstract kill(id: BashTaskId): boolean
-onTaskDone(listener: BashTaskListener): () => void
+abstract start(spec: BashExecSpec): BashProcess
 ```
 
-Types: [BashExecRequest](../core-data-structures/bash.md) · [BashExecSpec](../core-data-structures/bash.md) · [BashRunResult](../core-data-structures/bash.md) · [BashTask](../core-data-structures/bash.md) · [BashTaskRead](../core-data-structures/bash.md)
+Types: [BashExecRequest](../core-data-structures/bash.md) · [BashExecSpec](../core-data-structures/bash.md) · [BashRunResult](../core-data-structures/bash.md)
 
-Source: [`packages/bash/bash/src/index.ts:38`](../../packages/bash/bash/src/index.ts)
+Source: [`packages/bash/bash/src/index.ts:49`](../../packages/bash/bash/src/index.ts)
+
+## `ctx.bashEnv` — `BashEnvRegistry`
+
+Registry (`ctx.bashEnv`) for trusted, per-execution `DSH_*` variables. The namespace is rebuilt for every model bash call: ambient `DSH_*` values are discarded by the executor, then the registry's current snapshot is injected. Built-in shell facts remain owned by the registry itself while plugins can register additional, enumerable facts with effect-scoped disposal.
+
+```ts cordis-catalog
+register(contributor: BashEnvContributor): () => void
+collect(execution: ToolExecution): DshEnvironment
+list(): BashEnvVariableInfo[]
+```
+
+Types: [ToolExecution](../core-data-structures/tools.md)
+
+Source: [`packages/bash/tool-bash/src/index.ts:102`](../../packages/bash/tool-bash/src/index.ts)
 
 ## `ctx.codeRuntime` — `CodeRuntime` (abstract seam)
 
@@ -86,7 +101,7 @@ Source: [`packages/code-runtime/code-runtime/src/index.ts:30`](../../packages/co
 
 ## `ctx.compact` — `CompactService` (abstract seam)
 
-Abstract compaction service. Implementations own token estimation, retention, and summarization, but a successful run must replace the selected surface span with one summary node and prevent concurrent compaction of the same session. Load one implementation per context as `ctx.compact`.
+Abstract compaction service. Implementations own trigger policy, retention, and summarization, and may consume a separate measurement service. A successful run replaces the selected surface span with one summary node and prevents concurrent compaction of the same session. Load one implementation per context as `ctx.compact`.
 
 ```ts cordis-catalog
 abstract compactIfNeeded( agent: CompactAgentContext, fullSystemPrompt: string, sessionPrefix: readonly Message[], signal: AbortSignal, ): Promise<CompactionResult | null>
@@ -95,15 +110,16 @@ abstract compactRegion( session: Session, start: number, end: number, agent: Com
 
 Types: [Message](../core-data-structures/core.md)
 
-Source: [`packages/compact/compact/src/index.ts:36`](../../packages/compact/compact/src/index.ts)
+Source: [`packages/compact/compact/src/index.ts:38`](../../packages/compact/compact/src/index.ts)
 
 ## `ctx.fs` — `FileSystem` (abstract seam)
 
 Abstract filesystem provider. Targets must preserve identity across aliases; reads expose regular UTF-8 text or typed errors, listings are stable and content-free, and mutations are atomic. Optional guards add stale protection without changing the unguarded provider contract.
 
 ```ts cordis-catalog
-abstract resolve(path: string, opts?: { cwd?: string }): Promise<FsTarget>
+abstract resolve(path: string, opts?: { cwd?: string; signal?: AbortSignal }): Promise<FsTarget>
 abstract stat(target: FsTarget, signal?: AbortSignal): Promise<FsInfo | undefined>
+abstract lstat(path: string, opts?: { cwd?: string }, signal?: AbortSignal): Promise<FsPathInfo | undefined>
 abstract readText(target: FsTarget, signal?: AbortSignal): Promise<string>
 abstract streamText(target: FsTarget, signal?: AbortSignal): Promise<AsyncIterable<string>>
 abstract listDir(target: FsTarget, signal?: AbortSignal): Promise<FsDirEntry[]>
@@ -113,21 +129,22 @@ abstract editText(target: FsTarget, edit: FsEditRequest, expected?: { version: F
 
 Types: [FsEditOutcome](../core-data-structures/filesystem.md) · [FsEditRequest](../core-data-structures/filesystem.md) · [FsInfo](../core-data-structures/filesystem.md) · [FsTarget](../core-data-structures/filesystem.md) · [FsVersion](../core-data-structures/filesystem.md) · [FsWriteIntent](../core-data-structures/filesystem.md) · [FsWriteOutcome](../core-data-structures/filesystem.md)
 
-Source: [`packages/fs/fs/src/index.ts:78`](../../packages/fs/fs/src/index.ts)
+Source: [`packages/fs/fs/src/index.ts:80`](../../packages/fs/fs/src/index.ts)
 
 ## `ctx.llm` — `LlmService`
 
 The abstract `llm` service: an adapter registry plus a streaming model-call surface, interceptable via the `llm/stream` waterfall.
 
 ```ts cordis-catalog
-registerAdapter(models: string[], adapter: LlmAdapter): () => void
-models(): string[]
+registerAdapter(providers: string[], adapter: LlmAdapter): () => void
+listProviders(): LlmProviderInfo[]
+async listModels(provider: string): Promise<LlmModelInfo[]>
 stream(options: GenerateOptions): AsyncIterable<StreamChunk>
 ```
 
 Types: [GenerateOptions](../core-data-structures/core.md) · [StreamChunk](../core-data-structures/llm-streaming.md)
 
-Source: [`packages/llm/llm/src/index.ts:75`](../../packages/llm/llm/src/index.ts)
+Source: [`packages/llm/llm/src/index.ts:96`](../../packages/llm/llm/src/index.ts)
 
 ## `ctx.permission` — `PermissionService`
 
@@ -161,6 +178,7 @@ Source: [`packages/sandbox/sandbox/src/index.ts:111`](../../packages/sandbox/san
 Durable append-only session storage. Implementations preserve contiguous, losslessly JSON-serializable events; append resolves only after durability, and load balances a complete interrupted tail without rewriting committed events.
 
 ```ts cordis-catalog
+abstract locate(meta: SessionHeader): SessionLocation | undefined
 abstract create(meta: SessionHeader): Promise<void>
 abstract append(id: SessionId, events: readonly SessionEvent[]): Promise<void>
 abstract load(id: SessionId): Promise<{ meta: SessionHeader; events: SessionEvent[] }>
@@ -169,19 +187,21 @@ abstract list(): Promise<SessionHeader[]>
 
 Types: [SessionEvent](../core-data-structures/core.md)
 
-Source: [`packages/session-persistence/session-persistence/src/index.ts:30`](../../packages/session-persistence/session-persistence/src/index.ts)
+Source: [`packages/session-persistence/session-persistence/src/index.ts:42`](../../packages/session-persistence/session-persistence/src/index.ts)
 
 ## `ctx.sessionQuery` — `SessionQueryService`
 
-Live-preferred logical-corpus and exact-event read service.
+Live-preferred logical-corpus exact-read and relationship-tracing service.
 
 ```ts cordis-catalog
 listSessions(): Promise<SessionRecord[]>
 async listEvents(sessionId: SessionId): Promise<SessionEventRecord[]>
+async traceSession(sessionId: SessionId): Promise<SessionLineageTrace>
+async traceEvent(request: SessionEventTraceRequest): Promise<SessionEventTrace>
 async readEvent(request: SessionEventReadRequest): Promise<SessionEventWindow>
 ```
 
-Source: [`packages/session-query/session-query/src/index.ts:35`](../../packages/session-query/session-query/src/index.ts)
+Source: [`packages/session-query/session-query/src/index.ts:38`](../../packages/session-query/session-query/src/index.ts)
 
 ## `ctx.sessions` — `SessionStore`
 
@@ -200,7 +220,7 @@ list(): Session[]
 fork(source: SessionForkSource, boundary?: number, childSessionId?: SessionId): Session
 ```
 
-Source: [`packages/core/session/src/index.ts:564`](../../packages/core/session/src/index.ts)
+Source: [`packages/core/session/src/index.ts:580`](../../packages/core/session/src/index.ts)
 
 ## `ctx.skills` — `SkillService`
 
@@ -215,6 +235,22 @@ async get(name: string, options: SkillLookupOptions = {}): Promise<SkillDefiniti
 
 Source: [`packages/skill/skill/src/index.ts:141`](../../packages/skill/skill/src/index.ts)
 
+## `ctx.spillStore` — `SpillStore` (abstract seam)
+
+Abstract spill storage service. Subclass, implement saveText, and load the subclass as a plugin — it registers as `ctx.spillStore` (one implementation per context; loading a second throws, cordis' standard duplicate-service behavior).
+
+Semantics every implementation must honor:
+
+- saveText persists the FULL `content` verbatim and returns an opaque locator, exact byte length, and model-facing retrieval guidance.
+- Storage is scoped by the request's SaveTextSpill.owner session; the backend chooses a private (not world-readable) location and a collision-free name derived from — never equal to — the caller's `suggestedName`.
+- `saveText` REJECTS on a real storage failure (permissions, ENOSPC, backend unavailable); the caller decides how to degrade (the spill policy treats a rejection as best-effort and keeps the inline result).
+
+```ts cordis-catalog
+abstract saveText(input: SaveTextSpill): Promise<SpillRef>
+```
+
+Source: [`packages/spill/spill/src/index.ts:45`](../../packages/spill/spill/src/index.ts)
+
 ## `ctx.subagents` — `SubagentService`
 
 Named provider registry and capability-checked start surface.
@@ -226,7 +262,7 @@ list(): string[]
 async start(name: string, request: SubagentStartRequest): Promise<SubagentRun>
 ```
 
-Source: [`packages/subagent/subagent/src/index.ts:125`](../../packages/subagent/subagent/src/index.ts)
+Source: [`packages/subagent/subagent/src/index.ts:141`](../../packages/subagent/subagent/src/index.ts)
 
 ## `ctx.systemPrompt` — `SystemPrompt`
 
@@ -241,6 +277,38 @@ async assemble(context: AssembleContext = {}): Promise<PromptAssembly>
 
 Source: [`packages/core/system-prompt/src/index.ts:209`](../../packages/core/system-prompt/src/index.ts)
 
+## `ctx.tasks` — `TaskService`
+
+The `tasks` service: the runtime-global background task registry. See the module doc for the ownership, isolation, and lifecycle contracts.
+
+```ts cordis-catalog
+start(spec: TaskStart): TaskId
+list(caller?: Agent): TaskSnapshot[]
+get(id: TaskId, caller?: Agent): TaskSnapshot
+read(id: TaskId, caller?: Agent): TaskRead
+kill(id: TaskId, caller?: Agent, reason?: string): 'requested' | 'already-finished'
+async wait(id: TaskId, timeoutMs: number, caller?: Agent, signal?: AbortSignal): Promise<TaskSnapshot>
+onTaskDone(listener: TaskDoneListener): () => void
+attachSurface(name: string): () => void
+```
+
+Types: [Agent](../core-data-structures/core.md)
+
+Source: [`packages/tasks/tasks/src/index.ts:76`](../../packages/tasks/tasks/src/index.ts)
+
+## `ctx.tokenMeter` — `TokenMeterService`
+
+Replay owner for one service-wide estimator and isolated per-session folds.
+
+```ts cordis-catalog
+measure(session: Session, requestHeader?: EpochHeader): TokenMeasurement
+estimateMessage(message: Message): number
+```
+
+Types: [Message](../core-data-structures/core.md)
+
+Source: [`packages/llm/token-meter/src/index.ts:106`](../../packages/llm/token-meter/src/index.ts)
+
 ## `ctx.tools` — `ToolRegistry`
 
 Tool registry and execution pipeline. Scoped registrations shadow globals; one visibility resolver feeds presentation, lookup, and dispatch.
@@ -251,12 +319,13 @@ restrict(filter: ToolRestriction): () => void
 guard(guard: ToolGuard): () => void
 get(name: string, scope?: ScopeKey): ToolDefinition | undefined
 schemas(scope?: ScopeKey): ToolSchema[]
+executionMode(exec: ToolExecutionInput): ToolExecutionMode
 async execute(exec: ToolExecutionInput): Promise<ToolExecutionResult>
 ```
 
-Types: [ToolDefinition](../core-data-structures/tools.md) · [ToolExecutionInput](../core-data-structures/tools.md) · [ToolExecutionResult](../core-data-structures/tools.md)
+Types: [ToolDefinition](../core-data-structures/tools.md) · [ToolExecutionInput](../core-data-structures/tools.md) · [ToolExecutionMode](../core-data-structures/tools.md) · [ToolExecutionResult](../core-data-structures/tools.md)
 
-Source: [`packages/core/tools/src/index.ts:363`](../../packages/core/tools/src/index.ts)
+Source: [`packages/core/tools/src/index.ts:438`](../../packages/core/tools/src/index.ts)
 
 ## `ctx.userInteraction` — `UserInteractionService`
 
@@ -305,12 +374,12 @@ Source: [`packages/workflow/workflow/src/index.ts:159`](../../packages/workflow/
 
 The framework `ctx` surface every plugin also sees, beyond the harness services above. This is pinned vendor source ([vendoring policy](../../vendor/README.md)); it is summarized here so the page is a complete picture of what `ctx` offers, without elevating framework internals to the harness tier's prominence.
 
-- `ctx.on / ctx.once` — Register an event listener (disposable). ([`vendor/cordis/src/events.ts:29`](../../vendor/cordis/src/events.ts))
-- `ctx.emit / ctx.parallel / ctx.serial / ctx.bail / ctx.waterfall` — Dispatch an event (sync / awaited / first-bail / veto-chain). ([`vendor/cordis/src/events.ts:29`](../../vendor/cordis/src/events.ts))
-- `ctx.plugin / ctx.inject` — Load a plugin / declare required services. ([`vendor/cordis/src/registry.ts:144`](../../vendor/cordis/src/registry.ts))
+- `ctx.on / ctx.once` — Register an event listener (disposable). ([`vendor/cordis/src/events.ts:34`](../../vendor/cordis/src/events.ts))
+- `ctx.emit / ctx.parallel / ctx.serial / ctx.bail / ctx.waterfall` — Dispatch an event (sync / awaited / first-bail / veto-chain). ([`vendor/cordis/src/events.ts:34`](../../vendor/cordis/src/events.ts))
+- `ctx.plugin / ctx.inject` — Load a plugin / declare required services. ([`vendor/cordis/src/registry.ts:164`](../../vendor/cordis/src/registry.ts))
 - `ctx.effect` — Register a disposable side effect tied to the fiber. ([`vendor/cordis/src/fiber.ts:9`](../../vendor/cordis/src/fiber.ts))
 - `ctx.get / ctx.set / ctx.provide / ctx.accessor / ctx.mixin` — Low-level service-store access and binding. ([`vendor/cordis/src/reflect.ts:7`](../../vendor/cordis/src/reflect.ts))
-- `ctx.extend / ctx.isolate / ctx.intercept` — Derive a child context (scoped services / isolation / interception). ([`vendor/cordis/src/context.ts:35`](../../vendor/cordis/src/context.ts))
+- `ctx.extend / ctx.isolate / ctx.intercept` — Derive a child context (scoped services / isolation / interception). ([`vendor/cordis/src/context.ts:42`](../../vendor/cordis/src/context.ts))
 - `ctx.root / ctx.scope / ctx.fiber / ctx.registry / ctx.reflect / ctx.events / ctx.logger` — Ambient handles onto the running context graph. ([`vendor/cordis/src/context.ts:16`](../../vendor/cordis/src/context.ts))
 - `ctx.timer (+ interval / timeout / throttle / debounce / setTimeout / setInterval)` — Disposable timer helpers. The `timer` key is provided at runtime; the six helpers are mixed onto ctx directly (declared via Pick). ([`vendor/timer/src/index.ts:4`](../../vendor/timer/src/index.ts))
 - `ctx.loader` — The config Loader that booted the app (present under the loader). ([`vendor/loader/src/index.ts:30`](../../vendor/loader/src/index.ts))
diff --git a/docs/core-data-structures/bash.md b/docs/core-data-structures/bash.md
index a5efe1a504..c61c24add6 100644
--- a/docs/core-data-structures/bash.md
+++ b/docs/core-data-structures/bash.md
@@ -1,12 +1,24 @@
 # Bash Executor
 
-The bash execution seam — the canonical [capability seam](../rfc/implemented/architecture/2026-06-13-capability-seams.md) example, split across three packages: interface ([dsh-bash](../../packages/bash/bash), `ctx.bash`), implementation ([dsh-bash-local](../../packages/bash/bash-local), local subprocesses), and consumer ([dsh-tool-bash](../../packages/bash/tool-bash), the `bash`/`bash_output`/`bash_kill` tool schemas). Bash is **one optional capability**, not part of the agent-loop spine — so its vocabulary lives here, not in [core.md](core.md). A sandboxed, containerized, or remote backend is a sibling package implementing the same interface.
+The bash execution seam is split across interface ([dsh-bash](../../packages/bash/bash), `ctx.bash`), implementations ([dsh-bash-local](../../packages/bash/bash-local) and [dsh-bash-sandbox](../../packages/bash/bash-sandbox)), and consumer ([dsh-tool-bash](../../packages/bash/tool-bash), the `bash` schema). Generic background-task ids, ownership, and controls live in [tasks.md](tasks.md); this seam returns a task-free process handle.
 
 Source: [`packages/bash/bash/src/types.ts`](../../packages/bash/bash/src/types.ts)
 
+## Managed shell environment namespace
+
+`DSH_*` variables are Harness-owned child-process facts. The model-facing bash tool collects them through `ctx.bashEnv` and passes them through `BashExecRequest.dshEnv`; executors remove inherited `DSH_*` names before merging the current snapshot.
+
+```ts type-equiv
+type DshEnvironmentKey = `${typeof DSH_ENV_PREFIX}${string}`
+```
+
+```ts type-equiv
+type DshEnvironment = Readonly<Record<DshEnvironmentKey, string>>
+```
+
 ## Request vs. spec: the `resolve()` split
 
-The seam separates the **model-/plugin-facing request** (optional `workdir`/`timeoutMs`, filled from config) from the **fully-resolved spec** the executor acts on (those fields required). The tool layer calls `ctx.bash.resolve(request)` between them — this is the repo's "explicit > implicit at package seams" rule made concrete: the reader of a `BashExecSpec` never wonders where the working directory came from.
+The seam separates the **model-/plugin-facing request** (optional `workdir`/`timeoutMs`/`stdoutMaxBytes`, filled from config or request policy) from the **fully-resolved spec** the executor acts on (those fields required). The tool layer calls `ctx.bash.resolve(request)` between them — this is the repo's "explicit > implicit at package seams" rule made concrete: the reader of a `BashExecSpec` never wonders where the working directory or output budget came from.
 
 ```ts type-equiv
 interface BashExecRequest {
@@ -15,6 +27,13 @@ interface BashExecRequest {
   workdir?: string | undefined
   /** Timeout override in milliseconds (implementations cap it). */
   timeoutMs?: number | undefined
+  /**
+   * Foreground stdout capture budget in bytes. Absent uses the executor's
+   * default output cap. Trusted in-process consumers use this when they must
+   * parse complete stdout up to their own bounded limit; the model-facing bash
+   * tool does not expose it as a parameter.
+   */
+  stdoutMaxBytes?: number | undefined
   /** Abort signal — implementations kill the command when it fires. */
   signal?: AbortSignal | undefined
   /**
@@ -26,24 +45,20 @@ interface BashExecRequest {
    */
   stdin?: string | undefined
   /**
-   * Extra environment entries for the command, merged AFTER the
-   * implementation's credential scrub (so an explicit entry here is honored even
-   * when its name matches the scrub pattern — the caller named a value it holds,
-   * not the harness's ambient secret). Set by in-process plugins (the hooks
-   * bridges set `CLAUDE_PROJECT_DIR`, `CLAUDE_PLUGIN_ROOT`, …); the model-facing
-   * bash tool does not expose it as a parameter (a model that needs an env var
-   * uses shell syntax like `FOO=bar cmd`).
+   * Ordinary environment entries for the command, merged after the credential
+   * scrub. `DSH_*` is reserved for {@link dshEnv} and implementations reject it
+   * here. Set by in-process plugins (the hooks bridges set
+   * `CLAUDE_PROJECT_DIR`, `CLAUDE_PLUGIN_ROOT`, …); the model-facing bash tool
+   * does not expose it as a parameter.
    */
   env?: Record<string, string> | undefined
   /**
-   * Opaque OWNER token for a background task — the consumer's isolation key
-   * (the tool layer passes the owning agent's `session.header.id`). The
-   * executor stores it on the task and exposes it via {@link BashExecutor.ownerOf};
-   * the executor itself NEVER interprets it (no access policy lives in the
-   * seam — that is the consumer's job). Absent for foreground runs and for an
-   * ownerless background start (a non-agent caller).
+   * Harness-owned `DSH_*` variables for this execution. Executors discard
+   * ambient `DSH_*` entries before merging this snapshot, so an unavailable
+   * current fact cannot inherit a stale value from the harness process, and
+   * reject non-`DSH_*` names supplied through this managed channel.
    */
-  owner?: OwnerToken | undefined
+  dshEnv?: DshEnvironment | undefined
   /**
    * Explicit per-call sandbox-policy input, overriding the executor's
    * configured default mode for THIS call. Never a silent default: a
@@ -66,14 +81,17 @@ interface BashExecSpec {
   command: string
   workdir: string
   timeoutMs: number
+  /**
+   * Resolved foreground stdout capture budget in bytes. `run()` uses it for
+   * stdout; background tasks and stderr keep the executor's own output cap.
+   */
+  stdoutMaxBytes: number
   /** Abort signal — implementations kill the command when it fires. */
   signal?: AbortSignal | undefined
   /**
    * Bytes to write to the command's stdin (then close it), carried through
-   * verbatim from {@link BashExecRequest.stdin}. OPTIONAL on the resolved spec
-   * (unlike `owner`): it has no config default, so a missing one means "no
-   * stdin" — the safe, ordinary case — not a silent footgun, so it stays a
-   * plain optional rather than required-but-nullable (see the request field).
+   * verbatim from {@link BashExecRequest.stdin}. It has no config default, so
+   * a missing value means "no stdin" and remains an ordinary optional.
    */
   stdin?: string | undefined
   /**
@@ -84,20 +102,13 @@ interface BashExecSpec {
    * config default, absent means "no extra env".
    */
   env?: Record<string, string> | undefined
+  /** Managed `DSH_*` snapshot; implementations reject ordinary names. */
+  dshEnv?: DshEnvironment | undefined
   /**
-   * Opaque owner token, REQUIRED-but-nullable (mirrors `workdir`/`timeoutMs`
-   * being required on the resolved spec): {@link BashExecutor.resolve} carries
-   * the request's `owner` through, defaulting a missing one to `undefined`. A
-   * required field makes a forgotten owner a VISIBLE `undefined` rather than a
-   * silently-absent property that yields an unowned (cross-session-readable)
-   * task. `start()` stores it; `run()` (foreground) ignores it.
-   */
-  owner: OwnerToken | undefined
-  /**
-   * The sandbox mode this call executes under, REQUIRED-but-nullable for the
-   * same visibility reason as `owner`. A sandboxing executor's `resolve()`
-   * stamps the effective mode (the request's explicit override, else its
-   * configured default) so `run()`/`start()` read the spec, never the config;
+   * The sandbox mode this call executes under, required-but-nullable so every
+   * resolved spec states its policy. A sandboxing executor's `resolve()` stamps
+   * the effective mode (the request's explicit override, else its configured
+   * default) so `run()`/`start()` read the spec, never the config;
    * a non-sandboxing executor carries the request value through verbatim and
    * ignores it (`undefined` under such an executor means what its README says:
    * unconfined execution).
@@ -106,11 +117,9 @@ interface BashExecSpec {
 }
 ```
 
-The `owner` token is the isolation key: the executor stores it but never interprets it (access policy is the consumer's job), so a background task started by one agent isn't readable cross-session. A required-but-nullable field makes a forgotten owner a visible `undefined` rather than a silently-unowned task.
+`stdin` and `env` are trusted in-process plugin inputs and are not exposed by `dsh-tool-bash`. The local executor scrubs ambient credentials before merging explicit caller-supplied env. See [the bash-stdin-env RFC](../rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md).
 
-Trusted in-process plugins use `stdin` and `env` for hook payloads and hook-specific variables. The model-facing bash tool constructs requests from its named schema fields and exposes neither input because shell syntax already provides equivalent power; tests guard against a future `...args` spread. This is request-shape discipline, not a security boundary: `dsh-bash-local` scrubs ambient credentials regardless of these fields, then overlays explicit values already held by the caller. See [the bash stdin/env RFC](../rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md).
-
-Both ids the seam handles are [branded](core.md) (zero-cost `string` brands, the same machinery as `SessionId`/`AgentId`): `BashTaskId` (a tracked background task, generated `bash-N` by the local executor) and `OwnerToken` (the opaque isolation key). `OwnerToken` is deliberately a DISTINCT brand from `SessionId`, not an alias: the bash seam is a capability seam that must not know what an owner token *means*, so it never imports `dsh-session`'s vocabulary — the `dsh-tool-bash` consumer is the single boundary that casts the owning agent's `SessionId` into an `OwnerToken`. Branding both stops a raw `string` (or a `BashTaskId` where an `OwnerToken` is expected, or vice versa) from slipping through the type checker on the model-facing `task_id` path.
+`stdoutMaxBytes` is also trusted-plugin-only. It lets a foreground consumer request complete stdout up to a bounded parser budget without changing stderr, background tasks, or the model-facing bash tool's ordinary output cap.
 
 ## Foreground runs: `BashRunResult`
 
@@ -154,9 +163,9 @@ interface CollectedOutput {
 
 ## File sandbox: `BashSandboxInfo`
 
-A sandbox-consuming executor (`dsh-bash-sandbox`) exposes its configured fallback through `BashExecutor.sandboxMode`. The tool layer folds each agent session's durable `bash/sandbox-mode` override, stamps the effective mode onto the request, and may replace it for one user-approved strictly wider call. It deliberately neither states the standing mode nor narrates switches; a denial result names the mode that command actually ran under. The mode/enforcement vocabulary is owned and cataloged by the [`@deepseek-ai/dsh-sandbox` seam](sandbox.md), whose provider wraps the executor's argv; modes govern FILE effects only, not network or process visibility.
+A sandbox-consuming executor exposes its configured fallback through `BashExecutor.sandboxMode`. The tool layer folds each session's durable `bash/sandbox-mode` override and may replace it for one user-approved strictly wider call. The mode/enforcement vocabulary is owned by the [`@deepseek-ai/dsh-sandbox` seam](sandbox.md); modes govern file effects only.
 
-A sandboxed run always reports the facts it executed under on `BashRunResult.sandbox`: `denied` is the executor's conservative classification of a failure as sandbox-caused (a failed exit whose stderr carries a filesystem-permission signature — never a clean exit or a signal kill), read from the collected stderr tail; `enforcement` reports how completely the selected backend governs the mode's file effects (`SandboxEnforcement = 'full' | 'partial'` — `partial` when an older Landlock ABI governs only a subset of the requested accesses; absent under `danger-full-access`, where nothing is confined); `runnerFailed` marks the opposite of a denial — the sandbox RUNNER itself failed and the command never ran (stamped only on settled background tasks; a foreground run surfaces the same condition as the thrown `SANDBOX_UNAVAILABLE` error):
+A sandboxed run reports its mode, conservative denial classification, and enforcement completeness. `runnerFailed` marks a sandbox runner failure before the command ran; foreground execution throws `SANDBOX_UNAVAILABLE`, while a settled background process has only its facts channel.
 
 ```ts type-equiv
 interface BashSandboxInfo {
@@ -195,38 +204,40 @@ interface BashSandboxInfo {
 
 One more piece completes the vocabulary: the `SANDBOX_UNAVAILABLE` error code (owned by the [sandbox seam](sandbox.md)) is what the `ctx.sandbox` provider throws — and the executor propagates — when a confined mode has no usable backend. A selected runner refusing its profile reaches the same fail-closed foreground error; a settled background task records `runnerFailed`. The model receives denial/runner facts in results, learns the effective mode only when a denial marker names it, and can request a one-shot strictly wider retry through `sandbox_permissions` plus `justification`; `ctx.approval` must grant that exact call before anything executes. The complete policy and switching design is the [sandbox RFC](../rfc/implemented/feature/2026-07-06-sandbox.md).
 
-## Background tasks: `BashTask`
+## Background processes: `BashProcess`
 
-A long-running command started with `start()` is tracked as a `BashTask`. `BashTaskStatus` is `'running' | 'completed' | 'killed'`; `done` resolves when the underlying process closes and never rejects. A sandboxing executor stamps `sandbox` once the task settles — classification runs against the settled task's collected stderr — so the field is absent while running and under an unsandboxed executor.
+`start()` returns a handle with no id or owner. `dsh-tool-bash` adapts it into `ctx.tasks.start()` hooks; the generic runtime then owns task identity and lifecycle. `done` resolves when the process closes and never rejects, reads remain valid after settlement, and sandbox facts are stamped before `done` resolves.
 
 ```ts type-equiv
-interface BashTask {
-  readonly id: BashTaskId
-  status: BashTaskStatus
+interface BashProcess {
+  /** Process lifecycle state (settled exactly once). */
+  status: BashProcessStatus
   /** Exit code once finished (null = killed by signal / still running). */
   exitCode: number | null
   /** Terminating signal name, when signal-killed. */
   signal: NodeJS.Signals | null
-  /** Resolves when the underlying process closes (never rejects). */
+  /** Resolves when the underlying process closes (never rejects — a spawn failure settles as `killed` with the error on stderr). */
   readonly done: Promise<void>
-  /**
-   * Sandbox facts for this task's execution, stamped by a sandboxing executor
-   * once the task settles and BEFORE completion listeners are notified — an
-   * `onTaskDone` consumer and a `done` awaiter both see it. Denial
-   * classification runs against the settled task's collected stderr, so the
-   * field cannot exist earlier: absent while the task is running and under an
-   * executor that does not sandbox. See {@link BashSandboxInfo} for the
-   * `denied` semantics.
-   */
+  /** Sandbox facts, stamped once a confined process settles. */
   sandbox?: BashSandboxInfo
+  /**
+   * Read output produced since the previous read (consuming — consecutive
+   * reads never re-deliver). Reads that lost data flag `lossy` and point at
+   * full-stream spill files when available.
+   */
+  readOutput(): BashProcessRead
+  /**
+   * Kill the process group. Returns false when it had already finished
+   * (no-op); idempotent.
+   */
+  kill(): boolean
 }
 ```
 
-`readOutput()` returns an incremental `BashTaskRead` — the output produced since the previous read, with a `lossy` flag when truncation dropped unread bytes:
+`readOutput()` returns the incremental delta and spill recovery facts:
 
 ```ts type-equiv
-interface BashTaskRead {
-  task: BashTask
+interface BashProcessRead {
   /** Output produced since the previous read (stderr in a marked section). */
   delta: string
   /** True when truncation dropped unread bytes the delta cannot include. */
@@ -240,4 +251,4 @@ interface BashTaskRead {
 
 ## The service
 
-`BashExecutor` (`ctx.bash`, abstract — defined in [`packages/bash/bash/src/index.ts`](../../packages/bash/bash/src/index.ts)) mirrors the `LlmService`/`LlmAdapter` split: `resolve` (request → spec), `run` (foreground), `start` (background), `get`/`ownerOf`/`list`/`readOutput`/`kill`, and `onTaskDone` (a `BashTaskListener` completion callback). Spawned commands get a **scrubbed env** (dropping `*KEY*`/`*SECRET*`/`*TOKEN*`) and spill files use a private 0700 dir with random names and owner-only opens — model output never gets the ambient environment or a predictable path. The implementation that provides all this is `dsh-bash-local`; the model-facing `bash`/`bash_output`/`bash_kill` schemas that call it are in `dsh-tool-bash` (and present as terminals via the [tool-presentation vocabulary](tools.md#tool-presentation-ui-vocabulary)).
+`BashExecutor` owns `resolve`, foreground `run`, background-process `start`, and the `sandboxMode` capability fact. `dsh-bash-local` owns process groups, timeout/abort handling, bounded collectors, spill files, credential scrubbing, and disposal quiescence. `dsh-tool-bash` owns model-facing rendering and adapts background handles into the [generic task runtime](tasks.md).
diff --git a/docs/core-data-structures/compaction.md b/docs/core-data-structures/compaction.md
index 82bf512eeb..45a0ac310e 100644
--- a/docs/core-data-structures/compaction.md
+++ b/docs/core-data-structures/compaction.md
@@ -11,7 +11,7 @@ Compaction extends [`SessionEventMap`](session.md) with three event types via de
 | Event | Payload | Role |
 |---|---|---|
 | `compact/start` | `{ turn }` | acquires the log-recorded lock |
-| `compact/summary` | `{ summary, shadowedRange, shadowedSeqs, shadowedTokenCount, model, maxTokens? }` | provenance: the summary blocks, the shadowed surface-boundary pair (`start`/`end` seqs — a position span, not a numeric interval), the shadowed seqs in surface order, the estimated token count, and the summarize call's envelope (`model`, plus its generation cap when one applied) — logged so the one-shot request is reconstructable from log + code (the reconstructability RFC) |
+| `compact/summary` | `{ summary, shadowedRange, shadowedSeqs, shadowedTokenCount, provider, model, maxTokens? }` | provenance: the summary blocks, the shadowed surface-boundary pair (`start`/`end` seqs — a position span, not a numeric interval), the shadowed seqs in surface order, the estimated token count, and the summarize call's envelope (`provider`, `model`, plus its generation cap when one applied) — logged so the one-shot request is reconstructable from log + code (the reconstructability RFC) |
 | `compact/end` | `{ turn, error? }` | releases the lock (`error` set when summarization threw) |
 
 The lock brackets the **whole** operation: `compact/start` is appended first, then summarization, the `compact/summary` provenance record, and the `user/message` replacement all land, and only then `compact/end`. Releasing the lock last turns a crash mid-operation into a detectable orphaned lock (a `compact/start` with no matching `compact/end`) rather than a `compact/end` that falsely claims compaction finished.
@@ -50,6 +50,8 @@ interface CompactionResult {
 
 ## The service
 
-`CompactService` exposes `compactIfNeeded(...)` for pressure-triggered compaction, returning `null` when no compaction is needed, and `compactRegion(...)` for an explicit inclusive surface range. The pre-step caller supplies the agent, full prompt, session prefix, and abort signal; implementations must forward that signal to summarization. Estimation, retention, event sequencing, and summarization remain backend policy.
+`CompactService` exposes `compactIfNeeded(...)` for pressure-triggered compaction, returning `null` when no compaction is needed, and `compactRegion(...)` for an explicit inclusive surface range. The pre-step caller supplies the agent, full prompt, session prefix, and abort signal; implementations must forward that signal to summarization. The seam owns no pricing API: [`ctx.tokenMeter`](token-meter.md) directly owns estimation and replay, while `dsh-compact-basic` owns retention, event sequencing, routed summarization calls, and their configuration.
 
 Auto-compaction runs at serial `agent/pre-step`, before the step and request derivation, so it can replace surface nodes while keeping trace events outside the step. Region boundaries preserve tool-call/result pairing but do not preserve whole turns, allowing early closed steps of one oversized turn to compact. `dsh-compact-basic` owns the retention and failure details.
+
+The seam exports `toolPairingBalancedBefore(session, seq)` and `toolPairingBalancedAfter(session, seq)` for those edge checks. Both validate current surface membership and reject missing seqs and orphan results; the [package contract](../../packages/compact/compact/README.md#tool-pairing-boundaries) owns their cache semantics.
diff --git a/docs/core-data-structures/core.md b/docs/core-data-structures/core.md
index 8926a8a998..c1ecdcc6a5 100644
--- a/docs/core-data-structures/core.md
+++ b/docs/core-data-structures/core.md
@@ -16,15 +16,16 @@ Everything else is documented on a **sub-page**, not here. The rule that draws t
 | Sub-page | Owns |
 |---|---|
 | [llm-streaming.md](llm-streaming.md) | the `StreamChunk` wire protocol + adapter contract, `BlockAssembler`, the `LlmAdapter` seam |
+| [token-meter.md](token-meter.md) | immutable scalar and positional replay measurements with consumed-log revisions |
 | [scope.md](scope.md) | scoped registration identity, dispatch carriers, and the owned `Scope` context |
 | [session.md](session.md) | the full `SessionEventMap` variant catalog, `TurnTrigger`/`TurnEndReason`, `deriveMessages()`, the turn-enclosure invariant |
 | [persistence.md](persistence.md) | the durability seam: `SessionPersistence`, JSONL + SQLite backends, `session/flush`, crash recovery, `SessionHeader` |
-| [session-query.md](session-query.md) | logical session/event records and bounded exact-event reads |
+| [session-query.md](session-query.md) | logical records, bounded exact-event reads, and relationship traces |
 | [system-prompt.md](system-prompt.md) | per-assembly context, tool-provider results, prompt sections, and cooperative assembly |
 | [tools.md](tools.md) | `ToolDefinition` full fields, the schema DSL, `ToolExecution`/`ToolResult`, tool-presentation UI types, and the guarded execution pipeline |
 | [user-interaction.md](user-interaction.md) | the UI-backed human question/answer seam: `AskUserQuestionRequest`, answer/options vocabulary, provider API, error taxonomy |
 | [approval.md](approval.md) | the one-shot user-approval seam: `ApprovalRequest`, `ApprovalOutcome`, per-session policy, audit and answerer contracts |
-| [bash.md](bash.md) | the bash executor seam: `BashExecRequest`/`Spec`, `BashRunResult`, background `BashTask`s |
+| [bash.md](bash.md) | the bash executor seam: `BashExecRequest`/`Spec`, `BashRunResult`, background `BashProcess` handles |
 | [sandbox.md](sandbox.md) | the process-confinement seam: file-effect modes, `SandboxPolicy`, `ConfinedArgv`, enforcement and fail-closed errors |
 | [code-runtime.md](code-runtime.md) | the code-execution seam: `CodeRunRequest`/`Result`, binding namespaces, captured logs, the `CodeRunFailure` taxonomy |
 | [filesystem.md](filesystem.md) | the filesystem seam: `FsTarget`, read/write/edit outcomes, observed-file state, `FsErrorCode` |
@@ -32,6 +33,7 @@ Everything else is documented on a **sub-page**, not here. The rule that draws t
 | [compaction.md](compaction.md) | the compaction seam: the `compact/*` session events, `CompactionResult`, the `CompactService` interface |
 | [subagent.md](subagent.md) | the subagent seam: the named-provider registry, `SubagentStartRequest`/`Result`/`Run`, the start-time-vs-runtime capability split |
 | [web.md](web.md) | the web access seam: `WebSearchRequest`/`Result`, `WebFetchRequest`/`Result`, `WebFetchBody`, provider availability, `WebError` |
+| [spill.md](spill.md) | the spill storage seam: `SaveTextSpill`, `SpillOwner`/`SpillSource`, `SpillRef`, the branded `SpillLocator` |
 | [workflow.md](workflow.md) | the workflow seam: `WorkflowStartRequest`, `WorkflowMeta`, `WorkflowRun`/`Result`, the `workflow/*` event payloads, `WorkflowError` fatality |
 
 > Type definitions on this page are pasted **verbatim** from source and drift-checked by `pnpm run verify-type-equiv` (see [development.md](../development.md#documenting-types-verbatim-ts-type-equiv)). Inline JSDoc is omitted for readability; follow the source link for the full contracts.
@@ -76,7 +78,7 @@ Two large discriminated unions are the ones consumers `switch` over most: **`Str
 
 IDs that cross package boundaries are **branded** — structurally strings, but non-interchangeable at the type level (an `AgentId` can't be passed where a `CallId` is expected). Construction goes through a per-type factory; comparison, logging, and JSON behave as ordinary strings.
 
-The `Branded<B>` primitive lives in its own type-only package, [dsh-brand](../../packages/util/brand) (no runtime code, no harness-package dependency), so any package can brand the ids it owns without depending on an unrelated capability package (e.g. dsh-bash brands `BashTaskId`/`OwnerToken` via dsh-brand alone, never pulling in dsh-llm).
+The `Branded<B>` primitive lives in its own type-only package, [dsh-brand](../../packages/util/brand) (no runtime code, no harness-package dependency), so any package can brand the ids it owns without depending on an unrelated capability package.
 
 Source: [`packages/util/brand/src/index.ts`](../../packages/util/brand/src/index.ts)
 
@@ -84,7 +86,7 @@ Source: [`packages/util/brand/src/index.ts`](../../packages/util/brand/src/index
 type Branded<B extends string> = string & { readonly [BRAND]: B }
 ```
 
-The three core IDs: `CallId` (correlates a tool call with its result; dsh-llm), `SessionId` (dsh-session), `AgentId` (dsh-agent). Each is `Branded<'CallId'>` etc. plus a same-named factory function. Capability seams brand their own ids too — see `BashTaskId`/`OwnerToken` in [bash.md](bash.md).
+The three core IDs are `CallId`, `SessionId`, and `AgentId`. Capability packages brand their own ids too, such as `TaskId` in [tasks.md](tasks.md).
 
 ## Content blocks and messages
 
@@ -103,12 +105,29 @@ interface ContentBlockMap {
 
 The block interfaces (full fields in source): `TextBlock` (`text`), `ReasoningBlock` (thinking, distinct from visible text), `ToolCallBlock` (`id: CallId`, `name`, raw-JSON `arguments`), `ToolResultBlock` (`toolCallId`, nested `content: ContentBlock[]`, `isError?`). `ContentBlock = ContentBlockMap[ContentBlockType]`. The core set is limited to blocks every shipping path honors — multimodal content (images, audio, …) has no core block type; a feature that needs one adds it via the merge-extensible map together with the adapter/UI/compaction support that honors it.
 
-A `Message` is a role plus blocks:
+A `Message` is a role plus blocks. Loop-derived assistant messages carry their durable provider/model identity and optional adapter-private replay metadata:
+
+```ts type-equiv
+interface AssistantProvenance {
+  /** Provider route that produced the message. */
+  provider: string
+  /** Provider model id that produced the message. */
+  model: string
+  /**
+   * Lossless-JSON adapter state needed to replay the provider response.
+   * `LlmService` exposes it to a target adapter only when that adapter instance
+   * currently owns both this historical provider and the target provider.
+   */
+  replayState?: unknown
+}
+```
 
 ```ts type-equiv
 interface Message {
   role: 'system' | 'user' | 'assistant'
   content: ContentBlock[]
+  /** Present only on assistant messages produced by a routed adapter. */
+  provenance?: AssistantProvenance
 }
 ```
 
@@ -133,8 +152,34 @@ One model call is a fully-assembled `GenerateOptions`. The adapter answers with
 
 Source: [`packages/llm/llm/src/types.ts`](../../packages/llm/llm/src/types.ts)
 
+Provider and model discovery uses small provider-neutral descriptors. A model catalog is advisory: routing still keys on a registered provider, and an adapter may accept unlisted model ids.
+
+```ts type-equiv
+interface LlmProviderInfo {
+  /** Provider route key used by {@link GenerateOptions.provider}. */
+  id: string
+  /** Human-readable provider name for selectors and diagnostics. */
+  name: string
+}
+```
+
+```ts type-equiv
+interface LlmModelInfo {
+  /** Provider route that owns this model entry. */
+  provider: string
+  /** Model id passed to {@link GenerateOptions.model}. */
+  id: string
+  /** Human-readable model name for selectors. */
+  name: string
+  /** Optional user-facing distinction from otherwise similar models. */
+  description?: string
+}
+```
+
 ```ts type-equiv
 interface GenerateOptions {
+  /** Registered provider route selecting the adapter instance. */
+  provider: string
   model: string
   /**
    * Ordered conversation messages, exactly as the provider sees them (after
@@ -193,9 +238,9 @@ The model-facing `ToolSchema` is the wire shape; the registered `ToolDefinition`
 
 ### The request envelope: `LlmCallConfig` and the logged header
 
-The loop builds each request from logged state. `EpochHeader` records call config, rendered prompt, authoritative returned tool order (configured by `toolOrder`, or lexicographic when unset), and session prefix through `request/header` snapshots and deltas. Together with derived history, this makes the request reconstructable from the session log. See [session.md](session.md#the-request-header-events-requestheader-and-requestheader-delta) and the [reconstructability RFC](../rfc/implemented/architecture/2026-07-05-reconstructable-requests.md).
+The loop builds each request from logged state. `EpochHeader` records call config, rendered prompt, authoritative returned tool order (configured by `toolOrder`, or lexicographic when unset), and session prefix through full `request/header` snapshots. Together with derived history, this makes the request reconstructable from the session log. See [session.md](session.md#the-request-header-event-requestheader) and the [reconstructability RFC](../rfc/implemented/architecture/2026-07-05-reconstructable-requests.md).
 
-`agent/request` receives a frozen call-config seed and may return a replacement. `agent/session-prefix` composes request-only prefix messages once per loop instance, and the header records the exact result used. Requests reaching `llm/stream` are deep-frozen, so mutation throws.
+`agent/request` receives a frozen call-config seed and may return a replacement to switch provider, model, or sampling. `agent/session-prefix` composes request-only prefix messages once per loop instance, and the header records the exact result used. Requests reaching `llm/stream` are deep-frozen, so mutation throws.
 
 On the wire, a loop-built request reads in this order: the `system` slot (the rendered prompt assembly) → `messagePrefix` (the frozen session prefix) → the derived history — the boundary snapshot, whose tail is the newest `user/message` on a turn's first step and the previous step's tool results on later steps. The prefix never enters the derived history; its durable record is the header events, and the dev invariant recomputes exactly this equation against every loop-built request.
 
@@ -203,6 +248,7 @@ FIXME(call-config-shape): revisit the exact definition of this type — which fi
 
 ```ts type-equiv
 interface LlmCallConfig {
+  provider: string
   model: string
   temperature?: number
   maxTokens?: number
@@ -238,7 +284,7 @@ type SessionEvent<T extends SessionEventType = SessionEventType> = {
 }[T]
 ```
 
-The fifteen event variants (`turn/start`, `turn/end`, `step/start`, `step/end`, `user/message`, `prompt/blocked`, `context/message`, `assistant/chunk`, `assistant/message`, `tool/call`, `tool/result`, `steering/message`, `todo/write`, `request/header`, `request/header-delta`), the `deriveMessages()` projection rules, the `TurnTrigger`/`TurnEndReason` reasons, and the turn-enclosure invariant are on **[session.md](session.md)**. How the log is made durable — the `SessionPersistence` seam, JSONL/SQLite backends, the `session/flush` checkpoint, crash recovery, and `SessionHeader` — is on **[persistence.md](persistence.md)**.
+The fourteen event variants (`turn/start`, `turn/end`, `step/start`, `step/end`, `user/message`, `prompt/blocked`, `context/message`, `assistant/chunk`, `assistant/message`, `tool/call`, `tool/result`, `steering/message`, `todo/write`, `request/header`), the `deriveMessages()` projection rules, the `TurnTrigger`/`TurnEndReason` reasons, and the turn-enclosure invariant are on **[session.md](session.md)**. How the log is made durable — the `SessionPersistence` seam, JSONL/SQLite backends, the `session/flush` checkpoint, crash recovery, and `SessionHeader` — is on **[persistence.md](persistence.md)**.
 
 ## The agent handle
 
@@ -246,6 +292,15 @@ The fifteen event variants (`turn/start`, `turn/end`, `step/start`, `step/end`,
 
 Source: [`packages/core/agent/src/types.ts`](../../packages/core/agent/src/types.ts)
 
+`InjectOptions` extends ordinary message attribution with context-only framing and durable model-hidden JSON metadata:
+
+```ts type-equiv
+interface InjectOptions extends SendOptions {
+  envelope?: ContextEnvelope
+  meta?: JsonValue
+}
+```
+
 ```ts type-equiv
 interface Agent {
   readonly id: AgentId
@@ -282,8 +337,15 @@ interface Agent {
   /**
    * Inject in-session context (file-change notices, skill content, cron
    * notifications, …): appends a `context/message` session event the next model
-   * request sees at its chronological position, rendered as tagged synthetic
-   * context rather than a user prompt. Does not run the model.
+   * request sees at its chronological position, rendered as synthetic context
+   * rather than a user prompt. The default uses the canonical context tag;
+   * `options.envelope: 'raw'` preserves caller-owned framing. Does not run the
+   * model.
+   *
+   * In an open turn, inject appends at the current log position except while
+   * the current tool-call batch executes: accepted context waits FIFO until the
+   * batch settles, then appends after every recorded result and before turn
+   * close even when execution is interrupted.
    *
    * Turn-enclosure (the turn-enclosure RFC): an inject while a turn is open joins that turn;
    * an inject while idle wraps its `context/message` in a one-shot `injection`
@@ -293,11 +355,11 @@ interface Agent {
    * (inject is synchronous): a failing flush is reported via `agent/error`
    * (step `0`) and the logger, never thrown into the caller.
    *
-   * Live-adapter review has validated the tagged-envelope rendering against
-   * current DeepSeek behavior; provider-specific mismatches belong in that
-   * adapter, not in the canonical session vocabulary.
+   * Live-adapter review has validated the canonical tagged-envelope rendering
+   * against current DeepSeek behavior; provider-specific mismatches belong in
+   * that adapter, not in the canonical session vocabulary.
    */
-  inject(content: ContentBlock[], options?: SendOptions): void
+  inject(content: ContentBlock[], options?: InjectOptions): void
 
   /**
    * Cancel ALL pending work for the agent. `cancel()`:
@@ -347,13 +409,13 @@ interface Agent {
 }
 ```
 
-`AgentStatus` is `'idle' | 'running' | 'disposed'`, and `AgentId` is branded. `AgentOptions` is merge-extensible and currently includes `model?`. Persona belongs to `dsh-system-prompt`: an agent-scoped `deployment:persona` may shadow the global default.
+`AgentStatus` is `'idle' | 'running' | 'disposed'`, and `AgentId` is branded. `AgentOptions` is merge-extensible and currently includes `provider?` and `model?`; dispatch requires both after `agent/request`. Persona belongs to `dsh-system-prompt`: an agent-scoped `deployment:persona` may shadow the global default.
 
 The [event taxonomy](../architecture.md#event) owns the `agent/*` lifecycle, checkpoint, and waterfall contracts. Turn and step boundaries are durable session events rather than agent emits.
 
 ## Interception decisions
 
-Each `agent/*` interception waterfall returns a small, seam-specific typed union — the unified Decision idiom (the tool seams' `PreToolDecision`/`PostToolDecision` in [tools.md](tools.md) follow the same shape). A CC/Codex hook bridge maps its `permissionDecision`/`decision`/`continue`/`additionalContext` fields onto these; a native plugin returns them directly. They share one envelope for model-facing context, `HookContext`, which is `inject()`ed as a `context/message` and so carries a REQUIRED `source` (a missing source would default to `{kind:'user'}` and mislabel plugin context as a user prompt).
+Each `agent/*` interception waterfall returns a small, seam-specific typed union — the unified Decision idiom (the tool seams' `PreToolDecision`/`PostToolDecision` in [tools.md](tools.md) follow the same shape). A CC/Codex hook bridge maps its `permissionDecision`/`decision`/`continue`/`additionalContext` fields onto these; a native plugin returns them directly. Prompt and post-tool decisions share one model-facing context shape, `HookContext`, which is `inject()`ed as a `context/message` and therefore carries a REQUIRED `source` (a missing source would default to `{kind:'user'}` and mislabel plugin context as a user prompt). Its optional `envelope` selects the canonical context tag or caller-owned raw framing, while JSON `meta` persists plugin state without exposing it to the model. Both decisions carry `additionalContexts[]` so every entry preserves its own provenance, framing, and metadata. Continuation reasons are steering messages instead and deliberately use the narrower content/source shape.
 
 Source: [`packages/core/agent/src/types.ts`](../../packages/core/agent/src/types.ts)
 
@@ -361,23 +423,25 @@ Source: [`packages/core/agent/src/types.ts`](../../packages/core/agent/src/types
 interface HookContext {
   content: ContentBlock[]
   source: MessageSource
+  envelope?: ContextEnvelope
+  meta?: JsonValue
 }
 ```
 
-`agent/prompt-submit` returns a `PromptDecision` (allow a drained queued message — optionally rewriting its `content` or attaching `additionalContext` — or block it; a batch whose every prompt is blocked opens a zero-step turn that ends `rejected`):
+`agent/prompt-submit` returns a `PromptDecision` (allow a drained queued message — optionally rewriting its `content` or attaching `additionalContexts` — or block it; a batch whose every prompt is blocked opens a zero-step turn that ends `rejected`):
 
 ```ts type-equiv
 type PromptDecision =
-  | { kind: 'allow'; content?: ContentBlock[]; additionalContext?: HookContext }
+  | { kind: 'allow'; content?: ContentBlock[]; additionalContexts?: HookContext[] }
   | { kind: 'block'; reason: string }
 ```
 
-`agent/turn-continuation` returns a `ContinuationDecision` (the loop's default is `continue` when the step had tool calls or steering was injected, else `stop`; a `continue` `reason` is recorded as next-step steering in the same turn — the typed `/goal` pattern):
+`agent/turn-continuation` returns a `ContinuationDecision` (the loop's default is `continue` when the step had tool calls or steering was injected, else `stop`; a `continue` `reason` is recorded as next-step steering in the same turn and therefore carries no context envelope or metadata — the typed `/goal` pattern):
 
 ```ts type-equiv
 type ContinuationDecision =
   | { action: 'stop' }
-  | { action: 'continue'; reason?: HookContext }
+  | { action: 'continue'; reason?: { content: ContentBlock[]; source: MessageSource } }
 ```
 
 `agent/turn-stop` returns the stop-only `ContinuationStop` subset or `undefined`. The loop calls this serial checkpoint after folding the ordinary decision, its reason, and pending steering; a stop is terminal and discards pending steering.
diff --git a/docs/core-data-structures/filesystem.md b/docs/core-data-structures/filesystem.md
index 6c6dd3a130..6438b1ba03 100644
--- a/docs/core-data-structures/filesystem.md
+++ b/docs/core-data-structures/filesystem.md
@@ -37,6 +37,16 @@ interface FsInfo {
 }
 ```
 
+`lstat` is the path-level no-follow metadata primitive. It takes a path instead of an `FsTarget` because `resolve` intentionally follows symlinks to produce stable identity; consumers that need trust-boundary checks can call `lstat` first and reject `symlink` before resolving.
+
+```ts type-equiv
+interface FsPathInfo {
+  version: FsVersion
+  type: 'file' | 'directory' | 'symlink' | 'other'
+  size?: number
+}
+```
+
 `listDir` returns direct child entries in stable name order. Each entry carries the child basename, type, resolved target, and cheap metadata when the backend can report it. It must not read file contents, so `size` is only for regular files and `version` is metadata-derived. Broken or disappeared children may be returned as `other` without metadata; permission or backend I/O failures while listing or resolving child metadata fail the whole listing with `FS_PERMISSION_DENIED` or `FS_IO_ERROR`.
 
 ```ts type-equiv
@@ -144,4 +154,4 @@ type FsErrorCode =
 
 ## The service and the plugin
 
-`FileSystem` (`ctx.fs`, abstract) owns the provider primitives: `resolve`, `stat`, `readText`, `streamText`, `listDir`, `writeText`, and `editText`. `dsh-fs-policy` registers **no service** — it is a plugin that adds policy through the `fs/*` event gate: it decides the write/edit intent waterfalls (supplying `createIfAbsent`/`replaceIfVersion`/`{ version }` or throwing `FS_NOT_OBSERVED`) and records on `fs/observed`. The executor is `dsh-tool-fs`: it reads/writes/edits through `ctx.fs`, dispatches the waterfalls, and emits the recording event. The generated wiring catalog shows the exact `ctx.fs` signatures on [services.md](../cordis-catalog/services.md#ctxfs--filesystem-abstract-seam).
+`FileSystem` (`ctx.fs`, abstract) owns the provider primitives: `resolve`, `stat`, `lstat`, `readText`, `streamText`, `listDir`, `writeText`, and `editText`. `dsh-fs-policy` registers **no service** — it is a plugin that adds policy through the `fs/*` event gate: it decides the write/edit intent waterfalls (supplying `createIfAbsent`/`replaceIfVersion`/`{ version }` or throwing `FS_NOT_OBSERVED`) and records on `fs/observed`. The executor is `dsh-tool-fs`: it reads/writes/edits through `ctx.fs`, dispatches the waterfalls, and emits the recording event. The generated wiring catalog shows the exact `ctx.fs` signatures on [services.md](../cordis-catalog/services.md#ctxfs--filesystem-abstract-seam).
diff --git a/docs/core-data-structures/llm-streaming.md b/docs/core-data-structures/llm-streaming.md
index fff329310a..f9489a6d7f 100644
--- a/docs/core-data-structures/llm-streaming.md
+++ b/docs/core-data-structures/llm-streaming.md
@@ -16,7 +16,12 @@ type StreamChunk =
   | { type: 'tool-call-delta'; index: number; id: CallId; name?: string; argumentsDelta: string }
   | { type: 'block-end'; index: number; block: ContentBlock }
   | { type: 'usage'; usage: TokenUsage }
-  | { type: 'finish'; reason: FinishReason }
+  | {
+    type: 'finish'
+    reason: FinishReason
+    /** Adapter-private lossless-JSON state for replaying a successful response. */
+    replayState?: unknown
+  }
 ```
 
 ## The adapter contract
@@ -27,8 +32,9 @@ Every adapter MUST obey these, and every consumer may rely on them:
 - **Tool-call `arguments` stay raw JSON strings end-to-end.** Partial fragments stream via `argumentsDelta`; a provider that hands back parsed objects re-stringifies at `block-end`.
 - **Two sanctioned error paths.** A failure may either THROW from `stream()` (transport/protocol errors) **or** end the stream with `finish {kind:'error'|'aborted'}` (provider in-band errors, for adapters that can't throw mid-stream). Consumers must handle *both*. The agent loop translates a finish-error/aborted into a turn error — it never logs a normal completed assistant message for a failed step.
 - **Every provider HTTP request carries the app-attribution header.** Adapters send `attributionHeaders()` (below) - the `User-Agent` baseline - and prove it with a wire-level test (mock server asserting the received header, or the library's header hook for a library-backed adapter).
+- **Replay state is adapter-owned.** A successful `finish` may carry lossless-JSON state needed to reconstruct a native provider response. The loop stores it with the assembled assistant message unless an `agent/step-result` listener rewrote the content. On a later request, `LlmService` passes the state only when the historical provider and target provider are currently registered to the exact same adapter instance. That adapter validates the state and owns any cross-model or cross-provider conversion; other adapters receive the provider-neutral content and provenance without the private state.
 
-This contract is why two adapters exist as a deliberate pair: `dsh-llm-deepseek` (hand-rolled fetch/SSE) and `dsh-llm-pi-ai` (the same endpoint through `@earendil-works/pi-ai`). Two independent internals over one contract is what pinned the protocol down — the library-backed adapter can't throw mid-stream, so it exercises the finish-chunk error path the hand-rolled one might not.
+This contract was pinned down by two deliberately independent implementations: `dsh-llm-deepseek` (hand-rolled fetch/SSE) and `dsh-llm-pi-ai` (a generic multi-provider adapter through `@earendil-works/pi-ai`). The library-backed adapter cannot throw mid-stream, so it exercises the finish-chunk error path the hand-rolled one might not.
 
 ## `AppIdentity` — app attribution
 
@@ -58,11 +64,11 @@ interface TokenUsage {
 
 ## `BlockAssembler`
 
-`BlockAssembler` ([`packages/llm/llm/src/assembler.ts`](../../packages/llm/llm/src/assembler.ts)) is the single shared implementation that folds a `StreamChunk` stream back into `ContentBlock`s and a final `Message`. The loop logs the raw chunks (for replay fidelity) while feeding the same chunks through an assembler — so the canonical log keeps token-level detail and the derived message is rebuilt deterministically. A consumer that needs the assembled result without re-implementing the fold uses this.
+`BlockAssembler` ([`packages/llm/llm/src/assembler.ts`](../../packages/llm/llm/src/assembler.ts)) is the single shared implementation that folds a `StreamChunk` stream back into `ContentBlock`s, usage, finish reason, and replay state. The loop logs the raw chunks while feeding the same chunks through an assembler, then stores the assembled assistant content with its provider/model provenance. A consumer that needs the assembled result without re-implementing the fold uses this.
 
 ## The seam
 
-`LlmAdapter` is the provider seam: subclass, implement `stream()`, register with `ctx.llm.registerAdapter(models, adapter)`. The `block-start` / `block-end` `index` correlation and the assembler together mean an adapter only has to emit well-formed chunks — block reassembly is not each adapter's problem. The consumer surface (`ctx.llm.stream()`) and the `llm/stream` waterfall are described in [architecture.md § Content blocks and streaming](../architecture.md#content-blocks-and-streaming-dsh-llm).
+`LlmAdapter` is the provider seam: subclass, implement `stream()`, and register one adapter instance with `ctx.llm.registerAdapter(providers, adapter)`. `GenerateOptions.provider` selects the registered adapter; `GenerateOptions.model` is passed to that adapter and need not be registered at lifecycle start. Duplicate provider routes fail atomically. Optional `providerInfo()` and asynchronous `listModels()` methods feed `LlmService.listProviders()` / `listModels()` with detached selector metadata. That catalog is advisory rather than a request whitelist: the adapter remains authoritative and may accept unlisted model ids. Adapter lookup happens at the terminal continuation of the `llm/stream` waterfall, so a listener may short-circuit the call or route a mutable one-shot request before lookup. The `block-start` / `block-end` `index` correlation and the assembler together mean an adapter only has to emit well-formed chunks — block reassembly is not each adapter's problem. The consumer surface (`ctx.llm.stream()`) and the `llm/stream` waterfall are described in [architecture.md § Content blocks and streaming](../architecture.md#content-blocks-and-streaming-dsh-llm).
 
 `ContentBlockType` (the key set the `index`-correlated blocks carry) derives from `ContentBlockMap`:
 
diff --git a/docs/core-data-structures/persistence.md b/docs/core-data-structures/persistence.md
index 7bf102924b..c1e02ae743 100644
--- a/docs/core-data-structures/persistence.md
+++ b/docs/core-data-structures/persistence.md
@@ -2,7 +2,7 @@
 
 The **durability seam** for the event log. [session.md](session.md) describes the in-memory `Session` — the append-only `SessionEvent` log that is the source of truth. This page describes how that log is made durable: the abstract `SessionPersistence` service, its backends, the flush checkpoint, crash recovery, and the metadata header that travels alongside the log. The event vocabulary the log carries is enumerated, member by member, in the generated [persistence log event catalog](../persistence-catalog.md).
 
-The seam is a textbook [capability seam](../rfc/implemented/architecture/2026-06-13-capability-seams.md): one abstract service ([dsh-session-persistence](../../packages/session-persistence/session-persistence), `ctx.sessionPersistence`) defining create/append/load/list over the existing `SessionEvent` — **no parallel persisted type** — and two interchangeable backends that pass the same `runPersistenceContract` suite. See the [session-persistence RFC](../rfc/implemented/architecture/2026-06-14-session-persistence.md).
+The seam is a textbook [capability seam](../rfc/implemented/architecture/2026-06-13-capability-seams.md): one abstract service ([dsh-session-persistence](../../packages/session-persistence/session-persistence), `ctx.sessionPersistence`) defining locate/create/append/load/list over the existing `SessionEvent` — **no parallel persisted type** — and two interchangeable backends that pass the same `runPersistenceContract` suite. See the [session-persistence RFC](../rfc/implemented/architecture/2026-06-14-session-persistence.md).
 
 ## The flush checkpoint
 
@@ -12,6 +12,19 @@ The seam is a textbook [capability seam](../rfc/implemented/architecture/2026-06
 
 A backend that reloads a log crashed mid-turn finds an open `turn/start` with no `turn/end`. It does **not** truncate — a single turn can be huge in a long-horizon task (many steps, large tool output), and those events were durably appended before the crash. Instead it closes the orphaned turn with a synthetic `turn/end { reason: { kind: 'interrupted' } }`, keeping the log balanced and the turn-enclosure invariant intact. `interrupted` is the one `TurnEndReason` no loop emits (see [session.md](session.md#why-a-turn-ended-turnendreasonmap)).
 
+## `SessionLocation` — optional per-session artifact target
+
+`SessionPersistence.locate(meta)` synchronously resolves a backend-owned independent artifact without reading, creating, or flushing it. JSONL returns its absolute target path; SQLite returns `undefined` because sessions share one database. A returned path can therefore name a file that does not yet exist or lacks the current unflushed turn; it is a location hint, not authorization or a freshness guarantee.
+
+```ts type-equiv
+interface SessionLocation {
+  /** Backend-specific artifact kind, for example `jsonl`. */
+  readonly kind: string
+  /** Absolute path to this session's backend-owned artifact. */
+  readonly path: string
+}
+```
+
 ## `SessionHeader` — metadata beside the log
 
 Per-session metadata travels **separately** from the event log: format version, cwd, lineage, and the seed boundary are storage concerns, not conversation events, so they stay out of `SessionEventMap` and never reach `deriveMessages()`. The header is attached to a `Session` via `session.header`.
@@ -80,7 +93,7 @@ Replay/fork is therefore `ctx.sessions.create(id, { seed: seedEvents })`; resumi
 
 ## The backends
 
-Both implement the same abstract `SessionPersistence` (create/append/load/list over `SessionEvent`) and pass `runPersistenceContract`, proving the seam is genuinely backend-agnostic:
+Both implement the same abstract `SessionPersistence` (locate/create/append/load/list over `SessionEvent`) and pass `runPersistenceContract`, proving the seam is genuinely backend-agnostic:
 
 - **[dsh-session-persistence-jsonl](../../packages/session-persistence/session-persistence-jsonl)** — an append-only JSONL log per session with crash-safe atomic writes, the interrupted-turn crash recovery above, and a read/replay path.
 - **[dsh-session-persistence-sqlite](../../packages/session-persistence/session-persistence-sqlite)** — `node:sqlite`, one row per `SessionEvent`. The row shape `(session_id, seq, type, time, data, source_event_seqs, surface_op)` maps 1:1 onto the event, including optional surface metadata, so there is no parallel persisted schema to keep in sync.
diff --git a/docs/core-data-structures/session-query.md b/docs/core-data-structures/session-query.md
index ded8ca3f7e..86d2259f7f 100644
--- a/docs/core-data-structures/session-query.md
+++ b/docs/core-data-structures/session-query.md
@@ -1,6 +1,6 @@
 # Session Query
 
-Exact reads over the live-preferred logical session corpus. The [package contract](../../packages/session-query/session-query) owns source precedence, dynamic optional persistence, cloning, surface classification, bounded windows, and typed failures. Full-text search is a separate proposed SQLite phase.
+Exact reads and relationship traces over the live-preferred logical session corpus. The [package contract](../../packages/session-query/session-query) owns source precedence, dynamic optional persistence, cloning, surface classification, bounded windows, tracing validation, and typed failures. Full-text search is a separate proposed SQLite package.
 
 Source: [`packages/session-query/session-query/src/types.ts`](../../packages/session-query/session-query/src/types.ts)
 
@@ -30,6 +30,34 @@ export interface SessionEventRecord {
 }
 ```
 
+## Session lineage
+
+`SessionLineageTrace` carries known parents in immediate-to-outward order and a forest of recursively nested direct descendants. The completeness discriminant makes a known root and a missing parent mutually exclusive.
+
+```ts type-equiv
+export interface SessionLineageNode {
+  session: SessionRecord
+  descendants: SessionLineageNode[]
+}
+```
+
+```ts type-equiv
+export type SessionLineageTrace = {
+  target: SessionRecord
+  ancestors: SessionRecord[]
+  descendants: SessionLineageNode[]
+} & (
+  | {
+    complete: true
+    root: SessionRecord
+  }
+  | {
+    complete: false
+    unresolvedParentId: SessionId
+  }
+)
+```
+
 ## Bounded event reads
 
 The request addresses one raw seq and optional neighboring counts. The result carries a `SessionHeader` rather than availability flags so a known live target can remain independent of persistence health.
@@ -53,6 +81,28 @@ export interface SessionEventWindow {
 }
 ```
 
+## Event relationships
+
+Event traces distinguish positional surface replacement from logged provenance. Every seq list contains direct links except `replacementChain`, which follows immediate replacers from the target to the final positional replacement.
+
+```ts type-equiv
+export interface SessionEventTraceRequest {
+  sessionId: SessionId
+  seq: number
+}
+```
+
+```ts type-equiv
+export interface SessionEventTrace {
+  target: SessionEventRecord
+  replacedBy?: number
+  replacementChain: number[]
+  replacedEventSeqs: number[]
+  sourceEventSeqs: number[]
+  derivedEventSeqs: number[]
+}
+```
+
 ## Errors
 
 The closed code union distinguishes request validation, missing targets, malformed surface logs, optional-backend failure, and contradictory source metadata.
@@ -61,6 +111,7 @@ The closed code union distinguishes request validation, missing targets, malform
 export type SessionQueryErrorCode =
   | 'SESSION_QUERY_EVENT_NOT_FOUND'
   | 'SESSION_QUERY_INVALID_CONFIG'
+  | 'SESSION_QUERY_INVALID_LINEAGE'
   | 'SESSION_QUERY_INVALID_SURFACE'
   | 'SESSION_QUERY_INVALID_WINDOW'
   | 'SESSION_QUERY_PERSISTENCE_FAILED'
diff --git a/docs/core-data-structures/session.md b/docs/core-data-structures/session.md
index d8292c49b3..0e882bae33 100644
--- a/docs/core-data-structures/session.md
+++ b/docs/core-data-structures/session.md
@@ -4,6 +4,14 @@ The in-memory, event-sourced model of [dsh-session](../../packages/core/session)
 
 Source: [`packages/core/session/src/types.ts`](../../packages/core/session/src/types.ts)
 
+## Context framing
+
+`ContextEnvelope` selects the standard tagged projection or preserves a producer-owned complete frame. The latter changes framing only; the event remains a user-role `context/message` in chronological history.
+
+```ts type-equiv
+type ContextEnvelope = 'context' | 'raw'
+```
+
 ## `SessionEventMap` — the event vocabulary
 
 The append-only event types. Merge-extensible: a plugin declares extra event types via declaration merging — e.g. the [compaction seam](compaction.md) adds `compact/start` / `compact/summary` / `compact/end`, and `@deepseek-ai/dsh-hook-protocol` adds log-only `hook/invoked` / `hook/result` provenance for a hook bridge. Like `compact/*`, these are NOT `SurfaceEventType`s (no `surfaceOp`). The generated [persistence log event catalog](../persistence-catalog.md) enumerates every member — core and merged — with its payload, surface badge, and declaration site.
@@ -30,9 +38,16 @@ interface SessionEventMap {
   /**
    * In-session context injection (file-change notices, subdir AGENTS.md,
    * skill content, cron notifications, …). Rendered into the derived history
-   * as tagged synthetic context — NOT a user prompt.
+   * as synthetic context — NOT a user prompt. `envelope: 'raw'` lets a caller
+   * supply its own complete framing; `meta` is persisted JSON hidden from the
+   * model.
    */
-  'context/message': { content: ContentBlock[]; source: MessageSource }
+  'context/message': {
+    content: ContentBlock[]
+    source: MessageSource
+    envelope?: ContextEnvelope
+    meta?: JsonValue
+  }
   /** Raw stream chunk — token-level replay fidelity. */
   'assistant/chunk': { turn: number; step: number; chunk: StreamChunk }
   /**
@@ -41,7 +56,7 @@ interface SessionEventMap {
    * the model output and its accounting travel together (there is no separate
    * usage record). `usage` is absent when the adapter reported none.
    */
-  'assistant/message': { turn: number; step: number; content: ContentBlock[]; usage?: TokenUsage }
+  'assistant/message': { turn: number; step: number; content: ContentBlock[]; provenance: AssistantProvenance; usage?: TokenUsage }
   'tool/call': { turn: number; step: number; callId: CallId; name: string; arguments: string }
   'tool/result': { turn: number; step: number; callId: CallId; content: ContentBlock[]; isError: boolean; error?: { name: string; code: string }; meta?: unknown }
   /** Steering content injected between steps of a running turn. */
@@ -64,25 +79,14 @@ interface SessionEventMap {
    * Full snapshot of the {@link EpochHeader} the NEXT request is built under,
    * with the {@link RequestHeaderReason} it was recorded whole. Appended by
    * the loop inside the step, before dispatch, on a loop instance's first
-   * request-building step (`'initial'`/`'resume'`) or when a delta failed its
-   * round-trip guard (`'fallback'`); always records what the request actually
-   * used, post-`agent/request`. Anchors the header fold: reconstruction reads
-   * the latest snapshot and applies the deltas after it. NOT a
+   * request-building step (`'initial'`/`'resume'`) or when a later request's
+   * header changes (`'change'`); always records what the request actually
+   * used, post-`agent/request`. Reconstruction reads the latest snapshot. NOT a
    * {@link SurfaceEventType}: it produces no LLM message — it is the request
    * envelope, logged so every request is a pure function of the session log
    * (the reconstructability RFC).
    */
   'request/header': { header: EpochHeader; reason: RequestHeaderReason }
-  /**
-   * Amendment to the folded {@link EpochHeader}: system line-trim, name-keyed
-   * tools delta, whole replacement config, or whole replacement session
-   * prefix (an EMPTY array encodes the transition to "none"). The
-   * writer verifies `applyHeaderDelta(previous, delta)` reproduces the new
-   * header exactly and falls back to a `'fallback'` `request/header` snapshot
-   * when it cannot, so a logged delta ALWAYS round-trips. NOT a
-   * {@link SurfaceEventType}.
-   */
-  'request/header-delta': { system?: SystemDelta; tools?: ToolsDelta; config?: LlmCallConfig; messagePrefix?: Message[] }
 }
 ```
 
@@ -97,13 +101,13 @@ export interface TodoItem {
 }
 ```
 
-### The request header events: `request/header` and `request/header-delta`
+### The request header event: `request/header`
 
-The request envelope — the `EpochHeader` (call config + rendered system prompt + assembled tool schemas + the session prefix) — is logged session state, so every conversation request is a pure function of the log (the reconstructability RFC). A `request/header` snapshot (reason `'initial' | 'resume' | 'fallback'`) anchors the fold at conversation birth, process boundaries, and delta-encoding fallbacks; `request/header-delta` events amend it mid-run. `foldRequestHeader(events)` reconstructs the header any request was built under; the writer round-trip-verifies every delta before logging it, so a well-formed log always folds. Neither is a `SurfaceEventType` — they produce no LLM message.
+The request envelope — the `EpochHeader` (call config + rendered system prompt + assembled tool schemas + the session prefix) — is logged session state, so every conversation request is a pure function of the log (the reconstructability RFC). A full `request/header` snapshot with reason `'initial'` or `'resume'` records each loop-instance boundary; a later changed request records another full snapshot with reason `'change'`. `foldRequestHeader(events)` reconstructs the header by selecting the latest snapshot. The event is not a `SurfaceEventType`: it produces no LLM message.
 
 ```ts type-equiv
 export interface EpochHeader {
-  /** The conversation's call configuration (model + sampling scalars). */
+  /** The conversation's call configuration (provider + model + sampling scalars). */
   config: LlmCallConfig
   /** Rendered system prompt text; absent for a system-less request. */
   system?: string
@@ -120,7 +124,7 @@ export interface EpochHeader {
 }
 ```
 
-Canonical form: an empty system prompt, an empty tool list, and an empty session prefix are ABSENT fields, matching how requests are built. `messagePrefix` is the durable record of the `agent/session-prefix` waterfall's product (the request is `messagePrefix + derived history`); composed once per loop instance and anchored by that instance's snapshot, so the loop never produces a prefix delta in practice — the delta arm (whole-array replacement, an empty array encoding the transition back to absence) exists for codec totality. The other delta payloads (`SystemDelta` — a common-prefix/suffix line trim; `ToolsDelta` — name-keyed added/removed/changed) live beside the events in [`packages/core/session/src/types.ts`](../../packages/core/session/src/types.ts).
+Canonical form: an empty system prompt, an empty tool list, and an empty session prefix are absent fields, matching how requests are built. `messagePrefix` is the durable record of the `agent/session-prefix` waterfall's product (the request is `messagePrefix + derived history`); it is composed once per loop instance and included in every full snapshot that instance records. Legacy v0 logs containing the removed `request/header-delta` event or its full-snapshot `fallback` reason are rejected at seed, append, and persistence-load boundaries rather than replayed incompletely.
 
 ## `SessionEvent<T>` — one log entry
 
@@ -150,9 +154,11 @@ type SessionEvent<T extends SessionEventType = SessionEventType> = {
 
 `SessionEventType = keyof SessionEventMap`. Because `SessionEventMap` is merge-extensible, switches over `SessionEvent` must NOT use `assertNever` — a plugin-added variant is a valid unknown value; handle the known cases and fall through `default`.
 
+For `assistant/message`, a present `sourceEventSeqs: []` is a complete known-empty provider stream, while an absent field means legacy or otherwise unrecorded provenance. The loop writes the field for every successful model call; every other surface event requires a non-empty list when the field is present.
+
 ## Surface types
 
-The five message-producing types (`SurfaceEventType` — `user/message`, `assistant/message`, `tool/result`, `context/message`, `steering/message`) carry surface metadata declaring how they join the derived surface linked list. See the [session surface RFC](../rfc/implemented/architecture/2026-06-18-session-surface.md).
+The five message-producing types (`SurfaceEventType` — `user/message`, `assistant/message`, `tool/result`, `context/message`, `steering/message`) carry surface metadata declaring how they join the ordered derived surface. See the [session surface RFC](../rfc/implemented/architecture/2026-06-18-session-surface.md).
 
 ### `SurfaceEventType` — the message-producing subset of event types
 
@@ -173,7 +179,7 @@ export type SurfaceOp =
   | { op: 'replace'; start: number; end: number }
 ```
 
-`'append'` is the normal tail-append path. `replace` shadows surface nodes from `start` through `end` inclusive (both must be valid surface node seqs; `start === end` replaces a single node) and inserts the new node in their place.
+`'append'` is the normal tail-append path. `replace` shadows surface entries from `start` through `end` inclusive (both must be valid surface seqs; `start === end` replaces a single entry) and inserts the new event in their place.
 
 ### `SurfaceIntent` — the parameter to `session.append()`
 
@@ -186,19 +192,11 @@ export interface SurfaceIntent {
 
 Required for `SurfaceEventType` events — every message-producing event must declare how it joins the surface, the sole source of derived history. Non-surface types reject it at compile time.
 
-### `SurfaceNode` — a node in the surface linked list
-
-```ts type-equiv
-export interface SurfaceNode {
-  seq: number
-  prev: number | null
-  next: number | null
-}
-```
+The same provenance distinction applies here: only `assistant/message` may carry a present empty `sourceEventSeqs`; omission does not assert that its source stream was empty.
 
 ### `SurfaceFoldReplacement` and `SurfaceFoldResult` — a complete surface replay
 
-`foldSurface(events)` returns detached current nodes together with the actual node seqs shadowed by each declared replacement range. `SurfaceManager` uses the same transition functions for its incremental cache.
+`foldSurface(events)` returns detached current event sequences together with the actual sequences shadowed by each declared replacement range. `SurfaceManager` uses the same transitions for its incremental cache without retaining replacement history. Its `replaceGeneration` increments for each replacement so incremental consumers can distinguish pure tail growth from a rewrite.
 
 ```ts type-equiv
 export interface SurfaceFoldReplacement {
@@ -211,7 +209,7 @@ export interface SurfaceFoldReplacement {
 
 ```ts type-equiv
 export interface SurfaceFoldResult {
-  nodes: SurfaceNode[]
+  nodes: number[]
   replacements: SurfaceFoldReplacement[]
 }
 ```
@@ -221,11 +219,12 @@ export interface SurfaceFoldResult {
 `Session.deriveMessages()` projects the event log into the `Message[]` the model sees — cached (each surface node projected once, when first seen; a surface rewrite rebuilds) and frozen (a fresh array per call over shared, deep-frozen messages, so mutating logged history through a projection is unrepresentable). `deriveEventMessage(event)` is the per-node pure function the fold applies — public so external reconstructors and the dev invariant project a log prefix with exactly the same rules and cannot disagree with the cache. The projection rules:
 
 - `user/message` → a user message.
-- `assistant/message` → an assistant message. Raw `assistant/chunk` events are replay/UI data and are **skipped** in derivation (the assembled message is authoritative). An **empty-content** `assistant/message` is also skipped — a max-tokens step cut off with no content still records an `assistant/message` to host its `usage`, but a content-less assistant turn must not enter the provider transcript.
+- `assistant/message` → an assistant message with the event's provider/model provenance and optional adapter-private replay state. Raw `assistant/chunk` events are replay/UI data and are **skipped** in derivation (the assembled message is authoritative). An **empty-content** `assistant/message` is also skipped — a max-tokens step cut off with no content still records an `assistant/message` to host its usage/provenance, but a content-less assistant turn must not enter the provider transcript.
 - `tool/result` → a user message carrying a `tool-result` block.
-- `context/message`, `steering/message` → user-role messages wrapped in a tagged envelope (`<context source="…">…</context>`) at their chronological position — the "system-reminder" pattern; the model distinguishes them from real prompts by the envelope.
+- `context/message` → a user-role message at its chronological position. The default `envelope` is `context`, which wraps content as `<context source="…">…</context>`; `envelope: 'raw'` uses caller-owned framing verbatim. Optional JSON `meta` remains in the event log and is never rendered.
+- `steering/message` → a user-role message wrapped in `<steering source="…">…</steering>` at its chronological position.
 
-Everything else (`turn/*`, `step/*`) is structural and does not project into a message. Token usage is observed on `assistant/message.usage` (the step that produced it); an operational error's step number is on `turn/end.reason` for `kind: 'error'`.
+Everything else (`turn/*`, `step/*`) is structural and does not project into a message. Token usage is observed on `assistant/message.usage` (the step that produced it); an operational error's step number is on `turn/end.reason` for `kind: 'error'`. Because this unreleased format intentionally has no compatibility promise, seed/load validation rejects request headers without provider+model and assistant messages without provider/model provenance instead of guessing a route for historical data.
 
 ## Live-session fork API
 
diff --git a/docs/core-data-structures/spill.md b/docs/core-data-structures/spill.md
new file mode 100644
index 0000000000..4e8ced8258
--- /dev/null
+++ b/docs/core-data-structures/spill.md
@@ -0,0 +1,56 @@
+# Spill Storage
+
+The spill storage seam — a [capability seam](../rfc/implemented/architecture/2026-07-08-tool-output-spill-files.md) that persists a tool's oversized text and returns a model-facing locator plus retrieval guidance, split across packages: interface ([dsh-spill](../../packages/spill/spill), `ctx.spillStore`), implementation ([dsh-spill-local](../../packages/spill/spill-local), private session-scoped files on the host filesystem), and consumer ([dsh-spill-policy](../../packages/spill/spill-policy), the `tools/post-execute` policy). Spill is **one optional capability**, not part of the agent-loop spine — so its vocabulary lives here, not in [core.md](core.md). Preview mechanics stay in [dsh-retention](../../packages/util/retention); this seam only saves the final text the policy hands it.
+
+Source: [`packages/spill/spill/src/types.ts`](../../packages/spill/spill/src/types.ts)
+
+## The save request
+
+`saveText` is the whole seam: persist `content` verbatim, return an opaque locator, a backend-supplied retrieval hint, and the exact byte count. The request carries the save-time storage namespace (`owner`), WHERE it came from (`source`, descriptive provenance for naming and inspection — not access control), and a `suggestedName` the backend may use as a naming hint (it is not a path).
+
+```ts type-equiv
+interface SaveTextSpill {
+  owner: SpillOwner
+  source: SpillSource
+  suggestedName: string
+  content: string
+}
+```
+
+```ts type-equiv
+interface SpillOwner {
+  sessionId: SessionId
+}
+```
+
+`SpillOwner.sessionId` is the save-time storage namespace. Forked sessions inherit existing spill locators from the seeded log; those artifacts are not copied or re-owned, and spills produced after the fork use the child session id. A retention-period cleanup may expire old locators with other old session artifacts; the spill seam does not define a per-session cleanup policy.
+
+```ts type-equiv
+interface SpillSource {
+  toolName: string
+  callId: CallId
+  label: string
+}
+```
+
+## The result
+
+```ts type-equiv
+interface SpillRef {
+  locator: SpillLocator
+  bytes: number
+  retrievalHint: string
+}
+```
+
+`SpillLocator` is a [branded](core.md#branded-ids) model-facing handle returned by the backend. The local backend renders it as a filesystem path; a remote or database backend can render a URI, key, or command token. Consumers treat it as opaque and render it with `retrievalHint` instead of assuming `read` is always the right retrieval mechanism.
+
+```ts type-equiv
+type SpillLocator = Branded<'SpillLocator'>
+```
+
+## The service
+
+`SpillStore` (`ctx.spillStore`, defined in [`packages/spill/spill/src/index.ts`](../../packages/spill/spill/src/index.ts)) is a one-method abstract service: `saveText(input) → Promise<SpillRef>`. It persists the FULL `content` and REJECTS on a real storage failure (permissions, ENOSPC, backend unavailable). The seam owns storage only: no retention policy, no tool-result replacement, no retrieval/search API.
+
+The local backend ([dsh-spill-local](../../packages/spill/spill-local)) writes under `<root>/session-<hash>/<random>-<safeName>` — a configured or lazily-created private (0700) root, a `sha256(sessionId)` session subdir, and an exclusive owner-only (`open(path, 'wx', 0o600)`) write so a planted symlink cannot redirect it. Its `locator` is the local path and its `retrievalHint` tells the model to use `read` or `grep` on that path. The policy consumer ([dsh-spill-policy](../../packages/spill/spill-policy)) replaces an over-`maxInlineBytes` plain-text final result with a retention-library head/tail preview plus the spill reference, best-effort: a save failure keeps the original inline result rather than turning a successful call into an `isError`.
diff --git a/docs/core-data-structures/tasks.md b/docs/core-data-structures/tasks.md
new file mode 100644
index 0000000000..0482912af7
--- /dev/null
+++ b/docs/core-data-structures/tasks.md
@@ -0,0 +1,129 @@
+# Background Task Runtime
+
+Types shared by long-running producers, `ctx.tasks`, and task control surfaces. The [runtime RFC](../rfc/implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md) owns the design; this page records the literal shapes from [`packages/tasks/tasks/src/types.ts`](../../packages/tasks/tasks/src/types.ts).
+
+## Ids and status
+
+`TaskId` is a [branded id](core.md#branded-ids) generated as `<kind>-N`. Access control relies on owner authorization, not id secrecy. `TaskKind` derives from a merge-extensible map; the registry treats kinds as opaque id namespaces.
+
+```ts type-equiv
+interface TaskKindMap {
+  bash: 'bash'
+  subagent: 'subagent'
+}
+```
+
+`TaskStatus` is `'running' | 'stopping' | 'completed' | 'killed' | 'failed'`; producer-specific facts belong in `TaskSnapshot.detail`.
+
+## Producer contract
+
+`TaskStart` declares identity and a starter. The runtime finishes preflight before calling `run()` and commits without a later failable step. Producers own execution resources; the runtime owns identity, access, and lifecycle state.
+
+```ts type-equiv
+interface TaskStart {
+  /** Producer kind — also the id prefix (`bash`, `subagent`, …). */
+  kind: TaskKind
+  /** One-line model-facing label (the command; the delegation description). */
+  label: string
+  /**
+   * Owning live agent. Access is fenced by its session id, and agent disposal
+   * cancels and awaits the task. The instance must be the one currently
+   * registered under its agent id. Omitting the owner creates an unowned task,
+   * open to any caller until service disposal.
+   */
+  owner?: Agent
+  /**
+   * Start the work after preflight and synchronously return its hooks. Called
+   * once; a throw leaves nothing registered, and the producer must clean up any
+   * partially started resources.
+   */
+  run(): TaskHooks
+}
+```
+
+`TaskHooks.done` is the quiescence boundary. Optional `readOutput` distinguishes consuming stream tasks from final-output-only tasks.
+
+```ts type-equiv
+interface TaskHooks {
+  /**
+   * Request termination. Must be synchronous, idempotent, and eventually settle
+   * {@link done}; throws propagate. The optional reason is forwarded verbatim.
+   */
+  cancel(reason?: string): void
+  /**
+   * Resolves after the producer releases its resources, not merely when work
+   * finishes. Must not reject; the runtime converts a rejection to `failed`.
+   * If teardown cancellation throws, the runtime may force-fail only the
+   * registry record without claiming that the work stopped.
+   */
+  done: Promise<TaskOutcome>
+  /**
+   * Consume output produced since the previous call. The producer formats
+   * truncation and spill notices. Absence marks a final-output-only task; each
+   * task has one consuming cursor.
+   */
+  readOutput?(): string
+}
+```
+
+```ts type-equiv
+interface TaskOutcome {
+  /** How the task ended: finished (`completed`), cancelled (`killed`), or broke (`failed`). */
+  status: 'completed' | 'killed' | 'failed'
+  /** Kind-specific detail rendered into status lines ('exit code: 3', 'max-tokens'). */
+  detail?: string
+  /** Final output for tasks without `readOutput`; stream tasks leave it unset. */
+  output?: string
+}
+```
+
+## Consumer views
+
+Snapshots are fresh read-only projections. `ownerSession` carries the shared `SessionId` used for authorization; completion listeners separately receive the exact owner object used for lifecycle cleanup. `reported` suppresses a completion notice after another surface has delivered or committed to deliver the terminal state.
+
+```ts type-equiv
+interface TaskSnapshot {
+  /** The registry-issued id (`<kind>-N`). */
+  id: TaskId
+  /** The producer kind the task was registered with. */
+  kind: TaskKind
+  /** The producer-supplied one-line label. */
+  label: string
+  /**
+   * Owner session id used for authorization and correlation; absent for
+   * unowned tasks. Completion listeners receive the exact {@link Agent}
+   * separately through {@link TaskDoneListener}.
+   */
+  ownerSession?: SessionId
+  /** Current lifecycle state. */
+  status: TaskStatus
+  /** Kind-specific status detail, present once the producer supplied one (usually terminal). */
+  detail?: string
+  /** Epoch ms when the task was registered. */
+  startedAt: number
+  /** Epoch ms when the task settled; absent while `running`/`stopping`. */
+  finishedAt?: number
+  /**
+   * True when a kill, read, or wait has reported or committed to report the
+   * terminal state. Completion surfaces suppress redundant notices when set.
+   */
+  reported: boolean
+}
+```
+
+```ts type-equiv
+interface TaskRead {
+  /**
+   * Stream kinds: the consuming delta since the previous read. Final-output
+   * kinds: empty while live, the terminal {@link TaskOutcome.output} (or
+   * empty) once settled — idempotent, never consumed.
+   */
+  text: string
+  /** The task's state at read time. */
+  snapshot: TaskSnapshot
+}
+```
+
+## Service behavior
+
+[`TaskService`](../../packages/tasks/tasks/src/index.ts) provides atomic `start`, caller-scoped `get` and `list`, `read`, `kill`, bounded `wait`, contained `onTaskDone` listeners, and the `attachSurface` availability fence. Authorization compares owner sessions; owner cleanup selects the exact registered `Agent` instance. See [`dsh-tasks`](../../packages/tasks/tasks/README.md) for the package contract and [`dsh-tool-tasks`](../../packages/tasks/tool-tasks/README.md) for the model-facing surface.
diff --git a/docs/core-data-structures/token-meter.md b/docs/core-data-structures/token-meter.md
new file mode 100644
index 0000000000..e082ea08a9
--- /dev/null
+++ b/docs/core-data-structures/token-meter.md
@@ -0,0 +1,39 @@
+# Token Meter
+
+`@deepseek-ai/dsh-token-meter` exposes one detached replay snapshot for request pressure and positional surface pricing. `logRevision` is the number of durable events consumed for every field in the measurement.
+
+Source: [`packages/llm/token-meter/src/types.ts`](../../packages/llm/token-meter/src/types.ts)
+
+## `TokenMeasurement`
+
+```ts type-equiv
+interface TokenMeasurement {
+  /** Number of durable events consumed; equal to the next unread event seq. */
+  readonly logRevision: number
+  /** Provider or heuristic anchor used for this measurement. */
+  readonly baseline: TokenMeasurementBaseline
+  /** Signed repricing of current surface content relative to the baseline anchor. */
+  readonly surfaceDeltaTokens: number
+  /** Non-negative current request-and-response pressure. */
+  readonly totalTokens: number
+  /** Total heuristic tokens across the current surface. */
+  readonly surfaceTokens: number
+  /** Current surface nodes in positional head-to-tail order. */
+  readonly nodes: readonly TokenSurfaceNode[]
+}
+```
+
+`baseline.kind === 'usage'` means the latest successful provider call has the same canonical request envelope and its total is no lower than that call's full heuristic anchor. `estimated` means no reusable conservative usage anchor exists, so the service priced the complete envelope and surface with its fixed heuristic. A later successful request replaces the earlier anchor; signed `surfaceDeltaTokens` preserves growth and shrinkage relative to a matching anchor. `totalTokens` remains request-and-response pressure, while `surfaceTokens` is the surface-only heuristic total and equals the sum of the node prices.
+
+## `TokenSurfaceNode`
+
+```ts type-equiv
+interface TokenSurfaceNode {
+  /** Durable sequence number of the surface event. */
+  readonly seq: number
+  /** Heuristic tokens for the exact message projected by this node. */
+  readonly tokens: number
+}
+```
+
+Surface order is authoritative; replacement nodes can have higher durable seqs than later positional nodes. The snapshot is immutable and does not grow when the underlying replay fold advances.
diff --git a/docs/core-data-structures/tools.md b/docs/core-data-structures/tools.md
index 132d22d521..edb20586bc 100644
--- a/docs/core-data-structures/tools.md
+++ b/docs/core-data-structures/tools.md
@@ -6,11 +6,11 @@ Source: [`packages/core/tools/src/index.ts`](../../packages/core/tools/src/index
 
 ## `ToolDefinition` — a registered tool
 
-A `ToolSchema` (the model-facing fields) plus the `execute` function and optional UI presenters. The registry holds these; the loop dispatches calls through them. The registry's `schemas()` builds the model-facing `ToolSchema[]` by an explicit allowlist — `execute`/`presentCall`/`presentResult` must never leak into a model request.
+A `ToolSchema` (the model-facing fields) plus the `execute` function, host-only scheduler metadata, and optional UI presenters. The registry holds these; the loop dispatches calls through them. The registry's `schemas()` builds the model-facing `ToolSchema[]` by an explicit allowlist — `execute`/`timeoutMs`/`isConcurrencySafe`/`presentCall`/`presentResult` must never leak into a model request.
 
 ```ts type-equiv
 interface ToolDefinition extends ToolSchema {
-  execute(args: unknown, exec: ToolExecution): Promise<ToolExecuteReturn>
+  execute(args: unknown, exec: ToolRunContext): Promise<ToolExecuteReturn>
   /**
    * Cooperative tool-call timeout budget in milliseconds. Omit for no deadline.
    * Enforced by `@deepseek-ai/dsh-timeout-policy` (a `tools/execute` wrapper); it
@@ -19,6 +19,18 @@ interface ToolDefinition extends ToolSchema {
    * cooperative implementation that can reach quiescence when the signal aborts.
    */
   timeoutMs?: number
+  /**
+   * Pure synchronous classifier for overlap with sibling tool calls. Only
+   * `true` opts in; omission, exceptions, non-`true` returns, and invalid
+   * `defineTool` arguments are exclusive. This metadata is never model-visible.
+   *
+   * Opted-in executions must not mutate parent-owned state. Shared state must
+   * tolerate concurrent dispatch; recorder races are permitted only when they
+   * commute or fail closed. See the parallel-tool-call RFC for the full contract.
+   * @param args - parsed arguments; `defineTool` validates before calling.
+   * @returns Whether this call may join a parallel group.
+   */
+  isConcurrencySafe?(args: unknown): boolean
   /**
    * Optional: how to present the PENDING state of one call in a UI, derived from
    * the call's `args` (parsed arguments, `unknown` — the tool validates/narrows
@@ -120,6 +132,27 @@ interface ToolExecutionInput {
 }
 ```
 
+A tool body receives the runtime extension. `deferContext()` is the composite-tool channel: it records nested-dispatch context without injecting inside the still-open outer call.
+
+```ts type-equiv
+interface ToolRunContext extends ToolExecution {
+  /**
+   * Defer one nested-dispatch context until this tool's final result reaches
+   * the agent loop. Contexts retain their individual source, envelope, and
+   * metadata and are emitted in call order.
+   */
+  deferContext(context: HookContext): void
+}
+```
+
+The agent loop asks the registry for each pending call's execution mode and uses it to form exclusive barriers and rolling-pool parallel runs:
+
+```ts type-equiv
+type ToolExecutionMode =
+  | { kind: 'parallel' }
+  | { kind: 'exclusive' }
+```
+
 ```ts type-equiv
 interface ToolExecution extends ToolExecutionInput {
   /** Registry-assigned identity shared with nested calls only as their opaque `parent` token. */
@@ -146,16 +179,16 @@ interface ToolExecutionResult {
    */
   error?: ToolErrorInfo
   /**
-   * Extra model-facing context a `tools/post-execute` listener attached for the
-   * NEXT request (Claude Code's PostToolUse `additionalContext`). It is NOT part
-   * of this call's `content` — `content`/`feedback` shape the tool RESULT, but
-   * `additionalContext` is a SEPARATE `context/message`. A step can carry
-   * multiple tool calls, so the loop BUFFERS every call's `additionalContext`
-   * and appends them only AFTER all `tool/result`s for the step, keeping
-   * tool-call/result adjacency intact. Carried on the result purely to ferry it
-   * from `execute()` up to the loop's per-step buffer.
+   * Extra model-facing contexts deferred by a composite tool or attached by
+   * `tools/post-execute` listeners for the NEXT request. They are not part of
+   * this call's `content`: the loop accepts them into the active-batch FIFO and
+   * appends them after every recorded `tool/result` when the batch settles, even
+   * when execution is interrupted. The array preserves each context's source,
+   * envelope, metadata, and production order. An accepted outer call keeps
+   * deferred contexts before decision contexts; a block retains only contexts
+   * supplied by the blocking decision.
    */
-  additionalContext?: HookContext
+  additionalContexts?: HookContext[]
   /**
    * The tool-private presentation payload from a successful `execute` (the object
    * return form). Threaded onto the `tool/result` session event and back into
@@ -181,8 +214,8 @@ type PreToolDecision =
 
 ```ts type-equiv
 type PostToolDecision =
-  | { kind: 'accept'; content?: ContentBlock[]; additionalContext?: HookContext }
-  | { kind: 'block'; feedback: ContentBlock[]; additionalContext?: HookContext }
+  | { kind: 'accept'; content?: ContentBlock[]; additionalContexts?: HookContext[] }
+  | { kind: 'block'; feedback: ContentBlock[]; additionalContexts?: HookContext[] }
 ```
 
 Call `next()` for the default or return a decision to short-circuit. Pre-policy may deny or ask; only `allowed-once` proceeds, while a non-grant, missing approval channel or service, or agent-less request becomes a denial. Guards may still impose a final denial. Arguments cannot be rewritten because history, audit, UI, and execution must agree.
@@ -232,4 +265,4 @@ How a tool wants its call shown in a UI (an editor tool-call card, a CLI log lin
 
 `ToolCallKind` (`'read' | 'edit' | 'delete' | 'move' | 'search' | 'execute' | 'fetch' | 'other'`) picks an icon on a generic card. `FileLocation` (`{ path, line? }`) and `FileDiff` (`{ path, oldText, newText }`) are the shared file-card vocabulary. The design is pinned in [the render-intent-union RFC](../rfc/implemented/architecture/2026-07-02-tool-render-intent-union.md); the ACP bridge maps a `diff` card to a `{ type: 'diff' }` content block, a `terminal` card to the `_meta` terminal convention, and relativizes a file card's title against the session cwd.
 
-The full presentation field docs live in [`packages/core/tools/src/presentation.ts`](../../packages/core/tools/src/presentation.ts). The bash tool's own schemas (`bash`/`bash_output`/`bash_kill`) and the executor they drive are on [bash.md](bash.md).
+The full presentation field docs live in [`packages/core/tools/src/presentation.ts`](../../packages/core/tools/src/presentation.ts). The `bash` schema and executor are on [bash.md](bash.md); generic background controls are on [tasks.md](tasks.md).
diff --git a/docs/development.i18n.yaml b/docs/development.i18n.yaml
index cb4807c048..494eb09922 100644
--- a/docs/development.i18n.yaml
+++ b/docs/development.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-development.md: 376df72c14b59b8ac8e42b31f19442040d6a47a7
-development.zh.md: 676df50a8a1ab2eff3ec829b5d55904626fa3f5e
+development.md: b3c338f03f548b4de4b676850732731323d611f1
+development.zh.md: 20f5c585dd378a7b0a3bd6b0af8ebf7dc5e0fd3c
diff --git a/docs/development.md b/docs/development.md
index 376df72c14..b3c338f03f 100644
--- a/docs/development.md
+++ b/docs/development.md
@@ -19,7 +19,7 @@ Install dependencies from the repo root:
 pnpm install
 ```
 
-The install also runs the root `postinstall` script, which installs lefthook from the repo dev dependency through `scripts/install-lefthook.mjs`; the wrapper uses lefthook's reviewed `--force` mode so linked worktrees with an existing `core.hooksPath` do not fail normal `pnpm run …` commands.
+The install also runs the root `postinstall` script, which installs lefthook from the repo dev dependency through `scripts/install-lefthook.mjs`; the wrapper script uses lefthook's reviewed `--force` mode so linked worktrees with an existing `core.hooksPath` do not fail normal `pnpm run …` commands.
 
 If hooks are missing because dependencies were restored from cache or `postinstall` was skipped, install them manually:
 
@@ -59,7 +59,7 @@ DEEPSEEK_BASE_URL=https://... # optional
 lefthook is configured in `lefthook.yml` as an early local checkpoint before review:
 
 - `pre-commit` runs staged-file ESLint fixes, `pnpm run typecheck`, and the vendor manifest guard.
-- `pre-push` runs `pnpm run check:pre-push`, whose scheduler runs unit tests, snapshot tests, build, module-graph freshness, and the member gates of `pnpm run hygiene` and `pnpm run doc-sync` concurrently.
+- `pre-push` runs `pnpm run check:pre-push`, whose scheduler runs runtime-closure verification, unit tests, duplication detection, snapshot tests, build, module-graph freshness, and the member gates of `pnpm run hygiene` and `pnpm run doc-sync` concurrently.
 
 The vendor manifest guard checks that changes under `vendor/*/src` are staged with the matching `vendor/README.md` manifest update. See `vendor/README.md` before editing vendored code.
 
diff --git a/docs/development.zh.md b/docs/development.zh.md
index 676df50a8a..20f5c585dd 100644
--- a/docs/development.zh.md
+++ b/docs/development.zh.md
@@ -9,7 +9,7 @@
 - Node.js 支持 22.19+ 与 24+。CI 覆盖 22.19、24 和 26；见 [Node 引擎下限 RFC](rfc/implemented/process/2026-07-06-node-engine-floor.md)。
 - 启用了 Corepack 的 pnpm。仓库在 `package.json` 中固定使用 `pnpm@11.7.0`；如果 `pnpm --version` 无法通过 Corepack 解析，请先运行 `corepack enable`。
 - Git。
-- 可选：一个 DeepSeek API key，用于 REPL/ACP agent（智能体）演示和真实 API 的 e2e 测试。
+- 可选：一个 DeepSeek API key，用于 REPL/ACP（Agent Client Protocol） agent（智能体）演示和真实 API 的 e2e 测试。
 
 ## 首次搭建
 
@@ -58,8 +58,8 @@ DEEPSEEK_BASE_URL=https://... # optional
 
 lefthook 在 `lefthook.yml` 中配置，作为评审前的本地早期检查点：
 
-- `pre-commit` 运行对暂存文件的 ESLint 修复、`pnpm run typecheck` 和 vendor manifest 守卫；
-- `pre-push` 运行 `pnpm run check:pre-push`，其调度器并发运行单元测试、快照测试、构建、module-graph 新鲜度，以及 `pnpm run hygiene` 与 `pnpm run doc-sync` 的各成员门禁。
+- `pre-commit` 运行对暂存文件的 ESLint 修复、`pnpm run typecheck` 和 vendor manifest（元数据清单）守卫；
+- `pre-push` 运行 `pnpm run check:pre-push`，其调度器并发运行 runtime-closure 校验、单元测试、重复代码检查、快照测试、构建、module-graph 新鲜度，以及 `pnpm run hygiene` 与 `pnpm run doc-sync` 的各成员门禁。
 
 vendor manifest 守卫检查 `vendor/*/src` 下的改动是否连同对应的 `vendor/README.md` manifest 更新一起暂存。请在编辑 vendor 代码前先阅读 `vendor/README.md`。
 
diff --git a/docs/event-producer-consumer.md b/docs/event-producer-consumer.md
index ec2da1408d..acd90d4423 100644
--- a/docs/event-producer-consumer.md
+++ b/docs/event-producer-consumer.md
@@ -7,39 +7,39 @@ This matrix shows which packages dispatch each harness-owned event and which pac
 
 | Event | Mode | Declared in | Dispatchers | Listeners |
 | --- | --- | --- | --- | --- |
-| `agent/created` | `emit` | [`packages/core/agent/src/types.ts:139`](../packages/core/agent/src/types.ts) | [`agent`](../packages/core/agent) (`events.dispatch`) | [`jsonrpc`](../packages/ui/jsonrpc), [`stdio`](../packages/ui/stdio) |
-| `agent/disposed` | `emit` | [`packages/core/agent/src/types.ts:148`](../packages/core/agent/src/types.ts) | [`agent`](../packages/core/agent) (`events.dispatch`) | [`stdio`](../packages/ui/stdio) |
-| `agent/error` | `emit` | [`packages/core/agent/src/types.ts:283`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emit`) | - |
-| `agent/pre-step` | `serial` | [`packages/core/agent/src/types.ts:202`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`serial`) | [`compact-basic`](../packages/compact/compact-basic), [`user-approval`](../packages/ui/user-approval) |
-| `agent/prompt-submit` | `waterfall` | [`packages/core/agent/src/types.ts:212`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`acp`](../packages/ui/acp), [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex), [`repeat-tool-guard`](../packages/guard/repeat-tool-guard) |
-| `agent/queued` | `emit` | [`packages/core/agent/src/types.ts:167`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emit`) | - |
-| `agent/request` | `waterfall` | [`packages/core/agent/src/types.ts:224`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | - |
-| `agent/session-prefix` | `waterfall` | [`packages/core/agent/src/types.ts:239`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`tool-skill`](../packages/skill/tool-skill) |
-| `agent/session-start` | `emit` | [`packages/core/agent/src/types.ts:180`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emit`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex) |
-| `agent/status` | `emit` | [`packages/core/agent/src/types.ts:157`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emit`) | [`invariants`](../packages/support/invariants), [`repeat-tool-guard`](../packages/guard/repeat-tool-guard), [`stdio`](../packages/ui/stdio) |
-| `agent/step-result` | `waterfall` | [`packages/core/agent/src/types.ts:250`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | - |
-| `agent/turn-continuation` | `waterfall` | [`packages/core/agent/src/types.ts:260`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex) |
-| `agent/turn-stop` | `serial` | [`packages/core/agent/src/types.ts:270`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`serial`) | [`subagent-inprocess`](../packages/subagent/subagent-inprocess) |
+| `agent/created` | `emit` | [`packages/core/agent/src/types.ts:154`](../packages/core/agent/src/types.ts) | [`agent`](../packages/core/agent) (`events.dispatch`) | [`jsonrpc`](../packages/ui/jsonrpc), [`stdio`](../packages/ui/stdio) |
+| `agent/disposed` | `emit` | [`packages/core/agent/src/types.ts:163`](../packages/core/agent/src/types.ts) | [`agent`](../packages/core/agent) (`events.dispatch`) | [`stdio`](../packages/ui/stdio) |
+| `agent/error` | `emit` | [`packages/core/agent/src/types.ts:298`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emit`) | - |
+| `agent/pre-step` | `serial` | [`packages/core/agent/src/types.ts:217`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`serial`) | [`compact-basic`](../packages/compact/compact-basic), [`time-context`](../packages/context/time-context), [`user-approval`](../packages/ui/user-approval) |
+| `agent/prompt-submit` | `waterfall` | [`packages/core/agent/src/types.ts:227`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`acp`](../packages/ui/acp), [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex), [`repeat-tool-guard`](../packages/guard/repeat-tool-guard) |
+| `agent/queued` | `emit` | [`packages/core/agent/src/types.ts:182`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emit`) | - |
+| `agent/request` | `waterfall` | [`packages/core/agent/src/types.ts:239`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`acp`](../packages/ui/acp) |
+| `agent/session-prefix` | `waterfall` | [`packages/core/agent/src/types.ts:254`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`tool-skill`](../packages/skill/tool-skill), [`workspace-context`](../packages/context/workspace-context) |
+| `agent/session-start` | `emit` | [`packages/core/agent/src/types.ts:195`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emit`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex) |
+| `agent/status` | `emit` | [`packages/core/agent/src/types.ts:172`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`emit`) | [`invariants`](../packages/support/invariants), [`repeat-tool-guard`](../packages/guard/repeat-tool-guard), [`stdio`](../packages/ui/stdio) |
+| `agent/step-result` | `waterfall` | [`packages/core/agent/src/types.ts:265`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | - |
+| `agent/turn-continuation` | `waterfall` | [`packages/core/agent/src/types.ts:275`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`waterfall`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex) |
+| `agent/turn-stop` | `serial` | [`packages/core/agent/src/types.ts:285`](../packages/core/agent/src/types.ts) | [`agent-loop`](../packages/core/agent-loop) (`serial`) | [`subagent-inprocess`](../packages/subagent/subagent-inprocess) |
 | `approval/request` | `waterfall` | [`packages/ui/user-approval/src/index.ts:31`](../packages/ui/user-approval/src/index.ts) | [`user-approval`](../packages/ui/user-approval) (`waterfall`) | [`acp`](../packages/ui/acp) |
-| `fs/edit-intent` | `waterfall` | [`packages/fs/fs/src/index.ts:59`](../packages/fs/fs/src/index.ts) | [`tool-fs`](../packages/fs/tool-fs) (`waterfall`) | [`fs-policy`](../packages/fs/fs-policy) |
-| `fs/observed` | `emit` | [`packages/fs/fs/src/index.ts:68`](../packages/fs/fs/src/index.ts) | [`tool-fs`](../packages/fs/tool-fs) (`emit`) | [`fs-policy`](../packages/fs/fs-policy) |
-| `fs/write-intent` | `waterfall` | [`packages/fs/fs/src/index.ts:51`](../packages/fs/fs/src/index.ts) | [`tool-fs`](../packages/fs/tool-fs) (`waterfall`) | [`fs-policy`](../packages/fs/fs-policy) |
-| `llm/stream` | `waterfall` | [`packages/llm/llm/src/index.ts:39`](../packages/llm/llm/src/index.ts) | [`llm`](../packages/llm/llm) (`waterfall`) | [`invariants`](../packages/support/invariants), [`llm-replay`](../packages/support/llm-replay) |
-| `session/created` | `emit` | [`packages/core/session/src/index.ts:47`](../packages/core/session/src/index.ts) | [`session`](../packages/core/session) (`events.dispatch`) | [`invariants`](../packages/support/invariants), [`jsonrpc`](../packages/ui/jsonrpc), [`session-persistence`](../packages/session-persistence/session-persistence) |
-| `session/disposed` | `emit` | [`packages/core/session/src/index.ts:57`](../packages/core/session/src/index.ts) | [`session`](../packages/core/session) (`events.dispatch`) | - |
-| `session/event` | `emit` | [`packages/core/session/src/index.ts:69`](../packages/core/session/src/index.ts) | [`session`](../packages/core/session) (`events.dispatch`) | [`acp`](../packages/ui/acp), [`invariants`](../packages/support/invariants), [`jsonrpc`](../packages/ui/jsonrpc), [`session-persistence`](../packages/session-persistence/session-persistence), [`stdio`](../packages/ui/stdio) |
-| `session/flush` | `parallel` | [`packages/core/session/src/index.ts:79`](../packages/core/session/src/index.ts) | [`session`](../packages/core/session) (`events.dispatch`) | [`session-persistence`](../packages/session-persistence/session-persistence) |
-| `subagent/end` | `emit` | [`packages/subagent/subagent/src/index.ts:92`](../packages/subagent/subagent/src/index.ts) | [`subagent`](../packages/subagent/subagent) (`events.dispatch`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`jsonrpc`](../packages/ui/jsonrpc) |
-| `subagent/provider-added` | `emit` | [`packages/subagent/subagent/src/index.ts:66`](../packages/subagent/subagent/src/index.ts) | [`subagent`](../packages/subagent/subagent) (`emit`) | [`tool-subagent`](../packages/subagent/tool-subagent) |
-| `subagent/provider-removed` | `emit` | [`packages/subagent/subagent/src/index.ts:72`](../packages/subagent/subagent/src/index.ts) | [`subagent`](../packages/subagent/subagent) (`events.dispatch`) | [`tool-subagent`](../packages/subagent/tool-subagent) |
-| `subagent/start` | `emit` | [`packages/subagent/subagent/src/index.ts:83`](../packages/subagent/subagent/src/index.ts) | [`subagent`](../packages/subagent/subagent) (`events.dispatch`) | [`hooks-claude`](../packages/hooks/hooks-claude) |
-| `system-prompt/assemble` | `waterfall` | [`packages/core/system-prompt/src/index.ts:27`](../packages/core/system-prompt/src/index.ts) | [`system-prompt`](../packages/core/system-prompt) (`waterfall`) | - |
+| `fs/edit-intent` | `waterfall` | [`packages/fs/fs/src/index.ts:61`](../packages/fs/fs/src/index.ts) | [`tool-fs`](../packages/fs/tool-fs) (`waterfall`) | [`fs-policy`](../packages/fs/fs-policy) |
+| `fs/observed` | `emit` | [`packages/fs/fs/src/index.ts:70`](../packages/fs/fs/src/index.ts) | [`tool-fs`](../packages/fs/tool-fs) (`emit`) | [`fs-policy`](../packages/fs/fs-policy) |
+| `fs/write-intent` | `waterfall` | [`packages/fs/fs/src/index.ts:53`](../packages/fs/fs/src/index.ts) | [`tool-fs`](../packages/fs/tool-fs) (`waterfall`) | [`fs-policy`](../packages/fs/fs-policy) |
+| `llm/stream` | `waterfall` | [`packages/llm/llm/src/index.ts:40`](../packages/llm/llm/src/index.ts) | [`llm`](../packages/llm/llm) (`waterfall`) | [`invariants`](../packages/support/invariants), [`llm-replay`](../packages/support/llm-replay) |
+| `session/created` | `emit` | [`packages/core/session/src/index.ts:46`](../packages/core/session/src/index.ts) | [`session`](../packages/core/session) (`events.dispatch`) | [`invariants`](../packages/support/invariants), [`jsonrpc`](../packages/ui/jsonrpc), [`session-persistence`](../packages/session-persistence/session-persistence) |
+| `session/disposed` | `emit` | [`packages/core/session/src/index.ts:56`](../packages/core/session/src/index.ts) | [`session`](../packages/core/session) (`events.dispatch`) | [`session-persistence`](../packages/session-persistence/session-persistence) |
+| `session/event` | `emit` | [`packages/core/session/src/index.ts:68`](../packages/core/session/src/index.ts) | [`session`](../packages/core/session) (`events.dispatch`) | [`acp`](../packages/ui/acp), [`invariants`](../packages/support/invariants), [`jsonrpc`](../packages/ui/jsonrpc), [`session-persistence`](../packages/session-persistence/session-persistence), [`stdio`](../packages/ui/stdio), [`token-meter`](../packages/llm/token-meter), [`workspace-context`](../packages/context/workspace-context) |
+| `session/flush` | `parallel` | [`packages/core/session/src/index.ts:78`](../packages/core/session/src/index.ts) | [`session`](../packages/core/session) (`events.dispatch`) | [`session-persistence`](../packages/session-persistence/session-persistence) |
+| `subagent/end` | `emit` | [`packages/subagent/subagent/src/index.ts:108`](../packages/subagent/subagent/src/index.ts) | [`subagent`](../packages/subagent/subagent) (`events.dispatch`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`jsonrpc`](../packages/ui/jsonrpc) |
+| `subagent/provider-added` | `emit` | [`packages/subagent/subagent/src/index.ts:82`](../packages/subagent/subagent/src/index.ts) | [`subagent`](../packages/subagent/subagent) (`emit`) | [`tool-subagent`](../packages/subagent/tool-subagent) |
+| `subagent/provider-removed` | `emit` | [`packages/subagent/subagent/src/index.ts:88`](../packages/subagent/subagent/src/index.ts) | [`subagent`](../packages/subagent/subagent) (`events.dispatch`) | [`tool-subagent`](../packages/subagent/tool-subagent) |
+| `subagent/start` | `emit` | [`packages/subagent/subagent/src/index.ts:99`](../packages/subagent/subagent/src/index.ts) | [`subagent`](../packages/subagent/subagent) (`events.dispatch`) | [`hooks-claude`](../packages/hooks/hooks-claude) |
+| `system-prompt/assemble` | `waterfall` | [`packages/core/system-prompt/src/index.ts:27`](../packages/core/system-prompt/src/index.ts) | [`system-prompt`](../packages/core/system-prompt) (`waterfall`) | [`acp`](../packages/ui/acp) |
 | `system-prompt/change` | `emit` | [`packages/core/system-prompt/src/index.ts:33`](../packages/core/system-prompt/src/index.ts) | [`system-prompt`](../packages/core/system-prompt) (`emit`) | - |
 | `tools/change` | `emit` | [`packages/core/tools/src/index.ts:116`](../packages/core/tools/src/index.ts) | [`tools`](../packages/core/tools) (`emit`) | - |
 | `tools/execute` | `waterfall` | [`packages/core/tools/src/index.ts:89`](../packages/core/tools/src/index.ts) | [`tools`](../packages/core/tools) (`waterfall`) | [`timeout-policy`](../packages/timeout/timeout-policy) |
-| `tools/post-execute` | `waterfall` | [`packages/core/tools/src/index.ts:98`](../packages/core/tools/src/index.ts) | [`tools`](../packages/core/tools) (`waterfall`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex), [`repeat-tool-guard`](../packages/guard/repeat-tool-guard) |
+| `tools/post-execute` | `waterfall` | [`packages/core/tools/src/index.ts:98`](../packages/core/tools/src/index.ts) | [`tools`](../packages/core/tools) (`waterfall`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex), [`repeat-tool-guard`](../packages/guard/repeat-tool-guard), [`spill-policy`](../packages/spill/spill-policy), [`workspace-context`](../packages/context/workspace-context) |
 | `tools/pre-execute` | `waterfall` | [`packages/core/tools/src/index.ts:80`](../packages/core/tools/src/index.ts) | [`tools`](../packages/core/tools) (`waterfall`) | [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex) |
-| `tools/result` | `emit` | [`packages/core/tools/src/index.ts:106`](../packages/core/tools/src/index.ts) | [`tools`](../packages/core/tools) (`events.dispatch`) | [`subagent-inprocess`](../packages/subagent/subagent-inprocess) |
+| `tools/result` | `emit` | [`packages/core/tools/src/index.ts:106`](../packages/core/tools/src/index.ts) | [`tools`](../packages/core/tools) (`events.dispatch`) | [`subagent-inprocess`](../packages/subagent/subagent-inprocess), [`workspace-context`](../packages/context/workspace-context) |
 | `workflow/agent-end` | `emit` | [`packages/workflow/workflow/src/index.ts:81`](../packages/workflow/workflow/src/index.ts) | [`workflow`](../packages/workflow/workflow) (`events.dispatch`) | - |
 | `workflow/agent-start` | `emit` | [`packages/workflow/workflow/src/index.ts:70`](../packages/workflow/workflow/src/index.ts) | [`workflow`](../packages/workflow/workflow) (`events.dispatch`) | - |
 | `workflow/end` | `emit` | [`packages/workflow/workflow/src/index.ts:91`](../packages/workflow/workflow/src/index.ts) | [`workflow`](../packages/workflow/workflow) (`events.dispatch`) | - |
diff --git a/docs/i18n/README.i18n.yaml b/docs/i18n/README.i18n.yaml
index 43c64ddd90..ab1c9024ad 100644
--- a/docs/i18n/README.i18n.yaml
+++ b/docs/i18n/README.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-README.md: 3a5965ab60a7bdb6345b6abe6815f7090fa98fe1
-README.zh.md: 8e62b53ffefa25ad0bbf713889a2ec2ecdb7027a
+README.md: 17bb1eeb67b4f5119a698fca23f12490c9378a7f
+README.zh.md: c957a82bf420a942e2249942a2d9afc54ad950cf
diff --git a/docs/i18n/README.md b/docs/i18n/README.md
index 3a5965ab60..17bb1eeb67 100644
--- a/docs/i18n/README.md
+++ b/docs/i18n/README.md
@@ -17,16 +17,16 @@ This repo's documentation is read by people and agents both inside and outside t
 
   Blob hashes, not commit hashes, so the record is computable for files edited in the same PR (`git hash-object foo.md`) and consistency is a pure content comparison. The recorded hash also recovers the exact last-confirmed text of either side (`git cat-file -p <hash>`), so an out-of-sync pair is updated by diffing the edited side against its last-confirmed state and patching the counterpart minimally — never by re-translating whole files. After bringing the pair back in line, `pnpm run verify-translation-pairing --write` re-records both hashes; that yaml diff is the reviewable act of confirming consistency.
 - **Language switcher.** Both files link to each other immediately after their H1 heading: the English file carries `English | [中文](foo.zh.md)` and the Chinese file carries `[English](foo.md) | 中文`.
-- **Structure mirrors the counterpart.** Heading depths and order, list kinds, table columns, link targets, and verbatim code blocks match one to one across the pair — see [translation-rules.md](translation-rules.md) for the full preservation rules. Existing Markdown gates apply to `.zh.md` files unchanged (`verify-md-wrap`, `verify-md-links`).
+- **Structure mirrors the counterpart.** Heading depths and order, list kinds, ordered-list starts, list item counts, table row and column counts, link targets, and verbatim code blocks match one to one across the pair — see [translation-rules.md](translation-rules.md) for the full preservation rules. Existing Markdown gates apply to `.zh.md` files unchanged (`verify-md-wrap`, `verify-md-links`).
 
 ## The gate: verify-translation-pairing
 
 `pnpm run verify-translation-pairing` (part of `doc-sync`, so CI and the pre-push hook run it) enforces the contract mechanically:
 
 1. Every file listed as `required` in [scripts/translation-pairing.manifest.json](../../scripts/translation-pairing.manifest.json) has a complete pair.
-2. Every pair that exists at all — required or not — is complete and consistent: all three files present, each side's current blob hash equals the recorded one (editing either side without re-confirming the pair goes red), both sides carry the language switcher, and the structural signatures match in order — heading depths, verbatim code blocks (info string and content), table column counts, list kinds, and every link target apart from the switcher.
+2. Every pair that exists at all — required or not — is complete and consistent: all three files present, each side's current blob hash equals the recorded one (editing either side without re-confirming the pair goes red), both sides carry the language switcher, and the structural signatures match in order — heading depths, verbatim code blocks (info string and content), table row and column counts, list kinds, ordered-list starts, item counts, and every link target apart from the switcher.
 3. Files listed as `excluded` have no `.zh.md` and no `.i18n.yaml` at all.
-4. Every date-named document (`yyyy-mm-dd-*.md`) dated on or after the manifest's `requiredSince` cutoff has a complete pair — new documents merge bilingual from birth.
+4. Every date-named document (`yyyy-mm-dd-*.md`) dated on or after the manifest's `requiredSince` cutoff has a complete pair — new date-named RFCs merge bilingual from birth.
 
 `pnpm run verify-translation-pairing --list` prints the current pairing state of every document in scope — missing, out-of-sync, or ok — and is the work list for translation batches. It never fails; it reports.
 
@@ -49,4 +49,4 @@ The gate's limit, stated plainly: **a green gate means the pair was confirmed co
 
 ## Division of labor
 
-Counterparts here are produced by an agent running [dsh-translate-docs](../../.agents/skills/dsh-translate-docs/SKILL.md) and reviewed by a human — inference is cheap here, review attention is the scarce resource. The gate exists so that neither the agent nor the reviewer has to remember the contract: pair completeness, consistency, and structure are checked mechanically, and review attention goes to translation quality and terminology, where human judgment is the whole point.
+Counterparts here are produced by an agent running [dsh-translate-docs](../../.agents/skills/dsh-translate-docs/SKILL.md) and reviewed by a human — inference is cheap here, review attention is the scarce resource. The gate checks pair completeness, recorded hashes, switchers, and its documented structural signature. Review still owns translation quality, terminology, and structural requirements that the signature does not encode. The prompt contract is executable: [scripts/translation-prompt.ts](../../scripts/translation-prompt.ts) renders the canonical rules into either direction and strictly parses the three-field XML response, while `verify-translation-prompt` exercises both render directions, the checked-in example, and the CDATA split rule in `doc-sync`.
diff --git a/docs/i18n/README.zh.md b/docs/i18n/README.zh.md
index 8e62b53ffe..c957a82bf4 100644
--- a/docs/i18n/README.zh.md
+++ b/docs/i18n/README.zh.md
@@ -2,12 +2,12 @@
 
 [English](README.md) | 中文
 
-本仓库的文档会被公司内外的人和 agent（智能体）阅读，因此 README 与 docs 目录树以英文和简体中文双语维护。本页定义配对契约、强制门禁与推进策略；[translation-rules.md](translation-rules.md) 定义如何翻译；[terminology.md](terminology.md) 是术语真源。进仓的 agent 工作流见 [.agents/skills/dsh-translate-docs](../../.agents/skills/dsh-translate-docs/SKILL.md)。
+本仓库的文档会被公司内外的人和 agent（智能体）阅读，因此 README 与 docs 目录树以英文和简体中文双语维护。本页定义配对契约、强制门禁与推进策略；[translation-rules.md](translation-rules.md) 定义如何翻译；[terminology.md](terminology.md) 是术语真源。仓库内置的 agent 工作流见 [.agents/skills/dsh-translate-docs](../../.agents/skills/dsh-translate-docs/SKILL.md)。
 
 ## 配对契约
 
 - **两种语言同权。**一篇文档可以先用任一语言撰写和评审——先写中文的 RFC 与先写英文的一样正当——另一侧由它翻译而来。两个文件谁也不高于谁；约束它们的是二者必须说同样的话。
-- **一对文档是三个同目录文件。**英文 `foo.md`、中文 `foo.zh.md`，加一份一致性记录 `foo.i18n.yaml`，都在同一目录。不用语言目录，不用独立翻译仓库，不用中英混排的单文件。配对整体合入：PR 永远不会只带一种语言而缺其余两个文件。
+- **一对文档是三个同目录文件。**英文 `foo.md`、中文 `foo.zh.md`，加一份一致性记录 `foo.i18n.yaml`，都在同一目录。不用语言目录，不用独立翻译仓库，不用中英混排的单文件。配对整体合入：PR（Pull Request）永远不会只带一种语言而缺其余两个文件。
 - **一致性记录。**`foo.i18n.yaml` 保存两侧文件在上一次被确认「说同样的话」时各自的完整 git blob hash：
 
   ```yaml
@@ -17,15 +17,16 @@
 
   用 blob hash 而不是 commit hash，这样同一个 PR 里改动的文件也能算出记录（`git hash-object foo.md`），一致性是纯内容比较。记录的 hash 还能还原任一侧上次确认时的确切文本（`git cat-file -p <hash>`），所以失去同步的配对是「把被改的一侧与其上次确认状态做 diff、再最小化地修补另一侧」——从不整篇重译。两侧对齐后，`pnpm run verify-translation-pairing --write` 重新记录两个 hash；那份 yaml diff 就是「确认一致」这个动作本身，可以被评审。
 - **语言切换行。**两个文件在各自 H1 标题之后立即互链：英文文件带 `English | [中文](foo.zh.md)`，中文文件带 `[English](foo.md) | 中文`。
-- **结构与另一侧一一对应。**标题深度与顺序、列表类型、表格列、链接目标与逐字节一致的代码块在配对两侧一一对应——完整保持规则见 [translation-rules.md](translation-rules.md)。既有 Markdown 门禁对 `.zh.md` 文件原样生效（`verify-md-wrap`、`verify-md-links`）。
+- **结构与另一侧一一对应。**标题深度与顺序、列表类型、有序列表起始编号、列表项数量、表格行列数、链接目标与逐字节一致的代码块在配对两侧一一对应——完整保持规则见 [translation-rules.md](translation-rules.md)。既有 Markdown 门禁对 `.zh.md` 文件原样生效（`verify-md-wrap`、`verify-md-links`）。
 
 ## 门禁：verify-translation-pairing
 
-`pnpm run verify-translation-pairing`（`doc-sync` 的一环，因此 CI 和 pre-push 钩子都会运行）机械地强制执行这份契约：
+`pnpm run verify-translation-pairing`（`doc-sync`（文档同步门禁）的一环，因此 CI 和 pre-push 钩子都会运行）机械地强制执行这份契约：
 
 1. [scripts/translation-pairing.manifest.json](../../scripts/translation-pairing.manifest.json) 中 `required` 列出的每个文件都有完整配对。
-2. 任何已存在的配对——无论是否 required——都完整且一致：三个文件齐全、每一侧的当前 blob hash 等于记录值（改了任一侧而没重新确认配对就变红）、双方都带语言切换行、结构签名按序一致——标题深度、逐字节一致的代码块（信息字符串与内容）、表格列数、列表类型，以及除切换行之外的每个链接目标。
+2. 任何已存在的配对——无论是否 required——都完整且一致：三个文件齐全、每一侧的当前 blob hash 等于记录值（改了任一侧而没重新确认配对就变红）、双方都带语言切换行、结构签名按序一致——标题深度、逐字节一致的代码块（信息字符串与内容）、表格行列数、列表类型、有序列表起始编号、列表项数量，以及除切换行之外的每个链接目标。
 3. 列为 `excluded` 的文件完全没有 `.zh.md`，也没有 `.i18n.yaml`。
+4. 凡文件名符合 `yyyy-mm-dd-*.md` 且日期不早于 manifest（元数据清单）中 `requiredSince` 分界日期的文档，都必须有完整配对——新建的日期命名 RFC 从创建起便须配齐中英文。
 
 `pnpm run verify-translation-pairing --list` 打印范围内每篇文档的当前配对状态——missing、out-of-sync 或 ok——是翻译批次的工作清单。它从不失败；它只报告。
 
@@ -44,8 +45,8 @@
 - `docs/i18n/terminology.md` 与 [style-samples.md](style-samples.md)——二者本身即为中英对照文档。
 - [translation-prompt.md](translation-prompt.md)——自动翻译流水线的 prompt 模板；正文逐字进入模型请求，配对翻译会改变流水线行为。
 
-**推进**：以日期命名的文档（`yyyy-mm-dd-*.md`，即 RFC），只要标注日期等于或晚于 manifest 的 `requiredSince` 分界日期，合入时就必须配齐双语文件。更早日期的文件属于待翻清单（backlog），包括分界前夜创建的文件。RFC 文件名记录首次提出日期，因此倒填日期绕过分界属于评审可见的违规。manifest 中的 `required` 列表是当前执行红线，并非全量覆盖这一最终目标。翻译批次将路径加入 `required`，使门禁只向前收紧。未列入的文档仍可通过 `--list` 查看，而任何已存在的配对都受完整契约约束。后续修改必须同步更新两侧，因此 `required` 的扩展速度不能超过翻译评审的承载能力。
+**推进**：以日期命名的文档（`yyyy-mm-dd-*.md`，即 RFC），只要标注日期等于或晚于 manifest 的 `requiredSince` 分界日期，合入时就必须配齐双语文件。更早日期的文件属于 backlog（待翻清单），包括分界前夜创建的文件。RFC 文件名记录首次提出日期，因此倒填日期绕过分界属于评审可见的违规。manifest 中的 `required` 列表是当前执行红线，并非全量覆盖这一最终目标。翻译批次将路径加入 `required`，使门禁只向前收紧。未列入的文档仍可通过 `--list` 查看，而任何已存在的配对都受完整契约约束。后续修改必须同步更新两侧，因此 `required` 的扩展速度不能超过翻译评审的承载能力。
 
 ## 分工
 
-这里的对侧译文由运行 [dsh-translate-docs](../../.agents/skills/dsh-translate-docs/SKILL.md) 的 agent 产出、由人评审——在这里推理（inference）很便宜，评审注意力才是稀缺资源。门禁的存在让 agent 和评审者都不必记住契约：配对完整性、一致性和结构由机械检查兜底，评审注意力投向翻译质量与术语——这正是人的判断的用武之地。
+对侧译文由运行 [dsh-translate-docs](../../.agents/skills/dsh-translate-docs/SKILL.md) 的 agent 生成，再由人评审：在这里推理（inference）很便宜，评审注意力才是稀缺资源。门禁负责检查配对是否完整、记录的 hash、语言切换行以及本文列出的结构签名；翻译质量、术语和签名未涵盖的结构要求仍由评审把关。prompt 契约也有可执行实现：[scripts/translation-prompt.ts](../../scripts/translation-prompt.ts) 会把权威规则渲染到英译中或中译英的 prompt 中，并严格解析包含三个字段的 XML 响应；`doc-sync` 中的 `verify-translation-prompt` 会检查两个渲染方向、仓库内示例与 CDATA 拆分规则。
diff --git a/docs/i18n/style-samples.md b/docs/i18n/style-samples.md
index 6d9e82d184..786ebea0df 100644
--- a/docs/i18n/style-samples.md
+++ b/docs/i18n/style-samples.md
@@ -1,8 +1,8 @@
 # 翻译语体样例（style samples）
 
-本文件是翻译语体的校准锚点：每组样例是一段英文原文与一段人工定稿的中文译文，覆盖本仓库文档的主要文体。**译文的语体以这些样例为准**——它们的效力高于任何对语气的文字描述。翻译或评审时对照最接近的文体样例；样例与规则冲突时，样例胜出。本文件中英对照、自成双语，不参与配对（见 [README.md](README.md) 排除清单）。
+本文件是翻译语体的校准锚点：每组样例是一段英文原文与一段人工定稿的中文译文，覆盖本仓库文档的主要文体。**译文的语体以这些样例为准**——文体样例的效力高于对语气的文字描述，但术语表、忠实性与结构规则仍然优先。翻译或评审时对照最接近的文体样例。本文件中英对照、自成双语，不参与配对（见 [README.md](README.md) 排除清单）。
 
-维护方式：人工评审校准出新的金标段落后追加到对应文体；样例只增不改，改动需评审人签字（PR 评审即签字）。
+维护方式：人工评审校准出新的金标段落后追加到对应文体；发现语义、结构或术语错误时直接修正。新增或修正样例都需经过 PR 评审。
 
 ## ① 架构叙述
 
@@ -16,7 +16,7 @@
 
 > This document covers **behavior**; type shapes live in [core-data-structures/](../core-data-structures/core.md), the per-event/service reference in the [generated catalog](../cordis-catalog/events.md), per-package contracts in the package READMEs ([map](../../packages/README.md)).
 
-本文档描述整体行为逻辑；类型定义存放于 [core-data-structures/](../core-data-structures/core.md)；各类事件、服务的详细参考见[生成目录](../cordis-catalog/events.md)；各 package 对外约束协议写在对应包的 README（[索引](../../packages/README.md)）。
+本文档描述整体行为逻辑；类型定义存放于 [core-data-structures/](../core-data-structures/core.md)；各类事件、服务的详细参考见[生成目录](../cordis-catalog/events.md)；各包（package）的对外契约写在相应的 README 中（[索引](../../packages/README.md)）。
 
 ## ② 防御模式规则
 
@@ -26,11 +26,11 @@
 
 > **Dispose must reach quiescence, not just request it** — A teardown that issues kills/aborts but returns before the work stops leaves orphans. Make cleanup async and await the children's exit (kill → await `done`), and close listener/notification registries BEFORE killing so late completions stay silent. Tests prove disposal waited (pid gone right after `await fiber.dispose()`), not merely that the process eventually dies.
 
-**销毁操作必须等待所有任务完全停稳，不能仅下发终止指令就返回**——若销毁逻辑仅发送终止、中断信号，但不等任务停止就直接退出，会产生孤儿进程。清理逻辑需设为异步，等待所有子任务彻底退出（先下发终止信号，再等待执行完成）；在执行终止操作前先关闭监听器与通知注册表，让延迟到达的完成事件不再触发任何通知。测试要验证销毁流程确实完成等待：执行完 `await fiber.dispose()` 后进程 PID 立即消失，不能仅校验进程最终会自行消亡。
+**dispose（资源释放）必须等待所有任务完全停稳，不能仅下发终止指令就返回**：如果清理过程只发出终止或中断信号，却不等任务停止就返回，就会留下孤儿进程。清理应采用异步方式，等待所有子任务彻底退出（先发出终止信号，再等待退出）；发出信号前应先关闭监听器与通知注册表，使延迟到达的完成事件不再触发通知。测试要证明 dispose 的确等到清理完成：执行完 `await fiber.dispose()` 后进程 PID 立即消失，不能只检查进程最终会自行消亡。
 
 > **Async state is not synchronous state** — `agent.send()` does not flip status before returning; a background task's completion races turn boundaries; `reader.close()` fires for both EOF and disposal. Never gate control flow on a status you only just requested — drive lifecycle off the events/promises that actually fire (`agent/status`, `task.done`), and observe the transition (saw `running` THEN `idle`) rather than counting actions you assume map 1:1 to turns.
 
-**异步状态不等同于同步瞬时状态**——调用 `agent.send()` 不会在返回前同步更新状态；后台任务完成时机与轮次边界存在竞态；调用 `reader.close()` 既可能是读到文件末尾，也可能是资源销毁触发。切勿仅凭刚查询到的状态来阻断流程；生命周期逻辑应基于真实触发的事件与 promise 驱动（`agent/status`、`task.done`），观测完整状态切换（先 `running`、再 `idle`），而非主观认定操作和执行轮次一一对应（主循环会批量处理排队消息）。
+**异步状态不等同于同步瞬时状态**：调用 `agent.send()` 不会在返回前同步更新状态；后台任务的完成时间与轮次边界存在竞态；`reader.close()` 既会在读到文件末尾时触发，也会在资源释放时触发。切勿把刚刚发起的状态变更当成已经生效，据此控制流程；生命周期逻辑应以实际触发的事件和已完成的 promise（`agent/status`、`task.done`）为准，并观察完整的状态变化（先 `running`，再 `idle`），不要根据操作次数推断操作与轮次一一对应。
 
 ## ③ 测试政策清单
 
@@ -40,15 +40,15 @@
 
 > We are DeepSeek — do not ration real-API tests. A no-key test proves the plumbing; only a with-key run proves the agent works against a real model. Write many: real prompts that write files, multi-turn conversations, tool use, cancellation mid-stream. Cheapest and highest-value are **smoke tests** that boot the real example, send one real prompt, and check the world — they catch the "green unit tests, broken product" class that mocks structurally cannot. The self-skip exists only so secretless CI and keyless contributors aren't blocked; it is not a cost signal.
 
-我们是 DeepSeek：真实接口相关测试不得刻意缩减用例数量。无密钥测试仅能验证底层通路；只有携带有效密钥执行的用例，才能确认 agent 可正常对接真实模型。请大量编写此类测试：包含文件写入类真实提示词、多轮对话、工具调用、流式中途取消等场景。
+我们是 DeepSeek：真实接口相关测试不得刻意缩减用例数量。无密钥测试仅能验证底层通路；只有携带有效密钥执行的用例，才能确认 agent（智能体）可正常对接真实模型。请大量编写此类测试：包含文件写入类真实提示词、多轮对话、工具调用、流式中途取消等场景。
 
-成本最低、收益最高的是**冒烟测试**：拉起完整真实示例，发送一条真实提示并校验整体运行状态。这类用例能捕获一类问题——单元测试全部绿灯，但产品实际运行故障，单靠 mock 完全无法发现这类缺陷。
+成本最低、收益最高的是**冒烟测试**：拉起完整真实示例，发送一条真实提示，并检查文件、进程等外部可观察结果。这类用例能捕获一类问题——单元测试全部绿灯，但产品实际运行故障，单靠 mock 完全无法发现这类缺陷。
 
 自带自动跳过逻辑，仅用于保障无密钥 CI 环境、无权限贡献者不会被流程拦截，不代表可以以此为由削减真实接口测试投入。
 
 > **Prefer the real implementation over a mock** — Mock only the genuinely expensive or non-deterministic boundary (the LLM adapter, the network, the clock); keep everything downstream real. A hand-rolled stand-in proves the bridge moves bytes, not that the shipping tool behaves as asserted — the two drift while the test stays green.
 
-**优先使用真实实现，而非 mock 替身**——仅对开销极大、结果不确定的边界模块做 mock（LLM 适配器、网络、时钟），其余下游组件全部使用真实实现。手写的 mock 替身只能验证数据通路能传输字节，无法保证线上工具符合预期逻辑；长期下来业务逻辑与 mock 实现会出现偏差，但测试仍会显示通过。
+**优先使用真实实现，而非 mock 替身**——仅对开销极大、结果不确定的边界模块做 mock（LLM（大语言模型）适配器、网络、时钟），其余下游组件全部使用真实实现。手写的 mock 替身只能验证数据通路能传输字节，无法保证线上工具符合预期逻辑；长期下来业务逻辑与 mock 实现会出现偏差，但测试仍会显示通过。
 
 ## ④ 机制描述
 
@@ -60,7 +60,7 @@
 
 > The gate's limit, stated plainly: a green gate means the pair was confirmed consistent at these exact contents, not that the confirmation was sound. It checks hashes and shape; it cannot judge whether the two sides actually say the same thing — that is the reviewer's half of the contract. A re-recorded pair with a sloppy counterpart passes the gate; it must not pass review.
 
-明确门禁校验边界：门禁校验通过，仅代表两份文档哈希与结构完全匹配，不代表译文内容准确无误。门禁仅校验哈希与结构，无法判断双语表意是否统一——译文质量把关是评审人的责任。即便译文粗糙、表意偏差，只要哈希匹配，门禁就会放行，但这类 PR 绝不能通过人工评审。
+门禁的边界很明确：通过门禁只说明两侧文件当前的 blob hash 与伴随记录吻合，并且结构签名一致，也就是说，这组内容曾被确认一致；它不代表这次确认可靠。门禁无法判断两种语言是否真正表达了相同的意思；这部分契约要由评审人把关。即使译文粗糙、表意有误，重新记录配对后仍能通过门禁，但绝不能通过人工评审。
 
 ## ⑥ RFC 论证
 
@@ -70,9 +70,9 @@
 
 ## ⑦ 推进策略（长段拆分示范）
 
-> **Rollout**: new documents don't wait for a batch — a date-named document dated on or after the manifest's `requiredSince` cutoff must merge with its pair, so everything new is bilingual from birth. For the back-catalog, the `required` list in the manifest is the enforcement frontier, not the goal. […] Pairing a document is a commitment: every later edit to either side must carry the counterpart along, so grow the frontier at the pace translation review is actually resourced, not ahead of it.
+> **Rollout**: date-named RFCs don't wait for a batch — one dated on or after the manifest's `requiredSince` cutoff must merge with its pair, so each new date-named RFC is bilingual from birth. For the back-catalog, the `required` list in the manifest is the enforcement frontier, not the goal. […] Pairing a document is a commitment: every later edit to either side must carry the counterpart along, so grow the frontier at the pace translation review is actually resourced, not ahead of it.
 
-**推进**：新增文档不再走批量分批翻译流程。以日期命名的文档，若其标注日期等于或晚于 manifest 里 `requiredSince` 分界时间，提交合入时必须配套对应的双语译文文件——所有新文档从创建起就要求中英双语齐备。针对存量旧文档：manifest 内的强制翻译列表只是当下执行红线，并非最终目标。（……）文档完成双语配对等同于一份长期约束承诺：后续只要修改任一版本，就必须同步更新对应另一语种文件。因此强制翻译范围的推进节奏，要匹配翻译评审实际可投入人力，切勿超前铺开。
+**推进**：日期命名的 RFC 无需等待批量翻译。只要文件名中的日期不早于 manifest（元数据清单）的 `requiredSince` 分界日期，合入时就必须配齐中英文，因此此类 RFC 从创建起就要求双语齐备。对于存量文档，manifest 中的 `required` 列表只是当前的执行红线，并非最终目标。（……）一旦文档完成配对，后续修改任一侧都必须同步更新另一侧。因此，应根据实际可投入的翻译评审能力逐步扩展执行红线，不能超前。
 
 ## 从样例提炼的要点
 
diff --git a/docs/i18n/terminology.md b/docs/i18n/terminology.md
index d14d0f2fd8..120121e452 100644
--- a/docs/i18n/terminology.md
+++ b/docs/i18n/terminology.md
@@ -33,7 +33,7 @@
 | English | 中文 | 首次出现 | 不要译作 | 备注 |
 |---|---|---|---|---|
 | agent | agent | agent（智能体） | | |
-| agent harness | agent harness | | | agent 组合词（agent harness/workflow/loop/skill 等）整体保留英文；未括注过 agent 时首现按 agent 行处理 |
+| agent harness | agent harness | agent harness（智能体框架） | | agent 组合词（agent harness/workflow/loop/skill 等）整体保留英文；未括注过 agent 时首现按对应组合词或 agent 行处理 |
 | agent loop | agent loop | agent loop（智能体循环） | | |
 | backlog | backlog | backlog（待翻清单） | | 仅在双语翻译语境里括注`待翻清单` |
 | blob hash | blob hash | | | `git hash-object` 的结果 |
@@ -76,13 +76,16 @@
 | backend | 后端 | | | |
 | background task | 后台任务 | | | |
 | block | 块 | | | |
+| build target | 构建目标 | | | |
 | cancel | 取消 | | | |
-| capability | 能力 | | | |
+| feature | 功能 | | 能力 | SDK 产品与工程模型中的可管理产品单元 |
+| feature option | 功能选项 | | variant | 一项 SDK 功能内有限、可选择的实现或配置 |
 | checkpoint | 检查点 | | | |
 | chunk | 分片 | | | |
 | compaction | 压缩 | 压缩（compaction） | | |
 | companion tool | 配套工具 | | | |
-| config | 配置 | | | |
+| Cordis plugin config | Cordis 插件配置 | | | Cordis 插件公开的 `Config` 对象或配置结构 |
+| config key | 配置键 | | | Cordis 插件配置中的单个字段 |
 | consumer | 消费方 | | | |
 | content block | 内容块 | | | |
 | Cookbook | 实操手册 | | | 文档标题用语 |
@@ -90,10 +93,13 @@
 | counterpart | 对侧文件 | | 对应物、配对物 | 双语配对语境；泛指"另一侧"时可写「另一侧」 |
 | context compaction | 上下文压缩 | 上下文压缩（context compaction） | | |
 | contract | 契约 | | | 如：`pairing contract` →`配对契约` |
+| Cordis config entry | Cordis 配置项 | | | 指 `cordis.yml` 插件列表中的一项；插件实现本身写`Cordis 插件` |
+| Cordis plugin | Cordis 插件 | | | Cordis 加载的插件实现，不指 `cordis.yml` 中的一项配置 |
 | coverage | 覆盖率 | | | |
 | crash recovery | 崩溃恢复 | | | |
 | deploy root | 部署根目录 | | | |
 | durability | 持久性 | | | |
+| feature requirement | 功能依赖 | | | 功能或功能选项通过 `requires` 声明的关系 |
 | enforcement frontier | 执行红线 | | 强制边界 | i18n 配对机制用语：manifest `required` 清单所划的门禁生效范围；与金标样例（style-samples ⑦）一致 |
 | event | 事件 | | | |
 | event log | 事件日志 | | | |
@@ -104,7 +110,7 @@
 | extension point | 扩展点 | | | 注意与 `seam` 区分 |
 | fail-fast | 快速失败 | | | |
 | fenced code block | 围栏代码块 | | | 沿用 MDN 中文翻译 |
-| fingerprint | 指纹 | | | i18n 配对机制用语：`.zh.md` 首行记录英文源 blob hash 的注释 |
+| fingerprint | 指纹 | | | 通用内容指纹；双语配对机制使用 sidecar record 记录两侧 blob hash |
 | finish reason | 结束原因 | | | |
 | foreground run | 前台运行 | | | |
 | freshness | 新鲜度 | | | 沿用 MDN 中文翻译；在本项目中指译文相对源文的同步状态 |
@@ -122,6 +128,7 @@
 | mod | 模组 | | | |
 | model provider | 模型提供方 | | | |
 | module | 模块 | | | |
+| npm dependency | NPM 依赖 | | | `package.json` 中的包关系；`dependencies`、`devDependencies` 等字段保持原样 |
 | orphan | 遗留 | | 孤儿、孤立 | 指英文源已不存在的 `.zh.md`（如「遗留译文」）；进程语境按 OS 惯用语译「孤儿进程」 |
 | orphan branch | 孤立分支 | | 孤儿分支 | 沿用 git 官方中文翻译 |
 | package | 包 | 包（package） | | 指 npm 包（`@deepseek-ai/dsh-*`）；`package.json` 等代码标识保持原样 |
@@ -171,3 +178,5 @@
 | vocabulary | 词汇 | | | |
 | wire format | 协议格式 | 协议格式（wire format） | | |
 | workflow | 工作流 | | | |
+| wrapper | 包装层 | | | 软件层或 SDK 包装层 |
+| wrapper script | 包装脚本 | | | 可执行脚本包装层 |
diff --git a/docs/i18n/translation-prompt.md b/docs/i18n/translation-prompt.md
index 60a023979a..8bc15d6e8e 100644
--- a/docs/i18n/translation-prompt.md
+++ b/docs/i18n/translation-prompt.md
@@ -1,24 +1,27 @@
 # Translation prompt (pipeline asset)
 
-本文件是自动翻译流水线的 prompt 模板；自 `# Translation Prompt` 起的正文逐字进入模型请求，因此不参与双语配对（见 [README.md](README.md) 排除清单）。渲染时，[terminology.md](terminology.md) 整表填入 `{{terminology}}`。[style-samples.md](style-samples.md) 定义文体，模板内嵌的 Examples 仅抽样问题类型；两者冲突时以文体样例为准。修改本文件即修改翻译行为，需按正常 PR 评审。
+本文件是自动翻译流水线使用的 prompt 模板；从 `# Translation Prompt` 开始的正文会逐字进入模型请求，因此本文件不参与双语配对（见 [README.md](README.md) 排除清单）。渲染时会把 [translation-rules.md](translation-rules.md) 全文填入 `{{translation_rules}}`，把 [terminology.md](terminology.md) 整表填入 `{{terminology}}`，以免模板另存一份规则而日后失去同步。[style-samples.md](style-samples.md) 定义文体，模板中的 Examples 只用于说明典型问题；术语表、忠实性和结构规则优先于样例，样例只在这些硬性约束内决定文体。修改本文件会改变翻译行为，需正常经过 PR 评审。
 
 ## 占位符契约
 
-流水线渲染模板时替换以下占位符，除此之外不做任何文本处理：
+流水线渲染模板时替换以下占位符，除此之外不改写系统消息：
 
 | 占位符 | 填入内容 | 来源 |
 |---|---|---|
 | `{{source_lang}}` | 源语言名（`English` / `Chinese`） | 由改动侧文件推断：`.zh.md` 被改则为 `Chinese` |
 | `{{target_lang}}` | 目标语言名（`Chinese` / `English`） | 与 `{{source_lang}}` 相对 |
+| `{{translation_rules}}` | [translation-rules.md](translation-rules.md) 全文（Markdown 原文） | 渲染时读取仓库当前版本，不缓存 |
 | `{{terminology}}` | [terminology.md](terminology.md) 的完整表格（Markdown 原文） | 渲染时读取仓库当前版本，不缓存 |
-| `{{source_filename}}` | 源文档的 basename（如 `foo.md`） | 由流水线从待译文件路径取得 |
-| `{{source_filename_zh}}` | 中文侧 basename（如 `foo.zh.md`） | 由 `{{source_filename}}` 派生 |
+| `{{source_filename}}` | 源文档的 basename（如 `foo.md` 或 `foo.zh.md`） | 由流水线从待译文件路径取得 |
+| `{{source_filename_zh}}` | 中文侧 basename（如 `foo.zh.md`） | 英文源追加 `.zh`；中文源使用自身 basename |
 
-流水线仅支持上表占位符，并按整篇文档翻译。它不支持 `{{to}}`、`{{title_prompt}}`、`{{summary_prompt}}`、`{{terms_prompt}}`、`{{imt_style_guide}}` 或 `%%` 分段协议；输出采用下方三段 XML。
+例如，英译中时若源文件是 `foo.md`，`{{source_filename}}` 填 `foo.md`，`{{source_filename_zh}}` 填 `foo.zh.md`；中译英时若源文件是 `foo.zh.md`，两个占位符都填 `foo.zh.md`。
+
+流水线只识别上表中的占位符，并且一次翻译整篇文档。它不支持 `{{to}}`、`{{title_prompt}}`、`{{summary_prompt}}`、`{{terms_prompt}}`、`{{imt_style_guide}}` 或 `%%` 分段协议。输出必须是一个以 `<dsh-translation-response>` 为根元素的 XML 文档；三个子元素中的 Markdown 内容都放在 CDATA 中。内容出现 `]]>` 时写成 `]]]]><![CDATA[>`，XML 解析后仍会还原为原文。
 
 ## Few-shot 金标
 
-流水线的 few-shot 是**整文档级**的中英对照，不是模板内嵌的句子级正误例（那是最小抽样）。few-shot 集取自以下 5 组人工定稿的配对文档，以仓库当前版本为准、随仓库更新：
+流水线使用**整篇文档**的中英对照作为 few-shot，不是模板内嵌的句子级正误例。以下 5 组配对文档均经过人工评审，并以仓库当前版本为准，随仓库一同更新：
 
 - `README.md` ↔ `README.zh.md`
 - `docs/development.md` ↔ `docs/development.zh.md`
@@ -26,136 +29,58 @@
 - `docs/i18n/translation-rules.md` ↔ `docs/i18n/translation-rules.zh.md`
 - `docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.md` ↔ 对应 `.zh.md`
 
-注入方式：在系统消息（本模板）之后、待译文档之前，每组作为一轮示例对话——user 消息为源文档全文，assistant 消息为定稿译文全文（不带三段 XML 包装；只有真实请求要求三段输出）。上下文紧张时按上列顺序从后往前裁剪组数。这 5 组也是评审校准锚点（见 [style-samples.md](style-samples.md)），改动任何一组即改变流水线行为。
+注入时按当前翻译方向选择每组的源侧与目标侧：user 消息包含源文档全文，assistant 消息采用模板正文规定的 XML 协议；`translation` 与 `final` 都放入目标文档全文，`review` 填 `- [None] No corrections.`。CDATA 遵循上文的 `]]>` 拆分规则。上下文不足时，按上列顺序从后往前删减示例组数。这 5 组也是评审校准锚点；改动任何一组都会改变流水线行为。
 
 ## 模板正文
 
 ````text
 # Translation Prompt
 
-You are a senior technical translator specializing in LLM and agent development documentation. Your task is to translate the given source document from {{source_lang}} to {{target_lang}}, producing natural, professional technical prose.
+You are a senior technical translator specializing in LLM and agent development documentation. Translate the complete source document from {{source_lang}} to {{target_lang}} as natural, professional technical prose.
 
-## Quality Requirements
+## Binding Translation Rules
 
-### Structure and Format Preservation
-- Output a complete translated document that maintains exactly the same structure as the source: heading hierarchy, list shape, table columns, link targets, and code blocks.
-- Fenced code blocks must be byte-identical to the source, including ALL comments inside them. Do NOT translate comments inside code blocks. This is a hard rule with no exceptions.
-- Inline code spans (commands, flags, paths, API names, version numbers) must be kept verbatim. Never translate or reformat them.
-- Every relative link must point to the same target as in the source. Link text is translated; link targets are not.
-- Language switcher line: the source document's filename is `{{source_filename}}`. When translating into Chinese, write `[English]({{source_filename}}) | 中文` immediately after the H1 heading. When translating into English, write `English | [中文]({{source_filename_zh}})` immediately after the H1. Emit this line even when the source file has no switcher yet (a brand-new pair); when the source does have one, flip the link direction — never copy it unchanged.
-- After a closing bold marker `**`, insert a space before the next character when that character is a Latin letter or digit. Never insert a space before full-width (Chinese) punctuation.
+The canonical repository rules below are injected verbatim. Apply every direction-appropriate requirement. In those rules, the authored document is the source for this request and the generated document is its counterpart.
 
-### Tone and Style
-- The translation must read as if originally written in the target language by a native speaker. If an expression sounds like a word-for-word rendering from the source language, rephrase it.
-- Write in a professional, formal tone appropriate for developer documentation. Never use colloquial or casual expressions.
-- Use polite imperative forms where the text instructs the reader to do something.
-- Keep the author's register: concise stays concise, detailed stays detailed.
+{{translation_rules}}
 
-### Sentence Structure
-- Break long sentences with commas or semicolons. Avoid run-on sentences.
-- Prefer active voice. Convert passive constructions to active if it reads more naturally.
-- Translate meaning, not words. Restructure sentences where the target language grammar requires it.
-- Do not invent words or expressions that do not exist in natural technical writing of the target language.
+## Request-Specific Structure
 
-### Word Choice
-- Prefer precise, formal vocabulary over casual or colloquial alternatives.
-- When multiple synonyms exist, choose the one most commonly used in professional technical documentation of the target language.
-- Avoid slang, internal jargon, or overly literal translations that would not be recognized by the general developer audience.
-- Do not use the same word to translate two different source-language terms that carry distinct meanings.
-- Avoid repeating the same verb in close proximity; vary word choice for readability.
+- The source basename is `{{source_filename}}`. When translating into Chinese, write `[English]({{source_filename}}) | 中文` immediately after the H1. When translating into English, write `English | [中文]({{source_filename_zh}})` immediately after the H1.
+- Emit the switcher for a new pair and flip an existing switcher; never copy it unchanged.
 
-#### When translating into Chinese
-- When a number modifies a noun, always include a Chinese classifier or measure word (量词). For example: "three-package seam" → "由三个 package 构成的 seam", not "三 package seam".
+## Binding Terminology
 
-### Punctuation
-
-#### When translating into Chinese
-- Use full-width Chinese punctuation in prose: `，。：；？！（）「」`.
-- Strongly prefer replacing all em-dashes (——) with colons, periods, commas, or parentheses. Keep an em-dash only if no other punctuation works at all.
-- Use enumeration commas (、) between parallel items, not regular commas.
-- List item endings: use semicolons or no punctuation. Do not end list items with commas.
-- Put one half-width space between Chinese text and Latin words/numbers.
-- For RFC 2119 keywords (MUST, MUST NOT, SHOULD, MAY), render the corresponding Chinese term in italics: *必须*、*禁止*、*应当*、*可以*.
-
-#### When translating into English
-- Use half-width English punctuation and standard English spacing. Preserve full-width punctuation only in verbatim Chinese text.
-- Convert enumeration commas (、) to English commas; convert 「」 quotes to English double quotes.
-- Render the terminology table's English column the same way the Chinese column binds the other direction: listed terms use exactly the table's English form; first-occurrence glosses do not carry over (English prose never glosses an English term with Chinese).
-- Chinese topic-comment sentences and dropped subjects become explicit English subjects; prefer concise declaratives over nominalizations.
-- Do not transliterate Chinese engineering idioms literally: render the underlying concept (误报 → false positive, 执行红线 → enforcement frontier), consulting the terminology table first.
-- Keep the register of institutional developer documentation: contractions are acceptable, marketing language and hedging (very, quite, simply) are not.
-
-## Terminology
-
-A terminology table is provided below. Follow it strictly:
-- Render every listed term exactly as specified.
-- First occurrence: write as shown in the "首次出现" column (with parenthetical gloss). Subsequent occurrences: write only the part before the parentheses.
-- If a term has already been glossed as part of a compound term, do not gloss it again when it appears alone later.
-- NEVER use translations listed in the "不要译作" column.
-- For technical terms not in the table: keep them in the source language. Do not invent a translation. This rule applies to terminology only; for general prose, freely restructure and paraphrase for natural expression.
+Apply the current table below exactly as required by the injected translation rules.
 
 {{terminology}}
 
 ## Output Format
 
-Produce your output in three XML sections:
+Return exactly one well-formed XML document with this root and these three child elements. Do not wrap it in a Markdown code fence. Put all Markdown and review text inside CDATA. If any content contains the CDATA terminator, split it as `]]]]><![CDATA[>` so XML parsing reconstructs the original `]]>` sequence.
 
 ```xml
-<translation>
-(Complete translation of the source document)
-</translation>
-
-<review>
-(Self-review notes, one correction per line with category tag, e.g.)
-- [Tone] "旁挂记录" → "伴随记录"（生造词）
-- [Sentence] 第 3 段补充逗号断句
-- [Punctuation] 两处破折号替换为冒号
-- 无修正
-</review>
-
-<final>
-(Final translation after corrections)
-</final>
+<dsh-translation-response version="1">
+<translation><![CDATA[
+(Complete first-pass translation)
+]]></translation>
+<review><![CDATA[
+- [Tone] Replaced a literal rendering with the established target-language phrasing.
+- [Terminology] Applied the binding sidecar record term.
+]]></review>
+<final><![CDATA[
+(Complete corrected translation)
+]]></final>
+</dsh-translation-response>
 ```
 
 ## Self-Review Instructions
 
-After writing `<translation>`, re-read it in the target language only, without looking at the source. Check by category:
-
-**Structure**
-- Is the heading hierarchy, list shape, and code block content identical to the source?
-- Are ALL comments inside code blocks left untranslated (byte-identical to source)?
-- Is the language switcher line correctly flipped (not copied from source)?
-- Are link targets preserved and bold markers followed by a space?
-
-**Tone & Style**
-- Does every sentence read as if originally written by a native speaker?
-- Is there any colloquial, casual, or overly informal phrasing?
-
-**Sentence Structure**
-- Are there run-on sentences that need breaking?
-- Are there stiff passive constructions that should be converted to active voice?
-
-**Word Choice**
-- Are there overly literal translations that sound unnatural?
-- Is the same target-language word used to translate two distinct source concepts?
-- Is any slang or internal jargon present?
-
-**Terminology**
-- Are first-occurrence glosses correctly applied (not missing, not repeated)?
-- Are any "不要译作" forbidden translations present?
-- Are unlisted terms correctly kept in the source language?
-
-**Punctuation** (when target is Chinese)
-- Are there em-dashes that should be replaced with colons, periods, or commas?
-- Are list items ending with commas instead of semicolons?
-- Are RFC 2119 keywords rendered in italics?
-
-Record corrections in `<review>` with category tags. Then output the corrected version in `<final>`. If no corrections are needed, write "无修正" in `<review>` and copy the translation unchanged into `<final>`.
+After writing `<translation>`, re-read it in the target language without looking at the source. Then apply the injected translation rules as a clause-by-clause comparison against the source and record actual corrections in English inside `<review>`. Apply every recorded correction in `<final>`. If no correction is needed, write only `- [None] No corrections.` in `<review>` and copy `<translation>` unchanged into `<final>`.
 
 ## Examples
 
-Below are representative examples of common problems and their corrections. Follow the "Good" versions.
+Follow the Good versions; these sentence-level examples illustrate error categories, not the assistant-message wire format.
 
 ### Colloquial verb → Professional verb
 - Source: `The repo pins pnpm@11.7.0 in package.json`
@@ -177,40 +102,40 @@ Below are representative examples of common problems and their corrections. Foll
 - Bad: `旁挂记录两侧 blob hash，使一致性可检查`
 - Good: `伴随记录保存两侧 blob hash，使一致性可检查`
 
-### Em-dash → Colon/period
-- Source: `FIXME — an issue that should block a new release. A release should not ship with an open FIXME unless reviewers explicitly agree the change can be merged anyway.`
-- Bad: `FIXME——应当阻塞新版本发布的问题。除非评审者明确同意可以照常合入，发布不应带着未解决的 FIXME 出门。`
-- Good: `FIXME：应当阻塞新版本发布的问题。除非评审者明确同意该更改可以合并，否则发布版本不应包含未解决的 FIXME。`
-
 ### Overly literal → Meaningful rendering
 - Source: `awkward phrasing is easier to hear without the source anchoring you`
 - Bad: `没有源文锚着，别扭的表述更容易被听出来`
 - Good: `不对照原文时，更容易察觉别扭的表达`
 
-### Terminology — do not translate what should be kept in English
+### Terminology — keep the binding English form
 - Source: `typed service seams, and explicit extension points`
 - Bad: `类型化的服务 seam（扩展点）与显式扩展点`
 - Good: `类型化的服务 seam 与显式扩展点`
 
-### Slang/jargon → Professional phrasing
+### Slang → Professional phrasing
 - Source: `The committed agent workflow lives in .agents/skills/dsh-translate-docs`
 - Bad: `进仓的 agent 工作流见 .agents/skills/dsh-translate-docs`
 - Good: `仓库内置的 agent 工作流见 .agents/skills/dsh-translate-docs`
 
-### "For humans" — translate the intent, not the word
-- Source: `For humans, start with the development guide`
-- Bad: `对于人工读者，请先从开发指南开始`（"人工读者"生硬）
-- Good: `面向开发者：请先阅读开发指南`（"开发者"自然，且中文里冒号在此处更自然）
+### Chinese → English — idiomatic subject and predicate
+- Source: `门禁绿并不代表译文内容正确。`
+- Bad: `The gate green does not represent that the translation content is correct.`
+- Good: `A green gate does not mean the translation is correct.`
 
-### Code block comments — NEVER translate
+### Code block comments — never translate
 - Source code block contains: `# REPL agent demo (needs DEEPSEEK_API_KEY)`
 - Bad: `# REPL agent 演示（需要 DEEPSEEK_API_KEY）`
-- Good: `# REPL agent demo (needs DEEPSEEK_API_KEY)` (keep exactly as-is, byte-for-byte)
+- Good: `# REPL agent demo (needs DEEPSEEK_API_KEY)` (byte-identical)
 
-### Language switcher — flip direction
-- Source file (English) has: `English | [中文](README.zh.md)`
-- Bad (copying source unchanged): `English | [中文](README.zh.md)`
-- Good (flipped for Chinese file): `[English](README.md) | 中文`
+### Language switcher — English to Chinese
+- Source: `English | [中文](README.zh.md)`
+- Bad: `English | [中文](README.zh.md)`
+- Good: `[English](README.md) | 中文`
+
+### Language switcher — Chinese to English
+- Source: `[English](README.md) | 中文`
+- Bad: `[English](README.md) | 中文`
+- Good: `English | [中文](README.zh.md)`
 
 ---
 
diff --git a/docs/i18n/translation-rules.i18n.yaml b/docs/i18n/translation-rules.i18n.yaml
index 048ce5eef7..dea0883a4b 100644
--- a/docs/i18n/translation-rules.i18n.yaml
+++ b/docs/i18n/translation-rules.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-translation-rules.md: 3e99aa5432ccd904f702238e9b802a9ed9bf6832
-translation-rules.zh.md: 1d8dcb04f1015af2db46c2e25b44f5cd147b6ecc
+translation-rules.md: fb6aa9ac05bebe68ff9213af99f64457bdb1ad6f
+translation-rules.zh.md: 04dd0a704e19502c676ea0966437870c5af0624f
diff --git a/docs/i18n/translation-rules.md b/docs/i18n/translation-rules.md
index 3e99aa5432..fb6aa9ac05 100644
--- a/docs/i18n/translation-rules.md
+++ b/docs/i18n/translation-rules.md
@@ -6,27 +6,27 @@ How to translate between the two sides of a documentation pair in this repo. Bot
 
 ## Faithfulness
 
-- The counterpart MUST say what the authored side says — no added behavior, prerequisites, warnings, version claims, or examples, and no dropped ones. If the pair disagrees on substance, neither language wins by default: fix the side that is wrong, then bring the other along in the same change.
-- The counterpart SHOULD read as natural technical writing in its own language, not word-by-word gloss. Translate meaning, restructure sentences where the target grammar wants it, and keep the author's register — terse stays terse.
+- The counterpart *MUST* say what the authored side says — no added behavior, prerequisites, warnings, version claims, or examples, and no dropped ones. If the pair disagrees on substance, neither language wins by default: fix the side that is wrong, then bring the other along in the same change.
+- The counterpart *SHOULD* read as natural technical writing in its own language, not word-by-word gloss. Translate meaning, restructure sentences where the target grammar wants it, and keep the author's register — terse stays terse.
 - Do not translate the untranslatable: if a sentence resists natural rendering because it leans on an idiom of the source language, translate the idea, not the idiom.
 
 ## Voice
 
-- The register is calibrated by [style-samples.md](style-samples.md) — human-approved gold pairs, one per document genre. The counterpart MUST match the register of the nearest sample; where a sample and a prose rule here disagree, the sample wins. The target is institutional technical Chinese: complete sentences, declarative, neither chatty nor academic.
+- The register is calibrated by [style-samples.md](style-samples.md) — human-approved gold pairs, one per document genre. The counterpart MUST match the target-language side of the nearest sample; where its voice and a prose voice rule disagree, the sample wins. Chinese targets use institutional technical Chinese; English targets use concise professional developer prose.
 - Write as a native technical author restating the content, not as a translator transposing sentences. Then verify against the source clause by clause: nothing added, nothing dropped — fluency never justifies losing a clause.
-- Give sentences an explicit agent: where the English uses a passive or an abstract subject, name the actor (系统、门禁、评审人).
-- Prefer established Chinese engineering idiom over calques (误报／漏检 for false positive/negative, 执行红线 for enforcement frontier); localize metaphors instead of transplanting them, and unpack English noun chains into verb clauses.
+- Give sentences an explicit actor when the target language would otherwise obscure it; for Chinese, replace vague passives or abstract subjects with the actual actor (系统、门禁、评审人).
+- Prefer established target-language engineering idiom over calques (误报／漏检 for false positive/negative, 执行红线 for enforcement frontier); localize metaphors instead of transplanting them, and unpack noun chains where the target language requires it.
 - Split long paragraphs by semantic unit — one idea per paragraph. Paragraph boundaries MAY differ from the source; the structural signature does not count paragraphs.
-- Category nouns render in Chinese with a first-mention English annotation (实操手册（cookbook）); literal directory or file references stay code-formatted English.
+- When translating into Chinese, category nouns use Chinese with a first-mention English annotation (实操手册（cookbook）); when translating into English, use the conventional English category name. Literal directory or file references stay code-formatted English.
 
 ## Structure preservation
 
-Shape is enforced by the pairing gate, so the writer never trades fluency against it — write naturally inside the frame. The paired files MUST match one to one in:
+The pairing gate checks heading depths, fenced code blocks, table row and column counts, list kinds, ordered-list starts, list item counts, and link targets. Preserve the rest of the frame manually; the paired files MUST match one to one in:
 
 - heading hierarchy (same levels, same order — heading TEXT is translated),
 - list shape and numbering,
 - tables (same columns, same row order; header cells translated per terminology),
-- fenced code blocks — **byte-identical, including comments**; code is part of the verified surface (` ```ts ` blocks compile under `doc-typecheck`), and an edited comment is drift the fence-count gate cannot see,
+- fenced code blocks — **byte-identical, including comments**; the pairing signature compares their info strings and contents, and ` ```ts ` blocks compile under `doc-typecheck`,
 - inline code spans (commands, flags, config keys, file paths, event names, API names, version numbers) — verbatim, never translated or reformatted,
 - links and anchors: every relative link MUST point at the same target in both files — by convention the `.md` path, not the `.zh.md` sibling — so links never dangle when one pair lands before its neighbors. The ONLY zh-specific link is the language switcher. Link TEXT is translated; the target is not.
 
@@ -34,9 +34,9 @@ The repo's Markdown conventions apply to `.zh.md` files unchanged: one physical
 
 ## Terminology
 
-- [terminology.md](terminology.md) is the source of truth in both directions. Before translating, load it; while translating, every term it lists MUST be rendered exactly as it specifies, including its first-occurrence annotations (e.g. `agent（智能体）` on first mention, plain `agent` after) and its "不要译作" prohibitions. When the Chinese side is authored first, the English counterpart uses the table's English column the same way.
-- A technical term NOT in the table MAY be translated only when a major Chinese-language OSS or vendor doc has an established rendering for it (K8s/Vue/MDN Chinese docs, 微软简中风格指南, big-tech project docs). Cite the precedent in the PR.
-- A term with NO established precedent MUST stay in English in the translation and MUST be listed in the PR description under 「待定术语」(pending terms) with a suggested rendering for the reviewer to decide. MUST NOT invent a Chinese rendering inline — an unprecedented translation creates exactly the ambiguity the terminology table exists to prevent. Decided terms then land in [terminology.md](terminology.md) in the same PR or a follow-up.
+- [terminology.md](terminology.md) is the source of truth in both directions. Before translating, load it; every listed term MUST follow its row and its "不要译作" prohibitions. A Chinese target uses the "中文" column and its "首次出现" annotation; an English target uses the "English" column without adding a Chinese gloss.
+- For a Chinese target, an unlisted technical term MAY use an established rendering from a major Chinese-language OSS or vendor source (K8s/Vue/MDN Chinese docs, 微软简中风格指南, big-tech project docs), cited in the PR. Without such precedent it MUST stay in English and be listed under 「待定术语」(pending terms) with a suggested rendering.
+- For an English target, use the established English technical term. If the source term has no unambiguous established equivalent, preserve it with a short explanatory gloss and list it under pending terms. Neither direction may invent a rendering inline; a decided term enters [terminology.md](terminology.md) in the same PR or a follow-up.
 
 ## Typography
 
@@ -44,6 +44,7 @@ These rules govern the Chinese side; the English side follows the repo's normal
 
 - MUST put one half-width space between Chinese text and Latin words, and between Chinese text and numerals: `每个 plugin 注册 3 个 tool`。No space between a full-width punctuation mark and anything.
 - MUST use full-width (Chinese) punctuation in Chinese prose: `，。：；？！（）「」`. Half-width punctuation stays inside code spans, inside complete English sentences quoted as-is, and in numbers (`3.5`, `1,024`).
+- Chinese prose *SHOULD* prefer colons, periods, commas, or parentheses over em dashes. Keep an em dash only when no other punctuation preserves the sentence naturally.
 - Enumeration commas: a Chinese list of parallel items uses 顿号（、）, not commas.
 - MUST NOT use full-width digits or full-width Latin letters — `１２３` never, `123` always.
 - Proper nouns keep their canonical casing: GitHub, TypeScript, DeepSeek — never `github`/`Github` unless quoting code.
@@ -54,7 +55,7 @@ These rules govern the Chinese side; the English side follows the repo's normal
 
 - A pair is done when a bilingual engineer reading either file alone gets everything a reader of the other gets — same facts, same caveats, same tone — and nothing extra.
 - Before handing off, self-check the result against this file and re-read the counterpart ALONE, without the source side by side; awkward phrasing is easier to hear without the source anchoring you.
-- The mechanical contract (consistency record, switcher, structure, wrap, links) is checked by `pnpm run verify-translation-pairing` and the rest of `doc-sync` — run them; do not hand-verify what a gate covers.
+- Run `pnpm run verify-translation-pairing` and the rest of `doc-sync` for records, switchers, heading depths, code blocks, table row and column counts, list kinds, ordered-list starts, list item counts, links, and repository Markdown rules. Manually verify list and table order, noncanonical list numbering, inline code, emphasis, meaning, terminology, and tone.
 
 ## References
 
diff --git a/docs/i18n/translation-rules.zh.md b/docs/i18n/translation-rules.zh.md
index 1d8dcb04f1..04dd0a704e 100644
--- a/docs/i18n/translation-rules.zh.md
+++ b/docs/i18n/translation-rules.zh.md
@@ -12,21 +12,21 @@
 
 ## 行文
 
-- 语体以 [style-samples.md](style-samples.md) 为校准锚点——人工定稿的金标样例按文体各一组，译文必须对齐最接近的样例语体；样例与本文条款冲突时，以样例为准。目标语体是规范的技术制度文：完整主谓、确定语气，不口语化也不学术腔。
+- 语体以 [style-samples.md](style-samples.md) 为校准锚点。人工定稿的金标样例按文体各一组，译文必须参照文体最接近的样例，采用其中目标语言一侧的语体；如果样例与本文的行文规则冲突，以样例为准。译成中文时，采用规范的技术制度文；译成英文时，采用简洁、专业的开发者文档语体。
 - 以母语技术作者的身份重述内容，而不是逐句转写的译者。写完后逐句对照原文核验：不添加、不遗漏——流畅永远不是丢掉语义成分的理由。
-- 给句子补显式执行主体：英文的被动句和抽象主语，中文写成「系统、门禁、评审人」等实际执行者做主语。
-- 优先使用中文工程惯用语而非直译（false positive/negative→误报／漏检、enforcement frontier→执行红线）；隐喻做本地化替换而不是移植，英文名词链展开为动词句。
+- 如果直译会让执行主体含糊，请明确写出实际执行者；译成中文时，应由「系统、门禁、评审人」等实际执行者作主语，避免含糊的被动句或抽象主语。
+- 优先采用目标语言中通行的工程表达，避免生硬直译（false positive/negative→误报／漏检、enforcement frontier→执行红线）；隐喻应自然改写，名词链则按目标语言的习惯拆开。
 - 长段按语义单元拆分，一段一件事。段落边界可以与原文不同；结构签名不比对段落数。
-- 类别名词说中文并在首现括注英文（实操手册（cookbook））；指目录或文件本身时保留代码体英文。
+- 翻译为中文时，类别名词使用中文并在首现括注英文（实操手册（cookbook））；翻译为英文时，使用通行的英文类别名。指目录或文件本身时保留代码体英文。
 
 ## 结构保持
 
-结构一致性由配对门禁负责校验，译者无需为保持结构而牺牲行文的流畅：在既定框架内自然书写即可。配对的两个文件必须在以下方面一一对应：
+配对门禁会检查标题深度、围栏代码块、表格行列数、列表类型、有序列表起始编号、列表项数量与链接目标；门禁未覆盖的结构仍需人工核对。两个配对文件必须在以下方面一一对应：
 
 - 标题层级（相同级别、相同顺序；标题的**文字**要翻译）；
 - 列表形态与编号；
 - 表格（相同的列、相同的行序；表头单元格按术语表翻译）；
-- 围栏代码块：**逐字节一致，包括注释**。代码属于受验证的范围（` ```ts ` 块要通过 `doc-typecheck` 编译），而被改动的注释是代码块计数门禁看不见的漂移；
+- 围栏代码块：**逐字节一致，包括注释**。配对签名比对信息字符串与内容，` ```ts ` 块还要通过 `doc-typecheck` 编译；
 - 行内代码（命令、flag、配置键、文件路径、事件名、API 名、版本号）：原样保留，从不翻译或重排；
 - 链接与锚点：每个相对链接在两个文件中必须指向相同的目标（按约定是 `.md` 路径而非 `.zh.md` 兄弟文件），这样即使某对文档先于相邻文件落地，链接也不会悬空。唯一的 zh 特有链接是语言切换行。链接**文字**翻译；链接目标不翻。
 
@@ -34,18 +34,19 @@
 
 ## 术语
 
-- [terminology.md](terminology.md) 是双向的术语真源。翻译前请先加载它；翻译过程中，表内的每个术语都*必须*严格按表中规定的译法呈现，包括首次出现的括注（如首现写 `agent（智能体）`，之后写 `agent`）与「不要译作」的禁项。中文先行撰写时，英文对侧同样按表中英文列使用术语。
-- 表中**没有**的技术术语，只有当某个主要中文 OSS 或厂商文档已有成型译法时（K8s／Vue／MDN 中文文档、微软简中风格指南、大厂项目文档）才*可以*翻译。请在 PR 中注明先例出处。
-- **没有**成型先例的术语，译文中*必须*保留英文，并且*必须*在 PR 描述的「待定术语」下列出，附上建议译法交评审者定夺。*禁止*就地发明中文译法，因为无先例的翻译恰恰会制造术语表要防止的歧义。确定下来的术语随后在同一个 PR 或后续 PR 中进入 [terminology.md](terminology.md)。
+- [terminology.md](terminology.md) 是双向的术语真源。翻译前请先加载它；表内术语必须遵守对应行与「不要译作」禁项。译成中文时，采用「中文」列，并按「首次出现」列括注；译成英文时，采用「English」列，不加中文括注。
+- 译成中文时，术语表未收录的技术术语只有在主流中文 OSS 文档或厂商资料中已有通行译法时才可以翻译（K8s／Vue／MDN 中文文档、微软简中风格指南、大厂项目文档），并须在 PR 中注明出处；否则必须保留英文，并在 PR 描述的「待定术语」中给出建议译法。
+- 译成英文时，采用通行的英文技术术语。如果源术语没有明确的通行对应词，则保留原词、附上简短说明，并列入「待定术语」。两个方向都不得自行创造译法；确定后的术语须在同一个 PR 或后续 PR 中加入 [terminology.md](terminology.md)。
 
 ## 排版
 
 本节规则约束中文一侧；英文一侧遵循仓库常规的 Markdown 约定（根 `AGENTS.md`）。以下中西文混排规则遵循 [MDN 简体中文翻译指南](https://github.com/mdn/translated-content/blob/main/docs/zh-cn/translation-guide.md)、[Kubernetes 中文本地化指南](https://kubernetes.io/zh-cn/docs/contribute/localization_zh/)、[Vue.js 中文翻译须知](https://github.com/vuejs-translations/docs-zh-cn/wiki/%E7%BF%BB%E8%AF%91%E9%A1%BB%E7%9F%A5) 与[中文文案排版指北](https://github.com/sparanoid/chinese-copywriting-guidelines)的跨项目共识，其根据是 [W3C clreq](https://www.w3.org/TR/clreq/) 与 GB/T 15834—2011：
 
-- *必须*在中文与拉丁词之间、中文与数字之间各留一个半角空格：`每个 plugin 注册 3 个 tool`。全角标点与任何字符之间不加空格。
-- 中文行文*必须*使用全角（中文）标点：`，。：；？！（）「」`。半角标点保留在代码内、按原样引用的完整英文句子内、以及数字内（`3.5`、`1,024`）。
+- 必须在中文与拉丁词之间、中文与数字之间各留一个半角空格：`每个 plugin 注册 3 个 tool`。全角标点与任何字符之间不加空格。
+- 中文行文必须使用全角（中文）标点：`，。：；？！（）「」`。半角标点保留在代码内、按原样引用的完整英文句子内、以及数字内（`3.5`、`1,024`）。
+- 中文行文*应当*优先使用冒号、句号、逗号或括号，尽量不用破折号；只有其他标点都无法自然表达时才保留破折号。
 - 顿号：中文的并列项之间使用顿号（、），而非逗号。
-- *禁止*使用全角数字或全角拉丁字母：永远不写 `１２３`，永远写 `123`。
+- 禁止使用全角数字或全角拉丁字母：永远不写 `１２３`，永远写 `123`。
 - 专有名词保持规范大小写：GitHub、TypeScript、DeepSeek。除非引用代码，否则绝不写 `github`／`Github`。
 - 第二人称用「你」，不用「您」（与 Vue、Kubernetes 中文约定及本仓库的直接语气一致）。
 - 强调标记（`**加粗**`、`*斜体*`）落在与对侧相同的文字段上。中文没有斜体，渲染效果可能看不出差别，不要用引号或其他装饰替代。
@@ -54,7 +55,7 @@
 
 - 一对文档的完成标准：一位双语工程师只读其中任一文件，能获得与另一文件读者完全相同的信息（相同的事实、相同的告诫、相同的语气），并且没有任何多余的内容。
 - 交付前，请对照本文自查一遍，并**单独通读对侧文件**，不与源侧对照；不对照原文时，更容易察觉别扭的表达。
-- 机械契约（一致性记录、切换行、结构、折行、链接）由 `pnpm run verify-translation-pairing` 和 `doc-sync` 的其余门禁检查。请运行门禁；门禁已覆盖的内容无需手工核对。
+- 请运行 `pnpm run verify-translation-pairing` 与 `doc-sync` 的其余门禁。这些门禁会检查一致性记录、切换行、标题深度、代码块、表格行列数、列表类型、有序列表起始编号、列表项数量、链接及仓库 Markdown 规则；列表与表格的顺序、非常规列表编号、行内代码、强调标记、语义、术语和语体仍需人工核对。
 
 ## 参考资料
 
diff --git a/docs/module-graph.md b/docs/module-graph.md
index 5d5ef1addf..8b67c73d27 100644
--- a/docs/module-graph.md
+++ b/docs/module-graph.md
@@ -9,12 +9,16 @@ Inter-package dependencies among the `@deepseek-ai/dsh-*` harness packages, deri
 flowchart TD
   subgraph group_util["packages/util"]
     pkg_brand["brand"]
+    pkg_home["home"]
+    pkg_paths["paths"]
+    pkg_retention["retention"]
     pkg_timeout["timeout"]
   end
   subgraph group_llm["packages/llm"]
     pkg_llm["llm"]
     pkg_llm_deepseek["llm-deepseek"]
     pkg_llm_pi_ai["llm-pi-ai"]
+    pkg_token_meter["token-meter"]
   end
   subgraph group_core["packages/core"]
     pkg_agent["agent"]
@@ -35,6 +39,7 @@ flowchart TD
     pkg_fs_local["fs-local"]
     pkg_fs_policy["fs-policy"]
     pkg_tool_fs["tool-fs"]
+    pkg_tool_fs_search["tool-fs-search"]
   end
   subgraph group_skill["packages/skill"]
     pkg_skill["skill"]
@@ -62,6 +67,11 @@ flowchart TD
     pkg_web_search_exa["web-search-exa"]
     pkg_web_search_perplexity["web-search-perplexity"]
   end
+  subgraph group_spill["packages/spill"]
+    pkg_spill["spill"]
+    pkg_spill_local["spill-local"]
+    pkg_spill_policy["spill-policy"]
+  end
   subgraph group_timeout["packages/timeout"]
     pkg_timeout_policy["timeout-policy"]
   end
@@ -86,6 +96,7 @@ flowchart TD
   end
   subgraph group_support["packages/support"]
     pkg_acp_snapshot["acp-snapshot"]
+    pkg_agent_loop_testkit["agent-loop-testkit"]
     pkg_invariants["invariants"]
     pkg_llm_replay["llm-replay"]
     pkg_loader_smoke["loader-smoke"]
@@ -107,6 +118,7 @@ flowchart TD
   end
   subgraph group_context["packages/context"]
     pkg_time_context["time-context"]
+    pkg_workspace_context["workspace-context"]
   end
   subgraph group_examples["packages/examples"]
     pkg_acp_demo["acp-demo"]
@@ -124,6 +136,14 @@ flowchart TD
     pkg_sandbox["sandbox"]
     pkg_sandbox_local["sandbox-local"]
   end
+  subgraph group_sdk["packages/sdk"]
+    pkg_helper["helper"]
+    pkg_scripts["scripts"]
+  end
+  subgraph group_tasks["packages/tasks"]
+    pkg_tasks["tasks"]
+    pkg_tool_tasks["tool-tasks"]
+  end
   subgraph group_workflow["packages/workflow"]
     pkg_tool_workflow["tool-workflow"]
     pkg_workflow["workflow"]
@@ -131,6 +151,8 @@ flowchart TD
   end
   pkg_llm --> pkg_brand
   pkg_code_runtime_worker --> pkg_code_runtime
+  pkg_helper --> pkg_brand
+  pkg_scripts --> pkg_app_boot
   pkg_llm_deepseek --> pkg_llm
   pkg_llm_pi_ai --> pkg_llm
   pkg_session --> pkg_brand
@@ -142,17 +164,19 @@ flowchart TD
   pkg_fs --> pkg_llm
   pkg_web --> pkg_llm
   pkg_sandbox --> pkg_llm
+  pkg_token_meter --> pkg_llm
+  pkg_token_meter --> pkg_session
   pkg_agent --> pkg_brand
   pkg_agent --> pkg_llm
   pkg_agent --> pkg_scope
   pkg_agent --> pkg_session
   pkg_agent --> pkg_system_prompt
-  pkg_bash --> pkg_brand
   pkg_bash --> pkg_sandbox
   pkg_bash --> pkg_session
   pkg_fs_local --> pkg_fs
   pkg_fs_policy --> pkg_fs
   pkg_skill_local --> pkg_fs
+  pkg_skill_local --> pkg_home
   pkg_skill_local --> pkg_skill
   pkg_compact --> pkg_llm
   pkg_compact --> pkg_session
@@ -161,6 +185,9 @@ flowchart TD
   pkg_web_search_deepseek --> pkg_web
   pkg_web_search_exa --> pkg_web
   pkg_web_search_perplexity --> pkg_web
+  pkg_spill --> pkg_brand
+  pkg_spill --> pkg_llm
+  pkg_spill --> pkg_session
   pkg_session_persistence --> pkg_session
   pkg_llm_replay --> pkg_llm
   pkg_llm_replay --> pkg_session
@@ -172,6 +199,8 @@ flowchart TD
   pkg_compact_basic --> pkg_compact
   pkg_compact_basic --> pkg_llm
   pkg_compact_basic --> pkg_session
+  pkg_compact_basic --> pkg_token_meter
+  pkg_spill_local --> pkg_spill
   pkg_hook_protocol --> pkg_bash
   pkg_hook_protocol --> pkg_session
   pkg_session_persistence_jsonl --> pkg_session
@@ -194,7 +223,10 @@ flowchart TD
   pkg_user_interaction --> pkg_agent
   pkg_user_interaction --> pkg_llm
   pkg_time_context --> pkg_agent
-  pkg_time_context --> pkg_system_prompt
+  pkg_tasks --> pkg_agent
+  pkg_tasks --> pkg_brand
+  pkg_tasks --> pkg_session
+  pkg_tasks --> pkg_timeout
   pkg_workflow --> pkg_agent
   pkg_workflow --> pkg_brand
   pkg_workflow --> pkg_llm
@@ -225,9 +257,12 @@ flowchart TD
   pkg_agent_loop --> pkg_tools
   pkg_tool_bash --> pkg_agent
   pkg_tool_bash --> pkg_bash
+  pkg_tool_bash --> pkg_home
   pkg_tool_bash --> pkg_llm
   pkg_tool_bash --> pkg_sandbox
+  pkg_tool_bash --> pkg_session_persistence
   pkg_tool_bash --> pkg_system_prompt
+  pkg_tool_bash --> pkg_tasks
   pkg_tool_bash --> pkg_tools
   pkg_tool_bash --> pkg_user_approval
   pkg_tool_fs --> pkg_fs
@@ -235,6 +270,13 @@ flowchart TD
   pkg_tool_fs --> pkg_session
   pkg_tool_fs --> pkg_system_prompt
   pkg_tool_fs --> pkg_tools
+  pkg_tool_fs_search --> pkg_bash
+  pkg_tool_fs_search --> pkg_llm
+  pkg_tool_fs_search --> pkg_retention
+  pkg_tool_fs_search --> pkg_session
+  pkg_tool_fs_search --> pkg_spill
+  pkg_tool_fs_search --> pkg_system_prompt
+  pkg_tool_fs_search --> pkg_tools
   pkg_tool_skill --> pkg_agent
   pkg_tool_skill --> pkg_llm
   pkg_tool_skill --> pkg_skill
@@ -247,6 +289,11 @@ flowchart TD
   pkg_tool_web --> pkg_system_prompt
   pkg_tool_web --> pkg_tools
   pkg_tool_web --> pkg_web
+  pkg_spill_policy --> pkg_llm
+  pkg_spill_policy --> pkg_retention
+  pkg_spill_policy --> pkg_session
+  pkg_spill_policy --> pkg_spill
+  pkg_spill_policy --> pkg_tools
   pkg_timeout_policy --> pkg_llm
   pkg_timeout_policy --> pkg_timeout
   pkg_timeout_policy --> pkg_tools
@@ -259,7 +306,13 @@ flowchart TD
   pkg_hooks_codex --> pkg_hook_protocol
   pkg_hooks_codex --> pkg_llm
   pkg_hooks_codex --> pkg_session
+  pkg_hooks_codex --> pkg_session_persistence
   pkg_hooks_codex --> pkg_tools
+  pkg_agent_loop_testkit --> pkg_agent
+  pkg_agent_loop_testkit --> pkg_llm
+  pkg_agent_loop_testkit --> pkg_session
+  pkg_agent_loop_testkit --> pkg_system_prompt
+  pkg_agent_loop_testkit --> pkg_tools
   pkg_acp --> pkg_agent
   pkg_acp --> pkg_bash
   pkg_acp --> pkg_llm
@@ -267,16 +320,27 @@ flowchart TD
   pkg_acp --> pkg_sandbox
   pkg_acp --> pkg_session
   pkg_acp --> pkg_session_persistence
+  pkg_acp --> pkg_system_prompt
   pkg_acp --> pkg_tools
   pkg_acp --> pkg_user_approval
   pkg_acp --> pkg_user_interaction
   pkg_tool_ask_user --> pkg_agent
   pkg_tool_ask_user --> pkg_tools
   pkg_tool_ask_user --> pkg_user_interaction
+  pkg_workspace_context --> pkg_agent
+  pkg_workspace_context --> pkg_fs
+  pkg_workspace_context --> pkg_llm
+  pkg_workspace_context --> pkg_paths
+  pkg_workspace_context --> pkg_session
+  pkg_workspace_context --> pkg_tools
   pkg_repeat_tool_guard --> pkg_agent
   pkg_repeat_tool_guard --> pkg_tools
   pkg_mcp_client --> pkg_llm
   pkg_mcp_client --> pkg_tools
+  pkg_tool_tasks --> pkg_agent
+  pkg_tool_tasks --> pkg_system_prompt
+  pkg_tool_tasks --> pkg_tasks
+  pkg_tool_tasks --> pkg_tools
   pkg_tool_workflow --> pkg_agent
   pkg_tool_workflow --> pkg_llm
   pkg_tool_workflow --> pkg_system_prompt
@@ -295,11 +359,13 @@ flowchart TD
   pkg_tool_subagent --> pkg_agent
   pkg_tool_subagent --> pkg_llm
   pkg_tool_subagent --> pkg_subagent
+  pkg_tool_subagent --> pkg_tasks
   pkg_tool_subagent --> pkg_tools
   pkg_hooks_claude --> pkg_agent
   pkg_hooks_claude --> pkg_hook_protocol
   pkg_hooks_claude --> pkg_llm
   pkg_hooks_claude --> pkg_session
+  pkg_hooks_claude --> pkg_session_persistence
   pkg_hooks_claude --> pkg_subagent
   pkg_hooks_claude --> pkg_tools
   pkg_subagent_mock --> pkg_agent
@@ -312,15 +378,19 @@ flowchart TD
   pkg_jsonrpc --> pkg_subagent
   pkg_agent_spine_demo --> pkg_agent
   pkg_agent_spine_demo --> pkg_agent_loop
+  pkg_agent_spine_demo --> pkg_home
   pkg_agent_spine_demo --> pkg_invariants
   pkg_agent_spine_demo --> pkg_llm
   pkg_agent_spine_demo --> pkg_session
   pkg_agent_spine_demo --> pkg_skill
   pkg_agent_spine_demo --> pkg_skill_local
   pkg_agent_spine_demo --> pkg_system_prompt
+  pkg_agent_spine_demo --> pkg_tasks
   pkg_agent_spine_demo --> pkg_tool_bash
   pkg_agent_spine_demo --> pkg_tool_skill
+  pkg_agent_spine_demo --> pkg_tool_tasks
   pkg_agent_spine_demo --> pkg_tools
+  pkg_agent_spine_demo --> pkg_workspace_context
   pkg_workflow_workerthread --> pkg_agent
   pkg_workflow_workerthread --> pkg_brand
   pkg_workflow_workerthread --> pkg_llm
@@ -340,6 +410,7 @@ flowchart TD
   pkg_acp_demo --> pkg_session_persistence_jsonl
   pkg_acp_demo --> pkg_tools
   pkg_acp_demo --> pkg_user_interaction
+  pkg_acp_demo --> pkg_workspace_context
   pkg_stdio_demo --> pkg_agent
   pkg_stdio_demo --> pkg_agent_spine_demo
   pkg_stdio_demo --> pkg_app_boot
@@ -350,11 +421,15 @@ flowchart TD
   pkg_stdio_demo --> pkg_tool_ask_user
   pkg_stdio_demo --> pkg_tools
   pkg_stdio_demo --> pkg_user_interaction
+  pkg_stdio_demo --> pkg_workspace_context
 ```
 
 | Package | Group | Depends on |
 | --- | --- | --- |
 | [`brand`](../packages/util/brand) | `util` | — |
+| [`home`](../packages/util/home) | `util` | — |
+| [`paths`](../packages/util/paths) | `util` | — |
+| [`retention`](../packages/util/retention) | `util` | — |
 | [`timeout`](../packages/util/timeout) | `util` | — |
 | [`scope`](../packages/core/scope) | `core` | — |
 | [`skill`](../packages/skill/skill) | `skill` | — |
@@ -366,6 +441,8 @@ flowchart TD
 | [`jsonrpc-demo`](../packages/examples/jsonrpc-demo) | `examples` | — |
 | [`llm`](../packages/llm/llm) | `llm` | [`brand`](../packages/util/brand) |
 | [`code-runtime-worker`](../packages/code-runtime/code-runtime-worker) | `code-runtime` | [`code-runtime`](../packages/code-runtime/code-runtime) |
+| [`helper`](../packages/sdk/helper) | `sdk` | [`brand`](../packages/util/brand) |
+| [`scripts`](../packages/sdk/scripts) | `sdk` | [`app-boot`](../packages/ui/app-boot) |
 | [`llm-deepseek`](../packages/llm/llm-deepseek) | `llm` | [`llm`](../packages/llm/llm) |
 | [`llm-pi-ai`](../packages/llm/llm-pi-ai) | `llm` | [`llm`](../packages/llm/llm) |
 | [`session`](../packages/core/session) | `core` | [`brand`](../packages/util/brand), [`llm`](../packages/llm/llm), [`scope`](../packages/core/scope) |
@@ -373,21 +450,24 @@ flowchart TD
 | [`fs`](../packages/fs/fs) | `fs` | [`brand`](../packages/util/brand), [`llm`](../packages/llm/llm) |
 | [`web`](../packages/web/web) | `web` | [`llm`](../packages/llm/llm) |
 | [`sandbox`](../packages/sandbox/sandbox) | `sandbox` | [`llm`](../packages/llm/llm) |
+| [`token-meter`](../packages/llm/token-meter) | `llm` | [`llm`](../packages/llm/llm), [`session`](../packages/core/session) |
 | [`agent`](../packages/core/agent) | `core` | [`brand`](../packages/util/brand), [`llm`](../packages/llm/llm), [`scope`](../packages/core/scope), [`session`](../packages/core/session), [`system-prompt`](../packages/core/system-prompt) |
-| [`bash`](../packages/bash/bash) | `bash` | [`brand`](../packages/util/brand), [`sandbox`](../packages/sandbox/sandbox), [`session`](../packages/core/session) |
+| [`bash`](../packages/bash/bash) | `bash` | [`sandbox`](../packages/sandbox/sandbox), [`session`](../packages/core/session) |
 | [`fs-local`](../packages/fs/fs-local) | `fs` | [`fs`](../packages/fs/fs) |
 | [`fs-policy`](../packages/fs/fs-policy) | `fs` | [`fs`](../packages/fs/fs) |
-| [`skill-local`](../packages/skill/skill-local) | `skill` | [`fs`](../packages/fs/fs), [`skill`](../packages/skill/skill) |
+| [`skill-local`](../packages/skill/skill-local) | `skill` | [`fs`](../packages/fs/fs), [`home`](../packages/util/home), [`skill`](../packages/skill/skill) |
 | [`compact`](../packages/compact/compact) | `compact` | [`llm`](../packages/llm/llm), [`session`](../packages/core/session) |
 | [`web-fetch-local`](../packages/web/web-fetch-local) | `web` | [`timeout`](../packages/util/timeout), [`web`](../packages/web/web) |
 | [`web-search-deepseek`](../packages/web/web-search-deepseek) | `web` | [`web`](../packages/web/web) |
 | [`web-search-exa`](../packages/web/web-search-exa) | `web` | [`web`](../packages/web/web) |
 | [`web-search-perplexity`](../packages/web/web-search-perplexity) | `web` | [`web`](../packages/web/web) |
+| [`spill`](../packages/spill/spill) | `spill` | [`brand`](../packages/util/brand), [`llm`](../packages/llm/llm), [`session`](../packages/core/session) |
 | [`session-persistence`](../packages/session-persistence/session-persistence) | `session-persistence` | [`session`](../packages/core/session) |
 | [`llm-replay`](../packages/support/llm-replay) | `support` | [`llm`](../packages/llm/llm), [`session`](../packages/core/session) |
 | [`sandbox-local`](../packages/sandbox/sandbox-local) | `sandbox` | [`llm`](../packages/llm/llm), [`sandbox`](../packages/sandbox/sandbox) |
 | [`bash-local`](../packages/bash/bash-local) | `bash` | [`bash`](../packages/bash/bash), [`timeout`](../packages/util/timeout) |
-| [`compact-basic`](../packages/compact/compact-basic) | `compact` | [`agent`](../packages/core/agent), [`compact`](../packages/compact/compact), [`llm`](../packages/llm/llm), [`session`](../packages/core/session) |
+| [`compact-basic`](../packages/compact/compact-basic) | `compact` | [`agent`](../packages/core/agent), [`compact`](../packages/compact/compact), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`token-meter`](../packages/llm/token-meter) |
+| [`spill-local`](../packages/spill/spill-local) | `spill` | [`spill`](../packages/spill/spill) |
 | [`hook-protocol`](../packages/hooks/hook-protocol) | `hooks` | [`bash`](../packages/bash/bash), [`session`](../packages/core/session) |
 | [`session-persistence-jsonl`](../packages/session-persistence/session-persistence-jsonl) | `session-persistence` | [`session`](../packages/core/session), [`session-persistence`](../packages/session-persistence/session-persistence) |
 | [`session-persistence-sqlite`](../packages/session-persistence/session-persistence-sqlite) | `session-persistence` | [`session`](../packages/core/session), [`session-persistence`](../packages/session-persistence/session-persistence) |
@@ -395,36 +475,42 @@ flowchart TD
 | [`invariants`](../packages/support/invariants) | `support` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`scope`](../packages/core/scope), [`session`](../packages/core/session) |
 | [`user-approval`](../packages/ui/user-approval) | `ui` | [`agent`](../packages/core/agent), [`brand`](../packages/util/brand), [`llm`](../packages/llm/llm), [`scope`](../packages/core/scope), [`session`](../packages/core/session), [`system-prompt`](../packages/core/system-prompt) |
 | [`user-interaction`](../packages/ui/user-interaction) | `ui` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm) |
-| [`time-context`](../packages/context/time-context) | `context` | [`agent`](../packages/core/agent), [`system-prompt`](../packages/core/system-prompt) |
+| [`time-context`](../packages/context/time-context) | `context` | [`agent`](../packages/core/agent) |
+| [`tasks`](../packages/tasks/tasks) | `tasks` | [`agent`](../packages/core/agent), [`brand`](../packages/util/brand), [`session`](../packages/core/session), [`timeout`](../packages/util/timeout) |
 | [`workflow`](../packages/workflow/workflow) | `workflow` | [`agent`](../packages/core/agent), [`brand`](../packages/util/brand), [`llm`](../packages/llm/llm) |
 | [`tools`](../packages/core/tools) | `core` | [`agent`](../packages/core/agent), [`code-runtime`](../packages/code-runtime/code-runtime), [`llm`](../packages/llm/llm), [`scope`](../packages/core/scope), [`session`](../packages/core/session), [`system-prompt`](../packages/core/system-prompt), [`user-approval`](../packages/ui/user-approval) |
 | [`bash-sandbox`](../packages/bash/bash-sandbox) | `bash` | [`bash`](../packages/bash/bash), [`bash-local`](../packages/bash/bash-local), [`sandbox`](../packages/sandbox/sandbox) |
 | [`permission`](../packages/ui/permission) | `ui` | [`bash`](../packages/bash/bash), [`sandbox`](../packages/sandbox/sandbox), [`session`](../packages/core/session), [`user-approval`](../packages/ui/user-approval) |
 | [`stdio`](../packages/ui/stdio) | `ui` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`user-interaction`](../packages/ui/user-interaction) |
 | [`agent-loop`](../packages/core/agent-loop) | `core` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`scope`](../packages/core/scope), [`session`](../packages/core/session), [`session-persistence`](../packages/session-persistence/session-persistence), [`system-prompt`](../packages/core/system-prompt), [`tools`](../packages/core/tools) |
-| [`tool-bash`](../packages/bash/tool-bash) | `bash` | [`agent`](../packages/core/agent), [`bash`](../packages/bash/bash), [`llm`](../packages/llm/llm), [`sandbox`](../packages/sandbox/sandbox), [`system-prompt`](../packages/core/system-prompt), [`tools`](../packages/core/tools), [`user-approval`](../packages/ui/user-approval) |
+| [`tool-bash`](../packages/bash/tool-bash) | `bash` | [`agent`](../packages/core/agent), [`bash`](../packages/bash/bash), [`home`](../packages/util/home), [`llm`](../packages/llm/llm), [`sandbox`](../packages/sandbox/sandbox), [`session-persistence`](../packages/session-persistence/session-persistence), [`system-prompt`](../packages/core/system-prompt), [`tasks`](../packages/tasks/tasks), [`tools`](../packages/core/tools), [`user-approval`](../packages/ui/user-approval) |
 | [`tool-fs`](../packages/fs/tool-fs) | `fs` | [`fs`](../packages/fs/fs), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`system-prompt`](../packages/core/system-prompt), [`tools`](../packages/core/tools) |
+| [`tool-fs-search`](../packages/fs/tool-fs-search) | `fs` | [`bash`](../packages/bash/bash), [`llm`](../packages/llm/llm), [`retention`](../packages/util/retention), [`session`](../packages/core/session), [`spill`](../packages/spill/spill), [`system-prompt`](../packages/core/system-prompt), [`tools`](../packages/core/tools) |
 | [`tool-skill`](../packages/skill/tool-skill) | `skill` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`skill`](../packages/skill/skill), [`tools`](../packages/core/tools) |
 | [`subagent`](../packages/subagent/subagent) | `subagent` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`scope`](../packages/core/scope), [`tools`](../packages/core/tools) |
 | [`tool-web`](../packages/web/tool-web) | `web` | [`llm`](../packages/llm/llm), [`system-prompt`](../packages/core/system-prompt), [`tools`](../packages/core/tools), [`web`](../packages/web/web) |
+| [`spill-policy`](../packages/spill/spill-policy) | `spill` | [`llm`](../packages/llm/llm), [`retention`](../packages/util/retention), [`session`](../packages/core/session), [`spill`](../packages/spill/spill), [`tools`](../packages/core/tools) |
 | [`timeout-policy`](../packages/timeout/timeout-policy) | `timeout` | [`llm`](../packages/llm/llm), [`timeout`](../packages/util/timeout), [`tools`](../packages/core/tools) |
 | [`tool-todo`](../packages/todo/tool-todo) | `todo` | [`agent`](../packages/core/agent), [`session`](../packages/core/session), [`tools`](../packages/core/tools) |
 | [`tool-cordis`](../packages/cordis/tool-cordis) | `cordis` | [`scope`](../packages/core/scope), [`tools`](../packages/core/tools) |
-| [`hooks-codex`](../packages/hooks/hooks-codex) | `hooks` | [`agent`](../packages/core/agent), [`hook-protocol`](../packages/hooks/hook-protocol), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`tools`](../packages/core/tools) |
-| [`acp`](../packages/ui/acp) | `ui` | [`agent`](../packages/core/agent), [`bash`](../packages/bash/bash), [`llm`](../packages/llm/llm), [`permission`](../packages/ui/permission), [`sandbox`](../packages/sandbox/sandbox), [`session`](../packages/core/session), [`session-persistence`](../packages/session-persistence/session-persistence), [`tools`](../packages/core/tools), [`user-approval`](../packages/ui/user-approval), [`user-interaction`](../packages/ui/user-interaction) |
+| [`hooks-codex`](../packages/hooks/hooks-codex) | `hooks` | [`agent`](../packages/core/agent), [`hook-protocol`](../packages/hooks/hook-protocol), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`session-persistence`](../packages/session-persistence/session-persistence), [`tools`](../packages/core/tools) |
+| [`agent-loop-testkit`](../packages/support/agent-loop-testkit) | `support` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`system-prompt`](../packages/core/system-prompt), [`tools`](../packages/core/tools) |
+| [`acp`](../packages/ui/acp) | `ui` | [`agent`](../packages/core/agent), [`bash`](../packages/bash/bash), [`llm`](../packages/llm/llm), [`permission`](../packages/ui/permission), [`sandbox`](../packages/sandbox/sandbox), [`session`](../packages/core/session), [`session-persistence`](../packages/session-persistence/session-persistence), [`system-prompt`](../packages/core/system-prompt), [`tools`](../packages/core/tools), [`user-approval`](../packages/ui/user-approval), [`user-interaction`](../packages/ui/user-interaction) |
 | [`tool-ask-user`](../packages/ui/tool-ask-user) | `ui` | [`agent`](../packages/core/agent), [`tools`](../packages/core/tools), [`user-interaction`](../packages/ui/user-interaction) |
+| [`workspace-context`](../packages/context/workspace-context) | `context` | [`agent`](../packages/core/agent), [`fs`](../packages/fs/fs), [`llm`](../packages/llm/llm), [`paths`](../packages/util/paths), [`session`](../packages/core/session), [`tools`](../packages/core/tools) |
 | [`repeat-tool-guard`](../packages/guard/repeat-tool-guard) | `guard` | [`agent`](../packages/core/agent), [`tools`](../packages/core/tools) |
 | [`mcp-client`](../packages/mcp/mcp-client) | `mcp` | [`llm`](../packages/llm/llm), [`tools`](../packages/core/tools) |
+| [`tool-tasks`](../packages/tasks/tool-tasks) | `tasks` | [`agent`](../packages/core/agent), [`system-prompt`](../packages/core/system-prompt), [`tasks`](../packages/tasks/tasks), [`tools`](../packages/core/tools) |
 | [`tool-workflow`](../packages/workflow/tool-workflow) | `workflow` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`system-prompt`](../packages/core/system-prompt), [`tools`](../packages/core/tools), [`workflow`](../packages/workflow/workflow) |
 | [`subagent-acp`](../packages/subagent/subagent-acp) | `subagent` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`subagent`](../packages/subagent/subagent), [`subagent-subprocess`](../packages/subagent/subagent-subprocess) |
 | [`subagent-inprocess`](../packages/subagent/subagent-inprocess) | `subagent` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`subagent`](../packages/subagent/subagent), [`system-prompt`](../packages/core/system-prompt), [`tools`](../packages/core/tools) |
-| [`tool-subagent`](../packages/subagent/tool-subagent) | `subagent` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`subagent`](../packages/subagent/subagent), [`tools`](../packages/core/tools) |
-| [`hooks-claude`](../packages/hooks/hooks-claude) | `hooks` | [`agent`](../packages/core/agent), [`hook-protocol`](../packages/hooks/hook-protocol), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`subagent`](../packages/subagent/subagent), [`tools`](../packages/core/tools) |
+| [`tool-subagent`](../packages/subagent/tool-subagent) | `subagent` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`subagent`](../packages/subagent/subagent), [`tasks`](../packages/tasks/tasks), [`tools`](../packages/core/tools) |
+| [`hooks-claude`](../packages/hooks/hooks-claude) | `hooks` | [`agent`](../packages/core/agent), [`hook-protocol`](../packages/hooks/hook-protocol), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`session-persistence`](../packages/session-persistence/session-persistence), [`subagent`](../packages/subagent/subagent), [`tools`](../packages/core/tools) |
 | [`subagent-mock`](../packages/support/subagent-mock) | `support` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`subagent`](../packages/subagent/subagent) |
 | [`jsonrpc`](../packages/ui/jsonrpc) | `ui` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`llm-deepseek`](../packages/llm/llm-deepseek), [`session`](../packages/core/session), [`subagent`](../packages/subagent/subagent) |
-| [`agent-spine-demo`](../packages/examples/agent-spine-demo) | `examples` | [`agent`](../packages/core/agent), [`agent-loop`](../packages/core/agent-loop), [`invariants`](../packages/support/invariants), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`skill`](../packages/skill/skill), [`skill-local`](../packages/skill/skill-local), [`system-prompt`](../packages/core/system-prompt), [`tool-bash`](../packages/bash/tool-bash), [`tool-skill`](../packages/skill/tool-skill), [`tools`](../packages/core/tools) |
+| [`agent-spine-demo`](../packages/examples/agent-spine-demo) | `examples` | [`agent`](../packages/core/agent), [`agent-loop`](../packages/core/agent-loop), [`home`](../packages/util/home), [`invariants`](../packages/support/invariants), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`skill`](../packages/skill/skill), [`skill-local`](../packages/skill/skill-local), [`system-prompt`](../packages/core/system-prompt), [`tasks`](../packages/tasks/tasks), [`tool-bash`](../packages/bash/tool-bash), [`tool-skill`](../packages/skill/tool-skill), [`tool-tasks`](../packages/tasks/tool-tasks), [`tools`](../packages/core/tools), [`workspace-context`](../packages/context/workspace-context) |
 | [`workflow-workerthread`](../packages/workflow/workflow-workerthread) | `workflow` | [`agent`](../packages/core/agent), [`brand`](../packages/util/brand), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`subagent`](../packages/subagent/subagent), [`tools`](../packages/core/tools), [`workflow`](../packages/workflow/workflow) |
 | [`subagent-fork`](../packages/subagent/subagent-fork) | `subagent` | [`agent`](../packages/core/agent), [`session`](../packages/core/session), [`subagent`](../packages/subagent/subagent), [`subagent-inprocess`](../packages/subagent/subagent-inprocess) |
 | [`subagent-spawn`](../packages/subagent/subagent-spawn) | `subagent` | [`subagent`](../packages/subagent/subagent), [`subagent-inprocess`](../packages/subagent/subagent-inprocess) |
-| [`acp-demo`](../packages/examples/acp-demo) | `examples` | [`acp`](../packages/ui/acp), [`agent-spine-demo`](../packages/examples/agent-spine-demo), [`app-boot`](../packages/ui/app-boot), [`session-persistence-jsonl`](../packages/session-persistence/session-persistence-jsonl), [`tools`](../packages/core/tools), [`user-interaction`](../packages/ui/user-interaction) |
-| [`stdio-demo`](../packages/examples/stdio-demo) | `examples` | [`agent`](../packages/core/agent), [`agent-spine-demo`](../packages/examples/agent-spine-demo), [`app-boot`](../packages/ui/app-boot), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`session-persistence-jsonl`](../packages/session-persistence/session-persistence-jsonl), [`stdio`](../packages/ui/stdio), [`tool-ask-user`](../packages/ui/tool-ask-user), [`tools`](../packages/core/tools), [`user-interaction`](../packages/ui/user-interaction) |
+| [`acp-demo`](../packages/examples/acp-demo) | `examples` | [`acp`](../packages/ui/acp), [`agent-spine-demo`](../packages/examples/agent-spine-demo), [`app-boot`](../packages/ui/app-boot), [`session-persistence-jsonl`](../packages/session-persistence/session-persistence-jsonl), [`tools`](../packages/core/tools), [`user-interaction`](../packages/ui/user-interaction), [`workspace-context`](../packages/context/workspace-context) |
+| [`stdio-demo`](../packages/examples/stdio-demo) | `examples` | [`agent`](../packages/core/agent), [`agent-spine-demo`](../packages/examples/agent-spine-demo), [`app-boot`](../packages/ui/app-boot), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`session-persistence-jsonl`](../packages/session-persistence/session-persistence-jsonl), [`stdio`](../packages/ui/stdio), [`tool-ask-user`](../packages/ui/tool-ask-user), [`tools`](../packages/core/tools), [`user-interaction`](../packages/ui/user-interaction), [`workspace-context`](../packages/context/workspace-context) |
diff --git a/docs/persistence-catalog.md b/docs/persistence-catalog.md
index a27b19b9a7..3bc15253a5 100644
--- a/docs/persistence-catalog.md
+++ b/docs/persistence-catalog.md
@@ -1,13 +1,85 @@
 <!-- Generated by scripts/gen-persistence-catalog.ts — do not edit by hand.
      Run `pnpm run gen-persistence-catalog` to regenerate. -->
 
-# Persistence Log Event Catalog
+# Session Persistence Event Catalog
 
-Every event type that can appear in a session's durable event log: each member of the merge-extensible `SessionEventMap` — the owning vocabulary in `@deepseek-ai/dsh-session` plus every plugin declaration merge in this repo — with the payload it carries, its surface badge, and the declaration it comes from. It complements [session.md](core-data-structures/session.md) (the `SessionEvent` envelope, surface list, and `deriveMessages()` projection), [persistence.md](core-data-structures/persistence.md) (how the log is made durable), and the [cordis events catalog](cordis-catalog/events.md) (the live bus wiring — a log event is NOT a cordis event; it reaches listeners via the single `session/event` emit).
+Every event type that can appear in a session's durable event log: the complete persisted `SessionEvent` envelope and each member of the merge-extensible `SessionEventMap` — the owning vocabulary in `@deepseek-ai/dsh-session` plus every plugin declaration merge in this repo — with source JSDoc, full payload declaration, surface badge, and declaration site. It complements [session.md](core-data-structures/session.md) (surface ordering and the `deriveMessages()` projection), [persistence.md](core-data-structures/persistence.md) (how the log is made durable), and the [cordis events catalog](cordis-catalog/events.md) (the live bus wiring — a log event is NOT a cordis event; it reaches listeners via the single `session/event` emit).
 
-This file is GENERATED from source (`scripts/gen-persistence-catalog.ts`) and verified fresh by `pnpm run verify-persistence-catalog` (part of `doc-sync`) — do not edit it by hand. Payload blocks use a `ts persistence-catalog` fence (skipped by doc-typecheck, since a bare payload fragment is not standalone-compilable). Type names in a payload link to the page that documents them. See [the persistence-log-catalog RFC](rfc/implemented/process/2026-07-04-persistence-log-catalog.md).
+This file is GENERATED from source (`scripts/gen-persistence-catalog.ts`) and verified fresh by `pnpm run verify-persistence-catalog` (part of `doc-sync`) — do not edit it by hand. Declaration blocks retain the source declaration and nested property JSDoc, removing only the indentation imposed by a containing interface/module, and use a `ts persistence-catalog` fence (skipped by doc-typecheck because declarations reference types from their owning modules). Type names in a payload link to the page that documents them. See [the persistence-log-catalog RFC](rfc/implemented/process/2026-07-04-persistence-log-catalog.md).
 
-The on-disk envelope around every payload is `SessionEvent` — `type`, monotonic `seq`, epoch-ms `time`, the `data` documented here, plus `surfaceOp`/`sourceEventSeqs` on **surface** events only ([envelope](core-data-structures/session.md#sessioneventt--one-log-entry)). **surface** marks a `SurfaceEventType` member: it produces an LLM message and declares how it joins the surface list. **log-only** marks everything else: durable, replayable record with no derived-history contribution. Every payload is JSON-serializable (enforced at `Session.append`), and the whole format is pinned at `SESSION_FORMAT_VERSION = 0` — pre-release, no compatibility implied ([the version stance](core-data-structures/persistence.md)). Scope: the packages in this repo; a downstream plugin can merge further event types, which are outside this catalog by construction.
+The envelope declarations below compose each event's `type`, monotonic `seq`, epoch-ms `time`, `data`, and the conditional `surfaceOp`/`sourceEventSeqs` fields. **surface** marks a `SurfaceEventType` member: it produces an LLM message and declares how it joins the surface list. **log-only** marks everything else: a durable, replayable record with no derived-history contribution. Every payload is JSON-serializable (enforced at `Session.append`), and the whole format is pinned at `SESSION_FORMAT_VERSION = 0` — pre-release, no compatibility implied ([the version stance](core-data-structures/persistence.md)). Scope: the packages in this repo; a downstream plugin can merge further event types, which are outside this catalog by construction.
+
+## Event envelope
+
+```ts persistence-catalog
+/** The appendable event-type keys of {@link SessionEventMap}, plugin-merged extensions included. */
+export type SessionEventType = keyof SessionEventMap
+
+/**
+ * The subset of {@link SessionEventType} values whose events produce LLM
+ * messages and are eligible to appear on the ordered surface. Only these
+ * event types may carry {@link SurfaceOp} and {@link SessionEvent.sourceEventSeqs}.
+ */
+export type SurfaceEventType =
+  | 'user/message'
+  | 'assistant/message'
+  | 'tool/result'
+  | 'context/message'
+  | 'steering/message'
+
+/**
+ * How a session event entered the ordered surface. Only valid on
+ * {@link SurfaceEventType} events.
+ *
+ * - `'append'`: added to the tail — normal path for user/assistant/tool/context
+ *   messages.
+ * - `{ op: 'replace', start, end }`: replaces surface nodes from `start`
+ *   (inclusive) through `end` (inclusive) with this node. Both must exist as
+ *   surface nodes in the current surface. `start === end` replaces a single
+ *   node. The node's {@link SessionEvent.sourceEventSeqs} must include every
+ *   shadowed surface node. Used by compaction and possible other manipulations.
+ */
+export type SurfaceOp =
+  | 'append'
+  | { op: 'replace'; start: number; end: number }
+
+/**
+ * One immutable entry in the session log.
+ *
+ * A proper discriminated union over `type` (not independent `type`/`data`
+ * unions), so `switch (event.type)` narrows `event.data` without casts.
+ *
+ * The {@link sourceEventSeqs} and {@link surfaceOp} fields are conditional:
+ * they only exist on {@link SurfaceEventType} variants (`user/message`,
+ * `assistant/message`, `tool/result`, `context/message`, `steering/message`).
+ * Non-surface events (boundary markers, chunks, usage, errors) never carry
+ * surface metadata — the compiler enforces this at `Session.append()`
+ * call sites.
+ */
+export type SessionEvent<T extends SessionEventType = SessionEventType> = {
+  [K in SessionEventType]: {
+    type: K
+    /** Monotonic sequence number within the session. */
+    seq: number
+    /** Unix epoch milliseconds. */
+    time: number
+    data: SessionEventMap[K]
+  } & (K extends SurfaceEventType ? {
+    /**
+     * Seq numbers of events that are provenance sources of this event
+     * (e.g. the `assistant/chunk` seqs that built an `assistant/message`,
+     * or the surface nodes shadowed by a compaction replace node). An
+     * `assistant/message` may carry a present empty array for a known empty
+     * provider stream; omission means unrecorded provenance.
+     */
+    sourceEventSeqs?: number[]
+    /** How this event entered the surface; absent for non-surface events. */
+    surfaceOp?: SurfaceOp
+  } : object)
+}[T]
+```
+
+Sources: [`packages/core/session/src/types.ts:255`](../packages/core/session/src/types.ts) · [`packages/core/session/src/types.ts:262`](../packages/core/session/src/types.ts) · [`packages/core/session/src/types.ts:292`](../packages/core/session/src/types.ts) · [`packages/core/session/src/types.ts:324`](../packages/core/session/src/types.ts)
 
 ## Events
 
@@ -15,10 +87,21 @@ The on-disk envelope around every payload is `SessionEvent` — `type`, monotoni
 
 #### `approval/asked` — log-only
 
-An approval question was put to the answerer chain — log-only audit (like `hook/*`; NOT a surface event, carries no `surfaceOp`). `id` pairs it with the `approval/decided` that always follows; `toolName` is the tool the question is about, `callId` the exact tool call when the asker had one, `reason` the asker's human-readable explanation (e.g. a hook's permission-decision reason).
-
 ```ts persistence-catalog
-'approval/asked': { id: ApprovalRequestId; toolName: string; callId?: CallId; reason?: string }
+/**
+ * An approval question was put to the answerer chain — log-only audit
+ * (like `hook/*`; NOT a surface event, carries no `surfaceOp`). `id` pairs
+ * it with the `approval/decided` that always follows; `toolName` is the
+ * tool the question is about, `callId` the exact tool call when the asker
+ * had one, `reason` the asker's human-readable explanation (e.g. a hook's
+ * permission-decision reason).
+ */
+'approval/asked': {
+  id: ApprovalRequestId
+  toolName: string
+  callId?: CallId
+  reason?: string
+}
 ```
 
 Types: [CallId](core-data-structures/core.md)
@@ -27,19 +110,31 @@ Source: [`packages/ui/user-approval/src/index.ts:45`](../packages/ui/user-approv
 
 #### `approval/decided` — log-only
 
-The outcome of a prior `approval/asked` (same `id`) — log-only audit. Exactly one per ask, appended when the outcome is known: a decision, a cancellation, or the fail-closed `'unavailable'`.
-
 ```ts persistence-catalog
-'approval/decided': { id: ApprovalRequestId; outcome: ApprovalOutcome }
+/**
+ * The outcome of a prior `approval/asked` (same `id`) — log-only audit.
+ * Exactly one per ask, appended when the outcome is known: a decision, a
+ * cancellation, or the fail-closed `'unavailable'`.
+ */
+'approval/decided': {
+  id: ApprovalRequestId
+  outcome: ApprovalOutcome
+}
 ```
 
 Source: [`packages/ui/user-approval/src/index.ts:56`](../packages/ui/user-approval/src/index.ts)
 
 #### `approval/policy` — log-only
 
-The session's approval policy was switched — log-only, durable, replayable, never in the model transcript (the model learns the policy from the prompt section and the narrator's notices). The LAST such event is the session's override (effectiveApprovalPolicy); who asked for it is derivable from position (an event after the log's last `request/header*` was a runtime switch by the user).
-
 ```ts persistence-catalog
+/**
+ * The session's approval policy was switched — log-only, durable,
+ * replayable, never in the model transcript (the model learns the policy
+ * from the prompt section and the narrator's notices). The LAST such
+ * event is the session's override ({@link effectiveApprovalPolicy});
+ * who asked for it is derivable from position (an event after the log's
+ * last `request/header` was a runtime switch by the user).
+ */
 'approval/policy': { policy: ApprovalPolicy }
 ```
 
@@ -49,35 +144,41 @@ Source: [`packages/ui/user-approval/src/index.ts:68`](../packages/ui/user-approv
 
 #### `assistant/chunk` — log-only
 
-Raw stream chunk — token-level replay fidelity.
-
 ```ts persistence-catalog
+/** Raw stream chunk — token-level replay fidelity. */
 'assistant/chunk': { turn: number; step: number; chunk: StreamChunk }
 ```
 
 Types: [StreamChunk](core-data-structures/llm-streaming.md)
 
-Source: [`packages/core/session/src/types.ts:242`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:219`](../packages/core/session/src/types.ts)
 
 #### `assistant/message` — surface
 
-Assembled assistant message for one step (derived history uses this). Carries the step's `usage` when the adapter reported token accounting, so the model output and its accounting travel together (there is no separate usage record). `usage` is absent when the adapter reported none.
-
 ```ts persistence-catalog
-'assistant/message': { turn: number; step: number; content: ContentBlock[]; usage?: TokenUsage }
+/**
+ * Assembled assistant message for one step (derived history uses this).
+ * Carries the step's `usage` when the adapter reported token accounting, so
+ * the model output and its accounting travel together (there is no separate
+ * usage record). `usage` is absent when the adapter reported none.
+ */
+'assistant/message': { turn: number; step: number; content: ContentBlock[]; provenance: AssistantProvenance; usage?: TokenUsage }
 ```
 
 Types: [ContentBlock](core-data-structures/core.md) · [TokenUsage](core-data-structures/llm-streaming.md)
 
-Source: [`packages/core/session/src/types.ts:249`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:226`](../packages/core/session/src/types.ts)
 
 ### `bash/*`
 
 #### `bash/sandbox-mode` — log-only
 
-Durable log-only sandbox-mode override; never a surface event or model message. Execution and ACP option reporting fold the latest event through effectiveSandboxMode without adding a prompt notice.
-
 ```ts persistence-catalog
+/**
+ * Durable log-only sandbox-mode override; never a surface event or model
+ * message. Execution and ACP option reporting fold the latest event through
+ * {@link effectiveSandboxMode} without adding a prompt notice.
+ */
 'bash/sandbox-mode': { mode: SandboxMode }
 ```
 
@@ -87,19 +188,17 @@ Source: [`packages/bash/bash/src/session-mode.ts:20`](../packages/bash/bash/src/
 
 #### `compact/end` — log-only
 
-Marks the end of a compaction — log-only, releases the lock. `error` set if summarization failed.
-
 ```ts persistence-catalog
+/** Marks the end of a compaction — log-only, releases the lock. `error` set if summarization failed. */
 'compact/end': { turn: number; error?: string }
 ```
 
-Source: [`packages/compact/compact/src/types.ts:38`](../packages/compact/compact/src/types.ts)
+Source: [`packages/compact/compact/src/types.ts:40`](../packages/compact/compact/src/types.ts)
 
 #### `compact/start` — log-only
 
-Marks the start of a compaction — log-only, holds the lock until `compact/end`.
-
 ```ts persistence-catalog
+/** Marks the start of a compaction — log-only, holds the lock until `compact/end`. */
 'compact/start': { turn: number }
 ```
 
@@ -107,10 +206,30 @@ Source: [`packages/compact/compact/src/types.ts:15`](../packages/compact/compact
 
 #### `compact/summary` — log-only
 
-Provenance record of a completed summarization — log-only, no surfaceOp. The summary content is in `data.summary`; the actual surface replacement is performed by a subsequent `user/message` event that shadows the compacted range.
-
 ```ts persistence-catalog
-'compact/summary': { summary: ContentBlock[]; shadowedRange: { start: number; end: number }; shadowedSeqs: number[]; shadowedTokenCount: number; model: string; maxTokens?: number }
+/**
+ * Provenance record of a completed summarization — log-only, no surfaceOp.
+ * The summary content is in `data.summary`; the actual surface replacement
+ * is performed by a subsequent `user/message` event that shadows the
+ * compacted range.
+ */
+'compact/summary': {
+  summary: ContentBlock[]
+  shadowedRange: { start: number; end: number }
+  shadowedSeqs: number[]
+  shadowedTokenCount: number
+  /** The provider route that wrote the summary. */
+  provider: string
+  /**
+   * The model that wrote the summary — the summarize call's envelope,
+   * reported by the backend that made the call, logged so the one-shot
+   * request is reconstructable from log + code and "which model wrote
+   * this summary" has a durable answer (the reconstructability RFC).
+   */
+  model: string
+  /** The generation cap the summarize call sent, when one applied. */
+  maxTokens?: number
+}
 ```
 
 Types: [ContentBlock](core-data-structures/core.md)
@@ -121,34 +240,68 @@ Source: [`packages/compact/compact/src/types.ts:22`](../packages/compact/compact
 
 #### `context/message` — surface
 
-In-session context injection (file-change notices, subdir AGENTS.md, skill content, cron notifications, …). Rendered into the derived history as tagged synthetic context — NOT a user prompt.
-
 ```ts persistence-catalog
-'context/message': { content: ContentBlock[]; source: MessageSource }
+/**
+ * In-session context injection (file-change notices, subdir AGENTS.md,
+ * skill content, cron notifications, …). Rendered into the derived history
+ * as synthetic context — NOT a user prompt. `envelope: 'raw'` lets a caller
+ * own the complete model-facing frame; `meta` is durable JSON state omitted
+ * from the model projection.
+ */
+'context/message': {
+  content: ContentBlock[]
+  source: MessageSource
+  envelope?: ContextEnvelope
+  meta?: JsonValue
+}
 ```
 
 Types: [ContentBlock](core-data-structures/core.md) · [MessageSource](core-data-structures/core.md)
 
-Source: [`packages/core/session/src/types.ts:240`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:212`](../packages/core/session/src/types.ts)
 
 ### `hook/*`
 
 #### `hook/invoked` — log-only
 
-A hook command was invoked at a hook point — log-only provenance (like `compact/*`; NOT a SurfaceEventType, carries no `surfaceOp`). `dialect` is the bridge that ran it (`claude`/`codex`), `point` the hook point (`PreToolUse`, `Stop`, …), `matcher` the matcher-group pattern that selected it (absent for match-all), `handlerId` a stable id for the command (so an invoked/result pair correlates). `turn` is the open turn the invocation lives inside.
-
 ```ts persistence-catalog
-'hook/invoked': { turn: number; point: string; dialect: HookDialect; matcher?: string; handlerId: string }
+/**
+ * A hook command was invoked at a hook point — log-only provenance (like
+ * `compact/*`; NOT a {@link SurfaceEventType}, carries no `surfaceOp`).
+ * `dialect` is the bridge that ran it (`claude`/`codex`), `point`
+ * the hook point (`PreToolUse`, `Stop`, …), `matcher` the matcher-group
+ * pattern that selected it (absent for match-all), `handlerId` a stable id
+ * for the command (so an invoked/result pair correlates). `turn` is the open
+ * turn the invocation lives inside.
+ */
+'hook/invoked': {
+  turn: number
+  point: string
+  dialect: HookDialect
+  matcher?: string
+  handlerId: string
+}
 ```
 
 Source: [`packages/hooks/hook-protocol/src/types.ts:19`](../packages/hooks/hook-protocol/src/types.ts)
 
 #### `hook/result` — log-only
 
-Log-only outcome paired to `hook/invoked` by `handlerId`. Decision is the parsed permission result, `stop` for `continue:false`, or `pass`; exit code may be absent, stderr is bounded, and duration is wall-clock runtime.
-
 ```ts persistence-catalog
-'hook/result': { turn: number; point: string; handlerId: string; decision: string; exitCode?: number; stderrSummary?: string; durationMs: number }
+/**
+ * Log-only outcome paired to `hook/invoked` by `handlerId`. Decision is the
+ * parsed permission result, `stop` for `continue:false`, or `pass`; exit code
+ * may be absent, stderr is bounded, and duration is wall-clock runtime.
+ */
+'hook/result': {
+  turn: number
+  point: string
+  handlerId: string
+  decision: string
+  exitCode?: number
+  stderrSummary?: string
+  durationMs: number
+}
 ```
 
 Source: [`packages/hooks/hook-protocol/src/types.ts:31`](../packages/hooks/hook-protocol/src/types.ts)
@@ -157,9 +310,13 @@ Source: [`packages/hooks/hook-protocol/src/types.ts:31`](../packages/hooks/hook-
 
 #### `permission/preset` — log-only
 
-Records the selected preset as durable, log-only user intent. The knob events follow in the same turn and control execution; this event stays out of the model transcript and lets effectivePermissionPreset preserve a selection when bundles match.
-
 ```ts persistence-catalog
+/**
+ * Records the selected preset as durable, log-only user intent. The knob
+ * events follow in the same turn and control execution; this event stays
+ * out of the model transcript and lets {@link effectivePermissionPreset}
+ * preserve a selection when bundles match.
+ */
 'permission/preset': { preset: string }
 ```
 
@@ -169,162 +326,181 @@ Source: [`packages/ui/permission/src/index.ts:33`](../packages/ui/permission/src
 
 #### `prompt/blocked` — log-only
 
-Durable record of a prompt veto and its reason. It is log-only: the blocked prompt never enters the model-visible surface, including in a mixed batch.
-
 ```ts persistence-catalog
+/**
+ * Durable record of a prompt veto and its reason. It is log-only: the blocked
+ * prompt never enters the model-visible surface, including in a mixed batch.
+ */
 'prompt/blocked': { content: ContentBlock[]; source: MessageSource; reason: string }
 ```
 
 Types: [ContentBlock](core-data-structures/core.md) · [MessageSource](core-data-structures/core.md)
 
-Source: [`packages/core/session/src/types.ts:234`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:204`](../packages/core/session/src/types.ts)
 
 ### `request/*`
 
 #### `request/header` — log-only
 
-Full EpochHeader for the next request, appended inside its step before dispatch. It is log-only and anchors subsequent deltas.
-
 ```ts persistence-catalog
+/**
+ * Full header for the next request, appended inside its step before dispatch.
+ * It is log-only; the latest snapshot reconstructs the request header.
+ */
 'request/header': { header: EpochHeader; reason: RequestHeaderReason }
 ```
 
-Source: [`packages/core/session/src/types.ts:277`](../packages/core/session/src/types.ts)
-
-#### `request/header-delta` — log-only
-
-Log-only amendment to the folded EpochHeader. System and tools use their delta codecs; config and prefix replace whole, with an empty prefix encoding removal. Writers verify round-trip equality or log a fallback snapshot.
-
-```ts persistence-catalog
-'request/header-delta': { system?: SystemDelta; tools?: ToolsDelta; config?: LlmCallConfig; messagePrefix?: Message[] }
-```
-
-Source: [`packages/core/session/src/types.ts:283`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:251`](../packages/core/session/src/types.ts)
 
 ### `steering/*`
 
 #### `steering/message` — surface
 
-Steering content injected between steps of a running turn.
-
 ```ts persistence-catalog
+/** Steering content injected between steps of a running turn. */
 'steering/message': { turn: number; content: ContentBlock[]; source: MessageSource }
 ```
 
 Types: [ContentBlock](core-data-structures/core.md) · [MessageSource](core-data-structures/core.md)
 
-Source: [`packages/core/session/src/types.ts:267`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:244`](../packages/core/session/src/types.ts)
 
 ### `step/*`
 
 #### `step/end` — log-only
 
-Closes step `step` of turn `turn`.
-
 ```ts persistence-catalog
+/** Closes step `step` of turn `turn`. */
 'step/end': { turn: number; step: number }
 ```
 
-Source: [`packages/core/session/src/types.ts:227`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:197`](../packages/core/session/src/types.ts)
 
 #### `step/start` — log-only
 
-Opens step `step` of turn `turn` — one model call plus the tool executions it requested.
-
 ```ts persistence-catalog
+/** Opens step `step` of turn `turn` — one model call plus the tool executions it requested. */
 'step/start': { turn: number; step: number }
 ```
 
-Source: [`packages/core/session/src/types.ts:225`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:195`](../packages/core/session/src/types.ts)
 
 ### `todo/*`
 
 #### `todo/write` — log-only
 
-Whole-list snapshot; the latest write wins on replay. It is log-only UI state and never enters derived model history.
-
 ```ts persistence-catalog
+/** Whole-list snapshot; latest write wins on replay. Log-only UI state; never derived history. */
 'todo/write': { todos: TodoItem[] }
 ```
 
 Types: [TodoItem](core-data-structures/session.md)
 
-Source: [`packages/core/session/src/types.ts:272`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:246`](../packages/core/session/src/types.ts)
 
 ### `tool/*`
 
 #### `tool/call` — log-only
 
-The model requested one tool invocation: `name` with the raw `arguments` JSON string exactly as the model produced it (unparsed). `callId` pairs the call with its `tool/result`.
-
 ```ts persistence-catalog
+/**
+ * The model requested one tool invocation: `name` with the raw `arguments`
+ * JSON string exactly as the model produced it (unparsed). `callId` pairs the
+ * call with its `tool/result`.
+ */
 'tool/call': { turn: number; step: number; callId: CallId; name: string; arguments: string }
 ```
 
 Types: [CallId](core-data-structures/core.md)
 
-Source: [`packages/core/session/src/types.ts:255`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:232`](../packages/core/session/src/types.ts)
 
 #### `tool/code-dispatch` — log-only
 
-One bridged sub-dispatch from a `run_code` program: the parent `run_code` call id, the deterministic sub-call id (`<parent>:code:<n>`), the tool `name` with its JSON-normalized `arguments` — the exact value dispatched, normalized before dispatch, so this append can never fail on payload shape — whether the sub-call errored, and a bounded `resultSummary` of its model-facing text.
-
 ```ts persistence-catalog
+/**
+ * One bridged sub-dispatch from a `run_code` program: the parent
+ * `run_code` call id, the deterministic sub-call id
+ * (`<parent>:code:<n>`), the tool `name` with its JSON-normalized
+ * `arguments` — the exact value dispatched, normalized BEFORE dispatch,
+ * so this append can never fail on payload shape — whether the sub-call
+ * errored, and a bounded `resultSummary` of its model-facing text. Before
+ * bounding, occurrences of a non-root session workspace path are
+ * normalized to `.` so host-specific absolute path lengths cannot change
+ * the summary.
+ * Log-only: `deriveMessages()` ignores it, so sub-calls never re-enter
+ * model context; persistence and UIs get every call. Appended inside the
+ * parent `run_code`'s execution (the bridge drains its queue before
+ * returning), so the turn-enclosure invariant holds by construction.
+ */
 'tool/code-dispatch': { parentCallId: CallId; subCallId: CallId; name: string; arguments: unknown; isError: boolean; resultSummary: string }
 ```
 
 Types: [CallId](core-data-structures/core.md)
 
-Source: [`packages/core/tools/src/code-mode.ts:25`](../packages/core/tools/src/code-mode.ts)
+Source: [`packages/core/tools/src/code-mode.ts:34`](../packages/core/tools/src/code-mode.ts)
 
 #### `tool/result` — surface
 
-A completed tool call's model-facing result, plus an optional tool-private `meta` presentation payload. `meta` is opaque to the core (`unknown` — the producing tool owns its shape and reads it back in `presentResult`) but MUST be JSON-serializable: `Session.append` runtime-validates all event data with `isJsonValue`, so a non-serializable `meta` is rejected at the source, and the durable log reproduces the identical card on replay. Absent unless the tool attaches one (e.g. `dsh-tool-fs` carries its result-time contextual diff here).
-
 ```ts persistence-catalog
+/**
+ * A completed tool call's model-facing result, plus an optional tool-private
+ * `meta` presentation payload. `meta` is opaque to the core (`unknown` — the
+ * producing tool owns its shape and reads it back in `presentResult`) but MUST
+ * be JSON-serializable: `Session.append` runtime-validates all event data with
+ * `isJsonValue`, so a non-serializable `meta` is rejected at the source, and the
+ * durable log reproduces the identical card on replay. Absent unless the tool
+ * attaches one (e.g. `dsh-tool-fs` carries its result-time contextual diff here).
+ */
 'tool/result': { turn: number; step: number; callId: CallId; content: ContentBlock[]; isError: boolean; error?: { name: string; code: string }; meta?: unknown }
 ```
 
 Types: [CallId](core-data-structures/core.md) · [ContentBlock](core-data-structures/core.md)
 
-Source: [`packages/core/session/src/types.ts:265`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:242`](../packages/core/session/src/types.ts)
 
 ### `turn/*`
 
 #### `turn/end` — log-only
 
-Closes turn `turn` with the TurnEndReason that ended it. The loop fires the awaited `session/flush` checkpoint at every turn end, so the turn boundary is also the durable-commit boundary.
-
 ```ts persistence-catalog
+/**
+ * Closes turn `turn` with the {@link TurnEndReason} that ended it. The loop
+ * fires the awaited `session/flush` checkpoint at every turn end, so the turn
+ * boundary is also the durable-commit boundary.
+ */
 'turn/end': { turn: number; reason: TurnEndReason }
 ```
 
 Types: [TurnEndReason](core-data-structures/session.md)
 
-Source: [`packages/core/session/src/types.ts:223`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:193`](../packages/core/session/src/types.ts)
 
 #### `turn/start` — log-only
 
-Opens turn `turn`. `trigger` records what started it — a drained message batch or an idle-time injection. The turn is the durability/replay boundary: every event sits between a `turn/start` and its matching `turn/end` (the turn-enclosure invariant).
-
 ```ts persistence-catalog
+/**
+ * Opens turn `turn`. `trigger` records what started it — a drained message
+ * batch or an idle-time injection. The turn is the durability/replay
+ * boundary: every event sits between a `turn/start` and its matching
+ * `turn/end` (the turn-enclosure invariant).
+ */
 'turn/start': { turn: number; trigger: TurnTrigger }
 ```
 
 Types: [TurnTrigger](core-data-structures/session.md)
 
-Source: [`packages/core/session/src/types.ts:217`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:187`](../packages/core/session/src/types.ts)
 
 ### `user/*`
 
 #### `user/message` — surface
 
-A user-visible prompt (queued message drained at turn start).
-
 ```ts persistence-catalog
+/** A user-visible prompt (queued message drained at turn start). */
 'user/message': { content: ContentBlock[]; source: MessageSource }
 ```
 
 Types: [ContentBlock](core-data-structures/core.md) · [MessageSource](core-data-structures/core.md)
 
-Source: [`packages/core/session/src/types.ts:229`](../packages/core/session/src/types.ts)
+Source: [`packages/core/session/src/types.ts:199`](../packages/core/session/src/types.ts)
diff --git a/docs/rfc/INDEX.md b/docs/rfc/INDEX.md
index 65c637ebd5..942e6c17c8 100644
--- a/docs/rfc/INDEX.md
+++ b/docs/rfc/INDEX.md
@@ -9,10 +9,12 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand;
 | Title | First proposed |
 |---|---|
 | [Pre-tool input rewrite — a consistent design](proposed/feature/2026-06-30-pre-tool-input-rewrite.md) | 2026-06-30 |
+| [Recallable compaction — index checkpoints, a state checkpoint, and in-session history recall](proposed/feature/2026-07-06-recallable-compaction.md) | 2026-07-06 |
 | [Claude Code and Codex subagent backends (out-of-process delegation to external coding agents)](proposed/feature/2026-07-07-claude-code-and-codex-subagent-backends.md) | 2026-07-07 |
 | [Interactive side sessions and merge-back](proposed/feature/2026-07-08-interactive-side-sessions.md) | 2026-07-08 |
 | [SQLite FTS5 session search](proposed/feature/2026-07-10-sqlite-session-query-provider.md) | 2026-07-10 |
 | [Stream workflow progress through tool calls](proposed/feature/2026-07-13-stream-workflow-progress-through-tool-calls.md) | 2026-07-13 |
+| [Developer-owned SDK projects](proposed/feature/2026-07-14-sdk-developer-projects.md) | 2026-07-14 |
 
 ### Simplification
 
@@ -20,14 +22,13 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand;
 |---|---|
 | [Unify the agent id and the session id](proposed/simplification/2026-06-20-unify-agent-and-session-id.md) | 2026-06-20 |
 | [Prune dead public and result surface](proposed/simplification/2026-07-04-prune-dead-core-spine-surface.md) | 2026-07-04 |
-| [Simplify session-log representation](proposed/simplification/2026-07-12-simplify-session-log-representation.md) | 2026-07-12 |
 
 ### Architecture
 
 | Title | First proposed |
 |---|---|
 | [Runtime schemas for the event vocabulary (Zod vs the merge-extensible-map pattern)](proposed/architecture/2026-06-16-typed-event-schemas.md) | 2026-06-16 |
-| [Extract a generic long-running tool runtime](proposed/architecture/2026-06-20-generic-long-running-tool-runtime.md) | 2026-06-20 |
+| [SDK project editing architecture](proposed/architecture/2026-07-15-sdk-project-editing-architecture.md) | 2026-07-15 |
 
 ### Process
 
@@ -37,6 +38,7 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand;
 | [Architectural conformance — dependency rules and the adapter kit](proposed/process/2026-06-11-architectural-conformance.md) | 2026-06-11 |
 | [Supply chain checks and vendor drift verification](proposed/process/2026-06-11-supply-chain-and-vendor-drift.md) | 2026-06-11 |
 | [Discover package inventories instead of maintaining static lists](proposed/process/2026-06-20-discover-package-inventory.md) | 2026-06-20 |
+| [Periodic human-review maintenance for dsh-code-review](proposed/process/2026-07-13-human-review-skill-maintenance.md) | 2026-07-13 |
 
 ### Testing
 
@@ -59,6 +61,7 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand;
 | [Compaction as a capability seam (abstract contract + basic backend)](implemented/feature/2026-06-18-compaction-capability-seam.md) | 2026-06-18 |
 | [Subagent capability seam](implemented/feature/2026-06-21-subagent-capability-seam.md) | 2026-06-21 |
 | [ACP subagent backend (out-of-process delegation)](implemented/feature/2026-06-22-acp-subagent-backend.md) | 2026-06-22 |
+| [Workspace context instruction files](implemented/feature/2026-06-24-workspace-context.md) | 2026-06-24 |
 | [Ask-user question capability](implemented/feature/2026-06-25-ask-user-question.md) | 2026-06-25 |
 | [The `todo_write` tool — model task list as event-sourced session state](implemented/feature/2026-06-29-todo-write-tool.md) | 2026-06-29 |
 | [dsh-hooks-claude + dsh-hooks-codex — the Claude Code / Codex hook bridges](implemented/feature/2026-06-30-hook-bridges.md) | 2026-06-30 |
@@ -73,11 +76,17 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand;
 | [The subprocess sandbox — confinement seam, native runners, escalation, and per-session modes](implemented/feature/2026-07-06-sandbox.md) | 2026-07-06 |
 | [MCP client plugin — connect to external MCP servers and bridge their tools](implemented/feature/2026-07-07-mcp-client-plugin.md) | 2026-07-07 |
 | [The session prefix — request-only messages in front of the derived history](implemented/feature/2026-07-07-session-prefix.md) | 2026-07-07 |
+| [Background subagent tasks](implemented/feature/2026-07-08-background-subagent-tasks.md) | 2026-07-08 |
 | [Repeat-tool-call guard plugin](implemented/feature/2026-07-08-repeat-tool-guard.md) | 2026-07-08 |
 | [The self-referential cordis toolset](implemented/feature/2026-07-08-self-referential-cordis-toolset.md) | 2026-07-08 |
+| [Bash-backed grep and glob discovery tools](implemented/feature/2026-07-09-bash-backed-grep-glob-discovery.md) | 2026-07-09 |
+| [Expose agent session identity and JSONL location to tools and hooks](implemented/feature/2026-07-10-agent-session-identity-and-log-location.md) | 2026-07-10 |
+| [Parallel tool-call execution by per-call safety](implemented/feature/2026-07-10-parallel-tool-call-execution.md) | 2026-07-10 |
 | [Exact session query service](implemented/feature/2026-07-10-session-query-service.md) | 2026-07-10 |
 | [Configure subagent persona, tool visibility, and depth](implemented/feature/2026-07-12-subagent-persona-tool-filter-and-depth.md) | 2026-07-12 |
+| [Session query relationship tracing](implemented/feature/2026-07-13-session-query-tracing.md) | 2026-07-13 |
 | [Optional time-context plugin](implemented/feature/2026-07-14-time-context-plugin.md) | 2026-07-14 |
+| [Durable per-step time context](implemented/feature/2026-07-16-durable-per-step-time-context.md) | 2026-07-16 |
 
 ### Simplification
 
@@ -104,6 +113,7 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand;
 | [Trim unreachable ACP bridge surface — the branding knobs and the kind-sniffing fallback](implemented/simplification/2026-07-04-trim-acp-bridge-unreachable-surface.md) | 2026-07-04 |
 | [Drop unconsumed skill provider events](implemented/simplification/2026-07-12-drop-unconsumed-skill-provider-events.md) | 2026-07-12 |
 | [Prune unused web seam fields](implemented/simplification/2026-07-12-prune-unused-web-seam-fields.md) | 2026-07-12 |
+| [Simplify session-log representation](implemented/simplification/2026-07-12-simplify-session-log-representation.md) | 2026-07-12 |
 
 ### Architecture
 
@@ -123,10 +133,11 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand;
 | [Every session event is enclosed in a turn](implemented/architecture/2026-06-15-turn-enclosure-invariant.md) | 2026-06-15 |
 | [Filesystem capability seam — ctx.fs, local backend, and model-facing filesystem tools](implemented/architecture/2026-06-17-filesystem-capability-seam.md) | 2026-06-17 |
 | [Agent lifecycle and ownership seams](implemented/architecture/2026-06-18-agent-lifecycle-and-ownership-seams.md) | 2026-06-18 |
-| [Session surface — a linked list over the event log for LLM message derivation](implemented/architecture/2026-06-18-session-surface.md) | 2026-06-18 |
+| [Session surface — an ordered projection over the event log](implemented/architecture/2026-06-18-session-surface.md) | 2026-06-18 |
 | [Shared persistence write coordinator](implemented/architecture/2026-06-18-shared-persistence-write-coordinator.md) | 2026-06-18 |
 | [Branded IDs everywhere they belong](implemented/architecture/2026-06-20-branded-ids.md) | 2026-06-20 |
 | [Extract example apps into packages](implemented/architecture/2026-06-20-extract-example-app-packages.md) | 2026-06-20 |
+| [The background task runtime (`ctx.tasks`) and generic task control tools](implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md) | 2026-06-20 |
 | [Reorganize packages into a modular hierarchy](implemented/architecture/2026-06-20-package-hierarchy.md) | 2026-06-20 |
 | [Mandatory `User-Agent` attribution for provider requests](implemented/architecture/2026-06-21-mandatory-app-attribution-headers.md) | 2026-06-21 |
 | [Web capability seam - stable tools over multiple providers](implemented/architecture/2026-06-24-web-capability-seam.md) | 2026-06-24 |
@@ -141,10 +152,15 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand;
 | [Every LLM request is reconstructable from the session log](implemented/architecture/2026-07-05-reconstructable-requests.md) | 2026-07-05 |
 | [Subagent provider-lifecycle events — `subagent/provider-added` / `subagent/provider-removed`](implemented/architecture/2026-07-05-subagent-provider-lifecycle-events.md) | 2026-07-05 |
 | [A shared timeout/deadline primitive, with hard-kill left to each capability](implemented/architecture/2026-07-06-timeout-deadline-library.md) | 2026-07-06 |
+| [Tool result retention library](implemented/architecture/2026-07-06-tool-result-retention-library.md) | 2026-07-06 |
 | [Tool-call timeout policy as a plugin](implemented/architecture/2026-07-07-tool-call-timeout-policy.md) | 2026-07-07 |
 | [The agent is a registration scope](implemented/architecture/2026-07-08-agent-scope-contexts.md) | 2026-07-08 |
+| [Tool output spill policy](implemented/architecture/2026-07-08-tool-output-spill-files.md) | 2026-07-08 |
 | [Single-file executable SDK runtime distribution (single-exe)](implemented/architecture/2026-07-10-single-file-executable-sdk-runtime-distribution.md) | 2026-07-10 |
 | [Agent-scope runtime design and correctness](implemented/architecture/2026-07-12-agent-scope-runtime-design.md) | 2026-07-12 |
+| [Provider-routed LLM adapters and a generic pi-ai backend](implemented/architecture/2026-07-14-provider-routed-llm-adapters.md) | 2026-07-14 |
+| [Advisory LLM catalogs and per-session ACP model selection](implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.md) | 2026-07-15 |
+| [Replay token meter service](implemented/architecture/2026-07-15-replay-token-meter-service.md) | 2026-07-15 |
 
 ### Process
 
@@ -177,6 +193,7 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand;
 | [Package Model Experience contract](implemented/process/2026-07-12-package-model-experience-contract.md) | 2026-07-12 |
 | [Project canonical documentation into the website](implemented/process/2026-07-13-documentation-site-projection.md) | 2026-07-13 |
 | [TypeScript Program-backed semantic gates](implemented/process/2026-07-14-typescript-program-backed-semantic-gates.md) | 2026-07-14 |
+| [Run CI examples from built lib](implemented/process/2026-07-17-run-ci-examples-from-built-lib.md) | 2026-07-17 |
 
 ### Testing
 
diff --git a/docs/rfc/implemented/architecture/2026-06-13-capability-seams.md b/docs/rfc/implemented/architecture/2026-06-13-capability-seams.md
index 907e4cd86b..6887660de0 100644
--- a/docs/rfc/implemented/architecture/2026-06-13-capability-seams.md
+++ b/docs/rfc/implemented/architecture/2026-06-13-capability-seams.md
@@ -12,9 +12,9 @@ This is distinct from "who provides vs. needs a capability at runtime", which Co
 
 A swappable capability is **three packages**:
 
-1. **Interface** — an abstract service + the vocabulary types, owning the `ctx.<key>` and depending only on cordis (e.g. `dsh-bash`: `BashExecutor`, `BashRunResult`, `BashTask`).
+1. **Interface** — an abstract service + the vocabulary types, owning the `ctx.<key>` and depending only on its vocabulary dependencies (e.g. `dsh-bash`: `BashExecutor`, `BashRunResult`, `BashProcess`).
 2. **Implementation** — a concrete subclass loaded as a plugin (e.g. `dsh-bash-local`: subprocesses, process-group kills, spill-file truncation). Sandboxed/remote backends are sibling packages implementing the same interface.
-3. **Consumer** — what the model and plugins see (e.g. `dsh-tool-bash`: the `bash`/`bash_output`/`bash_kill` tool schemas). Consumers `inject` the interface key and never import implementation types.
+3. **Consumer** — what the model and plugins see (e.g. `dsh-tool-bash`: the `bash` schema, with background handles registered into the generic task runtime). Consumers `inject` the interface key and never import implementation types.
 
 Implementation and consumer then evolve independently: a sandboxed executor replaces `dsh-bash-local` without touching a tool schema.
 
diff --git a/docs/rfc/implemented/architecture/2026-06-15-turn-enclosure-invariant.md b/docs/rfc/implemented/architecture/2026-06-15-turn-enclosure-invariant.md
index e55cd0853e..a54f735219 100644
--- a/docs/rfc/implemented/architecture/2026-06-15-turn-enclosure-invariant.md
+++ b/docs/rfc/implemented/architecture/2026-06-15-turn-enclosure-invariant.md
@@ -18,7 +18,7 @@ In case 2, if the injected `context/message` is the last event before a flush/di
 **Every session event lives inside a turn** — between a `turn/start` and its matching `turn/end`. Concretely:
 
 - The loop appends queued `user/message` events **after** `turn/start` (inside the turn), not before it. `turn/end` is therefore owed the moment those messages are recorded, and the existing finalizer guarantees it.
-- An `agent.inject()` made while the agent is **running** appends its `context/message` into the already-open turn (unchanged).
+- An `agent.inject()` made while the agent is **running** joins the already-open turn. While the current step executes assistant tool calls, accepted context waits in arrival order until that batch settles, then appends after every recorded result and before the turn closes even when execution is interrupted.
 - An `agent.inject()` made while **idle** wraps its `context/message` in a one-shot turn: `turn/start{trigger:{kind:'injection'}}` → `context/message` → `turn/end{completed}`. A new `injection` variant joins the merge-extensible `TurnTriggerMap`.
 - The loop derives the next turn number from the log each iteration (`lastTurnNumber(session) + 1`) instead of keeping a private counter, so an idle injection's one-shot turn cannot collide with the next real turn's number.
 - The `dsh-invariants` plugin **enforces** the invariant in dev: a `user/message` / `context/message` / `steering/message` appended while no turn is open throws an `InvariantError`.
diff --git a/docs/rfc/implemented/architecture/2026-06-17-filesystem-capability-seam.md b/docs/rfc/implemented/architecture/2026-06-17-filesystem-capability-seam.md
index 71efeff826..0121a11fa0 100644
--- a/docs/rfc/implemented/architecture/2026-06-17-filesystem-capability-seam.md
+++ b/docs/rfc/implemented/architecture/2026-06-17-filesystem-capability-seam.md
@@ -81,7 +81,7 @@ Resolved targets must expose at least three concepts:
 - An opaque `targetKey`, used for stale guards and file-state lookup. The local backend might use a realpath-like key; a remote backend might use a workspace URI or file id. Consumers must not parse or assume this is a local absolute path.
 - A `displayPath`, used for model/UI-facing output. It may be a local absolute path, workspace-relative path, or remote URI depending on the backend.
 
-Read and mutation results must include an opaque file `version`. A local backend can use mtime/size or a hash-like token; a remote backend can use a revision id. The `dsh-fs-policy` plugin records versions for stale checks; consumers may display related metadata but must not interpret the version token.
+Read and mutation results must include an opaque file `version`. The local backend derives its token from bigint stat metadata (`dev`, `ino`, `size`, `mtimeNs`, and `ctimeNs`) so same-size rewrites and inode replacement invalidate consumers reliably; a remote backend can use a revision id or hash-like token. The `dsh-fs-policy` plugin records versions for stale checks; consumers may display related metadata but must not interpret the version token.
 
 The provider hands back decoded text: `readText` returns a whole regular text file, `streamText` streams the same text semantics for large files. Both own regular-file checks, bounded line/output handling is NOT theirs — line windowing, numbered-line rendering, and total-line accounting live in the executor (`dsh-tool-fs`), which reads through `ctx.fs` and renders the model-facing window. The provider owns UTF-8 decoding and binary/NUL rejection; it does not know about line windows or views.
 
diff --git a/docs/rfc/implemented/architecture/2026-06-18-session-surface.md b/docs/rfc/implemented/architecture/2026-06-18-session-surface.md
index 6ba0df41c4..7c6c5b00d0 100644
--- a/docs/rfc/implemented/architecture/2026-06-18-session-surface.md
+++ b/docs/rfc/implemented/architecture/2026-06-18-session-surface.md
@@ -1,4 +1,4 @@
-# RFC: Session surface — a linked list over the event log for LLM message derivation
+# RFC: Session surface — an ordered projection over the event log
 
 Status: implemented
 
@@ -8,13 +8,13 @@ The event log is authoritative, but history manipulation had no durable shared m
 
 ## Decision
 
-Add a **surface** — a derived, cached linked list of "surface nodes" (the subset of events that produce LLM messages) — maintained by `surfaceOp` markers in the event log.
+Add a **surface** — a derived, cached order of event sequences (the subset of events that produce LLM messages) — maintained by `surfaceOp` markers in the event log.
 
 ### Two new top-level fields on `SessionEvent`
 
 Every `SessionEvent` gains two optional fields (structural metadata, like `seq`/`time`):
 
-- **`sourceEventSeqs?: number[]`** — seq numbers of events that are provenance sources (e.g., the `assistant/chunk` seqs that built an `assistant/message`, or the surface nodes shadowed by a compaction marker). Provenance is a core design principle; without it, the replace-range operation cannot be validated on replay.
+- **`sourceEventSeqs?: number[]`** — seq numbers of events that are provenance sources (e.g., the `assistant/chunk` seqs that built an `assistant/message`, or the surface nodes shadowed by a compaction marker). A present `[]` is valid only on `assistant/message` and records a known empty provider stream; omission there means legacy or otherwise unrecorded provenance. Other surface events require a non-empty list when the field is present. Provenance is a core design principle; without it, the replace-range operation cannot be validated on replay.
 - **`surfaceOp?: SurfaceOp`** — how this event entered the surface. Absent for non-surface events.
 
 ### SurfaceOp: two operations
@@ -25,13 +25,13 @@ export type SurfaceOp =
   | { op: 'replace'; start: number; end: number }  // shadow [start, end] inclusive
 ```
 
-1. **Append** — add a new node to the tail. Used by `user/message`, `assistant/message`, `tool/result`, `context/message`, `steering/message`. The loop passes `surfaceOp: 'append'` on all such appends, and `sourceEventSeqs` where applicable (e.g., `assistant/message` records its `assistant/chunk` sources; `tool/result` records its `tool/call` source).
+1. **Append** — add the new event seq to the tail. Used by `user/message`, `assistant/message`, `tool/result`, `context/message`, `steering/message`. The loop passes `surfaceOp: 'append'` on all such appends and records `sourceEventSeqs` where applicable: every successful `assistant/message` records its complete `assistant/chunk` source set, including `[]`, while `tool/result` records its `tool/call` source.
 
-2. **Replace** — remove nodes from `start` through `end` (both inclusive) and insert a new node in their place. Both `start` and `end` must be valid surface node seqs in the current surface; `start === end` replaces a single node. The node's `sourceEventSeqs` must contain every shadowed surface node. The shadowed events remain in the log but are no longer on the surface.
+2. **Replace** — remove entries from `start` through `end` (both inclusive) and insert the new event seq in their place. Both `start` and `end` must be present in the current surface; `start === end` replaces one entry. The event's `sourceEventSeqs` must contain every shadowed surface seq. The shadowed events remain in the log but are no longer on the surface.
 
 ### SurfaceManager: delta-based, not full rebuild
 
-A `SurfaceManager` class (private to `Session`) maintains the cached linked list. It tracks `_lastProcessedSeq` and processes only the **delta** (new events since the last access) rather than rescanning the entire log. Because the log is append-only, prior events never change; a seeded log is simply the initial delta folded on first access.
+A `SurfaceManager` class (private to `Session`) maintains one ordered `number[]` of event seqs. It tracks `_lastProcessedSeq` and processes only the new events since the last access rather than rescanning the entire log. Because the log is append-only, prior events never change; a seeded log is simply the initial suffix folded on first access. Replace locates its inclusive endpoints by array position and splices the replacement seq into that range; no link objects or seq-to-node map duplicate the order.
 
 Delta processing is O(1) when no new events and O(new events) when new events arrive.
 
@@ -47,23 +47,24 @@ The `repair.ts` module synthesizes `tool/result` closers for orphaned tool calls
 
 ### Invariants
 
-The dev-mode invariants plugin validates: `sourceEventSeqs` references (non-empty, no duplicates, references earlier events, references known seqs) and `surfaceOp` (replace `start ≤ end`, both endpoints are on the tracked surface, the range is non-reversed in surface position, and `sourceEventSeqs` includes every node the range shadows).
+The dev-mode invariants plugin validates: `sourceEventSeqs` references (only `assistant/message` may use an empty list; otherwise no duplicates, references earlier events, and references known seqs) and `surfaceOp` (replace `start ≤ end`, both endpoints are on the tracked surface, the range is non-reversed in surface position, and `sourceEventSeqs` includes every node the range shadows).
 
 Every surface-eligible event must carry `surfaceOp` or it would disappear from derived history. Typed `append` overloads enforce this for literal event types; runtime checks in `append` and the seed constructor cover widened unions and loaded logs. Invalid seeds are rejected rather than upgraded under the pre-release format policy.
 
 ## Alternatives considered
 
 - **Per-plugin `agent/request` wrapping** (the pre-surface pattern for history manipulation) — listener-ordering fragility, no durable record of what was changed, and every new manipulation forces another change to core `deriveMessages()`.
-- **Half-open `[start, endExclusive)` replace ranges** — rejected: the surface is a doubly-linked list whose ends are naturally named by node seqs, and single-node replacement (`start === end`) reads naturally with inclusive semantics.
+- **Half-open `[start, endExclusive)` replace ranges** — rejected: endpoints are named by surface event seqs, and single-entry replacement (`start === end`) reads naturally with inclusive semantics.
+- **Linked node objects plus a seq map** — rejected: production did not read predecessor links, the only successor use was the next array position, and replacement already required linear `indexOf` lookup. A single seq array preserves the same asymptotic behavior with one representation to validate.
 - **Full rebuild behind a dirty flag** instead of delta processing — O(N²) over a session's lifetime: every single-event append would rescan all prior events.
 
 ## Consequences
 
-- **`packages/core/session`**: New `surface.ts` (`SurfaceManager`), new types (`SurfaceOp`, `SurfaceIntent`), new fields on `SessionEvent`, modified `append()` (third required `SurfaceIntent` param), refactored `deriveMessages()` (walks the surface as the sole derivation path), surface-aware `repair.ts`. The seed constructor rejects a surface-eligible seed event missing its `surfaceOp` marker (see § Invariants).
+- **`packages/core/session`**: `surface.ts` (`SurfaceManager`) maintains one ordered seq array; `SurfaceOp`/`SurfaceIntent` and the top-level session-event fields record how entries join it. `append()` requires a `SurfaceIntent` for surface events, `deriveMessages()` walks the surface as the sole derivation path, and `repair.ts` emits surface-aware closers. The seed constructor rejects a surface-eligible seed event missing its `surfaceOp` marker (see § Invariants).
 - **`packages/core/agent-loop`**: All surface-capable appends pass surface opts. Chunk seqs are collected for `assistant/message` provenance; `tool/call` seqs are captured for `tool/result` provenance.
 - **`packages/session-persistence/session-persistence-sqlite`**: Two new nullable TEXT columns (`source_event_seqs`, `surface_op`) on the `events` table; `SCHEMA_VERSION` bumped (bump-and-reject, no migration).
 - **`packages/support/invariants`**: Surface-related validation rules.
 - **`packages/session-persistence/session-persistence-jsonl`**: No changes required.
 - **`packages/session-persistence/session-persistence`**: Abstract interface unchanged.
 
-The surface is the foundation for future history manipulation. A compaction or tool-result-prune plugin appends one of the existing message-producing event types (a `user/message` carrying the summary, say) with `surfaceOp: { op: 'replace', start, end }` and `sourceEventSeqs` covering the shadowed nodes — the new node takes the range's place on the surface while the plugin's own trace events (e.g. `compaction/start`, `compaction/end`) stay off it. Replay preserves the decision deterministically.
+The surface is the foundation for future history manipulation. A compaction or tool-result-prune plugin appends one of the existing message-producing event types (a `user/message` carrying the summary, say) with `surfaceOp: { op: 'replace', start, end }` and `sourceEventSeqs` covering the shadowed entries — the new event takes the range's place on the surface while the plugin's own trace events (e.g. `compaction/start`, `compaction/end`) stay off it. Replay preserves the decision deterministically.
diff --git a/docs/rfc/implemented/architecture/2026-06-18-shared-persistence-write-coordinator.md b/docs/rfc/implemented/architecture/2026-06-18-shared-persistence-write-coordinator.md
index b1773b480d..fef6ad034c 100644
--- a/docs/rfc/implemented/architecture/2026-06-18-shared-persistence-write-coordinator.md
+++ b/docs/rfc/implemented/architecture/2026-06-18-shared-persistence-write-coordinator.md
@@ -12,6 +12,8 @@ Extract a backend-agnostic `PersistenceCoordinator` into `dsh-session-persistenc
 
 Composition, not inheritance. The coordinator is a concrete class the backend holds, not a base class the backend extends. The RFC's risk — "a coordinator must not make unusual backends fight an inheritance hierarchy" — is avoided: a backend exposes only the hooks; it cannot reach the coordinator's private orchestration state, and the public `SessionPersistence` service shape is unchanged, so a third-party backend MAY still implement the abstract service directly without the coordinator at all.
 
+The coordinator retires each live session from its `session/disposed` notification: it waits for that exact Session object's initialization, serializes a final drain, and then removes the owned state, buffer, and init entries. Failed drains retain their buffers for backend teardown to retry. Settled per-id chain tails remove themselves only when they are still the current tail, so a completion cannot erase a newer operation for the same id. Backend teardown unregisters the write-path listeners before awaiting all admitted retirements, remaining buffers, and chains, then closes the backend.
+
 ### The hook interface (`PersistenceBackend<TornMarker>`)
 
 Six methods (five required + an optional lifecycle hook) — the only seam between the coordinator and storage:
@@ -30,7 +32,7 @@ The single design choice that keeps the seam clean: the crash-repair "where is t
 
 ## Testing
 
-The shared `runPersistenceContract` (public-API contract) keeps running for every backend. A new `runCoordinatorContract` (`tests/coordinator-contract.ts`) holds the write-path orchestration — adoption, HMR, collision, dispose-drain, crash-tail repair — and runs once per backend through a `CoordinatorFixture` (an in-memory reference + jsonl + sqlite). The per-backend specs shrank to storage mechanics only (JSONL: path safety, fsync rollback, bucket listing; SQLite: schema version, `scanRows`, transaction rollback). A through-coordinator torn-tail→load→`commitRepair` test per real backend (via a `corruptTail` fixture hook) keeps the coordinator's torn-marker repair branch covered under the 100% per-file gate — the contract crash test only produces synthetic closers, never a torn marker, so it could not reach that branch.
+The shared `runPersistenceContract` (public-API contract) keeps running for every backend. `runCoordinatorContract` (`tests/coordinator-contract.ts`) holds the write-path orchestration — adoption, HMR, collision, session and backend disposal drains, and crash-tail repair — and runs once per backend through a `CoordinatorFixture` (an in-memory reference + jsonl + sqlite). Coordinator-specific tests pin retirement map cleanup, same-id chain-tail races, failed-drain retry, and close ordering. The per-backend specs retain storage mechanics only (JSONL: path safety, fsync rollback, bucket listing; SQLite: schema version, `scanRows`, transaction rollback). A through-coordinator torn-tail→load→`commitRepair` test per real backend (via a `corruptTail` fixture hook) keeps the coordinator's torn-marker repair branch covered under the 100% per-file gate — the contract crash test only produces synthetic closers, never a torn marker, so it could not reach that branch.
 
 ## Alternatives considered
 
@@ -39,4 +41,4 @@ The shared `runPersistenceContract` (public-API contract) keeps running for ever
 
 ## Consequences
 
-The coordinator adds one indirection and an opaque torn marker, but centralizes correctness-heavy orchestration previously duplicated by every backend. Its hook surface stays narrow: collision checks reuse `loadStored`, materialization stays atomic inside `appendBatch`, and listing bypasses the coordinator. New backends implement storage primitives rather than copy the event-buffer-flush lifecycle.
+The coordinator adds one indirection, an opaque torn marker, and detached session-retirement tasks, but centralizes correctness-heavy orchestration previously duplicated by every backend. Session disposal remains an observe-only event, so the session owner does not await persistence retirement; the coordinator contains failures, preserves uncommitted buffers, and makes backend teardown the quiescence boundary. Its hook surface stays narrow: collision checks reuse `loadStored`, materialization stays atomic inside `appendBatch`, and listing bypasses the coordinator. New backends implement storage primitives rather than copy the event-buffer-flush lifecycle.
diff --git a/docs/rfc/implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md b/docs/rfc/implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md
new file mode 100644
index 0000000000..c99eb9f0fe
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md
@@ -0,0 +1,128 @@
+# RFC: The background task runtime (`ctx.tasks`) and generic task control tools
+
+Status: implemented
+
+## Problem
+
+Background bash originally combined two responsibilities: the bash executor ran processes and also managed task ids, ownership, incremental reads, cancellation, completion listeners, and model-facing control tools. Adding background subagents required the same lifecycle and interaction contract. Implementing that contract independently for every long-running capability would duplicate isolation, cleanup, notification, and prompt behavior while teaching the model a different collect-and-stop protocol for each producer.
+
+The task registry, control tools, and completion notices form one harness capability. Bash and subagents should supply execution-specific hooks without owning generic task behavior.
+
+## Decision
+
+The `tasks/` package group owns background-task semantics:
+
+- `@deepseek-ai/dsh-tasks` registers running work as `ctx.tasks` and owns task ids, authorization, snapshots, reads, cancellation, waiting, completion listeners, and cleanup.
+- `@deepseek-ai/dsh-tool-tasks` exposes `task_output`, `task_list`, and `task_kill`, injects completion notices, and supplies the background-task system-prompt guidance.
+
+Long-running tools are producers. `dsh-tool-bash` adapts a `BashProcess` into incremental output and process cancellation; `dsh-tool-subagent` adapts a child run into final output and child disposal. The execution seams remain independent of sessions and the task registry.
+
+`TaskService` is a concrete, process-local service. TODO(task-service-backend): separate its public contract from the implementation when a second backend defines the required lifecycle; a systemd-backed runtime is one plausible driver, but this PR does not speculate about its durability, reconnect, ownership, or observation semantics.
+
+## Runtime contract
+
+The literal types live in the [task data-structure catalog](../../../core-data-structures/tasks.md). A producer calls `ctx.tasks.start()` with a kind, label, optional owning `Agent`, and a `run()` function. The runtime completes all failable preflight work before calling `run()` and invokes it once. After `run()` returns hooks, registration commits without another failable step; a producer cannot start work that lacks a collectable task id.
+
+The producer hooks define three responsibilities:
+
+- `cancel(reason?)` synchronously requests termination, is idempotent, and must cause `done` to settle.
+- `done` never rejects and settles only after the producer has released the task's resources.
+- Optional `readOutput()` returns the next consuming output delta. Omitting it declares a final-output task whose terminal result comes from `TaskOutcome.output`.
+
+Statuses are `running`, `stopping`, `completed`, `killed`, and `failed`. Producer-specific information such as an exit code or stop reason belongs in `detail`; the registry does not interpret it. Task kinds form a merge-extensible string union, and task ids are branded and generated as `<kind>-N`, with a counter per kind.
+
+The runtime attaches one continuation to `done`, records the first terminal outcome, resolves waiters, and invokes completion listeners with per-listener error containment. First-wins settlement matters during teardown: if `cancel` throws, the runtime force-fails the record and warns that work may be orphaned rather than waiting forever for a promise that may never settle. A later producer outcome cannot overwrite that diagnosis or notify twice. A `cancel` that returns without eventually settling `done` still blocks teardown because the runtime cannot distinguish it from a slow, valid stop.
+
+Task registrations are not effects of the producer tool fiber. Reloading a tool or control-surface plugin therefore does not kill work owned by an agent and backend. The task service's own disposal cancels all live tasks and awaits contract-compliant producers.
+
+## Authorization and owner lifecycle
+
+Task ids are runtime-global and predictable, so every access is authorized by the registry. `get`, `read`, `wait`, and `kill` accept the calling `Agent`; `list` returns only tasks visible to that caller. An owned task is accessible only to the exact owning session. Unowned tasks are open to non-agent callers and die with the task service.
+
+The snapshot stores the owner's branded `SessionId` for authorization, while lifecycle operations retain the exact live `Agent` instance. These identities serve different purposes: session equality grants access, but exact object identity selects cleanup and completion delivery. Reusing an agent or session id cannot redirect an old scope's cleanup or notices to a replacement.
+
+The first task for an owner attaches one asynchronous effect to `owner.ctx`. Agent-scope disposal cancels that owner's live tasks, awaits their terminal records, and removes their snapshots. This effect survives producer reloads and joins the agent's existing quiescence boundary. The task service retains the effect disposer so service reload can detach callbacks from still-live agent scopes after global teardown.
+
+For contract-compliant producers, `AgentHandle.dispose()` resolves only after owned background work has stopped. Work intended to outlive an agent must be started unowned; survival across runtime restarts requires a separate durable-job design.
+
+## Service surface
+
+`TaskService` provides:
+
+- `start(spec)` for preflighted, atomic registration.
+- `get(id, caller?)` and `list(caller?)` for non-consuming snapshots.
+- `read(id, caller?)` for a consuming stream delta or an idempotent final result.
+- `kill(id, caller?, reason?)` for cancellation.
+- `wait(id, timeoutMs, caller?, signal?)` for bounded terminal waiting.
+- `onTaskDone(listener)` for effect-scoped observation with exact-owner delivery and listener containment.
+- `attachSurface(name)` for the control-surface availability fence.
+
+`wait` returns the terminal snapshot when the task settles or the live snapshot when its timeout expires. Aborting a wait cancels only that wait. If settlement has already assigned terminal delivery to the waiter, the terminal snapshot still wins. Waiters unregister synchronously on abort so a same-tick settlement cannot suppress a completion notice on behalf of a reader that receives nothing.
+
+A producer loaded without any control surface would let callers start work they cannot collect or stop. `dsh-tool-tasks` therefore calls `attachSurface()` for its lifetime, and `start()` fails before producer execution when no surface is attached. This check occurs at start rather than plugin load because sibling plugins may activate concurrently. Custom non-model surfaces can attach themselves without teaching the registry tool names.
+
+## Model-facing control surface
+
+`dsh-tool-tasks` registers three kind-independent tools with generic ACP cards:
+
+- `task_output(task_id, wait?, timeout_ms?)` reads output and always appends `[status: ...]`. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Reads are non-blocking unless `wait: true`, whose timeout is defaulted and capped by plugin config. A wait timeout reports the still-running status and does not stop the task.
+- `task_list()` returns caller-visible tasks as `<id> [<kind>] <status> — <label>`, or `(no background tasks)`.
+- `task_kill(task_id, reason?)` requests cancellation immediately. The optional logged reason is forwarded to the producer. Terminal tasks report their existing status; a throwing producer cancel fails the call and leaves the task running.
+
+Stream reads share one task-scoped consuming cursor because the owning model is the intended reader. A UI or multiple independent readers need a separate non-consuming observation API; sharing this cursor would let readers consume one another's output.
+
+The system prompt tells the model to retain task ids, continue independent work instead of busy-polling or duplicating a running task, collect relevant tasks before its final answer, and kill work that no longer matters. Completion injects a logged `context/message` into the exact owner's session; it becomes durable context for the next request but does not wake an idle agent.
+
+The runtime marks a terminal task `reported` when a read or wait delivers it, when a live waiter has claimed delivery at settlement, or when the model explicitly kills it. Reported tasks do not inject redundant completion notices. Listener failures are logged independently, do not stop later listeners, and are not awaited by waiters or teardown.
+
+## Producer opt-in
+
+Each producer owns whether its schema exposes `run_in_background` through defaulted config. `dsh-tool-bash` and each `dsh-tool-subagent` instance use `enableRunInBackground`, defaulting to true. A disabled instance omits the parameter and also rejects a forced background argument at execution because the generic argument validator permits undeclared keys. Schema omission advertises the capability; the execution check enforces it.
+
+`ctx.tasks` does not rewrite producer schemas. A bundle forwards configuration only for producers it owns. If a background call reaches `start()` without an attached surface, the runtime fence fails before execution.
+
+## Producer integrations
+
+The bash seam exposes `resolve`, `run`, and `start`. `start(spec)` returns a `BashProcess` with incremental reads, cancellation, exit facts, and a non-rejecting quiescence promise. The local executor retains live handles only so its own disposal can kill and join processes. Foreground callers continue to use `resolve` and `run` directly.
+
+For background bash, `dsh-tool-bash` registers the calling agent as owner. Its hooks map `kill()` to cancellation, `done` to a completed or killed `TaskOutcome`, and `readOutput()` to the process's bounded incremental output plus spill and sandbox notices. Generic task tools own ids, status lines, listing, waiting, and completion notices.
+
+For background subagents, `dsh-tool-subagent` creates a task-owned `AbortController` and begins provider startup inside the task starter. Cancellation aborts the same signal before or after provider readiness. `done` awaits both the child result and child disposal, maps completed output to a final result, maps abort to `killed`, and maps other stop reasons or infrastructure failures to `failed`. Intermediate child history remains in the child session and is not exposed through `readOutput()`.
+
+## Alternatives considered
+
+### Per-capability control tools
+
+Separate bash and subagent output/stop tools duplicate ids, isolation, cleanup, notification, and guidance while increasing the model's schema and protocol burden. One runtime keeps execution-specific behavior in producers without cloning the task lifecycle.
+
+### An immediate abstract task-runtime backend
+
+The current `TaskStart.run()` contract passes in-process callbacks and exact `Agent` objects. A durable backend changes identity, restart, ownership, and observation semantics, so extracting an interface before a second implementation exists would freeze the wrong boundary.
+
+### Consumer-owned authorization or cleanup events
+
+Consumer-owned checks invite inconsistent or missing isolation on each new surface. A broadcast cleanup event makes every listener filter every agent and provides no registration disposer. Central authorization plus one owner-scoped effect gives every consumer the same fence and an awaited, removable lifecycle hook.
+
+### Blocking output or a separate wait tool
+
+Blocking by default would serialize the parent while background work runs. Waiting without reading would add another model call and schema without returning useful information. `task_output(wait: true)` makes blocking explicit and combines it with result delivery.
+
+The wait uses the shared deadline primitives but not the generic tool-timeout policy. A wait timeout is a successful observation that returns `[status: running]`; the generic policy would replace it with a timeout error. No tool-call timeout controls task lifetime after a task id has been returned.
+
+### Runtime-owned output sinks
+
+A push sink would centralize buffering, but bash already owns bounded buffers, truncation, and spill files behind its executor seam. Pulling formatted deltas preserves that ownership. A durable backend that owns storage may justify revisiting the producer interface.
+
+### Random ids, promotion, or lifecycle session events
+
+Authorization, not unguessability, is the access boundary, and ids do not derive filesystem paths; sequential branded ids keep transcripts readable. Foreground-to-background promotion requires a user interaction contract the SDK does not prescribe. Starts, reads, and notices are already logged as tool and context events, so dedicated task session events would duplicate model-visible facts.
+
+## Testing
+
+Unit coverage pins preflight atomicity, per-kind ids, stream and final reads, wait timeout and abort races, cancellation, first-wins settlement, listener containment, notice suppression, owner isolation, stale owner instances, owner cleanup, service teardown, and the no-surface fence. Producer tests cover bash process mapping, subagent startup cancellation, terminal mapping, and disposal. Snapshot coverage pins the control-tool schemas and prompt guidance.
+
+## Consequences
+
+Bash commands and subagents share one id vocabulary, listing, notice format, prompt habit, and set of control tools. New long-running producers implement execution hooks instead of another registry and tool family. The [tool cookbook](../../../cookbook/adding-a-tool.md) points producers to this contract.
+
+Owned background bash now stops with its agent instead of surviving it. Background processes have no executor timeout; callers must kill irrelevant work or rely on owner/service disposal. Stream reads support one consuming reader, completion notices do not wake idle agents, and a producer that returns from `cancel` without settling `done` can still stall teardown. Durable jobs, independent observation cursors, and foreground promotion remain separate designs.
diff --git a/docs/rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md b/docs/rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md
index de6428fb64..bc91d425b6 100644
--- a/docs/rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md
+++ b/docs/rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md
@@ -14,9 +14,9 @@ Add `stdin?: string` and `env?: Record<string, string>` to **both** `BashExecReq
 
 Three deliberate choices:
 
-1. **The model-facing tool omits `stdin` and `env`.** Shell syntax already covers those needs, so duplicate parameters would add surface without authority separation. The tool builds requests only from declared model arguments, signal, and owner; trusted in-process callers may set the seam fields directly.
+1. **The model-facing tool omits `stdin` and `env`.** Shell syntax already covers those needs, so duplicate parameters would add surface without authority separation. The tool builds requests only from declared model arguments, signal, and owner; trusted in-process callers may set the seam fields directly. Harness-owned variables use the separate `dshEnv` channel from the [managed environment decision](../feature/2026-07-10-agent-session-identity-and-log-location.md), so ordinary `env` cannot replace them.
 
-2. **`env` merges AFTER the credential scrub, so an explicit caller entry always wins** — even a credential-shaped name. This is correct because the scrub's job is narrow: stop the harness's *ambient* `process.env` credentials from leaking into a spawned command. A caller that explicitly sets a var has named a value it already holds (not the ambient secret), so the scrub is not a constraint on it. `childEnv(extra?)` layers `scrub(process.env)` → `ENV_OVERRIDES` (the model-friendly `TERM=dumb` etc.) → `extra`, last-wins.
+2. **`env` merges AFTER the credential scrub, so an explicit caller entry wins even on a credential-shaped name.** The later managed-namespace decision reserves `DSH_*`: ambient entries are removed, ordinary `env` cannot set them, and trusted `dshEnv` merges last. The complete order is `scrub(process.env, including DSH_*)` → `ENV_OVERRIDES` → ordinary `env` → `dshEnv`.
 
 3. **`stdin`/`env` are required-absent-OK (plain optional) on the resolved spec, NOT required-but-nullable like `owner`.** `owner` is required-but-nullable because a *silently* missing owner yields an unowned, cross-session-readable task — a security footgun that a visible `undefined` guards against. `stdin`/`env` have no such hazard: a missing one means "no stdin / no extra env", which is the safe, ordinary case (every model-driven call). So they stay plain optionals, matching `signal`.
 
diff --git a/docs/rfc/implemented/architecture/2026-07-02-fs-per-session-cwd.md b/docs/rfc/implemented/architecture/2026-07-02-fs-per-session-cwd.md
index fad851265d..f60607df4d 100644
--- a/docs/rfc/implemented/architecture/2026-07-02-fs-per-session-cwd.md
+++ b/docs/rfc/implemented/architecture/2026-07-02-fs-per-session-cwd.md
@@ -12,7 +12,7 @@ Filesystem resolution used one plugin-load cwd while bash used the session proje
 
 Thread the caller's session cwd into path resolution, exactly as `dsh-tool-bash` already does for `workdir`. The **caller** (the tool) supplies the cwd; the provider does not read a session or agent.
 
-- `FileSystem.resolve` widens to `resolve(path: string, opts?: { cwd?: string }): Promise<FsTarget>`. `opts.cwd` is the base a RELATIVE `path` resolves against; an absolute `path` ignores it; omitting `opts.cwd` uses the backend's own default. An options object (not a positional `cwd?`) leaves room for future resolution hints without another signature change.
+- `FileSystem.resolve` accepts `resolve(path: string, opts?: { cwd?: string; signal?: AbortSignal }): Promise<FsTarget>`. `opts.cwd` is the base a RELATIVE `path` resolves against; an absolute `path` ignores it; omitting `opts.cwd` uses the backend's own default. `opts.signal` cancels resolution when the backend performs I/O. The options object keeps both caller-owned resolution controls together without positional growth.
 - `dsh-fs-local.resolve` uses `resolveLocalTarget(opts?.cwd ?? this.config.cwd, path)`. `config.cwd` stays the default for a caller that supplies none (non-ACP / no-session use, and the single-session stdio demo where `process.cwd()` IS the workspace).
 - `dsh-tool-fs`'s `read`/`write`/`edit` derive the session cwd through a shared `sessionCwd(exec)` helper (`exec.agent?.session.header.cwd`, mirroring bash's `resolveWorkdir`) and pass it to `resolve`. A non-agent / headerless caller yields `undefined`, so the backend applies its default.
 
diff --git a/docs/rfc/implemented/architecture/2026-07-02-tool-render-intent-union.md b/docs/rfc/implemented/architecture/2026-07-02-tool-render-intent-union.md
index cdaf152d73..37507c0c55 100644
--- a/docs/rfc/implemented/architecture/2026-07-02-tool-render-intent-union.md
+++ b/docs/rfc/implemented/architecture/2026-07-02-tool-render-intent-union.md
@@ -43,7 +43,7 @@ interface TerminalResultView { card: 'terminal'; title?: string; output?: string
 ### Producer mapping
 
 - `dsh-tool-fs` read → `generic` (`kind:'read'`, a follow-along `location`); write → `diff` (`oldText:null`); edit → `diff` (`oldText:old_string || null`, `newText:new_string ?? ''`). This mirrors `claude-agent-acp`'s `toolInfoFromToolUse` Read/Write/Edit arms field-for-field.
-- `dsh-tool-bash` foreground → `terminal` call + `terminal` result; `run_in_background` and `bash_output`/`bash_kill` → `generic`.
+- `dsh-tool-bash` foreground → `terminal` call + `terminal` result; `run_in_background` → `generic`. The generic `task_*` controls own their own generic cards.
 - `dsh-tool-todo` → `generic`.
 
 ### Terminal fallback ownership
diff --git a/docs/rfc/implemented/architecture/2026-07-05-reconstructable-requests.md b/docs/rfc/implemented/architecture/2026-07-05-reconstructable-requests.md
index d95a709ada..a084ba47a4 100644
--- a/docs/rfc/implemented/architecture/2026-07-05-reconstructable-requests.md
+++ b/docs/rfc/implemented/architecture/2026-07-05-reconstructable-requests.md
@@ -18,13 +18,13 @@ Prefix-cache stability is corollary #1, not the headline: an append-only log pro
 
 ### The mechanism
 
-**Messages.** `Session.deriveMessages()` is cached: each surface node is projected exactly once, when first seen, through the public per-node function `deriveEventMessage(event)`; a surface rewrite (a compaction `replace` — `SurfaceManager.replaceGeneration`) rebuilds. Callers get a fresh array per call over shared, deep-frozen messages: mutating logged history through a projection is unrepresentable (it throws), replacing the old clone-per-call isolation. External reconstructors fold the same public function over a log prefix, so no two paths can disagree.
+**Messages.** `Session.deriveMessages()` is cached: each surface entry is projected exactly once, when first seen, through the public per-event function `deriveEventMessage(event)`; a surface rewrite (a compaction `replace` — `SurfaceManager.replaceGeneration`) rebuilds. Callers get a fresh array per call over shared, deep-frozen messages: mutating logged history through a projection is unrepresentable (it throws), replacing the old clone-per-call isolation. External reconstructors fold the same public function over a log prefix, so no two paths can disagree.
 
-`EpochHeader` records the request's non-history state: call config, rendered system prompt, tool schemas, and session prefix, with empty values canonicalized to absence. `request/header` writes a full initial, resume, or fallback snapshot. `request/header-delta` encodes system changes by common-prefix/suffix line trim, tools by name-keyed additions/removals/changes, and config or prefix by full replacement. `foldRequestHeader`, `diffHeader`, and `applyHeaderDelta` are the pure codec. Each loop instance writes a snapshot on its first request to anchor process boundaries. Deltas are only an optimization: the writer verifies round-trip equality and falls back to a full snapshot for unrepresentable changes such as pure tool reordering.
+`EpochHeader` records the request's non-history state: call config, rendered system prompt, tool schemas, and session prefix, with empty values canonicalized to absence. `request/header` always writes a full snapshot: the first loop instance uses reason `initial`, later instances use `resume`, and an in-instance change uses `change`. `foldRequestHeader` selects the latest snapshot. Legacy `request/header-delta` events and the removed `fallback` reason are rejected when appended or loaded.
 
 Each step rebuilds prompt assembly. On the instance's first step, `agent/session-prefix` extends a frozen empty seed with request-only opener messages; the result is frozen and cached for that loop instance. `agent/pre-step` then receives the composed prefix before messages are snapshotted immediately ahead of `step/start`. The first call config starts from explicit `AgentOptions`, preserving fork overrides and resume reconfiguration; later calls start from the folded header. `agent/request` may replace only that frozen config seed, while model-visible content enters through logged channels. The loop records the owed header event—the prefix's only durable home—builds `GenerateOptions` from prefix, snapshot, and header, and deep-freezes it while leaving `AbortSignal` live. Per-instance state is only the cached prefix and whether its anchoring snapshot has been written.
 
-**`step/start` is the reconstruction boundary.** A step derives messages from events before that sequence. Injection after the snapshot joins the next request, and reentrant appends are rejected during event publication. `agent/pre-step` is the seam for content needed by the current request. Header reconstruction folds through the step's own `request/header*` event, or carries the prior fold when no new header is written.
+**`step/start` is the reconstruction boundary.** A step derives messages from events before that sequence. Injection after the snapshot joins the next request, and reentrant appends are rejected during event publication. `agent/pre-step` is the seam for content needed by the current request. Header reconstruction selects the step's `request/header`, or carries the prior snapshot when no new header is written.
 
 **Enforcement.** In development, `dsh-invariants` independently rebuilds each loop request through a fresh `Session`, so the live cache cannot vouch for itself, then compares messages and folded header fields at `llm/stream`. Loop requests are identified by their frozen shape and session id; direct one-shots are excluded. Correctness depends on sequence-bounded reconstruction rather than listener order. A with-key e2e requires positive cache-read tokens after the first request; per-step usage is the production signal, and a header change or compaction appears as a cache-read drop on the next step.
 
@@ -39,15 +39,16 @@ Like MiniCode, the conversation advances append-only and resets only when model-
 - **Per-call request scalars** (a freely mutable config handed to each `agent/request` dispatch): a listener flips the model per call with zero accounting, silently abandoning the provider cache this design exists to protect. Config is per-conversation logged state; the waterfall proposes, the log records.
 - **Detect-and-report** (compare consecutive requests, warn on divergence): catches violations after the fact; a violating request is still constructible and ships. Rejected for interface-level unrepresentability.
 - **Event-driven assembly** (re-render only on change signals): a missed-signal bug class — a tool registered mid-session emits `tools/change`, not `system-prompt/change`, and a third-party provider may emit nothing. Per-step render + value compare is robust with zero signal discipline.
-- **Narrative fields on the header events** (a `reason`/`changed` list on deltas): derivable by diffing consecutive events — one home per fact; snapshots carry a reason because an anchor's cause is NOT derivable from the data.
+- **A custom header-delta codec** (system line edits, name-keyed tool edits, whole config/prefix replacements): reduced repeated bytes but duplicated the representation and its diff/apply/fallback machinery. Full snapshots retain one replay representation.
+- **Narrative changed-field lists on header snapshots**: derivable by comparing consecutive snapshots. The `reason` remains because an instance boundary is not derivable from the snapshot values.
 
 ## Consequences
 
 - A request that is not explained by the log cannot be constructed by accident — not by the loop, not by a listener; mutating a built request throws; every header change is a durable, diffable log event.
-- Choosing between the advisory channels is a change-frequency decision, and the design makes the stable one structural: an `agent/session-prefix` contribution is composed once per loop instance and reused verbatim, so it extends the cacheable prefix at zero marginal cost and CANNOT bust the provider cache mid-session; content that changes mid-session flows through the append-only history channels — `agent.inject()`, a `tools/post-execute` decision's `additionalContext`, prompt-submit `additionalContext` — each a durable `context/message` paid once and prefix-cached thereafter, at the price of accumulating in history and the log. Route session-frozen openers to the prefix and change notices to the history channels; a per-step request-only tail slot was deliberately dropped (no consumer, and a durable append covers every current update pattern).
-- What still costs full price at the provider is inherent and logged: compaction (its `compact/*` events and replace node), a real prompt/tool change (`request/header-delta`), a config switch (ditto), a process boundary with drift (`'resume'` snapshot differing from its predecessor). The provider's own reasoning-content exclusion is managed server-side.
+- Choosing between the advisory channels is a change-frequency decision, and the design makes the stable one structural: an `agent/session-prefix` contribution is composed once per loop instance and reused verbatim, so it extends the cacheable prefix at zero marginal cost and CANNOT bust the provider cache mid-session; content that changes mid-session flows through the append-only history channels — `agent.inject()` and tool/prompt-submit `additionalContexts` — each a durable `context/message` paid once and prefix-cached thereafter, at the price of accumulating in history and the log. Route session-frozen openers to the prefix and change notices to the history channels; a per-step request-only tail slot was deliberately dropped (no consumer, and a durable append covers every current update pattern).
+- What still costs full price at the provider is inherent and logged: compaction (its `compact/*` events and replacement entry), a real prompt, tool, or config change (`request/header` with reason `change`), or a process boundary with drift (a differing `resume` snapshot). The provider's own reasoning-content exclusion is managed server-side.
 - The `step/start`-listener behavior change (above) is the one observable semantics change for plugins; `agent/pre-step` is the current-request seam.
-- Tool-result trimming (planned) needs no new mechanism: a logged single-node surface replace (`start === end`) carrying a trimmed `tool/result` under the same `callId` — compaction-family, replay-correct, cache-bust batched by the same pressure logic.
-- Session logs grow one `request/header` snapshot per conversation (system + tool schemas: the dominant term), plus deltas on real changes — small next to `assistant/chunk` volume; `SESSION_FORMAT_VERSION` stays `0` (pre-release churn is absorbed, backends reject-not-migrate).
+- Tool-result trimming (planned) needs no new mechanism: a logged single-entry surface replace (`start === end`) carrying a trimmed `tool/result` under the same `callId` — compaction-family, replay-correct, cache-bust batched by the same pressure logic.
+- Session logs grow one `request/header` snapshot per loop instance plus snapshots on real changes. This is larger than a delta codec but small beside chunk-heavy logs and retains one replay representation. `SESSION_FORMAT_VERSION` stays `0`; legacy delta events are rejected rather than migrated.
 - Snapshot goldens changed once (every transcript gains its header events); the fs-writing fixtures are stored in the normalized authored form with cwd-relative tool arguments, because replay only round-trips cwd-independent argument paths.
 - FIXME(call-config-shape): revisit `LlmCallConfig`'s exact field set — which fields are genuinely epoch-level for cache purposes (`model` certainly; the sampling scalars sit there out of caution), and where provider-specific extras (reasoning options, extra body params) belong when an adapter needs them.
diff --git a/docs/rfc/implemented/architecture/2026-07-06-tool-result-retention-library.md b/docs/rfc/implemented/architecture/2026-07-06-tool-result-retention-library.md
new file mode 100644
index 0000000000..f35cf56e84
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-06-tool-result-retention-library.md
@@ -0,0 +1,155 @@
+# RFC: Tool result retention library
+
+Status: implemented
+
+## Problem
+
+Several model-facing tools already bound the amount of context they return, but each one owns a different local mechanism and vocabulary: bash keeps a tail plus spill files, web search caps source lists, web fetch caps body content, and `glob` / `grep` discovery needs an inline first page while keeping exact omission metadata for the full result set. A single `truncate(text)` helper cannot cover those cases: item tools need item counts and grouping outside the primitive, while text tools need byte budgets and UTF-8-safe head/tail cuts.
+
+The shared abstraction the tools need is **retention**, not generic collection. A caller feeds items or text chunks into a bounded object and later receives the retained content plus exact omission metadata. Tool-specific code still owns business semantics: file grouping, line numbering, exit codes, provider error states, spill files, and model-facing prose. The common library owns only the mechanical question "what did we keep, and what did we omit?"
+
+## Decision
+
+`@deepseek-ai/dsh-retention` lives under `packages/util/` (peer to `dsh-brand` and `dsh-timeout`) and owns bounded model-facing output. It is a library of pure classes and functions, **not** a Cordis service or plugin: it takes no `ctx`, registers nothing, holds no cross-call state, and emits no events. Tool packages import it directly when they need bounded output.
+
+The library has two independent retainers:
+
+- `ItemRetainer<T>` handles ordered logical units such as paths, grep matches, or search sources. It supports `head` retention only in v1, while keeping the retainer shape open to additional retention strategies later.
+- `TextRetainer` handles byte-oriented text streams such as bash stdout/stderr or web response bodies. It supports `head`, `tail`, and `headTail` retention while preserving UTF-8 boundaries at `finish()`.
+
+Both retainers return a small `PushDecision` after each `push()` so callers can tell whether that unit/chunk was fully retained and whether the accumulated result is now truncated. Omission counts are exact because callers keep feeding every observed item/chunk.
+
+```ts ignore-check
+/**
+ * How much content the retainer omitted.
+ *
+ * `unknown` is reserved for callers that omit without a count; the retainers
+ * themselves return `none` or `exact`.
+ */
+type Omitted =
+  | { kind: 'none' }
+  | { kind: 'exact'; count: number }
+  | { kind: 'unknown' }
+
+interface PushDecision {
+  kept: boolean
+  truncated: boolean
+}
+
+/**
+ * Final result for ordered logical units.
+ */
+interface RetainedItems<T> {
+  items: T[]
+  truncated: boolean
+  seen: number
+  kept: number
+  omitted: Omitted
+}
+
+/**
+ * Final result for text streams.
+ *
+ * The returned `text` is safe to send to a formatter; the retainer does not add
+ * tool-specific headers, exit markers, XML tags, or recovery instructions.
+ */
+interface RetainedText {
+  text: string
+  truncated: boolean
+  omittedBytes: Omitted
+}
+```
+
+### Strategies
+
+Item retention supports a head window. Text retention supports head, tail, and headTail byte windows.
+
+```ts ignore-check
+type ItemRetentionStrategy =
+  | {
+      /** Keep the first `maxItems` units. Use for `glob`, `grep`, and web sources. */
+      kind: 'head'
+      maxItems: number
+    }
+
+type TextRetentionStrategy =
+  | {
+      /** Keep the first `maxBytes` bytes. */
+      kind: 'head'
+      maxBytes: number
+    }
+  | {
+      /** Keep the final `maxBytes` bytes. Requires reading to the end. */
+      kind: 'tail'
+      maxBytes: number
+    }
+  | {
+      /** Keep a stable prefix and suffix, omitting the middle. Requires reading to the end. */
+      kind: 'headTail'
+      headBytes: number
+      tailBytes: number
+    }
+```
+
+### Tool mapping
+
+`read` is intentionally outside the v1 retention library. Its `read-render` helper owns a file-specific pagination contract: `offset` / `limit`, line numbers, `totalLines`, offset-out-of-range errors, per-line preview truncation, and a selected-output byte cap that can stop scanning mid-window. That is a line-window renderer, not a generic retention primitive. It may share future neutral notice helpers, but it should not pass its already-selected window through `ItemRetainer`.
+
+`FsGlobEntry` and `FlatGrepMatch` below are the intended discovery-tool item shapes, not existing retention-library exports. `FsGlobEntry` is one backend-derived path, and `FlatGrepMatch` is one ungrouped grep match before the backend groups retained matches by file.
+
+`glob` uses `ItemRetainer<FsGlobEntry>` with `{ kind: 'head', maxItems: globMaxResults }` after collecting the full sorted path list. The tool keeps the retained first page inline and may save the full list through the spill seam. Path mapping, skipped candidates, and `incomplete` stay outside the retainer.
+
+`grep` uses `ItemRetainer<FlatGrepMatch>` with `{ kind: 'head', maxItems: grepMaxMatches }` before grouping. The executor parses ripgrep output, maps paths, applies per-line preview truncation, and pushes flat matches. After `finish()`, the tool groups retained matches by file and can save the full match list through the spill seam when the inline result is capped. Grouping is not part of the retainer because the cap is total matches, not files; per-match preview truncation and `incomplete` are also separate from result-level retention.
+
+`bash` can use `TextRetainer` with `tail` or `headTail` and reads to process completion. The bash executor still owns spill files, exit status, signal, timeout, and background-task behavior; the retention helper only replaces ad hoc in-memory head/tail accounting where that behavior is desired. Long-running task ownership remains orthogonal to the [generic long-running tool runtime](2026-06-20-generic-long-running-tool-runtime.md).
+
+`web_fetch` can use `TextRetainer` with `head` or `headTail`, or keep provider-owned body caps when the provider must read and decode internally. Either way, the fetch result's `truncated` remains a provider/tool fact, and the library only supplies retained text and omission metadata.
+
+`web_search` can use `ItemRetainer<WebSearchSource>` with `head`. Current providers often return an array, so this is post-hoc but still standardizes notices.
+
+### Notices
+
+The library exposes a neutral notice shape and a tiny formatter hook, but tools provide the user-facing words. A grep footer says "Narrow the pattern, path, or include"; a web fetch footer says "Fetch a more specific URL or section"; bash may point to a spill file. The retainer cannot know those recovery actions.
+
+```ts ignore-check
+interface RetentionNotice {
+  scope: string
+  strategy: 'head' | 'tail' | 'headTail'
+  unit: 'items' | 'bytes' | 'chars' | 'lines'
+  limit: number | { head: number; tail: number }
+  kept: number
+  omitted: Omitted
+}
+
+const formatGrepNotice = (notice: RetentionNotice): string =>
+  formatRetentionNotice(
+    notice,
+    ({ kept }) => `Results capped at ${kept}. Narrow the pattern, path, or include to see more.`,
+  )
+```
+
+The formatter hook is deliberately small: a tool turns a `RetentionNotice` into its own footer text. The helper may standardize omission wording, but it does not own recovery guidance.
+
+`truncated` means the retainer omitted otherwise-available content because of a budget. It does not mean the upstream was incomplete. Tools keep separate fields for permission failures, skipped binary files, provider partial failures, unreadable candidates, invalid UTF-8, and any other "could not inspect" condition.
+
+## Consequences
+
+**What shipped.** `@deepseek-ai/dsh-retention` exports `ItemRetainer`, `TextRetainer`, the result types (`RetainedItems`, `RetainedText`), the strategy types (`ItemRetentionStrategy`, `TextRetentionStrategy`), `Omitted`, `PushDecision`, `RetentionNotice`, and the neutral notice helpers `describeOmitted` / `formatRetentionNotice` — with no dependency on Cordis or any tool package. Unit tests cover item-head retention with exact omission counts, text-head retention, text-tail retention, head-tail byte retention, zero budgets, UTF-8 boundary handling (2-, 3-, and 4-byte codepoints and invalid lead bytes at each cut), and unknown omission wording.
+
+**What is documented but not yet migrated.** `glob`, `grep`, `bash`, `web_fetch`, and `web_search` have their mappings documented in the [package README](../../../../packages/util/retention/README.md), but not every tool has been migrated onto the library in this change; migration is deliberately separate follow-up work. `read` is documented as intentionally out of scope: its `read-render` line-window contract (`offset`/`limit`, `totalLines`, offset-range errors, per-line preview truncation, a byte cap over the selected window) is not generic retention, and one `Omitted` count cannot represent both sides of a line window.
+
+**Boundaries the library holds.** `truncated` means the retainer omitted otherwise-available content because of a budget; it never means the upstream was incomplete. Tool-specific states — `incomplete`, permission failures, provider partial failures, binary skips, bash spill-path recovery, invalid UTF-8 — stay in tool-domain fields, outside the retainer. When a future change migrates a tool, that package's README and tests must prove the model-facing result text is unchanged except for deliberate notice wording.
+
+**Tradeoffs accepted.** The v1 surface deliberately supports only item `head` retention and text `head` / `tail` / `headTail`; windows, grouped budgets, sort-aware caps, and upstream-stop control wait until a second consumer proves the need. Text retention counts bytes for process/body safety, leaving character- and line-level preview budgets as separate tool-owned concerns.
+
+## Alternatives considered
+
+**Post-hoc `truncate(text)` only.** Rejected: it matches Codex's history/tool-output truncation use case but loses item counts, grouping boundaries, UTF-8-safe byte windows, and exact omission metadata.
+
+**One generic `Collector<T>` with pluggable callbacks.** Rejected for v1: it hides the two important resource modes. Logical item retention counts items; text retention counts bytes and preserves UTF-8 boundaries. Separate `ItemRetainer` and `TextRetainer` names make that difference explicit while keeping the API small.
+
+**Put `read` windowing behind `ItemRetainer`.** Rejected for v1: `read` is the only current window consumer, and its semantics are file pagination rather than generic retention. A single `Omitted` count cannot represent both sides of a line window, and `read` also carries `totalLines`, offset-range errors, per-line preview truncation, and a byte cap over selected output. Keeping `read-render` tool-owned avoids growing the shared library around one special case.
+
+**Make truncation part of `ToolExecutionResult`.** Rejected: the tool registry would have to understand tool-specific recovery guidance, grouping, line numbering, exit status, and provider semantics. Retention is a library used before a tool returns `ContentBlock[]`; the model-facing result remains tool-owned.
+
+**Expose limits in every model-facing tool schema.** Rejected as the default: Claude Code's grep exposes `head_limit` / `offset`, but this harness keeps routine budgets as deployment config unless the model genuinely needs pagination control. A future read-like continuation field can be added per tool; it does not belong in the shared retention primitive.
diff --git a/docs/rfc/implemented/architecture/2026-07-07-tool-call-timeout-policy.md b/docs/rfc/implemented/architecture/2026-07-07-tool-call-timeout-policy.md
index 7bd4e2462d..1e516a9aa0 100644
--- a/docs/rfc/implemented/architecture/2026-07-07-tool-call-timeout-policy.md
+++ b/docs/rfc/implemented/architecture/2026-07-07-tool-call-timeout-policy.md
@@ -78,7 +78,7 @@ No new session event is needed for reconstructability: `TOOL_TIMEOUT` is the fin
 
 `bash` stays on the current backend timeout path. `dsh-tool-bash` continues to expose `timeoutMs` and `run_in_background`; `dsh-bash-local` continues to use `@deepseek-ai/dsh-timeout` for `BASH_TIMEOUT`; hook bridges continue to call `runHook()` and pass `timeoutMs` through `ctx.bash`. This keeps foreground/background/hook behavior stable.
 
-`read`, `write`, `edit`, `todo_write`, `bash_output`, and `bash_kill` do not opt into tool-call timeout: they are local filesystem or short registry/session operations where a deadline would be best-effort only or unnecessary.
+`read`, `write`, `edit`, `todo_write`, `task_list`, and `task_kill` do not opt into tool-call timeout. `task_output` owns its bounded wait because a wait timeout is a successful live-status result, not a tool failure.
 
 A future model-facing grep/glob tool can be implemented on top of `ctx.bash` without importing `@deepseek-ai/dsh-timeout`: it forwards `exec.signal` to `ctx.bash`, and declares its own `timeoutMs` (from its plugin's config) for the enforcer to apply. If bash-local's backend timeout becomes a problem for such a tool, the bash seam can later add a caller-owned-deadline mode; that is outside this cut.
 
diff --git a/docs/rfc/implemented/architecture/2026-07-08-tool-output-spill-files.md b/docs/rfc/implemented/architecture/2026-07-08-tool-output-spill-files.md
new file mode 100644
index 0000000000..3a60c5c223
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-08-tool-output-spill-files.md
@@ -0,0 +1,189 @@
+# RFC: Tool output spill policy
+
+Status: implemented
+
+## Problem
+
+Tool outputs need bounded model-facing previews, but some oversized results are still useful later. A fetched page body or a verbose tool response should not consume the next model request in full, but the model should be able to inspect the complete formatted result later with existing file-reading tools.
+
+Before this change the behavior was uneven. `dsh-bash-local` already writes complete stdout/stderr streams to private temp spill files when its in-memory tail overflows, but ordinary text tool results were returned inline unless the tool hand-rolled its own cap. The [tool result retention library](./2026-07-06-tool-result-retention-library.md) owns preview mechanics, but it does not own storage or an execution-pipeline policy that applies those mechanics to final tool results.
+
+The shape matches the timeout policy design: a tool author normally returns the text result, and a policy plugin enforces the deployment's default context budget. Tool-specific early spill remains possible later for outputs that do not survive to the final `ToolExecutionResult`; the first cut proves the default final-result path.
+
+## Decision
+
+A thin spill storage seam plus a default spill policy plugin, in a new `packages/spill/` group:
+
+| Package | Role |
+|---|---|
+| `@deepseek-ai/dsh-spill` | Interface: `ctx.spillStore`, vocabulary types, no storage implementation. |
+| `@deepseek-ai/dsh-spill-local` | Local backend: private, session-scoped file storage on the host filesystem. |
+| `@deepseek-ai/dsh-spill-policy` | Tool-result policy plugin: wraps final text results after dispatch and replaces oversized results with a retained preview plus a spill locator. |
+
+There is no dedicated model-facing consumer package. The consumer is the existing `ctx.tools` execution pipeline: `dsh-spill-policy` consumes final tool results through the `tools/post-execute` waterfall, and the model follows the backend-supplied retrieval hint for the returned locator.
+
+### Spill seam
+
+The storage seam is minimal: save text and return a locator plus retrieval hint.
+
+```ts ignore-check
+interface SpillStore {
+  saveText(input: SaveTextSpill): Promise<SpillRef>
+}
+
+interface SpillSource {
+  toolName: string
+  callId: CallId
+  label: string
+}
+
+interface SaveTextSpill {
+  owner: { sessionId: SessionId }
+  source: SpillSource
+  suggestedName: string
+  content: string
+}
+
+type SpillLocator = Branded<'SpillLocator'>
+
+interface SpillRef {
+  locator: SpillLocator
+  bytes: number
+  retrievalHint: string
+}
+```
+
+`SpillLocator` is a [branded](../../../../packages/util/brand) model-facing handle returned by the backend. The local backend renders it as a filesystem path; a remote or database backend can render a URI, key, or command token. Consumers treat it as opaque and render it with `retrievalHint` instead of assuming `read` is always the right retrieval mechanism. `SpillOwner.sessionId` is the save-time storage namespace: forked sessions inherit existing spill locators from the seeded log without copying or re-owning them, and new spills after the fork use the child session id. A retention-period cleanup may expire old locators with other old session artifacts; the spill seam does not define a per-session cleanup policy.
+
+`dsh-spill-local` owns only storage details: session-scoped directory selection, safe names, path-traversal protection, the write, and returning `{ locator, bytes, retrievalHint }`. It does not own retention policy, tool-result replacement, search, or file inspection. Files land at `<root>/session-<hash>/<random>-<safeName>`, where `root` is a configured path or a lazily-created private (0700) per-process temp dir, the session subdir is a short `sha256(sessionId)` prefix, and the leaf is a random hex prefix plus the caller's `suggestedName` sanitized to one path segment (mirrors the JSONL backend's `encodeSegment`). The write is `open(path, 'wx', 0o600)` — exclusive and owner-only, so a planted symlink cannot redirect it. The locator is the path, and the retrieval hint tells the model it can use `read` or `grep` on that path.
+
+### Spill policy
+
+`dsh-spill-policy` is a `tools/post-execute` result transformer with one configuration knob:
+
+```ts ignore-check
+interface Config {
+  /** Omitted means no automatic spill policy. Present means apply to oversized plain text tool results. */
+  maxInlineBytes?: number
+}
+```
+
+When `maxInlineBytes` is omitted the plugin registers nothing (a true no-op). When set, it applies a default policy to final plain-text tool results:
+
+1. Let the tool run normally, delegating via `next()` so a downstream listener settles the result first.
+2. Flatten the accepted final `ContentBlock[]` only when it is entirely plain text; a result with any non-text block is left untouched.
+3. If its UTF-8 byte size is at or below `maxInlineBytes`, leave it unchanged.
+4. If it is larger, call `ctx.spillStore.saveText()` with the full final text.
+5. Replace the model-facing result with a retained head/tail preview plus the spill reference.
+
+The preview is an implementation default owned by the policy: a head/tail split of `maxInlineBytes` via the retention library's `TextRetainer`. Future config can expose preview sizing only after a second deployment needs it.
+
+The replacement text is intentionally generic because the policy only knows the final formatted tool result, not the tool's internal resource:
+
+```text
+<retained preview>
+
+(Omitted N bytes. Full formatted result stored at: /.../session-.../....txt. Use read with offset/limit, or grep this path to search within it.)
+```
+
+If `ctx.spillStore.saveText()` fails (permissions, ENOSPC, backend unavailable), or the call has no session owner, or no backend is loaded, the plugin logs the reason and returns the original result unchanged. Spill failure never turns a successful tool call into an `isError` result or hides the inline result.
+
+The policy skips `read` to avoid a circular `read -> spill file -> read again` loop. Additional opt-out configuration is deferred until a real second tool needs it.
+
+## Showcase: web_fetch
+
+`web_fetch` is the first showcase because it returns a naturally large text result and needs no tool-specific spill code. The tool is ordinary:
+
+```ts ignore-check
+ctx.tools.register(defineTool({
+  name: 'web_fetch',
+  async execute(args, exec) {
+    const result = await ctx.web.fetch({ url: args.url }, exec.signal ? { signal: exec.signal } : undefined)
+    return [{ type: 'text', text: formatFetchOutput(result) }]
+  },
+}))
+```
+
+With `dsh-spill-policy` configured, a large formatted fetch result is automatically retained and spilled. A deployment demonstrates the behavior by setting the provider resource cap higher than the policy cap:
+
+```yaml
+- id: web-fetch-local
+  name: '@deepseek-ai/dsh-web-fetch-local'
+  config:
+    maxBodyChars: 500000
+
+- id: spill-local
+  name: '@deepseek-ai/dsh-spill-local'
+
+- id: spill-policy
+  name: '@deepseek-ai/dsh-spill-policy'
+  config:
+    maxInlineBytes: 50000
+```
+
+This separation is important. `web-fetch-local` still owns resource caps (`maxResponseBytes`, `maxBodyChars`) to protect network, memory, and decoding work. `spill-policy` owns only the model-facing context cap after the result already exists. If the provider already returned `truncated: true`, the spill file contains the full formatted result the tool returned, not the full original webpage; the policy does not claim otherwise.
+
+## Relationship to retention and early spill
+
+Retention is separate from spill storage:
+
+- `@deepseek-ai/dsh-retention` owns preview mechanics (`TextRetainer`, `ItemRetainer`, and omitted metadata).
+- `@deepseek-ai/dsh-spill` owns saving final text and returning a locator plus retrieval hint.
+- `@deepseek-ai/dsh-spill-policy` applies the default final-result policy in the tool pipeline, composing the two.
+
+The final-result policy cannot replace tool-owned early spill. Some useful content is not present in final `ToolExecutionResult.content`:
+
+- `bash` final output is already a tail plus a temp spill path; the complete stdout/stderr streams live in executor files.
+- `subagent` final output is the child final answer, not the child rollout.
+- Future tools may produce runtime artifacts that are never represented by their final `ToolExecutionResult.content`.
+
+Those cases can consume `ctx.spillStore` directly in later work. They are not part of the first showcase.
+
+## Non-goals
+
+- No new model-facing `artifact_read` or `artifact_search` tool in v1.
+- No per-tool retention configuration in v1.
+- No model-facing timeout/truncation arguments.
+- No migration of `read` output into spill files.
+- No replacement for provider/resource caps such as `web-fetch-local.maxBodyChars`.
+- No bash temp-file normalization or subagent rollout capture in the first cut.
+
+## Deferred
+
+- `saveFile()` / `linkOrCopy` for existing executor spill files, needed for bash normalization.
+- Tool-owned spill for subagent rollouts (`await run.result`, read in-process child session before `run.dispose()`, save JSONL).
+- Per-tool opt-out or per-tool policy declarations if the built-in `read` skip is insufficient.
+- Remote or database storage backends for ACP or remote environments where a local path is not meaningful.
+- Cleanup and retention policy for old spill files, likely tied to session cleanup.
+
+## Testing
+
+- `dsh-spill` unit tests pin the seam contract: registration as `ctx.spillStore`, one-implementation-per-context, and disposal release.
+- `dsh-spill-local` unit tests cover `saveText`, `encodeSegment` sanitization (separators/tilde/whole-segment dots/empty), the session-hash directory, owner-only permissions, distinct paths per save, the configured/private root, and a storage-failure rejection.
+- `dsh-spill-policy` unit tests drive real tools through `ctx.tools.execute`: disabled-mode no-op, oversized-text replacement, small/non-text passthrough, `read` skip, best-effort fallback (save failure / no backend / no owner), and downstream-composition (bounding a replaced result, preserving `additionalContexts`).
+- `dsh-tool-web` integration drives `web_fetch` through `ctx.tools.execute` with the real `spill-local` backend + policy, proving the model-facing text changes only by the deliberate spill notice while the spill file holds the full formatted result.
+- The `coding-agent` example loads `spill-local` + `spill-policy`, so its keyless Loader smoke exercises the real load path (the namespace-plugin export shape + `inject`).
+
+## Consequences
+
+The default policy only sees final formatted text. It cannot preserve provider-internal content that was already capped or runtime artifacts that were never part of the result. This is acceptable for the first cut because the showcase is final-result spill, not early spill; tool-owned early spill remains deferred work.
+
+Returning real paths from the local backend keeps v1 simple and matches proven agent-tool behavior, while the seam itself only promises an opaque locator plus retrieval hint so remote backends can return non-file locators.
+
+The local-backend value proposition depends on the existing `read`/`grep` tools being able to inspect the returned local path, even when the spill directory is outside the session cwd. That holds today because the filesystem policy records observations and write guards but does not confine reads to the workspace. A future workspace-confinement policy must either allow local spill paths explicitly or use a non-file spill backend whose retrieval hint points at a supported reader.
+
+**Snapshot gap.** No ACP snapshot scenario covers the transcript-visible `web_fetch` spill notice yet. The ACP snapshot harness replays keyless and cannot hit the live web, and a `web_fetch` spill requires a real over-cap HTTP body; a deterministic scenario would need a seeded loopback fetch target the replay tree does not currently wire (the examples do not load `tool-web` at all). The behavior is covered instead by the `dsh-tool-web` integration test against a loopback server. Closing the gap is follow-up work: wire `tool-web` + a seeded fetch target into the ACP example, then record a `web-fetch-spill` scenario.
+
+The policy can become too large if it starts owning tool-specific semantics. It stays narrow: plain-text final results only. Tool-owned early spill remains future work.
+
+## Alternatives considered
+
+**Require each tool to opt in with a retention declaration.** Rejected for v1: the goal is a default behavior similar to Claude Code's generic tool-result persistence. A single `maxInlineBytes` deployment knob is enough to prove the shape.
+
+**Make `tool-results` a broad tool-result platform.** Rejected: a broad package name invites retention policy, result replacement, preview wording, search, and early spill into one seam. The shared storage part is smaller: save text and return a locator plus retrieval hint.
+
+**Use `ctx.fs.writeText` or the model-facing `write` tool.** Rejected: workspace filesystem writes carry project-file semantics, write/edit policy, observation state, and user-facing side effects. Spill files are runtime artifacts, not model-authored workspace edits. The existing `read` tool may inspect them later, but creation belongs to the runtime spill seam.
+
+**Let `web-fetch-local` fetch without caps and rely on spill-policy.** Rejected: spill-policy runs after the final tool result exists and cannot protect network, memory, or decoding resources. Provider resource caps stay mandatory.
+
+**Merge retention into spill.** Rejected: retention and spill have different responsibilities. `TextRetainer`/`ItemRetainer` decide what preview is kept and what was omitted; spill storage only saves the final text the policy asks it to save.
diff --git a/docs/rfc/implemented/architecture/2026-07-14-provider-routed-llm-adapters.i18n.yaml b/docs/rfc/implemented/architecture/2026-07-14-provider-routed-llm-adapters.i18n.yaml
new file mode 100644
index 0000000000..48055ef6b7
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-14-provider-routed-llm-adapters.i18n.yaml
@@ -0,0 +1,6 @@
+# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
+# side as of the last confirmed-consistent state. Both languages carry equal authority;
+# after editing either side, bring the other along and re-record with:
+#   pnpm run verify-translation-pairing --write
+2026-07-14-provider-routed-llm-adapters.md: 75ef047a7f95621d9a9c018b6dc57439e6f2bb22
+2026-07-14-provider-routed-llm-adapters.zh.md: 75ac9adbe5f8e96930a72a977c1969ff3a119ee8
diff --git a/docs/rfc/implemented/architecture/2026-07-14-provider-routed-llm-adapters.md b/docs/rfc/implemented/architecture/2026-07-14-provider-routed-llm-adapters.md
new file mode 100644
index 0000000000..75ef047a7f
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-14-provider-routed-llm-adapters.md
@@ -0,0 +1,91 @@
+# RFC: Provider-routed LLM adapters and a generic pi-ai backend
+
+Status: implemented
+
+English | [中文](2026-07-14-provider-routed-llm-adapters.zh.md)
+
+## Problem
+
+`dsh-llm` registered adapters by exact model name. A plugin supplied a model list at Cordis startup, `LlmService` stored one adapter per listed string, and `GenerateOptions.model` selected the adapter and the provider model at once. This worked while both shipping adapters targeted the same two DeepSeek models, but it conflated two independent decisions: which upstream provider owns a request, and which model that provider should run.
+
+The conflation prevents a provider gateway from serving an open-ended model catalog. OpenRouter, for example, is one provider with many model ids, while a private OpenAI-compatible endpoint may add models without changing the Harness plugin tree. Every newly selected model currently needs to have been registered during plugin startup. The same model id can also exist at multiple providers, so model-only registration cannot state which provider the caller intended.
+
+`dsh-llm-pi-ai` exposed none of pi-ai's provider abstraction. It constructed an inline DeepSeek `openai-completions` model, applied DeepSeek-specific payload patches, and stamped every replayed assistant message as DeepSeek. pi-ai itself has a provider/model catalog, selects APIs such as `openai-responses`, `anthropic-messages`, and `google-generative-ai`, and preserves provider-specific response ids and reasoning/tool signatures for later turns. The Harness conversion dropped that provenance, so simply replacing the inline model with a catalog lookup would have made same-model replay and cross-provider handoff incomplete.
+
+The adapter configuration also assumes one DeepSeek API key and endpoint. A generic backend needs independent credentials and endpoint overrides per provider while leaving AWS, Google ADC, OAuth, and other ambient authentication mechanisms to pi-ai.
+
+## Decision
+
+### Provider is the adapter registration key
+
+`GenerateOptions` and `LlmCallConfig` carry `provider: string` beside `model: string`; `AgentOptions` carries the corresponding optional creation field. A loop request is valid only after both values are non-empty, and both values are part of the logged request header. `agent/request` may return a replacement pair on any step, so a session can switch providers and models without changing the Cordis plugin lifecycle.
+
+`LlmService` registers and resolves adapters by provider. `registerAdapter(providers, adapter)` checks the entire provider list before mutating the registry, rejects a duplicate with `DUPLICATE_ADAPTER`, and disposes the whole registration as one effect. Model ids are not registration keys; the selected adapter still validates or forwards them. The later [LLM catalog and ACP selection RFC](2026-07-15-llm-model-catalog-and-acp-selection.md) added advisory `listProviders()` / `listModels()` discovery without turning model membership into request validation.
+
+A provider has exactly one adapter owner in a Cordis context. `dsh-llm-deepseek` registers `deepseek`; `dsh-llm-pi-ai` may also register `deepseek`, but loading both owners is a configuration error rather than an ordering rule or fallback. A deployment that wants the hand-rolled DeepSeek implementation excludes `deepseek` from the pi-ai profiles. A deployment that wants pi-ai's DeepSeek implementation does not mount `dsh-llm-deepseek`.
+
+`dsh-llm-deepseek` removes its model registration list and accepts any model string routed through provider `deepseek`. Its request serialization, `/chat/completions` endpoint, thinking options, SSE parsing, and error behavior remain unchanged; `options.model` is still sent verbatim.
+
+### Explicit pi-ai provider profiles
+
+`dsh-llm-pi-ai` takes one non-empty list of provider profiles. Provider names must be unique within the list and present in pi-ai's `getProviders()` result. Each profile contains the provider name plus optional `apiKey`, `baseURL`, headers, reasoning level and budgets, cache retention, transport, timeouts, and retry settings. Credentials are never global: an explicit key applies only to its profile, while an absent key lets pi-ai resolve its standard environment variable, OAuth token, AWS credential chain, Google ADC, or other provider-native ambient authentication. An explicitly empty key is invalid configuration rather than an environment fallback.
+
+The plugin registers all configured provider names against one `PiAiAdapter` in one all-or-nothing call. A request uses its provider to select the matching profile and finds its model in `getModels(provider)` to obtain the catalog descriptor. An unknown provider fails at plugin load; an unknown model fails before network I/O with `UNKNOWN_MODEL`. The catalog object is never mutated. When a profile supplies `baseURL`, the adapter clones the selected descriptor and overrides only `baseUrl`, so a private endpoint can retain pi-ai's API, capabilities, compatibility flags, context limits, and reasoning map. The private endpoint must implement the selected provider's protocol, and the model id must still exist in the installed pi-ai catalog.
+
+The adapter calls pi-ai's `streamSimple()` so each catalog model chooses its registered API implementation, including OpenAI Responses instead of Chat Completions where the descriptor says `openai-responses`. Harness temperature, maximum tokens, signal, session id, and the profile's common stream options flow through directly. Profile headers merge with the mandatory Harness attribution headers, with Harness attribution winning its reserved names. The adapter no longer maintains DeepSeek-specific payload rewrites or a provider-protocol matrix.
+
+pi-ai's common stream options do not expose stop sequences. `dsh-llm-pi-ai` rejects a defined Harness `stop` option with `UNSUPPORTED_OPTION` rather than silently ignoring it or growing a second provider-specific payload implementation. `dsh-llm-deepseek` continues to support `stop` through its native request serializer.
+
+### Durable assistant provenance and replay state
+
+Assistant messages carry provider-neutral provenance containing the request's `provider` and `model`, plus an optional JSON-serializable adapter replay state. A successful `assistant/message` session event records this provenance and `deriveMessages()` returns it with the assistant message. User, system, context, and tool-result messages carry no assistant provenance. The provider/model fields are authoritative loop data; an adapter owns only its opaque replay-state payload.
+
+A terminal successful `finish` chunk may carry replay state, and `BlockAssembler` retains it alongside usage and finish reason. The loop attaches it to the assistant provenance only when the post-`agent/step-result` content is structurally equal to the assembled provider output. A listener that rewrites content keeps the provider/model provenance but loses the now-stale replay state. Error and aborted responses do not produce a normal assistant message and therefore do not enter future model history.
+
+The pi-ai replay state is a versioned, minimal projection of its successful `AssistantMessage`: source API/provider/model, response id/model, stop reason, and index-aligned text, thinking, and tool-call signatures. It does not duplicate text or tool arguments already carried by Harness content blocks, and it omits diagnostics, timestamps, usage, and errors. On a later request, `LlmService` gives replay state to the target adapter only when the historical provider and target provider are currently owned by the same adapter instance. That adapter combines the logged Harness content with replay state when it can restore the historical response, and owns any required cross-model or cross-provider conversion. An adapter receiving replay state with an unknown version or mismatched block shape fails explicitly; a different adapter receives only provider-neutral content and provenance.
+
+This state is model-visible replay input and therefore follows the existing [reconstructable-request rule](../../implemented/architecture/2026-07-05-reconstructable-requests.md): it is present in both the terminal `finish` chunk and the assembled `assistant/message` provenance that drives derivation. Resume and fork preserve it verbatim. Compaction that shadows the assistant message also removes its replay state from the active surface; the summary is ordinary provider-neutral content.
+
+### Propagate the target through every request producer
+
+Every model-selection surface carries provider and model together: declarative agents, ACP and stdio app config, the JSON-RPC initialize request, subagent overrides and inheritance, workflow child overrides, and direct compaction summarization. Subagents inherit both fields from their parent before applying request overrides. The system-prompt variable set gains `provider` beside `model`.
+
+Compaction configuration gains `summarizationProvider` beside `summarizationModel`. Both are empty to inherit, or both are non-empty to select an explicit target; a half-configured pair fails load. Inheritance uses the last logged request target when one exists and falls back to the agent's creation options. `compact/summary` records both fields with the existing model-call envelope.
+
+The JSON-RPC runtime receives provider and model explicitly. Its convenience fallback mounts `dsh-llm-deepseek` only for provider `deepseek` when that provider has no registered owner; other missing providers fail without guessing an adapter.
+
+The on-disk session format remains the pre-release pinned version `0`, with no compatibility promise. Seed/load validation rejects request headers lacking provider and assistant messages lacking required provenance instead of accepting an old shape that can no longer reconstruct the request.
+
+## Alternatives considered
+
+**Keep model names as registry keys and add wildcard adapters.** A wildcard introduces fallback ordering between exact registrations and catch-all plugins, makes duplicate ownership dependent on listener order, and still cannot distinguish the same model id at two providers without another convention.
+
+**Encode provider and model into one string.** Values such as OpenRouter's `openai/gpt-*` already contain provider-like prefixes and slashes. A delimiter convention would leak routing syntax into every model selector and require escaping rules; two explicit fields are unambiguous and independently loggable.
+
+**Add `backend + provider + model`.** A backend key would allow `dsh-llm-deepseek` and pi-ai's DeepSeek implementation to coexist and switch per request. The accepted deployment rule is instead one adapter owner per provider: implementations of the same upstream are alternatives selected by plugin composition. A third routing dimension would burden every request and configuration for a capability with no current consumer.
+
+**Let `dsh-llm-pi-ai` automatically register every pi-ai provider.** This would claim ambient credentials and provider names the deployment never intended to expose, and would conflict with native adapters such as `dsh-llm-deepseek`. Explicit profiles make capability and credential scope reviewable.
+
+**Mount one pi-ai plugin instance per provider.** Separate instances isolate config but repeat plugin declarations and cannot make profile registration atomic. One adapter already receives provider on every request, so a validated profile map is the smaller lifecycle surface.
+
+**Accept arbitrary inline pi-ai model descriptors.** This would support catalog-external private model ids, but it exposes pi-ai's model and compatibility schema as Harness configuration and makes the adapter responsible for validating protocol-specific combinations. The first version supports custom endpoints by overriding `baseURL` on catalog models; custom descriptors require a separate decision after a real catalog-external deployment is identified.
+
+## Consequences
+
+- Provider names are deployment-wide route ownership keys: two providers may use the same model string, but mounting two adapters for one provider fails at load instead of creating fallback order.
+- Model selection no longer changes the Cordis plugin graph. Catalog-backed adapters can accept any installed catalog model selected after startup, while the native DeepSeek adapter forwards arbitrary DeepSeek model ids.
+- A custom `baseURL` preserves the selected catalog model's protocol and capabilities; it does not make catalog-external model ids valid. Private endpoints must implement that catalog entry's protocol.
+- pi-ai credentials and transport knobs are scoped per provider profile. An omitted key delegates to pi-ai ambient authentication, while an explicitly empty key is invalid.
+- `dsh-llm-pi-ai` rejects stop sequences because pi-ai's common stream API cannot express them; the native DeepSeek adapter retains its stop support.
+- Replay state is portable only within the adapter instance that owns both the historical and target providers. Cross-provider and cross-model restoration is an adapter responsibility, and another adapter receives provider-neutral history without the opaque state.
+- Current pre-release session JSONL requires provider/model request headers and assistant provenance. Older shapes remain version `0` but are rejected rather than migrated.
+
+## Testing
+
+- Unit coverage exercises registry conflicts, request reconstruction, session validation, profile resolution, option forwarding, native API selection including OpenAI Responses, conversion, replay validation, error mapping, cancellation, content rewrites, and same-instance versus different-instance replay dispatch.
+- Keyless loop/session tests and ACP snapshots exercise durable provider/model metadata, resume and fork propagation, workflow/subagent overrides, and unchanged user-visible transcripts; the key-gated DeepSeek e2e retains real provider streaming and tool follow-up coverage.
+- Public JSDoc, package READMEs, architecture and core-data-structure docs, generated catalogs, examples, session fixtures, and Python SDK pairs use provider/model targets consistently and are checked by the repository documentation and type-equivalence gates.
+
+## Risks
+
+This is a repo-wide pre-release API break: model-only request construction, adapter registration, app protocols, fixtures, and persisted version-0 event shapes all change together, with no compatibility aliases. The provider exclusivity rule deliberately prevents two implementations of the same upstream from coexisting in one context. A pi-ai dependency update can change the accepted provider/model catalog, so the lockfile and adapter e2e matrix define the tested set. Custom `baseURL` endpoints inherit the chosen catalog model's protocol assumptions and cannot repair an incompatible proxy. Catalog-external model descriptors and multimodal content remain unsupported. pi-ai replay state may contain opaque encrypted reasoning signatures; it is persisted because the provider requires it for continuity, but it is never rendered or logged outside the existing session record.
diff --git a/docs/rfc/implemented/architecture/2026-07-14-provider-routed-llm-adapters.zh.md b/docs/rfc/implemented/architecture/2026-07-14-provider-routed-llm-adapters.zh.md
new file mode 100644
index 0000000000..75ac9adbe5
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-14-provider-routed-llm-adapters.zh.md
@@ -0,0 +1,91 @@
+# RFC: 基于提供方路由的 LLM 适配器与通用 pi-ai 后端
+
+Status: implemented
+
+[English](2026-07-14-provider-routed-llm-adapters.md) | 中文
+
+## 问题
+
+`dsh-llm` 按精确模型名称注册适配器。插件在 Cordis 启动时提供模型列表，`LlmService` 为列表中的每个字符串保存一个适配器，`GenerateOptions.model` 同时选择适配器与提供方模型。两个正式适配器都只面向相同的两个 DeepSeek 模型时，这种方式可以工作，但它混淆了两个独立决策：由哪个上游提供方承接请求，以及该提供方应运行哪个模型。
+
+这种混淆使提供方网关无法提供开放的模型目录。例如，OpenRouter 是一个包含大量模型 ID 的提供方，私有 OpenAI 兼容端点也可能在不修改 Harness 插件树的情况下增加模型。目前，每个新选择的模型都必须在插件启动期间完成注册。同一个模型 ID 还可能存在于多个提供方中，因此仅按模型注册无法表达调用方预期使用的提供方。
+
+`dsh-llm-pi-ai` 没有暴露 pi-ai 的提供方抽象。它以内联方式构造 DeepSeek `openai-completions` 模型，应用 DeepSeek 专用的 payload 补丁，并将每条回放的助手消息标记为 DeepSeek。pi-ai 自身提供提供方/模型目录，能够选择 `openai-responses`、`anthropic-messages`、`google-generative-ai` 等 API，并保留提供方专用的响应 ID，以及后续轮次所需的推理和工具签名。Harness 转换丢弃了这些来源信息，因此仅将内联模型替换为目录查询，会导致同模型回放与跨提供方移交不完整。
+
+适配器配置同样假定只存在一个 DeepSeek API 密钥和端点。通用后端需要为各提供方分别配置凭据和端点覆盖，同时继续由 pi-ai 处理 AWS、Google ADC、OAuth 等环境认证机制。
+
+## 决策
+
+### 提供方作为适配器注册键
+
+`GenerateOptions` 与 `LlmCallConfig` 在 `model: string` 之外携带 `provider: string`，`AgentOptions` 则携带对应的可选创建字段。只有两个值都非空时，agent loop（智能体循环）请求才有效；两个值也都会写入请求头日志。`agent/request` 可以在任意步骤返回替换后的字段组合，因此会话可以切换提供方与模型，无需改变 Cordis 插件生命周期。
+
+`LlmService` 按提供方注册和解析适配器。`registerAdapter(providers, adapter)` 在修改注册表前检查整个提供方列表，遇到重复项时返回 `DUPLICATE_ADAPTER`，并将整组注册作为一个 effect 释放。模型 ID 不作为注册键；仍由选中的适配器负责验证或转发。后续的 [LLM 目录与 ACP 模型选择 RFC](2026-07-15-llm-model-catalog-and-acp-selection.md) 增加了建议性的 `listProviders()` / `listModels()` 发现接口，但不会把目录成员关系变成请求校验规则。
+
+在一个 Cordis 上下文中，一个提供方只能有一个适配器所有者。`dsh-llm-deepseek` 注册 `deepseek`；`dsh-llm-pi-ai` 也可以注册 `deepseek`，但同时加载两个所有者属于配置错误，不采用顺序规则或回退行为。若部署选择手写的 DeepSeek 实现，需从 pi-ai 配置中排除 `deepseek`；若部署选择 pi-ai 的 DeepSeek 实现，则不挂载 `dsh-llm-deepseek`。
+
+`dsh-llm-deepseek` 移除模型注册列表，接受通过 `deepseek` 提供方路由的任意模型字符串。其请求序列化、`/chat/completions` 端点、thinking 选项、SSE（Server-Sent Events）解析和错误行为保持不变；`options.model` 仍会原样发送。
+
+### 显式 pi-ai 提供方配置
+
+`dsh-llm-pi-ai` 接受一个非空的提供方配置列表。列表内的提供方名称必须唯一，并且存在于 pi-ai 的 `getProviders()` 结果中。每项配置包含提供方名称，以及可选的 `apiKey`、`baseURL`、headers、推理级别和预算、缓存保留设置、传输方式、超时和重试设置。凭据不设全局值：显式密钥仅对所属配置生效；未提供密钥时，pi-ai 使用标准环境变量、OAuth token、AWS 凭据链、Google ADC 或其他提供方原生环境认证。显式空密钥属于无效配置，不会回退到环境认证。
+
+插件通过一次全有或全无调用，将所有已配置的提供方名称注册到同一个 `PiAiAdapter`。请求按 provider 选择对应配置，并在 `getModels(provider)` 中查找模型以取得目录描述符。未知提供方会在插件加载时失败；未知模型会在网络 I/O 前以 `UNKNOWN_MODEL` 失败。适配器不会修改目录对象。当配置提供 `baseURL` 时，适配器复制选中的描述符，仅覆盖 `baseUrl`，使私有端点保留 pi-ai 的 API、能力、兼容标志、上下文限制与推理映射。私有端点必须实现所选提供方的协议，模型 ID 也仍须存在于已安装的 pi-ai 目录中。
+
+适配器调用 pi-ai 的 `streamSimple()`，因此每个目录模型会选择其注册的 API 实现；描述符为 `openai-responses` 时使用 OpenAI Responses，而非 Chat Completions。Harness 的 temperature、最大 token 数、signal、session ID，以及提供方配置中的通用流选项均直接传递。配置 headers 与 Harness 强制归因 headers 合并；发生保留名称冲突时，以 Harness 归因为准。适配器不再维护 DeepSeek 专用 payload 重写或提供方协议矩阵。
+
+pi-ai 的通用流选项不支持停止序列。若 Harness `stop` 选项已定义，`dsh-llm-pi-ai` 会以 `UNSUPPORTED_OPTION` 拒绝请求，不会静默忽略，也不会增加第二套提供方专用 payload 实现。`dsh-llm-deepseek` 继续通过原生请求序列化器支持 `stop`。
+
+### 持久化助手来源信息与回放状态
+
+助手消息携带提供方无关的来源信息，其中包含请求的 `provider` 和 `model`，以及可选的 JSON 可序列化适配器回放状态。成功的 `assistant/message` 会话事件记录这些来源信息，`deriveMessages()` 返回助手消息时也会包含这些信息。用户、system、context 与工具结果消息不携带助手来源信息。provider/model 字段是 agent loop 的权威数据；适配器仅拥有其不透明回放状态 payload。
+
+成功的终止 `finish` 分片可以携带回放状态，`BlockAssembler` 会将其与 token 用量和结束原因一起保留。只有当 `agent/step-result` 处理后的内容与提供方组装输出在结构上相等时，agent loop 才会把回放状态附加到助手来源信息。监听器重写内容后，provider/model 来源信息仍会保留，但已经陈旧的回放状态会被移除。错误或中止响应不会生成正常助手消息，因此不会进入后续模型历史。
+
+pi-ai 回放状态是其成功 `AssistantMessage` 的带版本最小投影，包含源 API/provider/model、响应 ID/model、停止原因，以及按索引对齐的文本、thinking 和工具调用签名。它不会重复 Harness 内容块中已有的文本或工具参数，也不包含诊断信息、时间戳、用量或错误。后续请求中，只有历史提供方和目标提供方当前归同一个适配器实例所有时，`LlmService` 才会把回放状态交给目标适配器。适配器在能够恢复历史响应时，将 Harness 记录的内容与回放状态组合，并负责所需的跨模型或跨提供方转换。适配器收到未知版本或块形状不匹配的回放状态时会显式失败；其他适配器只能收到提供方无关的内容与来源信息。
+
+该状态属于模型可见的回放输入，因此遵循现有的[请求可重建规则](../../implemented/architecture/2026-07-05-reconstructable-requests.md)：它同时存在于终止 `finish` 分片和驱动派生的已组装 `assistant/message` 来源信息中。恢复和 fork 会原样保留该状态。压缩（compaction）遮蔽助手消息时，也会从活动 surface 中移除其回放状态；摘要属于普通的提供方无关内容。
+
+### 在所有请求生产方中传播目标
+
+每个模型选择接口都同时携带 provider 与 model：声明式 agent、ACP（Agent Client Protocol）和 stdio 应用配置、JSON-RPC initialize 请求、subagent 覆盖与继承、工作流子 agent 覆盖，以及直接压缩摘要。subagent 先从父 agent 继承两个字段，再应用请求覆盖。系统提示词变量集合在 `model` 之外增加 `provider`。
+
+压缩配置在 `summarizationModel` 之外增加 `summarizationProvider`。两个值均为空时继承，均非空时选择显式目标；只配置其中一个会导致加载失败。继承优先使用最近一次记录的请求目标，没有时回退到 agent 创建选项。`compact/summary` 使用现有模型调用 envelope 记录两个字段。
+
+JSON-RPC 运行时显式接收 provider 与 model。仅当 `deepseek` 提供方没有注册所有者时，其便利回退才会挂载 `dsh-llm-deepseek`；其他缺失的提供方会直接失败，不会猜测适配器。
+
+磁盘会话格式仍使用预发布阶段固定的版本 `0`，且不承诺兼容性。seed/load 验证会拒绝缺少 provider 的请求头，以及缺少必需来源信息的助手消息，不会接受已无法重建请求的旧格式。
+
+## 考虑过的替代方案
+
+**继续以模型名称作为注册表键，并增加通配适配器。** 通配机制会在精确注册与兜底插件之间引入回退顺序，使重复所有权取决于监听器顺序；若不再增加其他约定，仍无法区分不同提供方中相同的模型 ID。
+
+**将提供方与模型编码到一个字符串中。** OpenRouter 的 `openai/gpt-*` 等值已经包含类似提供方的前缀和斜杠。分隔符约定会把路由语法泄漏到每个模型选择接口，并需要转义规则；两个显式字段更清晰，也可以分别记录日志。
+
+**增加 `backend + provider + model`。** backend 键可以让 `dsh-llm-deepseek` 与 pi-ai 的 DeepSeek 实现共存，并按请求切换。最终采用的部署规则是一个提供方对应一个适配器所有者：同一上游的不同实现属于由插件组合选定的替代项。第三个路由维度会增加每个请求与配置的负担，却没有当前消费方。
+
+**让 `dsh-llm-pi-ai` 自动注册所有 pi-ai 提供方。** 这种方式会占用部署无意暴露的环境凭据和提供方名称，并与 `dsh-llm-deepseek` 等原生适配器冲突。显式配置可以审查能力和凭据范围。
+
+**每个提供方挂载一个 pi-ai 插件实例。** 独立实例可以隔离配置，但会重复插件声明，也无法实现配置注册的原子性。每个请求本就向同一个适配器提供 provider，因此经过验证的配置映射具有更小的生命周期接口。
+
+**接受任意内联 pi-ai 模型描述符。** 这种方式可支持目录外的私有模型 ID，但会将 pi-ai 的模型与兼容性 schema 暴露为 Harness 配置，并要求适配器验证协议专用组合。当前版本通过覆盖目录模型的 `baseURL` 支持自定义端点；只有实际出现目录外部署需求后，才会另行决策是否支持自定义描述符。
+
+## 影响
+
+- 提供方名称是部署范围内的路由所有权键：两个提供方可以使用相同的模型字符串，但为同一个提供方挂载两个适配器会在加载时失败，不会形成回退顺序。
+- 模型选择不再改变 Cordis 插件图。目录型适配器可以接受启动后选择的任意已安装目录模型，原生 DeepSeek 适配器则会转发任意 DeepSeek 模型 ID。
+- 自定义 `baseURL` 会保留所选目录模型的协议与能力，但不会让目录外模型 ID 变为有效。私有端点必须实现该目录项对应的协议。
+- pi-ai 凭据与传输选项按提供方配置隔离。省略密钥时委托 pi-ai 使用环境认证；显式空密钥无效。
+- pi-ai 的通用流 API 无法表达停止序列，因此 `dsh-llm-pi-ai` 会拒绝停止序列；原生 DeepSeek 适配器仍支持停止序列。
+- 仅当历史提供方与目标提供方归同一个适配器实例所有时，回放状态才可移植。适配器负责跨提供方和跨模型恢复；其他适配器只接收不含不透明状态的提供方无关历史。
+- 当前预发布会话 JSONL 要求请求头包含 provider/model，助手消息包含来源信息。旧格式仍使用版本 `0`，但会被拒绝，不执行迁移。
+
+## 测试
+
+- 单元测试覆盖注册表冲突、请求重建、会话验证、配置解析、选项转发、包括 OpenAI Responses 在内的原生 API 选择、转换、回放验证、错误映射、取消、内容重写，以及同一实例与不同实例间的回放分发。
+- 无密钥的 agent loop/会话测试和 ACP 快照覆盖持久化 provider/model 元数据、恢复与 fork 传播、工作流/subagent 覆盖，以及不变的用户可见 transcript（文本记录）；密钥门控的 DeepSeek e2e 测试保留真实提供方的流式输出与工具后续调用覆盖率。
+- 公共 JSDoc、package README、架构与核心数据结构文档、生成目录、示例、会话 fixture（测试前置数据）和 Python SDK 配对文档统一使用 provider/model 目标，并由仓库文档与类型等价门禁校验。
+
+## 风险
+
+这是一次覆盖全仓库的预发布 API 破坏性变更：仅模型的请求构造、适配器注册、应用协议、fixture，以及持久化版本 0 事件格式会同时变化，不提供兼容别名。提供方排他规则有意禁止同一上游的两个实现共存于同一上下文。pi-ai 依赖升级可能改变可接受的提供方/模型目录，因此锁文件与适配器 e2e 矩阵定义已验证集合。自定义 `baseURL` 端点会继承所选目录模型的协议假设，无法修复不兼容的代理。目录外模型描述符与多模态内容仍不受支持。pi-ai 回放状态可能包含不透明的加密推理签名；提供方需要该信息维持连续性，因此系统会持久化该状态，但不会在现有会话记录之外渲染或记录它。
diff --git a/docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.i18n.yaml b/docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.i18n.yaml
new file mode 100644
index 0000000000..d1756c9186
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.i18n.yaml
@@ -0,0 +1,6 @@
+# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
+# side as of the last confirmed-consistent state. Both languages carry equal authority;
+# after editing either side, bring the other along and re-record with:
+#   pnpm run verify-translation-pairing --write
+2026-07-15-llm-model-catalog-and-acp-selection.md: d84fe9fdb75bd2d28a00269c84d29c4223798253
+2026-07-15-llm-model-catalog-and-acp-selection.zh.md: 019819c4aa5ab4ad281b5b32daa76a004c9d6466
diff --git a/docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.md b/docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.md
new file mode 100644
index 0000000000..d84fe9fdb7
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.md
@@ -0,0 +1,66 @@
+# RFC: Advisory LLM catalogs and per-session ACP model selection
+
+Status: implemented
+
+English | [中文](2026-07-15-llm-model-catalog-and-acp-selection.zh.md)
+
+## Problem
+
+Provider-routed adapters let every request choose `provider + model`, but `LlmService` exposed only routing and streaming. A UI could not discover which providers were registered or which models an adapter was prepared to recommend. ACP clients therefore received no `model` session config option, so Zed, JetBrains, and VS Code integrations had no model list even though the request seam already supported runtime switching.
+
+Model discovery cannot become request validation. The hand-written DeepSeek adapter deliberately forwards arbitrary model ids to a public or private endpoint, while pi-ai has a finite installed catalog that is authoritative for its own request resolution. Treating one shared catalog as a whitelist would remove the private-endpoint behavior that provider routing was designed to preserve.
+
+ACP selection must also preserve the provider dimension. The same model id may appear under multiple routes, and switching a global adapter or agent template would leak one editor session's choice into every other session. Prompt variables and request routing must change together; a selection that lands during asynchronous prompt assembly cannot make `{{model}}` name one model while the request reaches another.
+
+## Decision
+
+### Provider-neutral advisory discovery
+
+`LlmAdapter` gains `providerInfo(provider)` and asynchronous `listModels(provider)` methods. Their provider-neutral results are `LlmProviderInfo { id, name }` and `LlmModelInfo { provider, id, name, description? }`. The defaults preserve existing adapter behavior by naming a provider after its route and advertising no models.
+
+`LlmService.listProviders()` returns detached metadata in registration order. `LlmService.listModels(provider)` delegates to the route owner, validates non-empty ids and names, rejects a mismatched provider or duplicate model id with `INVALID_CATALOG`, and returns detached values. Unknown providers still fail with `NO_ADAPTER`. Provider metadata is validated atomically during `registerAdapter()` so a malformed display record cannot leave a partial registration.
+
+Catalog membership is advisory. It drives selectors and diagnostics but never changes `stream()` routing and never rejects an otherwise valid request. Provider ownership remains exclusive and lifecycle-bound; model ids remain request-time adapter input.
+
+`dsh-llm-pi-ai` maps the configured provider's installed `getModels(provider)` entries into the neutral catalog. Its existing request-time catalog lookup remains authoritative and still rejects unknown models with `UNKNOWN_MODEL`. `dsh-llm-deepseek` accepts an optional `models` config containing display entries, defaulting to `deepseek-v4-flash` and `deepseek-v4-pro`. An explicit list replaces those defaults and an empty list disables discovery. The entries improve selector UX for known public or private models, while every unlisted model id continues to pass through unchanged.
+
+### ACP session config option
+
+The ACP bridge advertises one select with `id: model` and `category: model` in `session/new` and `session/load` when the session has a complete target whose provider is registered. Each opaque option value encodes the full provider/model pair. Models are grouped by provider when multiple non-empty provider groups exist; a single group is flattened for clients that render simple selects better.
+
+The session's current target is added to the displayed options when its adapter omits it. This preserves custom DeepSeek and private-endpoint models while keeping the adapter catalog advisory. A target with an unregistered provider is not advertised, and a model-less agent remains available to another `agent/request` supplier.
+
+`session/set_config_option` accepts only values from the current catalog snapshot and updates a target reference owned by that ACP session. No global `LlmService` or `AgentOptions` state changes, so concurrent sessions may select different providers and models. The existing permission select remains independent, and every response returns the complete refreshed option state.
+
+### Prompt/request consistency and durability
+
+Agent setup installs scoped `system-prompt/assemble` and `agent/request` listeners. Prompt assembly snapshots the selected pair once per step, overwrites the assembled `provider` and `model` variables after downstream prompt listeners, and the request listener applies that same snapshot after downstream request listeners. A selection during asynchronous assembly therefore starts on the next step rather than splitting prompt text from routing. Other call-config fields remain untouched.
+
+The request header remains the durable source of truth. When a selected target is actually used, the existing full `request/header` snapshot records it. `session/load` initializes the ACP selection from the folded last request header before falling back to bridge config. A selection that is never used by a request is intentionally in-memory only because it never became model-visible state.
+
+ACP's experimental `providers/*` capability is not used. That draft surface configures provider base URLs, protocols, and headers, including secrets; it does not enumerate models and would give the UI authority to rewrite deployment-owned adapter configuration.
+
+## Alternatives considered
+
+**Return model strings only.** A model-only value loses the provider route and becomes ambiguous as soon as two providers expose the same id.
+
+**Make catalogs mandatory whitelists.** This conflicts with the hand-written adapter's arbitrary model pass-through and private deployments. The selected adapter already owns authoritative request validation.
+
+**Store selection in `AgentOptions` or `LlmService`.** Those are creation-wide or deployment-wide objects. Mutating them would couple concurrent ACP sessions and bypass the logged `agent/request` replacement path.
+
+**Persist a new model-selection session event immediately.** An unused UI selection has not affected a model request. Recording the existing request header when the target is consumed preserves the model-visible-if-and-only-if-logged rule without adding a second source of truth.
+
+**Use ACP `providers/*`.** That unstable API changes endpoint and authentication configuration rather than selecting a model for one session, and its lifecycle and secret-handling semantics do not match this feature.
+
+## Consequences
+
+- Any adapter can expose a dynamic model list without leaking provider-library types into the core seam.
+- Catalog consumers must treat absence as “not advertised,” never “invalid request.”
+- pi-ai-backed ACP deployments automatically inherit the installed pi-ai provider catalogs; hand-written DeepSeek deployments list known choices explicitly and retain arbitrary model support.
+- ACP clients receive a standard stable model config option, with provider-aware values and per-session isolation.
+- Request headers remain compatible with the provider-routed session shape; no new JSONL event or format version is required.
+- A catalog read can be asynchronous. ACP reads a detached snapshot before creating or resuming an agent, so discovery failure cannot leave a partially published session.
+
+## Testing
+
+Unit coverage validates catalog detachment and malformed metadata, pi-ai and DeepSeek catalog projection, ACP provider grouping, custom-current insertion, invalid values, provider/model request routing, prompt-variable alignment, concurrent-session isolation, model-less fallback, and load restoration from the request header. The existing ACP transport suites verify that the additional config option does not change prompt, cancellation, replay, approval, or tool-rendering behavior.
diff --git a/docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.zh.md b/docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.zh.md
new file mode 100644
index 0000000000..019819c4aa
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.zh.md
@@ -0,0 +1,66 @@
+# RFC: 建议性 LLM 目录与 ACP 会话级模型选择
+
+Status: implemented
+
+[English](2026-07-15-llm-model-catalog-and-acp-selection.md) | 中文
+
+## 问题
+
+基于提供方路由的适配器允许每次请求选择 `provider + model`，但 `LlmService` 只暴露路由和流式调用。UI 无法发现已注册的提供方，也无法知道适配器愿意推荐哪些模型。因此，ACP 客户端收不到 `model` 会话配置项；即使请求接缝已经支持运行时切换，Zed、JetBrains 和 VS Code 集成仍没有模型列表。
+
+模型发现不能变成请求校验。手写 DeepSeek 适配器会把任意模型 ID 原样转发给公开或私有端点，而 pi-ai 的有限安装目录则是其自身请求解析的权威依据。将共享目录视为白名单，会破坏提供方路由需要保留的私有端点能力。
+
+ACP 选择还必须保留提供方维度。同一个模型 ID 可能存在于多个路由下；切换全局适配器或 agent 模板会让一个编辑器会话的选择泄漏到其他会话。Prompt 变量与请求路由必须同时变化；如果选择发生在异步 prompt 组装期间，不能让 `{{model}}` 表示一个模型、实际请求却到达另一个模型。
+
+## 决策
+
+### 提供方中立的建议性发现
+
+`LlmAdapter` 增加 `providerInfo(provider)` 与异步 `listModels(provider)` 方法。其提供方中立结果分别为 `LlmProviderInfo { id, name }` 和 `LlmModelInfo { provider, id, name, description? }`。默认实现以路由名称作为提供方名称，并且不展示模型，从而保持现有适配器行为。
+
+`LlmService.listProviders()` 按注册顺序返回分离后的元数据。`LlmService.listModels(provider)` 委托给路由所有者，校验非空 ID 和名称，并在提供方不匹配或模型 ID 重复时以 `INVALID_CATALOG` 失败，最后返回分离后的值。未知提供方仍以 `NO_ADAPTER` 失败。提供方元数据在 `registerAdapter()` 期间进行原子校验，错误展示记录不会留下部分注册。
+
+目录成员关系仅提供建议。它驱动选择器与诊断，但不会改变 `stream()` 路由，也不会拒绝原本有效的请求。提供方所有权仍然具有排他性并绑定生命周期；模型 ID 仍是请求时传给适配器的输入。
+
+`dsh-llm-pi-ai` 将已配置提供方的安装目录 `getModels(provider)` 映射为中立目录。其现有请求时目录查询仍是权威依据，未知模型仍以 `UNKNOWN_MODEL` 失败。`dsh-llm-deepseek` 接受可选的 `models` 配置作为展示条目，默认包含 `deepseek-v4-flash` 和 `deepseek-v4-pro`。显式列表会替换这些默认值，空列表则关闭发现。这些条目改善已知公开或私有模型的选择体验，而所有未列出的模型 ID 仍会原样透传。
+
+### ACP 会话配置项
+
+当会话具有完整目标且目标提供方已注册时，ACP bridge 会在 `session/new` 与 `session/load` 中展示一个 `id: model`、`category: model` 的选择项。每个不透明选项值都编码完整的提供方/模型字段组合。存在多个非空提供方分组时按提供方分组；只有一个分组时将其展开，以便对简单选择器支持更好的客户端展示。
+
+如果适配器目录未包含会话当前目标，该目标仍会加入展示选项。这能保留自定义 DeepSeek 与私有端点模型，同时维持目录的建议性。提供方未注册的目标不会展示；缺少模型的 agent 仍可由其他 `agent/request` 提供者补齐。
+
+`session/set_config_option` 只接受当前目录快照中的值，并更新该 ACP 会话独占的目标引用。它不会修改全局 `LlmService` 或 `AgentOptions` 状态，因此并发会话可以选择不同的提供方和模型。现有权限选择项保持独立，每次响应都返回完整的刷新后配置项状态。
+
+### Prompt/请求一致性与持久化
+
+Agent setup 会安装作用域内的 `system-prompt/assemble` 与 `agent/request` 监听器。Prompt 组装为每个 step 只快照一次选中的字段组合，在下游 prompt 监听器完成后覆盖组装结果中的 `provider` 与 `model` 变量；请求监听器则在下游请求监听器完成后应用同一个快照。因此，异步组装期间发生的选择会从下一个 step 生效，不会导致 prompt 文本与路由分裂。其他调用配置字段保持不变。
+
+请求头仍是持久化事实来源。当选中目标被实际使用时，现有的完整 `request/header` 快照会记录它。`session/load` 先从折叠后的最后请求头初始化 ACP 选择，再回退到 bridge 配置。一个从未被请求使用的选择只保留在内存中，因为它从未成为模型可见状态。
+
+本功能不使用 ACP 的实验性 `providers/*` 能力。该草案接口配置提供方 base URL、协议和 headers，其中可能包含密钥；它不枚举模型，并且会赋予 UI 改写部署所有的适配器配置的权力。
+
+## 考虑过的替代方案
+
+**只返回模型字符串。** 仅模型值会丢失提供方路由；两个提供方暴露相同 ID 时立刻产生歧义。
+
+**将目录设为强制白名单。** 这与手写适配器的任意模型透传和私有部署冲突。请求的权威校验本就属于被选中的适配器。
+
+**将选择存入 `AgentOptions` 或 `LlmService`。** 这些对象分别面向创建过程或整个部署。修改它们会耦合并发 ACP 会话，并绕开带日志归因的 `agent/request` 替换路径。
+
+**立即写入新的模型选择会话事件。** 尚未使用的 UI 选择没有影响模型请求。目标被消费时记录现有请求头，既满足“模型可见当且仅当已记录”的规则，也不会引入第二个事实来源。
+
+**使用 ACP `providers/*`。** 该不稳定 API 用于修改端点与认证配置，而不是为单个会话选择模型；其生命周期和密钥处理语义都不适合本功能。
+
+## 结果
+
+- 任意适配器都能暴露动态模型列表，无需把提供方库类型泄漏到核心接缝。
+- 目录消费者必须把缺失理解为“未展示”，而不是“请求无效”。
+- 基于 pi-ai 的 ACP 部署会自动继承已安装的 pi-ai 提供方目录；手写 DeepSeek 部署显式列出已知选项，同时保留任意模型能力。
+- ACP 客户端会收到稳定标准的模型配置项，其中的值保留提供方信息，并按会话隔离。
+- 请求头继续使用基于提供方路由的会话结构；不需要增加 JSONL 事件或格式版本。
+- 目录读取可以是异步的。ACP 在创建或恢复 agent 前读取分离后的快照，因此发现失败不会留下部分发布的会话。
+
+## 测试
+
+单元测试覆盖目录分离与错误元数据、pi-ai 和 DeepSeek 目录投影、ACP 提供方分组、自定义当前模型补入、无效值、提供方/模型请求路由、prompt 变量一致性、并发会话隔离、无模型回退，以及从请求头恢复选择。现有 ACP 传输测试验证新增配置项不会改变 prompt、取消、回放、审批或工具展示行为。
diff --git a/docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.i18n.yaml b/docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.i18n.yaml
new file mode 100644
index 0000000000..e1d17fef0b
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.i18n.yaml
@@ -0,0 +1,6 @@
+# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
+# side as of the last confirmed-consistent state. Both languages carry equal authority;
+# after editing either side, bring the other along and re-record with:
+#   pnpm run verify-translation-pairing --write
+2026-07-15-replay-token-meter-service.md: 34df0383d1b8ae8047c4283eef3800de772c3cae
+2026-07-15-replay-token-meter-service.zh.md: 51f319f3c473fe247791e69133eef4280b768002
diff --git a/docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.md b/docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.md
new file mode 100644
index 0000000000..34df0383d1
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.md
@@ -0,0 +1,60 @@
+# RFC: Replay token meter service
+
+Status: implemented
+
+English | [中文](2026-07-15-replay-token-meter-service.zh.md)
+
+## Problem
+
+Context pressure is useful outside compaction. A compaction backend, an overflow guard, or a future request-policy plugin can all need the same answer: how much of the configured context window does the durable request consume? Keeping that fold inside `dsh-compact-basic` duplicates replay logic, makes measurement unavailable without compaction, and encourages callers to reuse stale accounting.
+
+Provider usage is not a complete answer. It describes one successful call under one exact request envelope, while the current surface can grow, shrink, or be replaced afterward. Sessions also switch providers and models, old logs can lack chunk provenance, and usage fields separate input, cache-read, cache-write, output, and reasoning counts. A useful service therefore combines the latest exact anchor with conservative heuristic repricing and exposes the log revision consumed by each result.
+
+## Decision
+
+### One concrete LLM-family service
+
+`@deepseek-ai/dsh-token-meter` is one concrete package under `packages/llm/` and registers `ctx.tokenMeter`. It is not split into an interface and backend before a second implementation exists. `TokenMeterService` itself exposes `contextWindow`, `measure(session, requestHeader?)`, and `estimateMessage(message)`; consumers call the singleton service directly.
+
+The service has one `contextWindow`, defaulting to 128,000 tokens and configurable as a positive integer. Estimation uses a fixed four-characters-per-token heuristic plus structural overhead. There are no model profiles, density settings, tokenizer backends, or language-specific strategies.
+
+### Per-session replay folds
+
+Each session owns one isolated incremental fold. Active folds advance from `session/event`; every read catches up through the durable tail, so listener ordering, seeded sessions, and service reload do not change the answer. The fold tracks canonical full request-header snapshots, step boundaries, surface appends and replacements, assistant usage, and assistant-chunk provenance. A malformed next event fails transactionally and remains unread rather than partially mutating state.
+
+`measure(session, requestHeader?)` synchronizes the fold once and returns scalar pressure together with positional per-node prices. `totalTokens` remains request-and-response pressure; `surfaceTokens` is the surface-only heuristic total and equals the sum of `nodes[].tokens`. A `requestHeader` override changes pressure pricing only, while the surface fields always describe the current session. `estimateMessage(message)` applies the fixed heuristic without session state. Each result is one detached, deeply immutable snapshot carrying one `logRevision`. Every measurement clones the current nodes and is therefore O(surface).
+
+Provider usage is reused only when the measured canonical request envelope equals the latest successful-call anchor. Any provider, model, system, prefix, tool, or call-config change causes complete heuristic repricing. Surface changes remain a signed delta from a matching anchor, including negative values after a shrinking replacement. A later successful request replaces the earlier anchor, including across provider or model switches.
+
+Usage sums the disjoint input, cache-read, cache-write, and output buckets. Reasoning is not added a second time. Every successful model call records an `assistant/message`, including content-less and max-token calls, with its exact earlier chunk seqs. An explicit empty provenance list means a known empty provider stream; absent legacy provenance conservatively treats the durable assistant output as provider output.
+
+### Compact-basic consumes, but does not own, measurement
+
+`dsh-compact-basic` requires `ctx.tokenMeter`; `CompactService` gains no token methods or types. The backend is factored into configuration, automatic triggering, region transaction, and summarizer modules, while `summarize()` remains its sole subclass hook. The singleton service consistently prices pressure, retention, shadowed content, provenance, and non-shrinking-summary rejection.
+
+Automatic compaction uses one unified measurement for each threshold-and-retention decision. The region transaction measures after appending its durable `compact/start` lock and again after asynchronous summarization; any intervening durable append changes `logRevision` and prevents replacement.
+
+Compact policy has service-wide defaults: threshold ratio `0.8`, retained tail `floor(contextWindow × 0.16)`, empty summarization provider/model, maximum summary output `8192`, one extra compaction attempt, and automatic triggering enabled. Top-level `thresholdRatio` and `retainTokens` override the pressure policy; retention must remain below the resulting threshold. `summarizationProvider` and `summarizationModel` must both be set or both be empty; an empty pair resolves the latest logged request target, then the `AgentOptions` pair.
+
+The pre-step trigger measures a provisional envelope: the current prompt and prefix override logged values, while the latest logged header supplies provider, model, tools, and other call config. A router-only agent without a complete provider/model pair skips that provisional check because `agent/request` can route later; any routed target can use the singleton estimator.
+
+## Testing
+
+Unit coverage pins service configuration, fixed estimation, envelope invalidation, latest-anchor replacement across provider/model switches, usage and missing-usage paths, seeded append/replace replay, signed deltas, provenance modes, malformed boundaries, unified snapshot detachment and deep immutability, surface-total equality, listener ordering, reload, compact defaults, routing fallback, one-call automatic decisions, retention, convergence, and log-revision rollback. A real Loader/Include YAML fixture loads the exact zero-config token-meter and compact-basic package names in dependency order.
+
+## Alternatives considered
+
+- **Keep estimation inside `CompactService`** — rejected because measurement has consumers and replay semantics independent of compaction; it would also force every compactor to expose the same unrelated API.
+- **Split a token-meter interface from a heuristic backend immediately** — rejected because only one implementation exists. One concrete service preserves the future seam without speculative packages or configuration.
+- **Keep model-keyed windows and density profiles** — rejected because the deployment currently has one context policy and one estimator. Model registries, unknown-model failures, and configurable density add branches without a second behavior to select.
+- **Keep separate scalar and surface measurements** — rejected because callers would need two reads and revision matching for one decision. A scalar-only read could avoid cloning nodes below threshold, but the split API introduces a caller-side race window; the unified snapshot accepts O(surface) cloning in exchange for coherence.
+- **Treat provider usage as portable between envelopes** — rejected because model, tools, prefixes, and call config are request facts. Mismatch reprices the whole current request.
+
+## Consequences
+
+- Token pressure has one replay-aware owner that compaction and future plugins can share.
+- The default makes the bundled composition usable with two zero-config plugin entries; deployments override one context capacity when needed.
+- Fixed heuristic pricing remains an estimate of provider behavior and is not an exact tokenizer or request serializer.
+- Every measurement clones the current positional surface and therefore costs O(surface), including pressure checks that finish below threshold.
+- Measurements fail loudly on malformed durable boundaries. This turns corrupted replay into a named integration failure instead of silently drifting pressure.
+- The pre-step compact integration can skip a router-only first check and can miss tool or routing changes applied later in request middleware.
diff --git a/docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.zh.md b/docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.zh.md
new file mode 100644
index 0000000000..51f319f3c4
--- /dev/null
+++ b/docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.zh.md
@@ -0,0 +1,60 @@
+# RFC: 回放式 token 计量服务
+
+Status: implemented
+
+[English](2026-07-15-replay-token-meter-service.md) | 中文
+
+## 问题
+
+上下文压力并不只对压缩有用。压缩后端、溢出保护或未来的请求策略插件都可能需要回答同一个问题：持久请求占用了已配置上下文窗口的多少容量？如果把该折叠逻辑留在 `dsh-compact-basic` 内部，就会重复实现回放逻辑，使未加载压缩的调用方无法使用计量，并诱使调用方复用陈旧的核算结果。
+
+提供方 usage 也不是完整答案。它只描述某个精确请求信封下的一次成功调用，而当前表层之后还可能增长、缩小或被替换。会话也可能切换提供方与模型，旧日志可能缺少分片来源，usage 字段还会分别报告输入、缓存读取、缓存写入、输出与推理计数。因此，可用的服务必须把最新精确锚点与保守的启发式重新定价结合起来，并公开每个结果已经消费的日志修订号。
+
+## 决策
+
+### 一个具体的 LLM 家族服务
+
+`@deepseek-ai/dsh-token-meter` 是 `packages/llm/` 下的单个具体包，并注册 `ctx.tokenMeter`。在第二种实现出现之前，它不会被拆成接口与后端。`TokenMeterService` 本身公开 `contextWindow`、`measure(session, requestHeader?)` 与 `estimateMessage(message)`；消费方直接调用这个单例服务。
+
+服务只有一个 `contextWindow`，默认值为 128,000 token，并允许配置为正整数。估算采用固定的每 token 四个字符启发式规则，并加上结构开销。服务不提供模型 profile、密度设置、分词器后端或语言专用策略。
+
+### 逐会话回放折叠
+
+每个会话都有一个隔离的增量折叠。活跃折叠通过 `session/event` 前进；每次读取都会追到持久日志尾部，因此监听器顺序、种子会话与服务重载不会改变答案。折叠跟踪规范的完整请求头快照、步骤边界、表层追加与替换、assistant usage，以及 assistant 分片来源。下一个畸形事件会以事务方式失败并保持未读，不会让状态只修改一半。
+
+`measure(session, requestHeader?)` 只同步一次折叠，并在返回标量压力的同时给出逐位置节点价格。`totalTokens` 仍表示请求与响应压力；`surfaceTokens` 是仅针对表层的启发式总量，并等于 `nodes[].tokens` 之和。`requestHeader` 覆盖只改变压力定价，表层字段始终描述当前会话。`estimateMessage(message)` 不依赖会话状态，直接应用固定启发式规则。每个结果都是一个分离且深度不可变的快照，只携带一个 `logRevision`。每次计量都会复制当前节点，因此成本为 O(surface)。
+
+只有当待计量的规范请求信封等于最近一次成功调用的锚点时，服务才复用提供方 usage。提供方、模型、系统提示词、前缀、工具或调用配置任一变化都会触发完整的启发式重新定价。表层变化相对匹配锚点保留有符号增量，包括缩小替换后的负值。后续成功请求会替换先前锚点，提供方或模型切换时也一样。
+
+Usage 会对互不重叠的输入、缓存读取、缓存写入与输出 bucket 求和，不会再次加入推理计数。每次成功模型调用都会记录 `assistant/message`，包括无内容调用与达到 token 上限的调用，并带上精确的更早分片 seq。显式空来源列表表示已知为空的提供方流；旧日志中缺失的来源则保守地把持久 assistant 输出视为提供方输出。
+
+### compact-basic 消费计量，但不拥有计量
+
+`dsh-compact-basic` 要求 `ctx.tokenMeter`；`CompactService` 不增加 token 方法或类型。后端拆分为配置、自动触发、区域事务与摘要器模块，而 `summarize()` 仍是唯一的子类 hook。单例服务一致用于压力、保留、被遮蔽内容、来源以及非缩小摘要拒绝的定价。
+
+自动压缩的每次阈值与保留联合决策只使用一次统一计量。区域事务先追加持久 `compact/start` 锁，再执行一次计量，并在异步摘要完成后再次计量；期间任何持久追加都会改变 `logRevision`，从而阻止替换。
+
+压缩策略采用服务级默认值：阈值比例 `0.8`、保留尾部 `floor(contextWindow × 0.16)`、空的摘要提供方/模型、摘要最大输出 `8192`、一次额外压缩尝试，以及启用自动触发。顶层 `thresholdRatio` 与 `retainTokens` 覆盖压力策略；保留值必须小于最终阈值。`summarizationProvider` 与 `summarizationModel` 必须同时设置或同时为空；空组合先解析最近记录的请求目标，再使用 `AgentOptions` 中的组合。
+
+pre-step 触发器计量临时请求信封：当前提示词与前缀覆盖日志值，最近记录的请求头给出提供方、模型、工具及其他调用配置。没有完整提供方/模型组合的纯路由 agent（智能体）会跳过该临时检查，因为 `agent/request` 仍可稍后路由；任意已路由目标都可使用这个单例估算器。
+
+## 测试
+
+单元覆盖固定服务配置、固定估算、信封失效、提供方/模型切换时替换最新锚点、有无 usage 的路径、种子追加/替换回放、有符号增量、来源模式、畸形边界、统一快照的分离性与深度不可变性、表层总量相等性、监听器顺序、重载、压缩默认值、路由回退、自动决策单次调用、保留、收敛与日志修订回滚。真实 Loader/Include YAML fixture 按依赖顺序加载精确的零配置 token-meter 与 compact-basic 包名称。
+
+## 考虑过的替代方案
+
+- **把估算保留在 `CompactService` 内**——不予采纳，因为计量拥有独立于压缩的消费方与回放语义；它还会强迫每个压缩器暴露同一套无关 API。
+- **立即把 token meter 拆成接口与启发式后端**——不予采纳，因为目前只有一种实现。单个具体服务保留未来接缝，同时避免推测性的包与配置。
+- **保留模型键控的窗口与密度 profile**——不予采纳，因为当前部署只有一种上下文策略与一个估算器。模型注册表、未知模型错误和可配置密度只增加分支，却没有第二种行为可供选择。
+- **保留独立的标量与表层计量**——不予采纳，因为消费方必须为一次决策执行两次读取并匹配修订号。仅读取标量可以避免在低于阈值时复制节点，但拆分 API 会在消费方引入竞态窗口；统一快照接受 O(surface) 复制成本，以换取结果一致性。
+- **在不同信封之间移用提供方 usage**——不予采纳，因为模型、工具、前缀与调用配置都是请求事实。不匹配时会重新定价完整当前请求。
+
+## 后果
+
+- Token 压力拥有一个可供压缩与未来插件共享的回放感知所有者。
+- 默认值让内置组合只需两个零配置插件条目即可使用；部署需要时只覆盖一个上下文容量。
+- 固定启发式定价仍然只是提供方行为的估计，并不是精确分词器或请求序列化器。
+- 每次计量都会复制当前的位置表层，因此成本为 O(surface)，低于阈值即可结束的压力检查也不例外。
+- 遇到畸形持久边界时，计量会明确失败。这会把损坏的回放转化为具名集成错误，而不是让压力静默漂移。
+- pre-step 压缩集成可能跳过纯路由的首次检查，也可能错过请求中间件稍后应用的工具或路由变化。
diff --git a/docs/rfc/implemented/feature/2026-06-14-acp-agent-client-protocol.md b/docs/rfc/implemented/feature/2026-06-14-acp-agent-client-protocol.md
index bfd0e6a10e..ea4b98c31d 100644
--- a/docs/rfc/implemented/feature/2026-06-14-acp-agent-client-protocol.md
+++ b/docs/rfc/implemented/feature/2026-06-14-acp-agent-client-protocol.md
@@ -48,7 +48,7 @@ The precise supported and deferred protocol rows live in [`packages/ui/acp/acp-f
 
 Editors can create, load, prompt, cancel, render, ask, and reconfigure multiple harness sessions over one ACP connection without a loop-specific dependency. The session event log remains the durable source for replay, prompt settlement, cwd, and per-session configuration. Tool presentation and human-answer channels remain extensible plugin contracts instead of ACP-specific behavior.
 
-The bridge deliberately does not implement session list/delete/resume/close capabilities, MCP passthrough, additional directories, image/audio/embedded-resource prompts, runtime model selection, plans, slash commands, usage updates, editor filesystem delegation, or the ACP terminal execution sub-protocol. The feature checklist records these as unsupported rather than silently accepting them.
+The bridge deliberately does not implement session list/delete/resume/close capabilities, MCP passthrough, additional directories, image/audio/embedded-resource prompts, plans, slash commands, usage updates, editor filesystem delegation, or the ACP terminal execution sub-protocol. Runtime model selection was added later through standard session config options by the [LLM catalog and ACP selection RFC](../architecture/2026-07-15-llm-model-catalog-and-acp-selection.md).
 
 An idle config selection is truthful in the live response but not durable until the next `agent/prompt-submit` anchors it inside the open turn. Crashing before that boundary loses the pending selection; this is the cost of keeping session events turn-enclosed and replay-safe.
 
diff --git a/docs/rfc/implemented/feature/2026-06-15-code-mode.md b/docs/rfc/implemented/feature/2026-06-15-code-mode.md
index 950c2b4bd7..3bc07b839f 100644
--- a/docs/rfc/implemented/feature/2026-06-15-code-mode.md
+++ b/docs/rfc/implemented/feature/2026-06-15-code-mode.md
@@ -38,11 +38,11 @@ Three decisions, each elaborated in its own section below:
 
 Under `'code'` and `'both'` the registry owns `run_code` as a reserved presentation transport with one required parameter, `{ code: string }`. It is represented by a normal `ToolDefinition` for dispatch but stays outside the filterable capability layers, so restrictions cannot accidentally remove Code Mode's only entry point. Calls traverse the complete tool pipeline — `tools/pre-execute` → monotonic guards → `tools/execute` around dispatch → `tools/post-execute` → immutable `tools/result` notification — exactly like native calls; a permission plugin can inspect the program text before it runs, and final-result observers see the normalized outer outcome. Its `execute(args, exec)`:
 
-1. **Build bindings.** One run-scoped signal follows outer cancellation and is aborted whenever the run settles. Each visible tool binding JSON-normalizes its arguments—rejecting lossy values before dispatch—waits on the serialization queue, executes with a deterministic call id and the outer token as `parent`, and logs `tool/code-dispatch`. Successful text becomes a string and non-text blocks become placeholders; tool errors reject the binding promise. Every sub-call retains its own immutable execution identity and traverses the full tool pipeline.
+1. **Build bindings.** One run-scoped signal follows outer cancellation and is aborted whenever the run settles. Each visible tool binding JSON-normalizes its arguments—rejecting lossy values before dispatch—waits on the serialization queue, executes with a deterministic call id and the outer token as `parent`, defers returned contexts through the outer execution, and logs `tool/code-dispatch`. Successful text becomes a string and non-text blocks become placeholders; tool errors reject the binding promise. Every sub-call retains its own immutable execution identity and traverses the full tool pipeline.
 2. **Runs the program**: `ctx.codeRuntime.run({ program: args.code, bindings: [{ global: 'tools', functions }], signal: runController.signal })`. The runtime receives the run-scoped signal, not only the caller's outer signal, so any way the outer run settles also aborts work inside the runtime.
 3. **Settle after quiescence.** When the runtime settles, the bridge aborts outstanding work and drains the dispatch queue before returning. Success returns captured output and presentation metadata. A runtime failure becomes `CodeRunFailedError`; backend rejection uses the registry's normal error boundary. Both produce structured error results, and no sub-call can append after `run_code` settles.
 
-**Sub-call `additionalContext` is omitted.** Injecting it during `run_code` would break parent call/result adjacency, while one program can produce many contexts. Supporting it requires a plural channel or loop-level sub-dispatch buffer.
+**Sub-call contexts are deferred through the parent.** Injecting inside `run_code` would break parent call/result adjacency, so `ToolRunContext.deferContext()` collects every sub-result `additionalContexts` entry in dispatch order. The registry carries that array even when the program later throws, and the loop appends each entry only after the outer result and every sibling result in the step. An outer post-execute block discards tool-deferred entries and exposes only contexts explicitly attached by the blocking decision.
 
 **Concurrency is serialized.** Each run owns a dispatch queue, so even `Promise.all` executes tool calls in submission order. Settlement abandons queued calls that have not started. Parallelism requires per-tool concurrency-safety metadata.
 
@@ -86,14 +86,14 @@ The SDK instructs the model to write an async erasable-TypeScript body, call too
 
 ## Consequences
 
-Deployments switching to `'code'` must update any native-only `toolOrder`. Assembly listeners own the integrity of any rewritten protocol surface. Sub-dispatch remains serialized, and the bridge does not propagate per-call `additionalContext` until those contracts are designed for Code Mode.
+Deployments switching to `'code'` must update any native-only `toolOrder`. Assembly listeners own the integrity of any rewritten protocol surface. Sub-dispatch remains serialized, while per-call contexts retain their source, envelope, and metadata through the outer result.
 
 ## Testing
 
 - **Worker runtime:** Real-worker tests cover output and value capture, failure kinds, compute and wall budgets, hostile binding traffic, empty environment, structured-clone fallback, output caps, and disposal to quiescence. A built-package test runs the worker entry under plain Node.
-- **Registry integration:** Tests cover code generation, all presentation modes, reserved-name and restriction rules, scoped visibility, authoritative assembly rewrites, `toolOrder`, runtime compatibility failures, full-pipeline sub-dispatch, parent-token correlation, serialization, cancellation and queue drain, JSON normalization, error propagation, log events, omitted `additionalContext`, and HMR cleanup.
-- **With-key e2e:** A real model composes two bash calls in one program; the test verifies the collapsed request header, correlated dispatch events, resulting file, and curated answer.
-- **Snapshot:** The `code-mode-turn` and `both-mode-turn` fixtures pin the SDK section, header tool list, dispatch events, and result card.
+- **Registry integration:** Tests cover code generation, all presentation modes, reserved-name and restriction rules, scoped visibility, authoritative assembly rewrites, `toolOrder`, runtime compatibility failures, full-pipeline sub-dispatch, parent-token correlation, serialization, cancellation and queue drain, JSON normalization, error propagation, log events, ordered context deferral across successful and failed programs, outer-block suppression, and HMR cleanup.
+- **With-key e2e:** A real model composes two bash calls in one program; another discovers nested workspace instructions through a Code Mode fs dispatch. The tests verify collapsed request headers, correlated dispatch events, resulting files, deferred context, and model behavior.
+- **Snapshot:** The `code-mode-turn`, `both-mode-turn`, and `code-mode-workspace-context` fixtures pin SDK text, header tool lists, dispatch events, deferred context, and result cards.
 
 ## Alternatives considered
 
diff --git a/docs/rfc/implemented/feature/2026-06-18-compaction-capability-seam.md b/docs/rfc/implemented/feature/2026-06-18-compaction-capability-seam.md
index 4a6586b159..864428bc58 100644
--- a/docs/rfc/implemented/feature/2026-06-18-compaction-capability-seam.md
+++ b/docs/rfc/implemented/feature/2026-06-18-compaction-capability-seam.md
@@ -6,9 +6,9 @@ Status: implemented
 
 A long-running agent conversation grows without bound. As the event log accumulates turns, the derived message history eventually approaches the model's context window — the model then truncates mid-response (`max-tokens`) or degrades. **Compaction** is the mitigation: replace a run of older history with a concise summary, keeping recent context intact.
 
-The [session surface](../../implemented/architecture/2026-06-18-session-surface.md) was built as the foundation for exactly this — a linked list over the event log with a `surfaceOp: { op: 'replace', start, end }` operation purpose-built to shadow a range of nodes and insert a replacement, with `sourceEventSeqs` recording provenance so the decision replays deterministically. What remained was the plugin that *decides what to compact and produces the summary*.
+The [session surface](../../implemented/architecture/2026-06-18-session-surface.md) was built as the foundation for exactly this — an ordered projection over the event log with a `surfaceOp: { op: 'replace', start, end }` operation purpose-built to shadow a range of entries and insert a replacement, with `sourceEventSeqs` recording provenance so the decision replays deterministically. What remained was the plugin that *decides what to compact and produces the summary*.
 
-Two forces shape the design. First, compaction is **swappable**: token counting can be a char/4 heuristic or a real tokenizer, and summarization can be a model call, a template, or a remote service — these vary independently of *when* and *which range* to compact. Second, `SurfaceEventType` is closed to five event types (`user/message`, `assistant/message`, `tool/result`, `context/message`, `steering/message`); only those may carry `surfaceOp`. A bespoke `compaction/*` event therefore **cannot** itself appear on the surface — the compiler rejects `surfaceOp` on it and the invariants plugin rejects it at runtime.
+Two forces shape the design. First, compaction policy and reusable token measurement vary independently: measurement belongs to the LLM-family [`ctx.tokenMeter` service](../../implemented/architecture/2026-07-15-replay-token-meter-service.md), while summarization can be a model call, a template, or a remote service. Second, `SurfaceEventType` is closed to five event types (`user/message`, `assistant/message`, `tool/result`, `context/message`, `steering/message`); only those may carry `surfaceOp`. A bespoke `compaction/*` event therefore **cannot** itself appear on the surface — the compiler rejects `surfaceOp` on it and the invariants plugin rejects it at runtime.
 
 ## Decision
 
@@ -17,7 +17,7 @@ Two forces shape the design. First, compaction is **swappable**: token counting
 Per the [capability-seams RFC](../../implemented/architecture/2026-06-13-capability-seams.md), compaction ships as separate packages so the contract, the algorithm, and (later) the consumer surface evolve independently:
 
 1. **Interface** — `@deepseek-ai/dsh-compact`: an abstract `CompactService` owning the `ctx.compact` key, the `CompactionResult` vocabulary, and the `compact/*` session events. It declares `compactIfNeeded()` and `compactRegion()` as **abstract** — the contract states *what* compaction does, not *how*.
-2. **Implementation** — `@deepseek-ai/dsh-compact-basic`: a concrete `BasicCompactService` that owns the entire algorithm — token estimation (chars per token — the `charsPerToken` config, default 4 — + per-block overhead), the tail→head retention walk, summarization via `ctx.llm.stream()`, the surface replacement, the lock, and the `agent/pre-step` auto-compaction listener. A tokenizer-based or template-based backend is a sibling package (or a subclass overriding the two protected estimation/summarization hooks).
+2. **Implementation** — `@deepseek-ai/dsh-compact-basic`: a concrete `BasicCompactService` that consumes `ctx.tokenMeter` and owns the tail→head retention walk, summarization via `ctx.llm.stream()`, the surface replacement, the lock, and the `agent/pre-step` auto-compaction listener. `summarize()` is its sole subclass hook; pricing and replay stay with the meter.
 3. **Consumer** — deferred. A `/compact` tool and slash command will `inject: ['compact']` and call the contract; they are intentionally out of scope here so the seam settles first.
 
 ### The contract depends on `dsh-session` and `dsh-llm` — a deliberate deviation
@@ -28,9 +28,9 @@ This is not a coupling smell — it is the contract's domain. The "only cordis"
 
 ### Abstract `compactIfNeeded` / `compactRegion`, algorithm in the backend
 
-An earlier draft put the full algorithm (the retention walk, token-summing, text extraction) as concrete methods on the interface, with only `estimateContentTokens()` and `summarize()` abstract. That recouples the contract to one strategy: a backend that wants a different retention policy or a different event-sequencing would have to fight inherited concrete code. Making both core methods abstract puts every *how* decision in the backend, where it belongs, and keeps the interface a pure statement of *what*. The backend remains internally factored — `estimateContentTokens()` and `summarize()` are `protected` hooks a sub-backend can override without reimplementing the walk — but that factoring is the backend's private concern, not the contract's.
+An earlier draft put the full algorithm (the retention walk, token-summing, text extraction) as concrete methods on the interface. That recouples the contract to one strategy: a backend that wants a different retention policy or event sequence would have to fight inherited concrete code. Making both core methods abstract puts every *how* decision in the backend and keeps the interface a statement of *what*. Token measurement is not a compaction hook at all; the singleton service lets multiple consumers share one per-session replay fold.
 
-`compactIfNeeded(agent, turn, step, fullSystemPrompt, signal)` takes **required** parameters (not the original all-optional shape). The auto-compaction seam (below) always supplies the agent, lifecycle context, assembled system prompt (counted toward the estimate), and the turn's abort signal, so optionality would only invite a hidden default at the seam. The session being compacted comes from the agent context. `compactRegion(session, start, end, agent, turn, step, signal?)` keeps an optional signal (a manual caller may omit it). Passing lifecycle context rather than a concrete model keeps router agents honest: the backend's summarization request can run through `agent/request`, where model-routing plugins already choose the actual model.
+`compactIfNeeded(agent, fullSystemPrompt, sessionPrefix, signal)` takes required pressure inputs and cancellation. The session comes from the agent. `compactRegion(session, start, end, agent, signal?)` keeps an optional signal for manual callers and requires `session === agent.session`; implementations reject mismatch before model resolution, lock acquisition, summarization, or log mutation. The pre-step integration resolves a provisional model from the latest logged request header, then `AgentOptions.model`; a model-less router-only first step skips pressure because `agent/request` can route later. The default summarizer resolves its model from explicit config, the latest logged routed model, then agent options.
 
 ### Auto-compaction runs on `agent/pre-step`, a dedicated surface-mutation seam
 
@@ -40,7 +40,7 @@ The fix is a dedicated loop seam, **`agent/pre-step`** (`@mode serial`), fired b
 
 ```
 assembly = ctx.systemPrompt.assemble()
-await ctx.serial('agent/pre-step', agent, turn, step, system, signal)  ⟵ compaction mutates the surface here
+await ctx.serial('agent/pre-step', agent, turn, step, system, prefix, signal)  ⟵ compaction mutates the surface here
 session('step/start')                 ⟵ the step opens AFTER the seam
 messages = session.deriveMessages()   ⟵ single derive, reflects the compaction
 request  = waterfall agent/request    ⟵ pure request transform (hooks, model switch)
@@ -52,11 +52,11 @@ The loop derives messages once after `agent/pre-step`. Running before `step/star
 
 Auto-compaction fires before **every** step, not once per turn. This is **load-bearing for runaway-turn survival**: a tool-heavy ReAct turn appends an `assistant/message` + a `tool/result` per step, so the surface grows *within* a turn. A single turn can grow past the window on its own (a "runaway turn") — and the only moment to rescue it before the next model call overflows is the next step's `pre-step` checkpoint. Gating compaction to a turn's first step (or, worse, retaining the whole in-flight turn verbatim) re-opens exactly the hole compaction exists to close: the harness would die when compaction is most needed.
 
-`compactIfNeeded` retains the smallest tail of whole surface units whose estimated size reaches `retainTokens` and compacts older nodes. A unit is a complete closed step or one no-step message. If the token cutoff lands inside a step, retention expands until the cut is tool-pairing balanced. Balance is checked on surface order, not log sequence, because replacement summaries have new sequence numbers at old surface positions. `compactRegion` rejects boundaries that split a tool call from its result. The in-flight turn receives no special retention.
+`compactIfNeeded` retains the smallest tail of whole surface units whose estimated size reaches `retainTokens` and compacts older nodes. A unit is a complete closed step or one no-step message. If the token cutoff lands inside a step, retention expands until the cut is tool-pairing balanced. Balance is checked on surface order, not log sequence, because replacement summaries have new sequence numbers at old surface positions. `dsh-compact` exports the before/after edge helpers; their per-session cache folds only appended surface-tail nodes while `replaceGeneration` is unchanged, does no event reads for log-only growth, and rebuilds current membership and balances after replacement. `compactRegion` rejects boundaries that split a tool call from its result. The in-flight turn receives no special retention.
 
 A runaway turn thus compacts exactly like any other history: its early *closed* steps get summarized while its recent steps stay verbatim. When the only compactable content left is an un-splittable open tail step (its tool-calls have no results yet), compaction declines (`null`) and retries once that step closes.
 
-**Single-unit overflow is out of scope, by design.** If a single retained unit — one closed step, or a large free node such as a pasted `user/message` — *alone* exceeds the budget, compaction cannot help and the next model call may go out over-budget. Bounding an individual unit's size is a separate concern (output truncation), handled elsewhere; compaction makes no promise about it, and the harness without such a mechanism can still break on a single oversized unit. This is named honestly rather than papered over.
+**Single-unit overflow is out of scope, by design.** If a single retained unit — one closed step, or a large free entry such as a pasted `user/message` — *alone* exceeds the budget, compaction cannot help and the next model call may go out over-budget. Bounding an individual unit's size is a separate concern (output truncation), handled elsewhere; compaction makes no promise about it, and the harness without such a mechanism can still break on a single oversized unit. This is named honestly rather than papered over.
 
 ### Head-anchoring: one auto checkpoint, always at the head
 
@@ -64,11 +64,11 @@ Auto-compaction always starts at the surface head, merging the prior checkpoint
 
 ### Approximate convergence invariant
 
-`resolveConfig` validates numeric knobs but does NOT reject based on a pretend summary-length invariant. Convergence is dynamic: provider output caps can be spent on hidden or surfaced reasoning tokens, and the model may emit a summary of unpredictable size. `maxTokens` is only the provider-side generation cap for the summarization call; reasoning blocks are stripped before the checkpoint is stored. If a compacted surface is still over threshold, `compactIfNeeded()` re-compacts the head checkpoint up to `compactionRetries` extra times, but each committed summary must be smaller than the content it shadows. The sole residual is the single-unit-overflow case above (a backward-rounded oversized step can push the retained tail over budget) — which is exactly the out-of-scope concern, not a thrash bug.
+`resolveConfig` supplies usable defaults: threshold ratio `0.8`, retained tail `floor(contextWindow × 0.16)`, empty summarization-model override, `maxTokens: 8192`, `compactionRetries: 1`, and `auto: true`. Optional top-level `thresholdRatio` and `retainTokens` override the policy for the token meter's single context window; retention must remain below the resulting threshold. Convergence remains dynamic because provider output caps can be spent on hidden or surfaced reasoning tokens and summary size is unpredictable. If the compacted surface remains over threshold, `compactIfNeeded()` re-compacts the head checkpoint up to the configured retry count, but each committed summary must be smaller than what it shadows.
 
 ### Surface replacement: `compact/*` events are log-only; one `user/message` carries the summary
 
-Because `SurfaceEventType` is closed, the summary cannot ride on a `compact/*` event. The backend instead appends a **single `user/message`** with `surfaceOp: { op: 'replace', start, end }` whose `content` is the (framed) summary and whose `sourceEventSeqs` covers the shadowed nodes *and* the bookkeeping events. The `compact/*` events are pure log records (lock + provenance). The surface mutation sits **inside** the lock — `compact/end` is the last event appended:
+Because `SurfaceEventType` is closed, the summary cannot ride on a `compact/*` event. The backend instead appends a **single `user/message`** with `surfaceOp: { op: 'replace', start, end }` whose `content` is the (framed) summary and whose `sourceEventSeqs` covers the shadowed entries *and* the bookkeeping events. The `compact/*` events are pure log records (lock + provenance). The surface mutation sits **inside** the lock — `compact/end` is the last event appended:
 
 ```
 compact/start    → log-only. Acquires the lock.
@@ -79,7 +79,7 @@ user/message     → surfaceOp { op:'replace', start, end }. THE surface mutatio
 compact/end      → log-only. Releases the lock (carries `error` on a recoverable failure).
 ```
 
-`deriveMessages()` then yields `[summary_as_user_message, ...retained_nodes]`. Reusing `user/message` is honest rather than a workaround: a summary genuinely *is* user-role context.
+`deriveMessages()` then yields `[summary_as_user_message, ...retained_entries]`. Reusing `user/message` is honest rather than a workaround: a summary genuinely *is* user-role context.
 
 ### Checkpoint framing + incremental merge (backend-private)
 
@@ -103,19 +103,19 @@ Two failure paths, both documented:
 
 ## Alternatives considered
 
-- **The full algorithm as concrete interface methods** (only estimation/summarization abstract) — the earlier draft; rejected because it recouples the contract to one retention strategy. Both core methods are abstract; the `protected` estimation/summarization hooks are the backend's private factoring, not the contract's.
+- **The full algorithm as concrete interface methods** — rejected because it recouples the contract to one retention strategy. Both core methods are abstract; reusable measurement is a separate LLM-family service and `summarize()` is basic's sole hook.
 - **Compaction on the `agent/request` waterfall** — the earlier cut; rejected for the double-derive it forced and for handing the listener context it structurally cannot compact. The dedicated `agent/pre-step` seam makes the layering correct by construction.
 - **A separate `compact/error` event** — rejected: `compact/end` keeps an `error?` field, mirroring `tool/result`'s self-contained error — one event tells success from failure without correlating a sibling.
 - **Teaching core turn-repair about `compact/*`** — rejected: the log-only orphan is inert, and a core module patched for every future `xxx/start … xxx/end` plugin pair is exactly the coupling the capability-seam architecture exists to avoid.
 
 ## Consequences
 
-- **New packages**: `packages/compact/compact` (interface) and a sibling `compact-basic` (backend) under `packages/compact/`, wired into the root tsconfigs. The consumer tier is deferred.
+- **Packages**: `packages/compact/compact` supplies the interface and `compact-basic` supplies the backend. `packages/llm/token-meter` owns replay-aware measurement independently. The consumer tier is deferred.
 - **New loop seam**: `agent/pre-step` (`@mode serial`) declared in `dsh-agent` and emitted by `dsh-agent-loop` after system assembly and before `step/start`. This is a documented change to the loop — `docs/architecture.md` records it and the generated cordis catalog carries its signature.
 - **`SessionEventMap`** gains `compact/start` / `compact/summary` / `compact/end` by declaration merging (merge-extensible); `SurfaceEventType` is **not** touched. These are session events, not cordis `Events`, so the event-taxonomy gate needs no entry.
-- **`dsh-session`** gains the tool-pairing balance predicate (`isToolPairingBalanced`, in `tool-pairing.ts`, exported from the package index) that `compactRegion`/`compactIfNeeded` use to keep a collapsed region from splitting a step's tool-call/result pair. The surface `replace` op and the surface-metadata runtime guard already existed and are reused.
-- **`dsh-invariants`** drops its `surface replace: start must be <= end` assertion: a head-anchored compaction lands a high-seq replacement node at an older range's *position*, so `start > end` numerically is normal and valid (the range is positional, validated by the surface's `indexOf` checks that remain). The turn-enclosure invariant is reused unchanged.
-- **Wiring**: `dsh-compact-basic` is loaded in `examples/coding-agent`'s `cordis.yml`, so the seam ships in the real demo (it was previously loaded nowhere).
+- **`dsh-compact`** owns `toolPairingBalancedBefore(session, seq)` and `toolPairingBalancedAfter(session, seq)`, the cached surface-edge checks that `compactRegion` and `compactIfNeeded` use to avoid splitting a tool-call/result pair. The cache validates current membership by seq and answers both edges from one per-cut balance sequence; stale or missing seqs and orphan results reject. `dsh-session` continues to own the surface `replace` operation, ordered event sequences, and rewrite generation.
+- **`dsh-invariants`** drops its `surface replace: start must be <= end` assertion: a head-anchored compaction lands a high-seq replacement entry at an older range's *position*, so `start > end` numerically is normal and valid (the range is positional, validated by the surface's `indexOf` checks that remain). The turn-enclosure invariant is reused unchanged.
+- **Wiring**: `examples/coding-agent/cordis.yml` loads zero-config `dsh-token-meter` before `dsh-compact-basic`; the service-wide window and compact defaults make the pair usable without repeated numeric policy.
 
 ## Testing
 
diff --git a/docs/rfc/implemented/feature/2026-06-21-subagent-capability-seam.md b/docs/rfc/implemented/feature/2026-06-21-subagent-capability-seam.md
index e2871532c8..83c7beb440 100644
--- a/docs/rfc/implemented/feature/2026-06-21-subagent-capability-seam.md
+++ b/docs/rfc/implemented/feature/2026-06-21-subagent-capability-seam.md
@@ -67,6 +67,6 @@ The seam is tested through the real Cordis Loader/export path, which catches the
 ## Consequences
 
 - **Recursion.** Without a bound, an in-process child can see the delegation tool and recurse. The in-process backends implement the optional absolute depth limit and scoped live-global `toolFilter`; ACP advertises both capabilities off and rejects such a request. The [subagent composition-controls RFC](2026-07-12-subagent-persona-tool-filter-and-depth.md) owns their exact semantics and security limits.
-- **Blocking the parent turn.** Synchronous collect holds the parent's `runStep` open for the child's full duration. This is acceptable for the first cut; **background / poll / spill semantics are deferred to a future redesign that unifies long-running-tool handling across subagents AND bash** (a sub-agent and a long `bash` background task pose the same "the model started something slow, how does it collect later" problem, and should share one mechanism rather than each inventing its own).
+- **Blocking the parent turn.** Foreground collection holds the parent's step open for the child's full duration. Background delegation uses the shared `ctx.tasks` runtime and generic `task_*` tools, the same collection mechanism as background bash; the subagent seam itself remains task-agnostic.
 - **Live progress.** This cut surfaces only lifecycle + final result; a per-chunk child→parent update stream is deferred with the background redesign.
 - **ACP client surface.** Proxying `fs`/`terminal` from the ACP child back to the parent (a shared-workspace mode) is future work; the first cut advertises neither, so the child self-serves in its own process.
diff --git a/docs/rfc/implemented/feature/2026-06-24-workspace-context.md b/docs/rfc/implemented/feature/2026-06-24-workspace-context.md
new file mode 100644
index 0000000000..cc11fd19ff
--- /dev/null
+++ b/docs/rfc/implemented/feature/2026-06-24-workspace-context.md
@@ -0,0 +1,87 @@
+# RFC: Workspace context instruction files
+
+Status: implemented
+
+## Problem
+
+Repository guidance such as `AGENTS.md` belongs in a coding session's effective context so project conventions, build commands, and review rules arrive without repeated user pasting. The stdio and ACP products need the same behavior, isolated by session cwd: a global system-prompt section leaks one workspace's files into another live ACP session.
+
+Neighboring products establish useful conventions but differ in details. Codex treats `AGENTS.md` as native, Claude Code uses `CLAUDE.md` and familiar system-reminder-style user context, and opencode supports both names with one winner per directory plus lazy nested discovery. The harness needs cross-tool compatibility without loading duplicate or contradictory files from the same scope.
+
+The lifecycle has two distinct classes of content. The initial applicable chain is stable enough to live in the request prefix and benefit from provider prefix caching. Nested files, edits, candidate switches, and removals happen after the session starts and belong in durable append-only history rather than the frozen prefix.
+
+## Decision
+
+The implementation lives in `packages/context/workspace-context` as `@deepseek-ai/dsh-workspace-context`. It is a request-context extension, not a core service or a filesystem backend. `@deepseek-ai/dsh-agent-core` mounts it for both product front doors and forwards its config. The plugin consumes `agent/session-prefix`, `tools/post-execute`, and the optional `ctx.fs` capability.
+
+The plugin does not statically inject `fs`. Providerless product trees therefore boot normally and the plugin no-ops until a filesystem provider exists. All production reads go through that provider. Candidate probes call `lstat` before `resolve`, so a repository-owned final-component symlink is rejected rather than followed outside the workspace. The session-prefix signal and dynamic tool execution signal propagate through resolution, metadata probes, and streaming reads, so cancellation does not wait for an unrelated filesystem scan. Once `lstat` identifies a regular-file winner, a provider exception or disagreement during resolve/stat is classified as unavailable: it is neither interpreted as a deletion nor allowed to fall through to a lower-priority candidate.
+
+### File Names And Precedence
+
+The default per-directory candidate list is `['AGENTS.md', 'CLAUDE.md']`. The list is configurable as `instructionFileCandidates`, and `AGENTS.md` is an ordinary first candidate rather than a hidden priority. In one directory, only the first existing regular-file candidate loads. With defaults, `AGENTS.md` is native and `CLAUDE.md` is a compatibility fallback.
+
+Candidate entries are same-directory file names. Empty entries, `.`/`..`, and entries containing `/` or `\` are ignored. Lowercase names, local variants, and other same-directory names can be opted into explicitly; rule directories and import semantics are outside this contract.
+
+The user-global file is fixed at `$DSH_HOME/AGENTS.md` and is not affected by `instructionFileCandidates`. `$DSH_HOME` defaults to `~/.dsh`, matching the harness-level home role of `~/.codex` or `~/.claude` rather than introducing a plugin-specific home. Tilde expansion and the default live in `dsh-paths` so future harness features share the same convention.
+
+### Baseline Prefix
+
+On the first request of an agent-loop instance, the plugin contributes one user-role message through `agent/session-prefix`. It loads the user-global file first, then finds the project root by walking upward from `agent.session.header.cwd` to a configured root marker (default `.git`), then loads one candidate from each directory from the root to the cwd. A `.git` file and a `.git` directory are both valid markers, covering linked worktrees and submodules. Without a marker, the cwd itself is the root.
+
+The plugin prepends its contribution before `await next()` returns, so session-prefix contributions appear in plugin registration order. In the product spine workspace instructions are registered before a skills catalog and therefore appear first. The loop deep-freezes the composed prefix, logs it in `EpochHeader.messagePrefix`, and reuses it verbatim for that instance. It is request state, not `Session.deriveMessages()` history.
+
+A resumed agent creates a new loop instance and recomposes the baseline from current files, with the new prefix anchored by the resume request header. This permits current baseline content on resume without mutating a prefix already used by an earlier instance.
+
+The baseline is a user-role `<system-reminder>` with `Instructions from: <path>` sections and explicit authority and precedence language. This familiar model-facing frame avoids a harness-specific XML vocabulary. Project paths are root-relative and the user-global path is `~/.dsh/AGENTS.md` for the default home or `$DSH_HOME/AGENTS.md` for a configured home. A literal `</system-reminder>` inside file content is escaped. The package README owns the exact current [prompt shape](../../../../packages/context/workspace-context/README.md#prompt-shape).
+
+### Dynamic Discovery And Refresh
+
+After a successful first-party `read`, `write`, or `edit` call, the `tools/post-execute` listener reconciles the touched descendant chain and every scope already known to the session. A newly reached scope is returned through `additionalContexts` for the next request using an `Additional instructions from: <path>` system-reminder. Under Code Mode, `run_code` defers sub-dispatch contexts onto its outer result, so the same update is appended only after the parent result rather than being injected mid-call.
+
+A content edit appends `Updated instructions from: <path>`, states that the new content replaces the previous content, and includes the complete current file. If precedence changes from one candidate to another, the message also names the previous path and says it no longer applies. If no candidate remains, the plugin appends `Instructions removed: <path>` and states that the previously loaded instructions no longer apply.
+
+Dynamic messages use a raw `context/message` envelope because the plugin owns the complete system-reminder framing. Core context injection therefore supports `envelope: 'raw'`; callers that omit it retain the canonical `<context source="...">` wrapper. `context/message.meta` carries opaque JSON state that is persisted but never rendered to the model.
+
+Shell commands are not discovery triggers. Local bash calls start fresh shells, and inferring reached paths from arbitrary command strings would require shell semantics the prompt plugin does not own.
+
+### Duplicate Suppression And Change Detection
+
+Every dynamic workspace context event stores versioned metadata with `{ action, scope, path, previousPath?, digest? }`, where `digest` is SHA-1 over the loaded content. The model-facing prompt has no HTML comments, hidden markers, or headings that are parsed back into state.
+
+At reconciliation time the plugin scans plugin-owned `context/message` events and derives the latest state for each visible scope. A short per-session pending map begins only after the immutable top-level `tools/result` proves an `additionalContexts` entry survived every post-execute listener, then covers the interval before the loop appends that context to the log. Each entry records the open `{ turn, step }`: an equal durable `context/message` at or after its sequence boundary confirms and removes it, while a matching `step/end` arriving first means the loop discarded its context buffer, so the plugin removes both the pending entry and its version-cache fast path. A nested Code Mode result stages its changes under the parent's opaque execution token so repeated sub-dispatches in one run do not duplicate them; the parent result rolls that provisional state back and commits only contexts retained by outer policy.
+
+An unchanged path and digest is suppressed. A logged removal is a tombstone, so a reappearing candidate becomes a new `set`. Resume works from persisted metadata. If compaction removes an instruction event from the visible surface, that state no longer suppresses a later load, matching the fact that the model can no longer see it. Only changes actually included under the byte budget enter metadata or pending state, so an omitted file remains eligible on a later touch.
+
+The frozen baseline keeps an in-memory path/digest map for comparison. A later successful filesystem touch appends baseline edits or removals as dynamic messages; it never rewrites the prefix. During resumed prefix composition the plugin also reconciles visible dynamic scopes, so nested changes made while the agent was offline can append an update before the first resumed request.
+
+There is intentionally no watcher. Detection occurs at the next successful structured filesystem touch or resumed prefix composition. A provider failure produces no removal; absence is only accepted when all configured candidates in that scope were probed successfully.
+
+### Byte Budget And Bounded Reads
+
+`maxBytes` is required and applies separately to a rendered baseline or one dynamic reconciliation batch; there is no implicit or unbounded render budget. Non-positive and non-finite values disable loading. When content exceeds the budget, broader files are omitted before the most-specific file is truncated. A visible `Workspace instruction budget ...` notice names omitted and truncated paths and byte counts, and output never exceeds the configured bytes.
+
+`maxSourceBytes` is a positive per-file cap with a 1 MiB default. The loader checks reported size before reading and still consumes content through `streamText()` with a running UTF-8 byte count, so missing/stale metadata cannot force an unbounded allocation. An oversized winning candidate is unavailable rather than a reason to fall through to another same-directory name. The plugin deliberately keeps no process-wide cache and never retains instruction prose. It keeps only `{ path, version, digest }` per effective scope in a `WeakMap<Session, Map<scope, state>>`: a matching provider `FsVersion` plus matching effective prompt state skips the read, while a changed version triggers a bounded read and SHA-1 confirmation. SHA-1 remains the cross-provider content identity persisted in visible structured metadata; provider versions are only an in-memory invalidation fast path. Cache transitions for model-visible changes commit only when the corresponding context survives the complete tool-result policy chain, and are invalidated if that accepted context is later dropped with its aborted step before reaching the log.
+
+## Alternatives considered
+
+**Use a global `ctx.systemPrompt.section()`.** Rejected because one Cordis context can host sessions with different cwd values, while repository-owned text is lower-authority context rather than top-authority provider system content.
+
+**Inject the baseline on every `agent/pre-step`.** Rejected because repeated history injection wastes tokens, complicates duplicate state, and prevents a structurally stable provider prefix. Prefix composition gives a frozen, logged, per-instance baseline while dynamic append-only messages handle changes.
+
+**Load both `AGENTS.md` and `CLAUDE.md` in one directory.** Rejected because repositories in transition commonly duplicate guidance across both files. Ordered candidates make precedence explicit and configurable.
+
+**Parse rendered headings or hidden comments to recover loaded state.** Rejected because instruction prose can contain the same text, causing silent false positives. Persisted JSON metadata provides an unambiguous state channel that is invisible to the model.
+
+**Summarize files with a model.** Rejected because instruction files are already curated summaries; another model call is nondeterministic and can erase edge-case requirements. Deterministic full text with byte budgeting is simpler.
+
+## Consequences
+
+Workspace guidance is isolated per session and shared by both product front doors and every tool presentation mode. Initial instructions benefit from stable prefix caching, while nested and changed content remains durable and replayable. The generic session/agent context contract includes optional raw framing and JSON metadata, both propagated through prompt-submit and post-tool `additionalContexts` arrays without flattening entries.
+
+Repository text remains untrusted input. Lower-authority user-role framing, explicit precedence language, delimiter escaping, and symlink rejection reduce risk but do not eliminate prompt injection. Permission and sandbox layers treat workspace files as data rather than authority.
+
+The system is event-driven rather than watch-driven. Edits are not visible at the exact filesystem mutation instant unless that mutation goes through a structured tool; externally changed files are noticed on the next successful structured touch or resume. This keeps the design deterministic and provider-neutral.
+
+## Deferred
+
+Bash-derived path reporting, recursive startup scans, file watchers, lowercase defaults, `.claude/CLAUDE.md`, `.claude/rules/*.md`, import directives, ACP `additionalDirectories`, trust acknowledgements, and model-generated summaries are deferred. Same-directory private variants can be configured today; directory rule systems and imports need their own precedence and trust designs.
diff --git a/docs/rfc/implemented/feature/2026-06-29-todo-write-tool.md b/docs/rfc/implemented/feature/2026-06-29-todo-write-tool.md
index 126c69f1c0..06bc6daf6a 100644
--- a/docs/rfc/implemented/feature/2026-06-29-todo-write-tool.md
+++ b/docs/rfc/implemented/feature/2026-06-29-todo-write-tool.md
@@ -20,7 +20,7 @@ The list is appended as a `todo/write` event carrying the full `{ todos }` snaps
 
 ### NOT a surface event
 
-`todo/write` is deliberately excluded from `SurfaceEventType`. The surface is the projection that produces the LLM message history (`deriveMessages()`); a todo write produces no conversation message. So it carries no `surfaceOp`, never joins the surface linked list, and never reaches `deriveMessages()` — it is durable, replayable *UI* state that travels alongside the conversation without being part of it. (The dev-mode invariants still require it to sit inside an open turn, which it always does: it is appended mid-step during a tool call.)
+`todo/write` is deliberately excluded from `SurfaceEventType`. The surface is the projection that produces the LLM message history (`deriveMessages()`); a todo write produces no conversation message. So it carries no `surfaceOp`, never joins the ordered surface, and never reaches `deriveMessages()` — it is durable, replayable *UI* state that travels alongside the conversation without being part of it. (The dev-mode invariants still require it to sit inside an open turn, which it always does: it is appended mid-step during a tool call.)
 
 ### Priority synthesized only at the ACP boundary
 
diff --git a/docs/rfc/implemented/feature/2026-06-30-hook-bridges.md b/docs/rfc/implemented/feature/2026-06-30-hook-bridges.md
index 2d285fb152..af5c7e928d 100644
--- a/docs/rfc/implemented/feature/2026-06-30-hook-bridges.md
+++ b/docs/rfc/implemented/feature/2026-06-30-hook-bridges.md
@@ -12,8 +12,8 @@ The framing that shapes the whole design: **a bridge is a compatibility adapter,
 
 Two independent plugins in the `packages/hooks/` group, each a function/namespace plugin (`name`/`inject`/`Config`/`apply`, NO default export — see [postmortem 0001](../../../postmortem/0001-acp-default-export-drops-inject.md)) injecting only `bash`:
 
-- **`dsh-hooks-claude`** — the CC dialect. Seven of Claude Code's current hook points: `SessionStart`, `UserPromptSubmit`, `PreToolUse`, `PostToolUse`, `Stop`, `SubagentStart`, and `SubagentStop`. Owns CC-shaped per-event stdin payloads (a base of `session_id`/`cwd`/`hook_event_name` plus per-event fields), `CLAUDE_PROJECT_DIR` plus `${CLAUDE_PLUGIN_ROOT}`/`${CLAUDE_PROJECT_DIR}` substitution, and the literal-or-regex matcher mode. A CC hook's stdin carries a **trailing newline**.
-- **`dsh-hooks-codex`** — five of Codex's current hook points: `PreToolUse`, `PostToolUse`, `SessionStart`, `UserPromptSubmit`, and `Stop`. It uses an always-regex matcher, Codex-shaped snake_case payloads with `turn_id`/`model`/`permission_mode` extras written WITHOUT a trailing newline, no Codex plugin-env injection or config-time placeholder substitution, and no pre-tool approval or rewrite path. A tool call's payload carries the real `tool_name` in the bridge's reduced `tool_input: { command }` shape.
+- **`dsh-hooks-claude`** — the CC dialect. Seven of Claude Code's current hook points: `SessionStart`, `UserPromptSubmit`, `PreToolUse`, `PostToolUse`, `Stop`, `SubagentStart`, and `SubagentStop`. Owns CC-shaped per-event stdin payloads (a base of `session_id`/`transcript_path`/`cwd`/`hook_event_name` plus per-event fields), `CLAUDE_PROJECT_DIR` plus `${CLAUDE_PLUGIN_ROOT}`/`${CLAUDE_PROJECT_DIR}` substitution, and the literal-or-regex matcher mode. `transcript_path` is the persistence locator result or `''`; stdin carries a **trailing newline**.
+- **`dsh-hooks-codex`** — five of Codex's current hook points: `PreToolUse`, `PostToolUse`, `SessionStart`, `UserPromptSubmit`, and `Stop`. It uses an always-regex matcher, Codex-shaped snake_case payloads with `turn_id`/`model`/`permission_mode` extras written WITHOUT a trailing newline, no Codex plugin-env injection or config-time placeholder substitution, and no pre-tool approval or rewrite path. `transcript_path` is the same locator result or `null`; tool payloads carry the real `tool_name` in the reduced `tool_input: { command }` shape.
 
 ### Outcome → Decision mapping
 
@@ -35,9 +35,9 @@ The CC bridge's `ask` result is a real permission path, not a terminal bridge de
 
 `agent.inject()` defaults a missing `MessageSource` to `{ kind: 'user' }`, so every bridge `inject()` and `HookContext` passes `{ kind: 'plugin', plugin: 'hooks-claude' | 'hooks-codex' }`. Unit coverage pins the resulting `context/message.source` as the plugin rather than the user.
 
-### Adding context is not a veto — delegate, then fold
+### Adding context is not a veto — delegate, then prepend
 
-A context-only hook must call `next()` and then fold its `additionalContext` into the downstream decision; returning allow or accept directly would bypass later policy listeners. Post-tool block and accept decisions both preserve added context. Prompt allow preserves it, while prompt block drops it because the prompt never reaches the model. Only an explicit hook denial or block short-circuits the waterfall.
+A hook that only attaches `additionalContext` (no block/deny) is NOT a decision the bridge should return on its own: returning `allow`/`accept` from a waterfall listener WITHOUT calling `next()` short-circuits every later `agent/prompt-submit` / `tools/post-execute` listener, so a policy/sandbox plugin registered after the bridge would never see the prompt. Each bridge therefore delegates via `next()` before adding its context to the downstream decision. Both seams carry ordered `additionalContexts` arrays, so the bridge prepends its separately sourced entry while preserving every downstream source, envelope, and metadata field; a downstream prompt block still drops all context because the prompt never reaches the model, while post-tool block semantics may explicitly retain contexts. Code Mode ferries the same array through the outer `run_code` result. Only a real `deny`/`block` from the hook itself short-circuits. Tests assert a later listener can still block a prompt a context-only hook allowed and that retained prompt and post-tool contexts remain separate.
 
 ### CLAUDE_PROJECT_DIR defaults to the session workspace
 
diff --git a/docs/rfc/implemented/feature/2026-06-30-interception-seams.md b/docs/rfc/implemented/feature/2026-06-30-interception-seams.md
index ea371ad55a..1356afa06b 100644
--- a/docs/rfc/implemented/feature/2026-06-30-interception-seams.md
+++ b/docs/rfc/implemented/feature/2026-06-30-interception-seams.md
@@ -14,9 +14,9 @@ The canonical surface separates transformable policy, around-dispatch control, a
 
 **Agent events** (`dsh-agent`):
 - `agent/session-start(agent, source)` — emit, once before turn 1, carrying a `SessionStartSource` (`startup` for a fresh/forked create, `resume` for a reloaded persisted session; `clear`/`compact` reserved). A pure notification — it CANNOT block startup (a deliberate gap: a bridge logs/injects, it does not gate startup). A listener seeds context via `agent.inject()`.
-- `agent/prompt-submit(agent, content, source, next) → PromptDecision` — waterfall, fired per drained queued message inside the open turn, before the `user/message` append. `allow` (optionally rewriting the prompt `content` or attaching `additionalContext`) or `block` (dropping the prompt; the loop appends a durable `prompt/blocked` in its place — see the dispatch note below).
+- `agent/prompt-submit(agent, content, source, next) → PromptDecision` — waterfall, fired per drained queued message inside the open turn, before the `user/message` append. `allow` (optionally rewriting the prompt `content` or attaching separately sourced `additionalContexts[]`) or `block` (dropping the prompt; the loop appends a durable `prompt/blocked` in its place — see the dispatch note below).
 
-**`agent/turn-continuation`** receives and returns a `ContinuationDecision`. A `{action:'continue', reason?}` may carry model-facing context recorded as next-step steering in the same turn — the typed twin of the `/goal` step-end-steer pattern.
+**`agent/turn-continuation`** receives and returns a `ContinuationDecision`. A `{action:'continue', reason?}` may carry model-facing content and source recorded as next-step steering in the same turn — the typed twin of the `/goal` step-end-steer pattern. It is not a `context/message`, so its type does not offer a context envelope or durable context metadata.
 
 ### The tool pipeline gives each phase one kind of authority
 
@@ -25,7 +25,7 @@ Every call follows `tools/pre-execute` → guards → `tools/execute` → dispat
 - **`tools/pre-execute`** is the extensible waterfall gate. Its `PreToolDecision` allows, denies, or asks. Deny skips `tools/execute` and core dispatch. Ask resolves through the optional approval seam: only `allowed-once` continues through guards and dispatch; rejection, cancellation, an unavailable channel, a missing approval service, or an agent-less call becomes a normalized denial. Every outcome still reaches post-policy and final observers.
 - **`ctx.tools.guard()`** installs synchronous scope-aware policy after the whole pre-execute waterfall. A guard may deny or abstain, never force-allow, so listener ordering cannot resurrect an operation that a final invariant forbids.
 - **`tools/execute`** is the around-dispatch waterfall for timeout, retry, and metrics plugins. A wrapper delegates to core dispatch with `next()`, may add, replace, or remove only `exec.signal` before doing so, and receives the already-normalized result of a thrown or unknown tool; returning its own valid result short-circuits dispatch.
-- **`tools/post-execute`** is the inspect/transform waterfall. Its `PostToolDecision` accepts, blocks with feedback, optionally replaces content, or attaches `additionalContext`; in-place mutation of the result is not a transform channel, because the registry rebuilds the outcome from a protected snapshot plus the returned decision.
+- **`tools/post-execute`** is the inspect/transform waterfall. Its `PostToolDecision` accepts, blocks with feedback, optionally replaces content, or attaches `additionalContexts`. The returned decision is the supported transform channel; after the waterfall, the registry materializes the complete outcome once before final observation.
 - **`tools/result`** is the synchronous contained notification after every transform, lossless-JSON materialization, and the outer error boundary. It receives the same frozen execution identity and an immutable snapshot of the authoritative result; observer failures are contained per listener and cannot change or reject `ToolRegistry.execute()`'s returned outcome.
 
 Core dispatch and the tool body sit inside normalization boundaries, so tool, listener, malformed-result, non-JSON result, and identity-shape failures resolve as JSON-safe `isError` results rather than escaping the turn. A post-execute listener can therefore inspect a thrown tool, and a final observer sees exactly what the caller receives and the session log can persist.
@@ -34,9 +34,9 @@ Core dispatch and the tool body sit inside normalization boundaries, so tool, li
 
 ### Three load-bearing loop decisions
 
-1. **Open the turn before prompt policy.** A fully blocked batch becomes a zero-step `rejected` turn, preserving enclosure and giving ACP a durable terminal event. Every veto also records `prompt/blocked` with the original prompt and reason, so mixed batches retain blocked inputs. Allowed `additionalContext` is injected into the open turn.
+1. **Open the turn before prompt policy.** A fully blocked batch becomes a zero-step `rejected` turn, preserving enclosure and giving ACP a durable terminal event. Every veto also records `prompt/blocked` with the original prompt and reason, so mixed batches retain blocked inputs. Every allowed `additionalContexts` entry is injected into the open turn.
 
-2. **Post-tool `additionalContext` is buffered and appended AFTER all `tool/result`s.** `content`/`feedback` shape the result `execute()` returns, but `additionalContext` is a SEPARATE `context/message`, and a single step can carry multiple tool calls. Appending context right after each result would interleave `result(c1) → context → result(c2)` and break tool-call/result adjacency. So `execute()` surfaces `additionalContext` on its `ToolExecutionResult`, and the loop buffers every per-call context for the step and appends them as `context/message`(s) only after every `tool/result` is appended.
+2. **Post-tool `additionalContexts` and asynchronous injections enter the active-batch FIFO and append when that batch settles.** `content`/`feedback` shape the result `execute()` returns, but each context is a separate `context/message`, and a single step or composite tool can produce many. Appending context immediately would interleave `result(c1) → context → result(c2)` or place nested context before its outer result, breaking tool-call/result adjacency. `ToolRunContext.deferContext()` therefore collects nested-dispatch context through failures, `execute()` surfaces the ordered array on `ToolExecutionResult`, and the loop accepts it into the same FIFO as `agent.inject()` calls made during execution. The FIFO appends after every recorded result when the batch settles, including before an interrupted turn closes. An accepted outer call preserves deferred contexts before decision contexts; an outer block discards deferred contexts and exposes only contexts explicitly supplied by the blocking decision.
 
 3. **A forced `continue` `reason` is enqueued through the steering channel**, so the next step's top-of-loop drain records it as steering for the continued turn — next-*step* steering within the SAME turn, not a next-*turn* prompt (matching the existing `hasSteering` force-continue override).
 
diff --git a/docs/rfc/implemented/feature/2026-07-06-explicit-tool-order.md b/docs/rfc/implemented/feature/2026-07-06-explicit-tool-order.md
index c56e1eb0ba..e841de8973 100644
--- a/docs/rfc/implemented/feature/2026-07-06-explicit-tool-order.md
+++ b/docs/rfc/implemented/feature/2026-07-06-explicit-tool-order.md
@@ -38,7 +38,8 @@ Config plumbing follows the `persona` precedent, and `toolOrder` sits beside it:
 
 - Every registry-built assembly starts with a deterministic tool order on every host; absent an expert listener that deliberately changes it, every `request/header` event and model request inherits that order. The CI-vs-local registration-order flip is structurally gone, and the default is lexicographic.
 - The initial `PromptAssembly.tools` is canonical, so waterfall listeners start from the model-facing order; provider registration order is observable nowhere before that cooperative seam.
-- A pure tool reordering between steps is representable only as a `request/header` `'fallback'` snapshot (the name-keyed `ToolsDelta` cannot express it); with a stable canonical order such reorders no longer occur in practice, so the fallback path stays a safety valve.
+- The snapshot suite's single pinned request-header fixture (`text-turn`) carries the new canonical tool order; every other ACP snapshot keeps the header bulk scrubbed as `{{system}}`/`{{tools}}`, per the pinned-header design.
+- A pure tool reordering between steps is logged like any other header change: a full `request/header` snapshot with reason `'change'`. Stable canonical order prevents registration timing from creating such changes in the ordinary path.
 - The `toolOrder` key rides the app → `agent-core` → `SystemPrompt` forwarding chain, so deployments set it next to `persona` in the app config; `dsh-llm` and the agent loop are untouched.
 - A misspelled or unloaded tool name in `toolOrder` fails the turn at prompt assembly, not the boot: the loop assembles inside the turn (after `turn/start`, before `step/start`), so the rejection reaches the turn's outer catch — the turn closes balanced with an `error` reason carrying the message, `agent/error` mirrors it, no step opens, no `request/header` is logged, no request reaches the adapter, and the agent returns to idle. Every turn fails identically until the config is fixed; the process itself stays up (matching the repo rule that explicit config references must not be silently ignored — the enforcement point is the assembly because no earlier universal moment exists).
 - A tool provider that returns the reserved rest-entry name has the same prompt-assembly failure shape as an unknown listed name. This keeps the sentinel from becoming an ambiguous real tool and preserves the "never drops a tool" ordering contract.
diff --git a/docs/rfc/implemented/feature/2026-07-06-sandbox.md b/docs/rfc/implemented/feature/2026-07-06-sandbox.md
index 28f2066ee9..e236154cdc 100644
--- a/docs/rfc/implemented/feature/2026-07-06-sandbox.md
+++ b/docs/rfc/implemented/feature/2026-07-06-sandbox.md
@@ -70,7 +70,7 @@ Backend profiles share the mode contract but differ in necessary host grants. La
 
 #### The bash consumer
 
-`dsh-bash-sandbox` reuses local process execution and asks `ctx.sandbox` to wrap the exact bash argv. A kernel denial is a result fact independent of exit status and is inferred only from the selected wrap's stderr dialect. Runner failure outranks denial because it means the command never ran: foreground calls throw `SANDBOX_UNAVAILABLE`, while settled background tasks set `sandbox.runnerFailed` for `bash_output`. This keeps broken confinement distinct from both task failure and an enforced denial.
+`dsh-bash-sandbox` extends `LocalBashExecutor` and hands `ctx.sandbox` the exact `['bash', '-c', command]` argv it is about to spawn. A denial is an orthogonal result fact, conservatively classified from the active runner's stderr dialect. A runner failure outranks denial: foreground execution throws `SANDBOX_UNAVAILABLE`; a settled `BashProcess` stamps `sandbox.runnerFailed`, and the bash producer renders it through generic `task_output`.
 
 The model's view is result facts only: the static tool description explains the denial marker (`[sandbox: file access denied under <mode> mode]`), encourages attempting commands that may be denied, and forbids retrying around a denial; when the escalation fields are advertised, a denied result additionally carries the escalation hint itself, so the sanctioned same-turn retry is prompted at the decision point rather than depending on the model recalling the description (§ Escalation). No prompt section states the sandbox mode (§ Per-session modes).
 
@@ -78,13 +78,13 @@ The model's view is result facts only: the static tool description explains the
 
 `BashExecRequest.sandboxMode` is an optional per-call input; resolved specs make the field explicit. `BashExecutor.sandboxMode` advertises whether the mounted executor can honor it, so only a confining composition exposes escalation. The seam accepts any explicit mode; the tool owns the wider-only escalation rule. Non-sandboxing executors remain honestly unconfined.
 
-`SandboxBashExecutor.resolve()` stamps the effective mode — escalation grant > session override > configured default — so `run()`/`start()` read the spec, never the config. The `danger-full-access` branch, the confine call, and the result facts all key off the spec's mode, and the per-task facts map carries each task's mode alongside its wrap facts (`notifyTaskDone()` stamps from the map entry): one escalated call — foreground or background — reports the mode it ACTUALLY ran under while every neighbor keeps its own.
+`SandboxBashExecutor.resolve()` stamps the effective mode — escalation grant > session override > configured default — so `run()`/`start()` read the spec, never the config. Per-process wrap facts are keyed by the returned `BashProcess`; `onProcessDone()` classifies stderr and stamps that handle before `done` resolves, so overlapping processes retain their own modes and runner dialects.
 
 When a confining executor is mounted, `bash` advertises paired `sandbox_permissions` and `justification` fields. The schema exposes the full closed escalation vocabulary because effective mode is per-session; execution rejects any target that is not strictly wider than that call's effective mode. Approval resolves before execution. `allowed-once` stamps the granted mode onto only that request, while `rejected`, `cancelled`, `unavailable`, a missing approval service, or a missing agent all fail closed with distinct results. No grant is persisted.
 
 Escalation is a same-turn retry of the denied command with the narrowest sufficient `sandbox_permissions` and a `justification`; the approval prompt is the consent step. It must be grounded in an actual denial, except when the session already observed the same denied access, and a disabled or rejected approval ends that command. The retry, approval decision, and result use existing tool and approval events. `dsh-tool-bash` owns the ask because the executor seam has neither the agent nor call id required for user interaction.
 
-Left open, recorded for the phase that picks them up: what a grant's scope identity is beyond the sandbox mode — the exact call, a path, a command prefix, the session, a time window — the question `allow_always` grant storage must answer before that option can be advertised; and how escalation is defined for `run_in_background` denials that arrive via `bash_output`.
+Left open: what a durable grant's scope identity is beyond the sandbox mode — exact call, path, command prefix, session, or time window — before an `allow_always` option can be advertised.
 
 #### Per-session modes: the session log as the store
 
@@ -155,7 +155,7 @@ Each phase gets its full design when picked up, validated against the code at th
 - **A generic `env/state` facts map with an owner service** — rejected: approval and sandbox compose independently, so neither's state may drag in a third package; single-key folds are one `findLast` each, dissolving the owner service; no invariant spans the knobs, so atomic multi-key patches bought nothing.
 - **Narrate via `agent/user-message` + a bus event** — rejected: it presupposes a turn-entry seam that does not exist (the real seam is `agent/prompt-submit`), and pre-step's position serves both the coalesced turn-entry notice and the mid-turn immediacy bound with one listener.
 - **A standing prompt statement of the sandbox mode (+ a switch narrator)** — shipped first, then removed on live evidence: with `Bash commands run under the "read-only" file sandbox.` in every request, the model refused to ATTEMPT denied-then-escalatable work (five of twelve turns in the first manual session ended with zero tool calls), turning the sandbox into a soft lockout. The denial marker names the mode at the moment it matters and the escalation fields carry the recovery; the approval knob keeps its statement because an auto-rejection is behaviorally indistinguishable from a human "no".
-- **Track "last told" with its own bookkeeping events** — rejected: the `request/header*` fold already records the exact prompt the model saw; parsing the closed candidate sentences back replaces a second bookkeeping stream — events are needed only where they ARE the store.
+- **Track "last told" with its own bookkeeping events** — rejected: the `request/header` fold already records the exact prompt the model saw; parsing the closed candidate sentences back replaces a second bookkeeping stream — events are needed only where they ARE the store.
 - **ACP session modes instead of config options** — rejected: the preset is already one deployment-defined config-option select, and modes are slated for removal in ACP v2.
 
 ## Consequences
@@ -194,8 +194,8 @@ Costs and accepted limits:
 - **`bwrap` is installed on my host but unusable (disabled unprivileged userns, an LSM denying `mount`) — what happens?** The chain probe is functional — it builds and enforces a real profile rather than checking `--version` — so a present-but-unusable `bwrap` fails its probe, selection falls to the registry-installed Landlock launcher, and the verdict is cached for the provider's lifetime.
 - **Does the sandbox restrict network or process visibility?** No — `SandboxMode` claims FILE effects only; the bwrap profile deliberately does not unshare pid, and no backend claims network. Whether network restriction becomes its own knob is left open in § The seam.
 - **Which tools actually run confined?** OS subprocesses through `ctx.bash` — the bash tools, and hook commands transitively. fs/web/todo execute in-process, where an `execve` wrapper is mechanically meaningless; their `read-only` semantics arrive with the cross-family deferred phase, and until then the contract says bash-only honestly.
-- **Does a granted escalation persist, or cover background tasks?** Neither: the grant is consumed by the very call that asked (foreground or background), that one call reports the mode it actually ran under, and every neighbor keeps its own. How escalation should be DEFINED for a background denial that only surfaces later via `bash_output` is left open in § Escalation.
-- **When does an editor's mode switch take effect?** Mid-turn: appended immediately, honored by the very next call's stamp. Idle: held on the bridge's session record, anchored at the next `agent/prompt-submit` inside its open turn, with N flips coalescing to at most one event (none if net-zero); a crash before anchoring reverts it and `session/load` reports the truth. The model is not told — its next command simply behaves under the new mode.
+- **Does a granted escalation persist?** No. The grant is consumed by the exact foreground or background call that asked; every neighboring call keeps its own effective mode. A later background denial surfaces through `task_output` and may ground a new exact-command retry.
+- **When does an editor's mode switch take effect?** Mid-turn: appended immediately, honored by the very next call's stamp. Idle: held on the bridge's session record, anchored at the next turn's `agent/prompt-submit`, with N flips coalescing to at most one event (none if net-zero); a crash before anchoring reverts it and `session/load` reports the truth. The model is not told — its next command simply behaves under the new mode.
 - **What survives a restart — and what if the operator changed the config default while the process was down?** Overrides replay from the session log (`effective = fold ?? config`), so a resumed session keeps its modes with zero catch-up machinery; a default that drifted offline changes behavior the same way a switch does (the approval policy, being stated, is additionally narrated with operator/config attribution).
 - **What does `enforcement: 'partial'` on a result mean?** The selected backend enforces the subset its kernel ABI governs — e.g. Landlock before ABI v3 does not govern path truncate — and says so structurally instead of refusing the host; the probe's report line distinguishes the cases. The bwrap and Seatbelt profiles govern every promised file effect by construction, so they always report `full`.
 
@@ -204,7 +204,7 @@ Costs and accepted limits:
 In-repo precedents this design copies or contrasts with:
 
 - [The capability-seams RFC](../architecture/2026-06-13-capability-seams.md) — the interface/implementation/consumer split and the "don't split preemptively" timing rule the second consumer satisfied.
-- The `dsh-bash` request/spec split and its `owner` field ([the bash vocabulary catalog](../../../core-data-structures/bash.md)) — the per-call carrier template `sandboxMode` rides, and the explicit-`resolve()` defaulting convention.
+- The `dsh-bash` request/spec split ([the bash vocabulary catalog](../../../core-data-structures/bash.md)) — the per-call carrier template `sandboxMode` rides, and the explicit-`resolve()` defaulting convention.
 - [The approval seam RFC](2026-07-06-approval-seam.md) — the channel escalation asks through; its answerer waterfall, audit pair, and one-package rationale are recorded there.
 - [Event-sourced sessions](../architecture/2026-06-11-event-sourced-sessions.md) and [the turn-enclosure invariant](../architecture/2026-06-15-turn-enclosure-invariant.md) — the log-as-store foundation the per-session modes fold over, and the commit boundary the anchoring design obeys.
 - [The interception-seams RFC](2026-06-30-interception-seams.md) — the `tools/pre-execute` vocabulary the escalation gate deliberately does not reuse (an escalating call has no pre-execute moment of its own).
diff --git a/docs/rfc/implemented/feature/2026-07-07-session-prefix.md b/docs/rfc/implemented/feature/2026-07-07-session-prefix.md
index f0d458368a..05f254a21e 100644
--- a/docs/rfc/implemented/feature/2026-07-07-session-prefix.md
+++ b/docs/rfc/implemented/feature/2026-07-07-session-prefix.md
@@ -15,28 +15,27 @@ The obvious third option — let a plugin edit the request's `messages` on the w
 Three properties carry the design:
 
 - **Request-only, header-logged.** `deriveMessages()` never returns the prefix; its one durable record is `EpochHeader.messagePrefix` on the instance's anchoring `request/header` snapshot — the channel the reconstructable-requests RFC already owns for the request's non-history half, so no new session event exists. The dev invariant ([dsh-invariants](../../../../packages/support/invariants/src/index.ts)) recomputes `messagePrefix + boundary derivation` against every loop-built request; an unlogged prefix cannot reach the wire.
-- **Frozen per instance.** Reuse is structural, not disciplined: the cached product cannot change mid-session, so the provider's prompt cache holds by construction and the prefix extends the cacheable region at zero marginal cost per step. A process restart or `ctx.agents.resume()` is a new instance: it recomposes, and any drift lands attributably on the `'resume'` header snapshot. This is the routing rule the seam creates: session-frozen openers ride the prefix; content that changes mid-session rides the append-only history channels (`agent.inject()`, a `tools/post-execute` decision's `additionalContext`, prompt-submit `additionalContext` — [the interception-seams RFC](2026-06-30-interception-seams.md)), each a durable `context/message` paid once and prefix-cached thereafter.
+- **Frozen per instance.** Reuse is structural, not disciplined: the cached product cannot change mid-session, so the provider's prompt cache holds by construction and the prefix extends the cacheable region at zero marginal cost per step. A process restart or `ctx.agents.resume()` is a new instance: it recomposes, and any drift lands attributably on the `'resume'` header snapshot. This is the routing rule the seam creates: session-frozen openers ride the prefix; content that changes mid-session rides the append-only history channels (`agent.inject()` or tool/prompt-submit `additionalContexts` — [the interception-seams RFC](2026-06-30-interception-seams.md)), each a durable `context/message` paid once and prefix-cached thereafter.
 - **Composed before the pressure gate.** Composition precedes the instance's first `agent/pre-step`, and the seam hands the composed value through: `agent/pre-step` carries a `sessionPrefix` parameter and `CompactService.compactIfNeeded(agent, fullSystemPrompt, sessionPrefix, signal)` counts it in its token-pressure estimate — a gate reading the previous instance's folded prefix instead would under-gate a resumed or forked instance whose contributor grew, skipping compaction and shipping an over-window first request. A composition interrupted by a cancel/dispose landing inside the waterfall is discarded, never cached: an abort-aware listener's degraded fallback cannot leak into later requests, and the next turn recomposes under a live signal.
 
 Because composition runs before the boundary snapshot, a composing listener's session append joins the CURRENT request's derived history. Compaction structurally cannot touch the prefix (or the system prompt): it rewrites surface nodes, and header state never enters the surface.
 
 ## Testing
 
-[Interception tests](../../../../packages/core/agent-loop/tests/interception.spec.ts) pin compose-once reuse with no header deltas, prepend order, empty-prefix omission, immutability, and composition before pre-step; [cancellation tests](../../../../packages/core/agent-loop/tests/cancel.spec.ts) pin discard and recomposition. Session codec, invariant, and compaction tests cover header round trips, request reconstruction, and prefix-aware pressure accounting. Snapshot normalization preserves prefix counts, while the [pinned-header scenario](../testing/2026-07-06-pin-request-header-content-in-one-scenario.md) owns content and the default example remains prefix-free. No prefix-specific e2e is needed because the seam is deterministic and provider-independent; the with-key [request-cache e2e](../../../../packages/core/agent-loop/tests/request-cache.e2e.ts) covers its cache economics.
+**Unit** — [interception.spec.ts](../../../../packages/core/agent-loop/tests/interception.spec.ts) pins compose-once across turns and steps (one composition and no changed headers), canonical prepend ordering, empty-prefix omission from the header, the frozen seed (in-place push throws), held-reference mutation immunity, and composition-precedes-pre-step with the seam receiving the composed value; [cancel.spec.ts](../../../../packages/core/agent-loop/tests/cancel.spec.ts) pins cancel/dispose landing inside the composition window and the discard-and-recompose stale-cache guard; dsh-session header tests cover canonical prefix snapshots and latest-snapshot folding; dsh-invariants tests pin the `messagePrefix + derivation` equation; dsh-compact-basic tests pin that the pressure estimate counts the handed prefix. **Snapshot** — the acp-snapshot normalizer scrubs header prefixes to count-preserving `{{messagePrefix}}` tokens (unit-covered in dsh-acp-snapshot); header content itself is pinned per [the pinned-header scenario RFC](../testing/2026-07-06-pin-request-header-content-in-one-scenario.md), and the example tree loads no prefix contributor, so live goldens stay prefix-free. **e2e** — none prefix-specific: the seam is provider-independent and deterministic; the with-key cache measurement in [request-cache.e2e.ts](../../../../packages/core/agent-loop/tests/request-cache.e2e.ts) already proves the cacheable-prefix economics the design rests on.
 
 ## Alternatives considered
 
-- **Per-request `before`/`after` slots recomputed every step** (the shape first proposed: a waterfall firing on every request, contributing frozen `before` messages ahead of the history and fresh `after` messages behind it) — rejected. A per-step `before` recompose invites silent drift — nothing anchors it to the log short of logging a header delta per step — and an `after` slot sits behind the growing history, so its tokens re-pay on every request and everything after it is uncacheable. Measured against the alternatives, every current update pattern is served cheaper by a durable append (paid once, cache-read thereafter), and the only content with no home was the session-stable opener — which wants freezing, not recomputation.
-- **A system-prompt section** (`system-prompt/assemble`) — rejected for this content: the assembly renders to the single `system` string, so message-shaped openers do not fit, and the system prompt is deliberately re-assembled per step (with header deltas when it changes) while the opener wants instance-frozen semantics.
+- **Per-request `before`/`after` slots recomputed every step** (the shape first proposed: a waterfall firing on every request, contributing frozen `before` messages ahead of the history and fresh `after` messages behind it) — rejected. A per-step `before` recompose invites drift that must be logged as a full changed header, and an `after` slot sits behind the growing history, so its tokens re-pay on every request and everything after it is uncacheable. Measured against the alternatives, every current update pattern is served cheaper by a durable append (paid once, cache-read thereafter), and the only content with no home was the session-stable opener — which wants freezing, not recomputation.
+- **A system-prompt section** (`system-prompt/assemble`) — rejected for this content: the assembly renders to the single `system` string, so message-shaped openers do not fit, and the system prompt is deliberately re-assembled per step (with a full changed header when it changes) while the opener wants instance-frozen semantics.
 - **A durable history opener** (`inject()` at session start) — rejected: permanent history is the failure mode in the problem statement — replayed everywhere, compactable, stale across resumes.
-- **Compose per turn instead of per instance** — rejected: a turn-boundary recompose either desyncs silently from the log or forces a header delta per change, and it busts the provider cache exactly as often as it fires; the legitimate refresh point is the instance boundary, where the `'resume'` snapshot already records drift attributably.
+- **Compose per turn instead of per instance** — rejected: a turn-boundary recompose either desyncs silently from the log or forces a full changed header, and it busts the provider cache exactly as often as it fires; the legitimate refresh point is the instance boundary, where the `'resume'` snapshot already records drift attributably.
 - **Compose lazily at the first request and let compaction read the folded header** (the shape as first merged) — superseded in review: the fold matches the live prefix only from the instance's second request on, so on a resumed/forked instance's first step the pressure gate read the PREVIOUS instance's prefix and could under-gate. Composing before the first pre-step and handing the live value through the seam makes the estimate exact at every step.
-- **A dedicated session event carrying the prefix** — rejected: the header events are the request's non-history record by design; a second event would be a second home for the same fact and another codec to keep total.
+- **A dedicated session event carrying the prefix** — rejected: request headers are the request's non-history record by design; a second event would be a second home for the same fact.
 
 ## Consequences
 
 - `agent/pre-step` and `CompactService.compactIfNeeded` carry a `sessionPrefix` parameter: every pre-step listener and compaction backend sees the real per-instance value (all in-repo implementations updated in the same change, per the pre-release stance).
 - A contributor whose content changes mid-session is not re-read until the next instance — by design. A deployment needing mid-session catalog updates routes the change notice through the append-only history channels and pays one durable `context/message`.
 - The dropped `after` slot leaves no request-only channel near the request tail; nothing in the repo needs one, and adding it back would re-open the every-step re-pay cost the design exists to avoid.
-- The `request/header-delta` `messagePrefix` arm (whole-array replacement, empty array encoding transition to absence) exists for codec totality; the loop never exercises it, because the cached prefix cannot change within an instance.
 - An empty composition is canonical absence: no-contributor deployments log no extra header bytes and their requests are the bare derivation.
diff --git a/docs/rfc/implemented/feature/2026-07-08-background-subagent-tasks.md b/docs/rfc/implemented/feature/2026-07-08-background-subagent-tasks.md
new file mode 100644
index 0000000000..830d93577b
--- /dev/null
+++ b/docs/rfc/implemented/feature/2026-07-08-background-subagent-tasks.md
@@ -0,0 +1,62 @@
+# RFC: Background subagent tasks
+
+Status: implemented
+
+## Problem
+
+The [subagent seam](2026-06-21-subagent-capability-seam.md) returns a `SubagentRun`, but the model-facing tool originally collected every run synchronously. Independent, slow delegations therefore held the parent call open or ran serially.
+
+Subagents need the same start, collect, list, stop, ownership, notification, and cleanup behavior as other long-running tools without adopting process-stream semantics. The child session remains the detailed trace; the parent needs the final answer and task status. A background child also outlives its starting tool call, so its cancellation and owner-disposal contracts must be explicit.
+
+## Decision
+
+Each `dsh-tool-subagent` instance may expose `run_in_background`, controlled by `enableRunInBackground` and enabled by default. A disabled instance omits the parameter and rejects a forced background argument at execution. Provider selection remains deployment configuration, so one instance still registers one distinctly named tool for one provider.
+
+Background subagents use the [generic background task runtime](../architecture/2026-06-20-generic-long-running-tool-runtime.md). Collection, listing, cancellation, completion notices, and prompt guidance come from `task_output`, `task_list`, and `task_kill`; there are no subagent-specific companion tools.
+
+Foreground calls retain their synchronous contract: await provider startup and `run.result`, return final text only for `completed`, map other terminal reasons to an errored tool result, and always dispose the run before returning.
+
+For a background call, the tool validates the parent and refuses an already-aborted execution signal before calling `ctx.tasks.start()`. The task runtime preflights the control surface and owner cleanup before invoking the producer starter. That starter creates an independent `AbortController` and begins `ctx.subagents.start()`; after the id is returned, the tool-call signal no longer owns the child.
+
+The task registration maps the subagent seam as follows:
+
+- `kind` is `subagent`, `label` is the model-supplied description, and `owner` is the parent agent.
+- `cancel(reason?)` aborts the task-owned controller. The same signal covers pending provider startup and the ready child.
+- `done` awaits provider startup, the child result, and `run.dispose()`. Completed runs return final text, aborted runs become `killed`, and other stop reasons become `failed`. Startup, result, and disposal failures become failed outcomes rather than rejected task promises.
+- `readOutput` is absent. While live, `task_output` returns status only; after settlement, it returns final output idempotently. Intermediate child activity remains in the child session.
+
+## Lifecycle
+
+A background subagent belongs to its parent agent and is not durable across owner closure. The task runtime attaches cleanup to the exact owner's scope. Agent disposal cancels the task and awaits startup rollback or child disposal before `AgentHandle.dispose()` resolves, preventing leaked child agents and sessions.
+
+Completion notices target the exact owner captured at start. If owner teardown has already disposed the injection target, the notice is dropped; cleanup, not notification, is the lifecycle guarantee.
+
+## Model guidance
+
+The generic task prompt teaches the shared habit: retain ids, continue independent work instead of busy-polling, collect relevant tasks before answering, and kill irrelevant work. The subagent schema adds only that background mode returns a task id and that `task_output` collects the result. Authorization and owner cleanup enforce the runtime boundary independently of prompt compliance.
+
+## Alternatives considered
+
+### Subagent-specific wait, output, and stop tools
+
+Capability-specific tools would duplicate the task protocol, teach another collect-and-stop habit, and complicate multiple provider instances. The generic runtime provides the required behavior without changing the tool's one-provider-per-instance shape.
+
+### Survival after owner closure
+
+Survival requires persistent task state, child-session recovery, a late-result delivery channel, and policy for abandoned owners. Owner-scoped cleanup gives process-local work a clear lifetime. Durable jobs require a separate design.
+
+### No owner checks for isolated clients
+
+Agents and logs may be session-scoped, but the task registry and predictable ids are runtime-global. The generic owner fence therefore applies to subagents like every other producer.
+
+### Incremental child transcript output
+
+Streaming child history into the parent would blur the log boundary and make provider behavior diverge. This surface exposes final output only; richer observation belongs to session or UI tooling.
+
+## Testing
+
+Unit coverage pins stop-reason mapping, dispose-before-report behavior, startup and result failures, pre-aborted refusal, detachment from the starting call's signal, cancellation before and after provider readiness, collection through the real task tools, the no-surface preflight fence, missing-runtime failure, and per-instance schema gating. Snapshot coverage pins the model-facing schemas.
+
+## Consequences
+
+The parent can fan out slow delegations and collect them through the same task controls used by bash. Child work no longer occupies the starting tool call, but it can consume resources until collected, killed, or owner-disposed. Prompt guidance encourages collection; owner cleanup provides the hard lifetime boundary. Deployments that require synchronous delegation can disable background mode per tool instance.
diff --git a/docs/rfc/implemented/feature/2026-07-08-repeat-tool-guard.md b/docs/rfc/implemented/feature/2026-07-08-repeat-tool-guard.md
index 324aa37256..4996fc2935 100644
--- a/docs/rfc/implemented/feature/2026-07-08-repeat-tool-guard.md
+++ b/docs/rfc/implemented/feature/2026-07-08-repeat-tool-guard.md
@@ -14,7 +14,7 @@ The guard is a loop-hygiene plugin, not a model-facing tool. It counts consecuti
 
 The plugin is `@deepseek-ai/dsh-repeat-tool-guard` at `packages/guard/repeat-tool-guard/`, opening the `guard/` group for loop-hygiene plugins (single-package groups have precedent: [the todo-write RFC](2026-06-29-todo-write-tool.md) shipped `todo/tool-todo`). It registers three listeners and holds all state in plugin-local maps keyed by `AgentId` — the tool registry is a context-level singleton whose waterfalls interleave every agent's calls (subagents run on the same context), so per-agent keying is correctness, not polish.
 
-- **`tools/post-execute` (waterfall)** — the one detection point. The listener receives `(exec, result)` together, so counting and reminder delivery need no cross-event pending map (the pi extension needs one only because its `tool_call`/`tool_result` hooks are separate events). It always delegates via `next()` and, when a threshold is hit, folds a reminder onto the downstream decision's `additionalContext` — the observe-and-enrich posture [the hooks bridges](2026-06-30-hook-bridges.md) already use, honoring the waterfall contract. Counting happens here rather than in `tools/pre-execute` because post-execute also runs for denied calls (`ToolRegistry.execute` routes a deny through the same pipeline), and a model hammering a denied call is exactly the loop worth breaking.
+- **`tools/post-execute` (waterfall)** — the one detection point. The listener receives `(exec, result)` together, so counting and reminder delivery need no cross-event pending map (the pi extension needs one only because its `tool_call`/`tool_result` hooks are separate events). It always delegates via `next()` and, when a threshold is hit, prepends a reminder to the downstream decision's `additionalContexts` — the observe-and-enrich posture [the hooks bridges](2026-06-30-hook-bridges.md) already use, honoring the waterfall contract. Counting happens here rather than in `tools/pre-execute` because post-execute also runs for denied calls (`ToolRegistry.execute` routes a deny through the same pipeline), and a model hammering a denied call is exactly the loop worth breaking.
 - **`agent/prompt-submit` (waterfall)** — pure reset hook: delegate via `next()`, clear the submitting agent's chain. A user interjection changes the context; repetition across it is not a loop.
 - **`agent/status` (emit)** — on `disposed`, drop the agent's state, bounding the maps over harness lifetime.
 
@@ -29,7 +29,7 @@ Two deliberate rules, both documented in [the package README](../../../../packag
 
 ### Reminder delivery
 
-Reminders use `additionalContext` with the plugin source, preserving the original `tool/result`. The first threshold emits a short nudge; later thresholds include the tool, count, and a bounded argument preview while comparison still uses the full canonical string. Existing downstream context is concatenated under the guard's source because `HookContext` supports one source.
+Reminders ride `additionalContexts` as their own entries (source `{kind: 'plugin', plugin: 'repeat-tool-guard'}` — the label is load-bearing per `HookContext`), never a `content` replacement: the `tool/result` event stays the tool's own output for audit, and the loop appends buffered contexts as `context/message`s after the step's results, which the session renders as tagged synthetic-user envelopes and derived history replays. Thresholds escalate: the first configured threshold gets a short "you are repeating yourself, analyze the previous result" nudge; each later threshold gets the detailed form naming the tool, the repeat count, and the canonical arguments (head-truncated at `argumentsPreviewChars`, default 500 — a looping `write`-sized payload must not ride into the next request unbounded; the chain key always compares the full canonical string), and stating that the calls made no progress. The pi original hardcodes the gentle text to the literal count 3; the guard keys it to `thresholds[0]`, fixing that bug in the port. A downstream hook bridge contribution remains a separate array entry, so both plugins retain their source, envelope, and metadata.
 
 ### Config
 
@@ -53,7 +53,7 @@ Reminders use `additionalContext` with the plugin source, preserving the origina
 
 ## Alternatives considered
 
-- **Append the reminder into the tool result** (`accept` with replaced `content` — the pi extension's mechanism, which patches result content because that is the only channel its API offers) — rejected: it makes the logged `tool/result` lie about what the tool returned, and `additionalContext` exists precisely as the separate sanctioned channel for post-execute commentary, with loop-level buffering that preserves call/result adjacency.
+- **Append the reminder into the tool result** (`accept` with replaced `content` — the pi extension's mechanism, which patches result content because that is the only channel its API offers) — rejected: it makes the logged `tool/result` lie about what the tool returned, and `additionalContexts` is the separate sanctioned channel for post-execute commentary, with loop-level buffering that preserves call/result adjacency.
 - **Count in `tools/pre-execute` with a pending-reminder map** (the pi two-phase shape) — rejected: post-execute alone sees `(exec, result)` together and also fires for denied calls, so one listener with no cross-event state covers strictly more attempts with less machinery.
 - **Escalate to `block` at the highest threshold** — rejected for the initial scope: a blocked call punishes legitimate identical repeats (polling a long-running terminal, re-checking a file the agent expects to change), and an advisory reminder keeps the model in control. Revisit with evidence; the decision shape (`PostToolDecision`) already supports it.
 - **A per-deployment external hook via the CC/Codex bridges** (a `PostToolUse` script) — rejected as the answer: it works for one deployment, but a shipped, unit-tested, `cordis.yml`-configurable plugin is the harness-native form, without per-call subprocess cost.
@@ -65,7 +65,8 @@ Reminders use `additionalContext` with the plugin source, preserving the origina
 
 - The reminder is advisory by design: idempotent polling patterns that repeat identical calls on purpose still receive nudges past the thresholds, and the pressure valves are config (`thresholds`, `exclude`) plus reminder text that explicitly allows finishing when enough evidence has been gathered. Each trigger costs reminder tokens on the next request; thresholds bound the frequency.
 - Chain state is in-memory only: a session resumed from persistence starts with a fresh chain, so a loop spanning a resume draws its reminders later than a live one — accepted, the guard is a heuristic nudge, not a logged invariant, and persisting counter state would buy little for real complexity.
-- When multiple post-execute producers attach context on one call, the fold concatenates under the guard's `source`; ordering between plugins follows listener registration order. The seam cannot represent mixed provenance — a limit inherited from `HookContext`, not owned by this plugin.
+- When multiple post-execute producers attach context on one call, each contribution stays a separate `HookContext`; ordering follows waterfall nesting and each entry retains its own provenance.
+- Implementing the snapshot tier surfaced a hidden assumption in the suite kit: the fixture guard equated "authored model scenario" with "override-driven". The `Scenario` table now carries an explicit `overridden` flag, and the sidecar's presence is checked BOTH ways against it (an unregistered stray sidecar would silently replace the derived script) — the suite kit is stricter than it was before this plugin existed.
 
 ## Deferred
 
diff --git a/docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md b/docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md
index 7ea5f4390e..4d592af42b 100644
--- a/docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md
+++ b/docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md
@@ -52,7 +52,7 @@ Freshness is gated like every generated artifact: `pnpm run verify-cordis-api` (
 
 The plugin exposes one config field, validated by schemastery and documented in [the config catalog](../../../config-catalog.md): `vmTimeoutMs` (default 5000), the millisecond bound on the synchronous portion of mount-code evaluation. Tool names, the `cordis-dynamic` group name, and the `dyn-` id prefix are structural vocabulary and stay fixed. All three tools render as `generic` cards per [the tool cookbook](../../../cookbook/adding-a-tool.md) (`cordis_inspect` a `read`, `cordis_mount` an `execute` carrying the code as `rawInput`, `cordis_unmount` a `delete`), with no `presentResult` overrides.
 
-Model-visible ⟺ logged holds with no new session event type: a mount or unmount is visible only through its own `tool/call` / `tool/result` pair, which the loop logs, and the changed tool set a mount induces is logged by the request-header delta the loop already emits when schemas change between steps. There is deliberately no `cordis/mount` provenance event — it would duplicate what the tool-call pair records. Dynamic mounts are process-lifetime, not session state: resuming a persisted session rehydrates the conversation but does not re-mount plugins.
+Model-visible ⟺ logged holds with no new session event type: a mount or unmount is visible only through its own `tool/call` / `tool/result` pair, which the loop logs, and the changed tool set a mount induces is logged by the full changed request header the loop emits when schemas change between steps. There is deliberately no `cordis/mount` provenance event — it would duplicate what the tool-call pair records. Dynamic mounts are process-lifetime, not session state: resuming a persisted session rehydrates the conversation but does not re-mount plugins.
 
 ## Alternatives considered
 
@@ -71,7 +71,7 @@ The correctness investment therefore goes where it pays for every capability at
 
 **A hand-maintained service/event reference in the tool.** The first cut of the inspect tool carried a hand-written table of service method signatures. It was replaced by the generated `api-catalog.ts` because a hand table drifts from the JSDoc the moment a signature changes and nothing gates the drift, whereas the generated artifact is freshness-checked against the same AST the docs use.
 
-**A new `cordis/mount` session event.** A durable provenance event recording each mount (source, name) has clear precedent (`hook/invoked`, `compact/start`). It was declined for v1: mount and unmount are already visible as `tool/call` / `tool/result` pairs and the tool-set change is already logged as a request-header delta, so a dedicated event would only duplicate the record. It remains addable if an audit use case needs mount provenance separable from the tool call.
+**A new `cordis/mount` session event.** A durable provenance event recording each mount (source, name) has clear precedent (`hook/invoked`, `compact/start`). It was declined for v1: mount and unmount are already visible as `tool/call` / `tool/result` pairs and the tool-set change is already logged as a full changed request header, so a dedicated event would only duplicate the record. It remains addable if an audit use case needs mount provenance separable from the tool call.
 
 **A hardened / capability-restricted sandbox.** Trapping Node built-ins and handing mount code a whitelist façade rather than the raw context might suggest an intent to sandbox for safety. It is explicitly not that: the traps and the façade narrow the *surface* mount code sees — steering it onto cordis services and away from leak-prone Node built-ins and framework internals — for correctness and to close the unguarded-context escape, but the capabilities the façade exposes (`ctx.bash`, `ctx.fs`, `ctx.web`) reach the real runtime, so it is not a security boundary. A real one (separate process, permission prompts) was out of scope for a dev/opt-in toolset and would fight the entire point — handing the model the live runtime.
 
diff --git a/docs/rfc/implemented/feature/2026-07-09-bash-backed-grep-glob-discovery.md b/docs/rfc/implemented/feature/2026-07-09-bash-backed-grep-glob-discovery.md
new file mode 100644
index 0000000000..bc8a452c91
--- /dev/null
+++ b/docs/rfc/implemented/feature/2026-07-09-bash-backed-grep-glob-discovery.md
@@ -0,0 +1,166 @@
+# RFC: Bash-backed grep and glob discovery tools
+
+Status: implemented
+
+## Problem
+
+The harness needs model-facing `glob` and `grep` tools, but making them `ctx.fs` provider methods turns a local product convenience into a universal filesystem backend contract. Local workspace discovery is naturally a process-backed `rg` workflow; remote or virtual filesystem backends may expose their own search API, may not share a local `ripgrep` view, or may not support discovery at all. The v1 should not require every filesystem backend to implement search before the file read/write/edit seam has proven that need.
+
+Search output also has two distinct budgets. The tool needs enough raw `rg` output to compute a stable logical result, but the model should receive only a bounded preview plus a recovery path when the formatted result is larger than the inline budget. The generic spill policy only sees the final tool result, so it cannot recover matches that a search tool already omitted. Search therefore needs tool-owned retention and best-effort formatted-result spill.
+
+## Decision
+
+`glob` and `grep` are model-facing tools in `@deepseek-ai/dsh-tool-fs-search`, backed by the bash seam, not by new `ctx.fs` provider methods. The package registers model-facing filesystem discovery tools, but execution uses `ctx.bash.resolve(request)` followed by `ctx.bash.run(spec)` with fixed `rg` command templates assembled by the tool. The tool layer owns schemas, argument validation, shell quoting, result parsing, result formatting, retention, formatted-result spill handoff, and timeout declaration. The bash executor owns request defaulting/capping, subprocess execution, process-group termination, environment scrubbing, raw output capture, and backend substitution across local, sandboxed, or remote bash implementations.
+
+The tools do not use `ctx.bash.start()` and do not create model-visible background tasks. They run as ordinary foreground tools from the agent loop's perspective: the tool call returns only after the `rg` command exits, times out, is aborted, or fails. `defineTool({ timeoutMs })` declares the cooperative tool-call budget, `@deepseek-ai/dsh-timeout-policy` enforces it through `exec.signal`, and the tool forwards that signal into the bash request before `resolve()` / `run()`. The bash backend's own timeout remains a second safety cap; whichever aborts first wins.
+
+The tools align `path` with Claude Code's search tools while binding resolution to the bash workdir, not to `ctx.fs`. The tool derives the bash request workdir from `exec.agent?.session.header.cwd`, mirroring `dsh-tool-bash` and `dsh-tool-fs`; when no session cwd exists, it omits `request.workdir` so the bash implementation applies its configured cwd or process cwd through `resolve()`. For `grep`, `path` is an optional ripgrep target and may be a file or directory; omitted means the resolved bash workdir. For `glob`, `path` is an optional directory search root; omitted means the resolved bash workdir. Relative `path` values resolve against that workdir. Returned paths are displayed relative to the resolved bash workdir when possible and are intended to be follow-up-readable only in co-located deployments where the bash workdir and filesystem `read` root are the same workspace. v1 documents that deployment requirement but does not perform runtime cross-service validation. Remote or virtual filesystem search is deferred until there is a shared workspace/root contract or a provider-specific search backend.
+
+The package does not inject `fs`. It injects `tools`, `systemPrompt`, and `bash`; it deliberately reads `spillStore` with `ctx.get('spillStore')` instead of static inject because formatted-result spill is optional. Existing `@deepseek-ai/dsh-tool-fs` deployments that only want `read` / `write` / `edit` do not need to load bash.
+
+### Package shape
+
+The v1 package stays small. Inside `@deepseek-ai/dsh-tool-fs-search`, the source layout is:
+
+```text
+src/index.ts
+src/glob.ts
+src/grep.ts
+src/search-core.ts
+src/shell-quote.ts
+```
+
+`glob.ts` and `grep.ts` own their parameter validation, command construction, result parsing, formatting, and registration. `shell-quote.ts` is one shared helper because shell quoting is the safety boundary both tools must use; `search-core.ts` is the other (an implementation-time amendment to the original four-file plan): the `SEARCH_*` error vocabulary, the bash-run + raw-output acquisition, the formatted-spill handoff, and workdir-relative display are byte-identical between the two tools, and duplicating that delicate plumbing per tool is exactly the missed extraction the symmetry convention flags. Command builders must not hand-roll quoting or concatenate unquoted model-controlled values into the shell command.
+
+### Schemas and config
+
+`glob` exposes the small discovery shape:
+
+```ts
+interface GlobArgs {
+  pattern: string
+  path?: string
+}
+```
+
+`grep` exposes the OpenCode-style minimal shape:
+
+```ts
+interface GrepArgs {
+  pattern: string
+  path?: string
+  include?: string
+}
+```
+
+Routine budgets stay out of the model-facing schema. `@deepseek-ai/dsh-tool-fs-search` owns these defaulted, validated config fields:
+
+| Field | Default | Role |
+|---|---:|---|
+| `globMaxResults` | `100` | Max paths retained inline; matches Claude Code's default `GlobTool` result limit. |
+| `grepMaxMatches` | `250` | Max flat matches retained inline; matches Claude Code's default `GrepTool` `head_limit`. |
+| `grepMaxLineBytes` | `2000` | Max bytes retained for one matched-line preview, applied with `TextRetainer({ kind: 'head', maxBytes: grepMaxLineBytes })`. |
+| `rawOutputMaxBytes` | `20000000` | Max complete raw `rg` stdout the tool will parse; matches Claude Code's ripgrep raw buffer. |
+| `timeoutMs` | `30000` | Tool-call timeout attached to both tool definitions and enforced by `@deepseek-ai/dsh-timeout-policy`. |
+
+`globMaxResults` and `grepMaxMatches` use `ItemRetainer({ kind: 'head' })`. `grepMaxLineBytes` uses `TextRetainer({ kind: 'head', maxBytes: grepMaxLineBytes })` for each matched line so preview cuts preserve UTF-8 boundaries. This follows the [tool result retention library](../../implemented/architecture/2026-07-06-tool-result-retention-library.md) mapping for discovery items: collect the complete result, retain head items inline, and keep path mapping, grouping, and per-line preview outside the retainer. `grep` does not expose `case_insensitive`, `head_limit`, `offset`, `count`, multiline, context lines, output modes, or file type filters in v1. A model that needs surrounding context reads the matched file with `read`; a model that needs later results follows the returned spill locator's retrieval hint.
+
+The Claude Code values are reference points for the two-layer budget, not model-facing schema precedent. Its dedicated search tools buffer raw ripgrep output up to 20 MB for internal processing, use a 20-second ripgrep timeout on non-WSL platforms (60 seconds on WSL), then apply search-specific caps before the model sees a result: `GrepTool` defaults to `head_limit = 250` and persists formatted results above 20,000 characters, while `GlobTool` defaults to 100 paths and persists formatted results above 100,000 characters. This RFC mirrors the raw-buffer and inline-count defaults, chooses a 30-second default search timeout, and uses this harness's `ctx.spillStore.saveText()` path for formatted-result recovery.
+
+The `path` field follows the same split as Claude Code: `grep.path` is a file-or-directory ripgrep target, while `glob.path` is a directory search root. v1 does not expose a separate cwd/workdir argument on these tools.
+
+`include` is one positive glob filter, not a list and not an exclude syntax. Reject comma-separated or negated include patterns up front with a structured argument error. Every model-controlled value used in a shell command, including `pattern`, `path`, and `include`, must pass through the package-private shell quoting helper.
+
+### Execution
+
+`glob` builds a fixed `rg --files` command rooted at the resolved directory search root (`path` when supplied, else the bash workdir): `rg --files --glob <pattern> --sort=modified --no-ignore --hidden`, plus VCS metadata excludes for `.git`, `.svn`, `.hg`, `.bzr`, `.jj`, and `.sl`. This aligns with Claude Code on hidden/ignored-file discovery and modified-time ordering while keeping VCS internals out of broad searches. The tool parses one path per line, maps results back to paths relative to the bash workdir when possible, pushes each path into `ItemRetainer({ kind: 'head', maxItems: globMaxResults })`, and formats the full sorted path list for a spill artifact when the retained result is capped.
+
+`grep` builds a fixed line-oriented `rg --json` command against the supplied file/directory target (`path` when supplied, else the bash workdir) so file path, line number, and line text are parsed without colon-splitting ambiguity. It consumes `match` records, treats malformed JSON or malformed match records as `SEARCH_FAILED`, maps result paths relative to the bash workdir when possible, applies per-line preview retention with `grepMaxLineBytes`, pushes each match into `ItemRetainer({ kind: 'head', maxItems: grepMaxMatches })`, then groups only the retained preview matches by file for inline output. The spill artifact stores the full formatted match list, not only the omitted tail, so the retrieval hint points at the same logical result the model saw.
+
+Raw `rg` stdout is an internal transport detail. The tool requests `stdoutMaxBytes: rawOutputMaxBytes` through `ctx.bash.resolve()` and parses `stdout.text` only when the executor returns untruncated stdout within that cap. If stdout is larger than `rawOutputMaxBytes`, or the executor still returns `stdout.truncated`, the tool fails with a clear search error telling the model to narrow `pattern`, `path`, or `include`. The tool never exposes raw `rg` output or bash raw spill paths to the model.
+
+Only stdout is a parse source. Stderr is diagnostic text for invalid patterns, missing `rg`, and search failures; if bash truncates stderr, the tool uses the retained stderr tail with a truncation note and does not read `stderr.spillPath`.
+
+If `ctx.bash.run()` reports `aborted` because the tool timeout or caller cancellation fired, the tool returns a structured failure rather than pretending there were no matches. If bash reports its own timeout first, the tool likewise fails with a clear timeout message. Nonzero ripgrep exit semantics are tool-owned: exit 0 is success with matches, exit 1 is success with no matches, invalid pattern / missing `rg` / inaccessible search workdir are failures.
+
+Search failures use a package-owned `HarnessError` subclass with `SEARCH_*` codes, not `FsErrorCode`, because these tools are not `ctx.fs` provider operations. The v1 vocabulary is `SEARCH_INVALID_PATTERN`, `SEARCH_FAILED`, `SEARCH_RAW_OUTPUT_OVERFLOW`, and `SEARCH_ABORTED`. Model argument validation failures such as missing required fields, blank strings, or unsupported negated/list `include` values remain ordinary tool argument errors.
+
+### Formatted result spill
+
+`ctx.spillStore` is optional and used only for model-facing formatted results. This is the first tool-owned spill call pattern in the codebase, and it is intentional because search retention is item-level policy: `globMaxResults` caps paths and `grepMaxMatches` caps matches while the tool still holds the complete logical result. The generic `dsh-spill-policy` caps final text bytes on `tools/post-execute`; by then a search tool would already have omitted later paths or matches, so the policy cannot recover them.
+
+When a search produces more logical results than the inline cap and `ctx.spillStore` is present, the tool saves the complete formatted result with `saveText()`. The spill owner is the calling agent's session header id (`exec.agent?.session.header.id`); without that owner, the search keeps the inline result and reports that the complete result could not be saved. The spill source is the tool execution identity: `{ toolName: exec.name, callId: exec.callId, label: 'result' }`. The suggested filenames are `grep-results.txt` and `glob-results.txt`; the spill backend still treats them as hints, never paths.
+
+When spill storage is absent, the call has no session owner, or saving fails, the tool still returns the inline page and a footer explaining that the complete result could not be saved. Search success must not turn into an `isError` result solely because formatted-result spill storage is unavailable.
+
+The bash raw output stream and the formatted search spill artifact are different artifacts. Raw `rg` stdout is parsed only in memory within the requested bash stdout cap; the formatted spill artifact is the stable model-facing recovery locator produced by `ctx.spillStore.saveText()`.
+
+### Result shape
+
+A capped `glob` result with successful formatted spill returns the inline page and a spill notice:
+
+```text
+<first N paths>
+
+(Showing N of M paths. Full sorted result stored at: /.../session-abc123/9f8e7d-glob-results.txt. Use read with offset/limit, or grep this path to search within it.)
+```
+
+A capped `grep` result with successful formatted spill returns grouped preview matches and a spill notice:
+
+```text
+Found N of M matches
+
+<file>
+Line 12: ...
+
+(Full grep result stored at: /.../session-abc123/9f8e7d-grep-results.txt. Use read with offset/limit, or grep this path to search within it.)
+```
+
+If the complete logical result fits under the inline cap, no formatted spill artifact is created. If the complete logical result is too large but formatted spill is unavailable, the footer says that the result was capped and the complete result could not be saved. The `truncated` / omitted count is a budget fact, not an incomplete-search fact; timeout, invalid regex, missing `rg`, inaccessible workdirs, raw-output overflow, binary skips, and parse failures stay in tool-domain error or incomplete fields.
+
+## Alternatives considered
+
+**Put `glob` / `grep` on `ctx.fs`.** Rejected for v1: it forces every filesystem backend to grow a search API and makes local ripgrep behavior part of the provider seam. Search is useful product behavior, but it is not a universal text-storage primitive like `readText` or `writeText`.
+
+**Directly spawn ripgrep from `dsh-fs-local`.** Rejected for this RFC's v1: direct spawn gives the cleanest argv boundary, stdout/stderr control, and early-stop control, but it duplicates process execution concerns that the bash seam already owns: environment scrubbing, process-group kill, timeout propagation, sandbox/remote executor substitution, and bounded output capture. It remains a reasonable optimization if bash-backed search proves too shell-string-sensitive or if foreground streaming becomes necessary.
+
+**Use `ctx.bash.start()` for streaming early stop.** Rejected: `start()` creates model-visible background task semantics: task ids, owner tokens, `bash_output`, `bash_kill`, completion notifications, and no built-in timeout. `grep` needs a foreground tool result, not a background bash workflow. If streaming search becomes necessary, the right abstraction is a foreground streaming process handle on the bash/process seam, not borrowing the public background-task API.
+
+**Expose bash raw spill paths to the model.** Rejected: a bash raw spill path contains raw `rg` stdout (`rg --json` records for grep), not the stable formatted search result. Search parses raw stdout only as an internal transport; model recovery uses a formatted result saved through `ctx.spillStore.saveText()`.
+
+**Add `spillStore.saveFile()` for bash output normalization first.** Rejected for this RFC's v1: `saveFile()` would help a future bash normalization pass move existing executor spill files into session-scoped spill storage, but search only needs bounded in-memory raw `rg` stdout before producing the model-facing artifact. `saveText()` is sufficient for the formatted search result.
+
+**Rely on the generic `dsh-spill-policy`.** Rejected: generic post-execute spill sees only the final tool result. If `grep` / `glob` return the first page inline, the generic policy cannot recover omitted results. The search tools must save the complete formatted result themselves before returning the bounded model-facing text.
+
+**Expose Claude Code's full `GrepTool` schema.** Rejected for v1: `output_mode`, context flags, multiline, `head_limit`, `offset`, `case_insensitive`, and type filters make the model-facing surface into a ripgrep wrapper. This harness keeps routine budgets and continuation mechanics in deployment policy and spill artifacts.
+
+**Keep early-stop search and skip formatted spill artifacts.** Rejected for this proposal: early stop is more efficient but gives the model no path to inspect later results. The chosen v1 optimizes result recoverability and implementation simplicity, with `timeoutMs`, `rawOutputMaxBytes`, bash backend caps, and formatted spill artifacts as safety backstops.
+
+**Expand the bash seam with a raw-output reader first.** Rejected: a portable `readRawOutput(ref, maxBytes)` API would add reference lifetime, permission, and backend storage semantics. A per-run `stdoutMaxBytes` request is the narrower seam: search either receives complete stdout within `rawOutputMaxBytes` or fails clearly.
+
+## Testing
+
+- Tests prove an aborted `exec.signal` reaches the bash backend (same-reference spec assertion plus the `SEARCH_ABORTED` result), and cover command construction/quoting (malicious patterns, paths with spaces, leading-dash values, quotes, newlines, glob metacharacters — unit assertions plus a real `bash -c` round-trip for every hostile value), `grep.path` as file and directory targets, `glob.path` as a directory search root, invalid pattern handling, no matches, malformed `rg --json` output, matched-line preview truncation, raw-output overflow, timeout/abort, formatted spill success/failure, the package-owned `SEARCH_*` error codes, and the no-background-task invariant.
+- The first-party tool-owned spill precedent is covered directly: spill backend present, spill backend absent, `saveText()` failure, and missing spill owner.
+- The package has real Loader-path coverage for the namespace plugin export shape (`name`, `inject`, `Config`, and `apply`, with no default export).
+- A real-executor integration suite (`dsh-bash-local` + a real `rg`) verifies the world: hostile patterns stay inert, per-session cwd resolution, VCS-metadata exclusion, modification-time ordering, and real ripgrep stderr classification. It self-skips where `rg` is not on PATH (a CI accommodation mirroring the keyless e2e skip); the fake-executor suite alone carries the per-file 100% coverage gate.
+- Snapshot gap note for the transcript-visible spill notice: this landed with the gap note, not a snapshot. The snapshot tier replays the acp-agent tree, and adding the search plugin there changes the assembled system prompt — every golden would need re-recording with a real key, which the implementing environment did not hold. The spill notice's exact transcript text is pinned by unit tests (`formatGlobOutput`/`formatGrepOutput` and the through-the-registry spill tests); wiring the plugin into the acp-agent tree plus a `test:snapshot:record` pass is the follow-up for the next key-holding session.
+
+## Consequences
+
+- `glob` and `grep` are model-facing tools in `@deepseek-ai/dsh-tool-fs-search`, not `ctx.fs` provider methods and not part of the existing `@deepseek-ai/dsh-tool-fs` root plugin. The package injects `tools`, `systemPrompt`, and `bash`; it does not inject `fs`, and `ctx.spillStore` stays optional via `ctx.get('spillStore')`.
+- The schemas are exactly `glob(pattern, path?)` and `grep(pattern, path?, include?)`; search caps and timeout are defaulted, validated Config fields (`globMaxResults`, `grepMaxMatches`, `grepMaxLineBytes`, `rawOutputMaxBytes`, `timeoutMs`).
+- The tools execute through `ctx.bash.resolve(request)` → `ctx.bash.run(spec)`, forward `exec.signal`, never call `ctx.bash.start()`, and never expose a bash task id. The bash request workdir comes from `exec.agent?.session.header.cwd` when available; the resolved `spec.workdir` drives execution and relative-path display.
+- The tools request `stdoutMaxBytes: rawOutputMaxBytes` from the bash seam, parse only untruncated stdout within that cap, and treat over-cap or still-truncated raw output as a clear search failure; raw `rg` output is never exposed to the model.
+- Oversized complete formatted results are saved through `ctx.spillStore.saveText()` when available while inline results stay bounded; spill failure, a missing backend, or a missing owner preserves the inline result and reports the unsaved remainder — never an `isError`.
+- The package README, the generated config catalog, and exported JSDoc document the Config fields and `SEARCH_*` codes; the coding-agent example ships the tools (the acp-agent tree waits on the snapshot re-record above); the fs group README records the co-located bash/filesystem deployment requirement.
+
+## Risks
+
+Full-run `grep` can be slower than an early-stop search on broad patterns. The v1 accepts that cost for simpler implementation and complete-result recovery, bounded by tool timeout, bash timeout, `rawOutputMaxBytes`, and output caps. If this proves too slow, the direct-ripgrep or foreground-streaming alternatives remain available.
+
+Shell command construction is the sharpest safety edge. Because `ctx.bash` accepts a command string rather than an argv vector, the implementation must centralize shell quoting and test malicious patterns, paths with spaces, leading-dash patterns, quotes, newlines, and glob metacharacters.
+
+The v1 assumes a co-located bash/filesystem deployment. If bash searches one workspace and the `read` tool resolves paths against another, returned paths may not be follow-up-readable. The package documents this requirement but does not verify it at runtime.
+
+Spill locators are backend-owned. The current local backend returns local filesystem paths and works in deployments where `read`/`grep` can open those files; remote or workspace-confined deployments can use a backend whose locator and retrieval hint point at a supported retrieval mechanism.
diff --git a/docs/rfc/implemented/feature/2026-07-10-agent-session-identity-and-log-location.md b/docs/rfc/implemented/feature/2026-07-10-agent-session-identity-and-log-location.md
new file mode 100644
index 0000000000..118009dedb
--- /dev/null
+++ b/docs/rfc/implemented/feature/2026-07-10-agent-session-identity-and-log-location.md
@@ -0,0 +1,85 @@
+# RFC: Expose agent session identity and JSONL location to tools and hooks
+
+Status: implemented
+
+## Problem
+
+An agent can identify its workspace through `session.header.cwd`, but a model using bash cannot reliably identify the session that owns the call or the durable transcript that records it. Searching `./.sessions` guesses deployment config and JSONL layout; custom roots, alternate persistence backends, resume, forks, and concurrent parent/child agents make that guess unreliable. Hooks have the same need for transcript location, while future plugins may need to expose other harness-owned environment facts to shell commands.
+
+The boundary must preserve two properties: the owner of a fact decides how to resolve it, and every child receives a per-execution snapshot rather than process-global mutable state. In particular, a nested harness must not leak its ambient `DSH_*` values into a child whose current agent, persistence backend, or configuration differs.
+
+## Decision
+
+Extend the [`SessionPersistence`](../../implemented/architecture/2026-06-14-session-persistence.md) seam with a synchronous, side-effect-free location query:
+
+```ts
+import type { SessionHeader } from '@deepseek-ai/dsh-session'
+
+interface SessionLocation {
+  readonly kind: string
+  readonly path: string
+}
+
+interface SessionPersistence {
+  locate(meta: SessionHeader): SessionLocation | undefined
+}
+```
+
+`path` is an absolute local path to the backend's dedicated log for `meta`; `kind` identifies the representation. JSONL returns `{ kind: 'jsonl', path }` using its resolved root and path helpers. SQLite and any backend without an honest local per-session artifact return `undefined`. The query creates and flushes nothing, so it can report a lazy target path before that file exists.
+
+The model-facing bash package owns a `ctx.bashEnv` registry. A contributor declares its stable name, every `DSH_*` key it may return, a description for each key, and `resolve(execution: ToolExecution)`. Duplicate contributor names, duplicate key ownership, reserved keys, malformed declarations, undeclared runtime output, and non-string output fail loudly. Registration is a Cordis effect and is removed with the contributing plugin fiber. `list()` exposes declarations without running resolvers, keeping the environment surface enumerable for diagnostics and future prompt/UI consumers.
+
+The registry rebuilds a trusted overlay for every foreground and background bash `ToolExecution`:
+
+- `DSH_HOME` is always the absolute configured Harness home. The standalone [`@deepseek-ai/dsh-home`](../../../../packages/util/home/README.md) utility owns its precedence: explicit `dshHome`, then ambient `$DSH_HOME`, then `~/.dsh`.
+- `DSH_SHELL=1` is always present and identifies a model bash child managed by DeepSeek Harness.
+- `DSH_SESSION_ID` is present when the execution has an agent and equals `agent.session.header.id`.
+- The built-in persistence translator contributes `DSH_SESSION_JSONL` only when `ctx.sessionPersistence.locate(header)` returns `kind: 'jsonl'`.
+
+Session persistence remains the fact owner: JSONL does not depend on tool-bash or register shell variables itself, and hooks continue to consume `locate()` directly. Tool-bash is the translation layer from the persistence fact into a shell convention. Other plugins that need shell-visible facts depend on the registry and register their own keys; they do not modify `process.env`.
+
+The bash seam exports `DSH_ENV_PREFIX` as the single namespace source and derives `DshEnvironmentKey` from its `typeof`. Tool-bash derives built-in names and model guidance from that constant, while executors use it for filtering and channel validation. The seam carries the managed overlay separately as `BashExecRequest.dshEnv` / `BashExecSpec.dshEnv`. Ordinary `env` remains the general in-process plugin surface used by hooks, but cannot contain managed keys; symmetrically, `dshEnv` cannot contain ordinary keys. The local executor rejects either wrong channel before spawn, removes every inherited ambient managed key, applies its ordinary scrub/terminal environment/explicit `env`, and finally merges the trusted `dshEnv` snapshot. This guarantees that a missing value means absent now rather than inherited from an outer or previous harness. The model-facing tool still ignores model-supplied `env`/`stdin` arguments.
+
+The bash tool description teaches only the durable convention: current harness environment facts are available through managed `$DSH_*` variables and may be inspected when needed. It does not enumerate persistence-specific keys or add a permanent system-prompt section. Tool schemas are already logged in request headers and tool output is logged as `tool/result`, so no new session event is required.
+
+The [Claude Code and Codex hook bridges](../../implemented/feature/2026-06-30-hook-bridges.md) resolve transcript location from the same persistence seam when constructing payloads. Codex uses `transcript_path: string | null`; Claude Code preserves its string field and falls back to `''`. Hook lookup neither materializes nor flushes a session.
+
+## Peer product findings
+
+Peer products separate stable identity from physical storage. Codex injects stable `CODEX_THREAD_ID` into spawned shells while recorder and hook surfaces own transcript paths. Claude Code supplies `session_id` and `transcript_path` as structured hook/status input. OpenCode carries identity in structured tool context; Kimi Code expands a session placeholder; Reasonix keeps the active session path on its controller. The portable rule is to inject identity at the invocation boundary, let storage resolve location, and never use a process-global current-session variable in a concurrent harness.
+
+## Lifecycle and persistence semantics
+
+A fresh session receives its id before the first turn, so its first bash call can read `DSH_SESSION_ID` and a JSONL target. The JSONL file may still be absent until the first successful turn-end checkpoint, and during an open turn it contains only the last flushed prefix. `DSH_SESSION_JSONL` is a location hint, not an authorization credential or freshness guarantee.
+
+Resume reuses the loaded header and therefore the same id and location. Fork and spawn create new session ids and locations. Parent and child calls resolve from their own `ToolExecution.agent`; each command receives an immutable snapshot even when calls overlap. A persistence service replacement affects later collections because the translator queries `ctx.get('sessionPersistence')` at execution time; the registry itself is effect-scoped and HMR-safe.
+
+`dshHome` is session-independent deployment context. Agent-core resolves one value through `@deepseek-ai/dsh-home` and routes it to both tool-bash and local skill discovery; standalone consumers call the same resolver. If top-level `dshHome` and `skills.local.dshHome` are both supplied and resolve differently, composition fails instead of exposing contradictory homes. Persistence may change independently without freezing its facts into the session prefix.
+
+## Testing
+
+Unit coverage pins registry declaration validation, effect disposal, per-execution collection, the `dshHome` precedence, and the local executor's `DSH_*` scrub/rebuild order. Request-recording tests cover foreground/background snapshots, no-agent calls, absent/JSONL persistence, ignored model `env`, and parent/child isolation. JSONL/SQLite locator contract tests and both hook bridge suites pin available and unavailable transcript dialects.
+
+A keyless full-loop integration drives the real agent loop, JSONL persistence, tool-bash, and bash-local on the first turn. The child prints `DSH_HOME`, `DSH_SHELL`, session id, JSONL target, and an inherited stale sentinel; the test verifies current values, absence of the stale variable, pre-flush file absence, and the eventual persisted header. Snapshot coverage pins the generic bash description in the recorded request header. No with-key test is required because the contract is deterministic local execution rather than model choice.
+
+## Alternatives considered
+
+**Only an id plus `find`.** Search cannot know a custom root or backend layout and races under multiple sessions.
+
+**Only an absolute path.** A path can be unavailable, lazy, or representation-specific and is not stable session identity.
+
+**Global `process.env`.** Concurrent agents would overwrite one another and nested harnesses would inherit stale current-session values.
+
+**Put persistence instructions in the session prefix.** A session prefix is frozen while the active service can change across HMR or future backend switching; persistence-specific guidance would become stale.
+
+**A typed waterfall event.** Listeners cannot declare ownership without running, and later listeners can silently overwrite keys. A registry detects key conflicts at registration and remains enumerable.
+
+**Have each persistence backend register bash env directly.** That reverses the dependency from storage into one consumer and forces bash into deployments that do not use it. `locate()` is also still required by hooks.
+
+**A model-facing `session_info` tool.** It adds schema and another call while bash already supplies the query surface; the registry generalizes to future environment facts without one tool per fact.
+
+## Consequences
+
+Every model bash child receives current Harness home and shell identity, and agent calls additionally receive stable session identity. JSONL-backed calls get an optional target path; non-file persistence omits it honestly. The complete `DSH_*` namespace inside these children is managed by the harness: ambient values are removed, current trusted values are re-added, and ordinary callers cannot use `env` to bypass ownership checks.
+
+The namespace is discoverable but not secret. Paths can reveal configured roots, lazy targets can be absent or stale, and a command can override variables inside its own shell syntax. Consumers treat them as correlation and environment facts, verify transcript metadata when attribution matters, and rely on sandbox/filesystem policy rather than variable secrecy for authorization.
diff --git a/docs/rfc/implemented/feature/2026-07-10-parallel-tool-call-execution.md b/docs/rfc/implemented/feature/2026-07-10-parallel-tool-call-execution.md
new file mode 100644
index 0000000000..90fbc1cfb4
--- /dev/null
+++ b/docs/rfc/implemented/feature/2026-07-10-parallel-tool-call-execution.md
@@ -0,0 +1,101 @@
+# RFC: Parallel tool-call execution by per-call safety
+
+Status: implemented
+
+## Problem
+
+An assistant message may contain several sibling `tool-call` blocks. Running them serially adds the latency of independent reads and web requests even though the model has already requested them together.
+
+Concurrency is a host scheduling concern, not model-facing tool metadata. The loop needs to decide which calls may overlap without hardcoding tool names or exposing scheduler policy in the JSON schema.
+
+The session log remains authoritative: every started call has an audit event, every started call receives a result, and model history observes results in the original call order regardless of completion order.
+
+## Decision
+
+Each tool may provide an optional `isConcurrencySafe(args)` classifier. It is synchronous and pure: it examines only the current call's parsed arguments and performs no I/O or mutation. Only an explicit `true` opts in; a missing classifier, invalid arguments, a thrown classifier, or any other return value makes the call exclusive. The canonical type contract lives in the [tool data structures](../../../core-data-structures/tools.md).
+
+The classifier is deliberately unary. Returning `true` is the tool's promise that this call may overlap with any sibling call that also returns `true`; the scheduler does not compare calls or prove that their resource accesses are compatible.
+
+The unary classifier remains input-sensitive. A tool may classify a read-only operation as parallel and a mutating operation as exclusive. The interface cannot express relational rules such as "these writes are safe only when their paths differ," so a call whose safety depends on a sibling remains exclusive.
+
+`defineTool()` validates arguments before invoking a typed classifier. Invalid arguments classify as exclusive and produce the ordinary argument error only if the call executes. `ctx.tools.executionMode(exec)` resolves the live tool definition and returns the tagged `parallel` or `exclusive` mode; unknown tools fail closed to exclusive.
+
+A tagged mode, rather than a public boolean scheduler API, keeps resource-aware variants representable without changing the classifier contract.
+
+## Scheduling and ordering
+
+The loop waits for the complete assistant message, parses every call once, creates a distinct `ToolExecution` for each call, and scans them in model order. Consecutive parallel calls form one group; every exclusive call forms a singleton group and an ordering barrier. Groups execute sequentially. Classification is lazy: the scheduler resolves the next call after each barrier and reclassifies every later call before replenishing a parallel pool. If a registry mutation makes that call exclusive, the current pool drains before the call starts as the next barrier.
+
+For example:
+
+```text
+[parallel read(A), parallel read(B), exclusive write(A), parallel read(C)]
+
+→ [read(A), read(B)]
+→ [write(A)]
+→ [read(C)]
+```
+
+`read(A)` and `read(B)` may overlap. `write(A)` starts after both finish, and `read(C)` starts after the write finishes.
+
+Every group uses a rolling pool bounded by `maxParallelToolCalls`: the loop starts calls in model order up to the cap and starts another whenever one settles. An exclusive group is a pool of one. A cap of `1` preserves serial execution.
+
+Only dispatch and the tool body overlap. `tools/pre-execute` and `tools/post-execute` run in model order because middleware may maintain ordering-sensitive state. `tools/execute` wrappers run around concurrent dispatches and therefore must be reentrant across distinct executions.
+
+Each started call appends `tool/call` immediately before its pre-execute gate. Completed dispatches occupy model-order slots, and a commit cursor appends `tool/result` and collects `additionalContexts` only when the next slot is ready. Live surfaces may show several pending calls, but results and post-tool context remain model-ordered.
+
+An abort before a group starts records no calls from that group. An abort during a group stops replenishment, waits for already-started calls, commits their results in order, drains accepted batch context after those results, and then ends the step through the existing abort path. Calls that never start have no audit event.
+
+Code Mode remains outside this scheduler because the model emits one native `run_code` call. `run_code` and its internal dispatch queue remain serial; native sibling calls in `mode: 'both'` use the normal scheduler.
+
+## Safety contract
+
+A tool that returns `true` promises that its body is safe to run at the same time as other parallel calls. It must not directly mutate the parent session or other parent-owned state; it returns its outputs to the loop, which commits them in model order.
+
+Any shared state touched during execution must be concurrency-safe. This includes tool wrappers and providers: they may serialize internally or enforce their own capacity, but they must support concurrent dispatch without corrupting state.
+
+## Configuration and declarations
+
+`maxParallelToolCalls` is a positive AgentLoop deployment cap shared by every agent the factory creates. It defaults to `10`; `1` preserves serial execution. Exact fields and defaults live in the generated [configuration catalog](../../../config-catalog.md).
+
+The shipped declarations are conservative. Web search, web fetch, and filesystem read opt in. Filesystem writes and edits, bash tools, subagent delegation, workflow, user interaction, todo mutation, Code Mode, and Cordis mutation tools remain exclusive. A subagent may share its parent's workspace or external resources, and the unary classifier cannot prove that sibling delegations have disjoint effects. Bash has no proven input-sensitive classifier and remains exclusive.
+
+Filesystem read relies on a narrow recorder exception: its synchronous observation updates may settle out of order, but write and edit re-check the observed version before mutation, so stale state only produces `FS_STALE_VERSION`.
+
+## Verification
+
+Unit coverage pins fail-closed classification, typed argument validation, grouping, barriers, live reclassification after registry replacement, the rolling cap, distinct execution objects, middleware order, ordered results and context, and abort draining. First-party tests pin each parallel declaration.
+
+Snapshot coverage pins the visible multi-call transcript: pending calls may overlap while completed results remain model-ordered. Code Mode coverage pins its serial boundary. No provider-backed e2e is required because scheduling is deterministic loop behavior.
+
+## Alternatives considered
+
+**Keep serial execution.** This avoids new ordering and abort cases but retains unnecessary latency for independent sibling calls.
+
+**Use one tool-level boolean.** A fixed `supportsParallelToolCalls` flag is smaller but cannot distinguish a tool's read-only and mutating operations. The argument-sensitive classifier preserves that distinction.
+
+**Use stateful classification.** Giving the classifier a live agent, registry, or I/O access makes the decision depend on when it runs and creates a gap between classification and dispatch. Mutable authorization and stale-state checks remain execution-time responsibilities.
+
+**Use sibling-aware or resource-aware classification.** The scheduler could compare calls pairwise or let each call declare resource read/write claims. This can parallelize non-conflicting writes, but it requires shared resource identity and conflict semantics across unrelated tools. The unary contract instead gives up that concurrency and fails closed when safety is relational.
+
+**Parallelize the complete tool pipeline.** This keeps the loop on the public one-call API but runs pre- and post-execute middleware concurrently. Existing guards and hook bridges may carry ordered state, so only dispatch overlaps.
+
+**Expose staged methods or a scheduling waterfall.** Public `prepare` / `dispatch` / `finalize` methods or a `tools/execution-mode` event add extension surface before another consumer needs it. The loop uses an internal scheduler view, while `executionMode(exec)` leaves an insertion point for a policy seam.
+
+**Start calls while the model streams.** This may reduce latency further but changes assistant-message authority, replay, and call/result pairing. The scheduler starts only after the assistant message is complete.
+
+**Use fixed-size windows.** Waiting for every call in one window before starting the next leaves capacity idle behind a slow call. The rolling pool preserves the cap without that delay.
+
+**Expose concurrency metadata to the model.** The model can already emit sibling calls. Host scheduling metadata would enlarge requests without improving tool choice.
+
+## Consequences
+
+The design is fail-closed and simple for tool authors, but it cannot exploit concurrency whose safety depends on comparing siblings. A tool that opts in too broadly can expose latent shared-state races.
+
+Parallel calls may begin in cases where serial execution would have aborted before reaching them. The scheduler therefore records only started calls, drains them on abort, and never starts replacements after cancellation.
+
+Ordered commits may hold a fast result behind a slow earlier sibling. This preserves replay and model-history order while live surfaces still show pending progress.
+
+Concurrent external calls can compete for quota or process capacity. Providers own their capacity controls; the loop cap only limits calls from one agent step.
+
+Tool registration is a scheduling boundary. Registry mutations affect not-yet-started calls because the scheduler reclassifies after each barrier and before every pool replenishment. Already-started calls retain the scheduling decision under which they entered the pool.
diff --git a/docs/rfc/implemented/feature/2026-07-10-session-query-service.md b/docs/rfc/implemented/feature/2026-07-10-session-query-service.md
index 7e13256669..ee2acc62e1 100644
--- a/docs/rfc/implemented/feature/2026-07-10-session-query-service.md
+++ b/docs/rfc/implemented/feature/2026-07-10-session-query-service.md
@@ -4,13 +4,13 @@ Status: implemented
 
 ## Problem
 
-Session history exists in two places: current `SessionStore` objects and an optional persistence backend. Consumers that need exact inspection would otherwise duplicate live-versus-persisted precedence, persistence lifecycle handling, raw-event surface classification, and defensive cloning. Durable state can lag the live log between checkpoints, so persistence alone is not a truthful current source.
+Session history exists in two places: current `SessionStore` objects and an optional persistence backend. Consumers that need exact inspection would otherwise duplicate live-versus-persisted precedence, persistence lifecycle handling, raw-event surface classification, relationship tracing, and defensive cloning. Durable state can lag the live log between checkpoints, so persistence alone is not a truthful current source.
 
 Full-text search is related but materially larger. Designing provider registration, extraction, synchronization, invalidation, ranking, and cursor contracts before a real backend exists creates two speculative state machines: one in the interface service and another in the eventual database package.
 
 ## Decision
 
-`@deepseek-ai/dsh-session-query` owns `ctx.sessionQuery`, a small trusted exact-read service over one logical corpus. It exposes `listSessions()`, `listEvents(sessionId)`, and bounded `readEvent(request)`. It does not expose filters, lineage or provenance traversals, text extractors, search requests, provider registration, or derived-index synchronization.
+`@deepseek-ai/dsh-session-query` owns `ctx.sessionQuery`, a small trusted exact-inspection service over one logical corpus. It exposes `listSessions()`, `listEvents(sessionId)`, bounded `readEvent(request)`, `traceSession(sessionId)`, and `traceEvent(request)`. It does not expose filters, text extractors, search requests, provider registration, or derived-index synchronization. The separate [tracing decision](2026-07-13-session-query-tracing.md) owns lineage and event-relationship semantics.
 
 The service observes the optional `ctx.sessionPersistence` binding dynamically but retains no persisted cache or invalidation listener. Each cross-corpus list asks the active backend for authoritative metadata, then overlays a fresh live-store list. Matching ids become one `SessionRecord`: the live header wins and `live`/`persisted` independently report source availability. Immutable header disagreement is `SESSION_QUERY_SOURCE_CONFLICT`.
 
@@ -18,13 +18,13 @@ An exact target read first checks the live store and snapshots the live header a
 
 ## Surface semantics
 
-`dsh-session` exports `foldSurface(events)`, and `SurfaceManager` uses the same transition functions for its incremental cache. The fold returns detached current nodes and each replacement's actual removed seqs. `listEvents()` uses that result to classify every raw event as `current`, `shadowed`, or `log-only`, so inspection cannot disagree with model-history derivation about positional replacement semantics.
+`dsh-session` exports `foldSurface(events)`, and `SurfaceManager` uses the same transition functions for its incremental cache. The fold returns detached current event sequences and each replacement's actual removed seqs. `listEvents()` and `traceEvent()` use that result to classify every raw event, so inspection cannot disagree with model-history derivation about positional replacement semantics.
 
 `readEvent()` returns the complete target plus raw neighbors by contiguous seq. `before` and `after` default to zero and are independently bounded by `readWindowMax`, default 50. The result carries a cloned `SessionHeader`, not a source-availability record, because determining a live target's persisted flag would violate the guarantee that live exact reads do not depend on persistence health.
 
 ## Security boundary
 
-The service is context-wide trusted infrastructure, not an authorization layer. A future model-facing history tool or human UI applies explicit caller/session scope. This phase adds no model-facing tool and changes no transcript or snapshot surface.
+The service is context-wide trusted infrastructure, not an authorization layer. A future model-facing history tool or human UI applies explicit caller/session scope. The service adds no model-facing tool and changes no transcript or snapshot surface.
 
 ## Alternatives considered
 
@@ -32,10 +32,9 @@ The service is context-wide trusted infrastructure, not an authorization layer.
 - **Query only persistence** — rejected because checkpoints can lag the current live log.
 - **Cache persisted metadata and listen for writes/removals** — rejected because exact reads can ask the authoritative sources directly, while cache invalidation adds lifecycle and concurrency state before scale requires it.
 - **Define a provider-neutral search protocol now** — rejected because no provider consumes it. The first SQLite FTS package should own one reconciliation/transaction state machine; a smaller shared seam can be extracted later only when a second implementation proves the boundary.
-- **Include lineage, provenance, and generic filters in phase one** — rejected because no current consumer requires them and canonical logs remain sufficient to add them with evidence later.
 
 ## Consequences
 
-Phase one has one source-resolution state variable: the currently mounted persistence service. There are no provider queues, fingerprints, extractor registries, observation generations, or derived index updates. Exact reads remain usable in live-only deployments and deterministic when persistence is present.
+The service has one source-resolution state variable: the currently mounted persistence service. There are no provider queues, fingerprints, extractor registries, observation generations, or derived index updates. Exact reads and event traces remain usable in live-only deployments and deterministic when persistence is present.
 
-Cross-corpus listing and persisted exact reads perform backend I/O on each call. That is deliberate: correctness comes from current authoritative state, and scale-oriented search belongs to the phase-two database. Full-text search is unavailable until that package defines and implements its complete contract.
+Cross-corpus listing, lineage tracing, and persisted event operations perform backend I/O on each call. That is deliberate: correctness comes from current authoritative state, and scale-oriented search belongs to the proposed database package. Full-text search is unavailable until that package defines and implements its complete contract.
diff --git a/docs/rfc/implemented/feature/2026-07-13-session-query-tracing.md b/docs/rfc/implemented/feature/2026-07-13-session-query-tracing.md
new file mode 100644
index 0000000000..f16f543ac5
--- /dev/null
+++ b/docs/rfc/implemented/feature/2026-07-13-session-query-tracing.md
@@ -0,0 +1,34 @@
+# RFC: Session query relationship tracing
+
+Status: implemented
+
+## Problem
+
+Session relationships are encoded across immutable headers, positional surface operations, and logged provenance arrays. A consumer reconstructing those relationships directly would need to duplicate corpus precedence, surface folding, malformed-log handling, deterministic lineage ordering, and cloning. Positional replacement and provenance are different graphs, so collapsing them into one generic edge type would also lose meaning.
+
+## Decision
+
+`ctx.sessionQuery` exposes `traceSession(sessionId)` and `traceEvent({ sessionId, seq })` alongside its exact reads. Both are one-shot views over the existing live-preferred corpus: session tracing consumes one complete corpus listing, while event tracing consumes one loaded logical log and one canonical surface fold. The service retains no lineage, reverse-index, or replacement state after a call.
+
+`SessionLineageTrace` returns the target, known parents in immediate-to-outward order, and recursive descendant trees whose siblings sort by creation time and then session id. `complete: true` carries the known root; `complete: false` carries the first unresolved parent id. A cycle connected to the target fails with `SESSION_QUERY_INVALID_LINEAGE`.
+
+`SessionEventTrace` keeps positional and provenance relationships separate. `replacedBy` is the immediate positional replacer, `replacementChain` follows replacers to the final node, and `replacedEventSeqs` lists the actual surface nodes directly removed by the target. `sourceEventSeqs` preserves direct logged source order, while `derivedEventSeqs` lists later direct reverse references in log order. Provenance is not expanded transitively.
+
+## Validation boundary
+
+Event tracing checks target existence before surface analysis. Both event listing and tracing then use `dsh-session`'s one-pass surface fold, which accepts or rejects the loaded log as a whole: event seqs are zero-based and contiguous, surface markers obey event-type eligibility, provenance belongs only to surface event types, present arrays are nonempty and duplicate-free, every source is an earlier seq, and every positional replacement names and cites all surface nodes it removes. Every contract failure uses `SESSION_QUERY_INVALID_SURFACE`; there is no weaker classification-only surface standard.
+
+All returned records and arrays are detached. A known live event trace never consults persistence; persisted event traces preserve the exact-read list/load consistency check. Session lineage is necessarily a cross-corpus operation and therefore preserves cross-corpus persistence failure semantics.
+
+## Alternatives considered
+
+- **Expose standalone tracing helpers** — rejected because the source-precedence and detachment boundary belongs to `ctx.sessionQuery`; public helpers would invite callers to bypass it.
+- **Combine replacement and provenance edges** — rejected because a positional replacement can shadow surface nodes while also citing non-surface construction inputs, and consumers need to distinguish those meanings.
+- **Return transitive provenance closure** — rejected because it obscures logged direct evidence, increases result size, and lets one malformed distant edge alter otherwise local output.
+- **Best-effort traces over malformed provenance** — rejected because a structurally plausible partial graph would look authoritative. Exact inspection fails loudly when the canonical relationship contract is broken.
+
+## Consequences
+
+Consumers receive deterministic relationship views without a cache or second corpus. Event tracing performs whole-log validation and allocation on each call, while lineage tracing lists the complete logical corpus on each call. Those costs keep the source of truth explicit and are separate from the content-bearing full-text-search and filtering API.
+
+The feature has unit and service-level coverage but no snapshot or end-to-end fixture because it introduces no model-facing consumer, transcript change, or cross-process protocol.
diff --git a/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.i18n.yaml b/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.i18n.yaml
index cb5d12c562..f1d38d7d95 100644
--- a/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.i18n.yaml
+++ b/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-2026-07-14-time-context-plugin.md: 105bf53550f087fdefb1e6fe0ec493f8628d3e18
-2026-07-14-time-context-plugin.zh.md: 60e9004b1453e75e1bcd84870ad7f18d200a95d8
+2026-07-14-time-context-plugin.md: b8b54156e08aa1212866d46500ad1ca65b4f4f14
+2026-07-14-time-context-plugin.zh.md: 0af261c66a9b52cdf250294b4bfdc240176c8434
diff --git a/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.md b/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.md
index 105bf53550..b8b54156e0 100644
--- a/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.md
+++ b/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.md
@@ -6,6 +6,8 @@ English | [中文](2026-07-14-time-context-plugin.zh.md)
 
 ## Problem
 
+The dynamic system-prompt storage and refresh decision in this record is superseded by [Durable per-step time context](2026-07-16-durable-per-step-time-context.md). The opt-in package, zoned formatting, and validation remain; the follow-up owns the current model-visible and durability contract.
+
 An agent request has no live clock unless a deployment puts one in prompt text or gives the model a query tool. Static text becomes stale, while a tool call adds overhead to ordinary reasoning about dates, deadlines, or idle time. Without elapsed time, the model cannot distinguish an immediate follow-up from one sent hours after the preceding message.
 
 Prompt assembly can derive both facts per step from durable session timestamps, and request-header logging can record the exact rendered value. Accumulating stale readings in conversation history or waking idle agents would violate the existing request lifecycle.
@@ -30,11 +32,11 @@ When `timeZone` is omitted, `Intl.DateTimeFormat` resolves the Node process's sy
 
 ### Logging and token shape
 
-The loop records the temporal block through `request/header` and `request/header-delta` before transmission, satisfying the [reconstructable-requests contract](../architecture/2026-07-05-reconstructable-requests.md). Each request carries one current block; earlier readings do not remain in conversation history. The plugin owns the fact and contributes it through the prompt registry, following the [prompt-variables RFC](../architecture/2026-07-05-prompt-variables-and-tool-guidance-ownership.md) without a loop special case.
+The loop records the temporal block in full `request/header` snapshots before transmission, satisfying the [reconstructable-requests contract](../architecture/2026-07-05-reconstructable-requests.md). Each request carries one current block; earlier readings do not remain in conversation history. The plugin owns the fact and contributes it through the prompt registry, following the [prompt-variables RFC](../architecture/2026-07-05-prompt-variables-and-tool-guidance-ownership.md) without a loop special case.
 
 ## Testing
 
-Unit tests pin formatting, baselines, refresh policy, validation, per-agent state, disposal, and load-time system-zone capture. A real agent-loop test pins the transmitted prompt and `request/header-delta`. A keyless subprocess e2e boots a test-only `cordis.yml` through the real Loader and stdio app, omits `timeZone` under a controlled `TZ`, drives two turns, and verifies the persisted request headers externally. Default snapshot compositions omit the plugin, so their transcript fixtures contain no temporal block.
+Unit tests pin formatting, baselines, refresh policy, validation, per-agent state, disposal, and load-time system-zone capture. A real agent-loop test pins the transmitted prompt and full `request/header` snapshots. A keyless subprocess e2e boots a test-only `cordis.yml` through the real Loader and stdio app, omits `timeZone` under a controlled `TZ`, drives two turns, and verifies the persisted request headers externally. Default snapshot compositions omit the plugin, so their transcript fixtures contain no temporal block.
 
 ## Alternatives considered
 
@@ -52,6 +54,6 @@ Unit tests pin formatting, baselines, refresh policy, validation, per-agent stat
 
 - Opted-in models receive a zoned clock and inter-turn duration without a tool call. The system-prompt cost is fixed per request instead of growing with the session.
 - An omitted `timeZone` follows the process's `TZ`, host, or container zone as observed at plugin load. Operators must configure an explicit zone when the deployment environment does not represent the intended user.
-- A refresh changes the request header and can add a `request/header-delta`. `refreshIntervalMs` trades freshness against durable deltas; `0` records a new value on every step whose whole-second rendering changes.
+- A refresh changes the request header and can add a full `request/header` snapshot with reason `change`. `refreshIntervalMs` trades freshness against the number and size of durable full snapshots; `0` records a new value on every step whose whole-second rendering changes.
 - No request exists solely to refresh time. A long-running tool leaves the prior reading until the next step assembles.
 - Duration reflects harness processing time at durable append boundaries, not client-network latency before logging. Preserving a client-origin timestamp requires a separate durable input contract.
diff --git a/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.zh.md b/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.zh.md
index 60e9004b14..0af261c66a 100644
--- a/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.zh.md
+++ b/docs/rfc/implemented/feature/2026-07-14-time-context-plugin.zh.md
@@ -6,13 +6,15 @@ Status: implemented
 
 ## 问题
 
+本记录中的动态系统提示词存储和刷新决策已由[持久的逐步骤时间上下文](2026-07-16-durable-per-step-time-context.md)取代。需要显式启用的包（package）、分区时间格式和校验仍然保留；后续 RFC 负责当前的模型可见与持久性契约。
+
 如果部署方既未在提示词中提供时钟，也未给模型提供查询工具，agent（智能体）请求就无法获得实时准确的时间。静态文本会变得陈旧，而对于日期、截止时间或闲置时长等常规推理，调用工具会增加开销。缺少已经过去的时长时，模型无法区分紧接着发送的消息与上一条消息几小时后才发送的消息。
 
 提示词组装流程可以在每个步骤中根据持久会话时间戳派生这两项信息，请求头日志则可以记录实际渲染的确切值。在会话历史中累积陈旧读数或唤醒空闲 agent 都会违反现有请求生命周期。
 
 ## 决策
 
-`@deepseek-ai/dsh-time-context` 是位于 `packages/context/time-context/`、需要显式启用的函数插件。`context/` 产品分组用于容纳既不定义工具、也不定义服务的有界请求上下文增强。`dsh-agent-spine-demo` 和仓库提供的示例都不会加载该 package；只有当 token 与信息披露成本可接受时，部署方才显式挂载它。
+`@deepseek-ai/dsh-time-context` 是位于 `packages/context/time-context/`、需要显式启用的函数插件。`context/` 产品分组用于容纳既不定义工具、也不定义服务的有界请求上下文增强。`dsh-agent-spine-demo` 和仓库提供的示例都不会加载该包；只有当 token 与信息披露成本可接受时，部署方才显式挂载它。
 
 该插件注册顺序值为 10 的全局系统提示词区段 `context:time`，位置在部署方角色设定之后、工具指导之前。对于活跃轮次，它会输出带数字 UTC 偏移和 IANA 时区、形似 ISO 的时间戳，以及从轮次开始前最后一条模型可见消息起算的紧凑整秒时长。未绑定 agent 或 agent 处于空闲状态时，该区段为空。
 
@@ -30,11 +32,11 @@ Status: implemented
 
 ### 日志与 token 形态
 
-agent loop（智能体循环）会在发送前通过 `request/header` 和 `request/header-delta` 记录时间区块，从而满足[可重建请求契约](../architecture/2026-07-05-reconstructable-requests.md)。每个请求只携带一个当前区块；先前的读数不会保留在会话历史中。该插件拥有时间信息，并按照[提示词变量 RFC](../architecture/2026-07-05-prompt-variables-and-tool-guidance-ownership.md)通过提示词注册表贡献该信息，无需为循环添加特殊分支。
+agent loop（智能体循环）会在发送前通过完整的 `request/header` 快照记录时间区块，从而满足[可重建请求契约](../architecture/2026-07-05-reconstructable-requests.md)。每个请求只携带一个当前区块；先前的读数不会保留在会话历史中。该插件拥有时间信息，并按照[提示词变量 RFC](../architecture/2026-07-05-prompt-variables-and-tool-guidance-ownership.md)通过提示词注册表贡献该信息，无需为循环添加特殊分支。
 
 ## 测试
 
-单元测试固定格式化、基线、刷新策略、校验、逐 agent 状态、资源释放行为，以及系统时区在加载时的捕获行为。使用真实 agent loop 的测试固定实际发送的提示词和 `request/header-delta`。无密钥子进程端到端测试通过真实 Loader 和 stdio 应用启动测试专用 `cordis.yml`，在受控 `TZ` 下省略 `timeZone`，驱动两个轮次，并从外部校验持久请求头。默认快照组合不包含该插件，因此其中的 transcript（文本记录）fixture（测试前置数据）不包含时间区块。
+单元测试固定格式化、基线、刷新策略、校验、逐 agent 状态、资源释放行为，以及系统时区在加载时的捕获行为。使用真实 agent loop 的测试固定实际发送的提示词和完整的 `request/header` 快照。无密钥子进程端到端测试通过真实 Loader 和 stdio 应用启动测试专用 `cordis.yml`，在受控 `TZ` 下省略 `timeZone`，驱动两个轮次，并从外部校验持久请求头。默认快照组合不包含该插件，因此其中的 transcript（文本记录）fixture（测试前置数据）不包含时间区块。
 
 ## 考虑过的替代方案
 
@@ -46,12 +48,12 @@ agent loop（智能体循环）会在发送前通过 `request/header` 和 `reque
 - **省略配置时仍默认使用 UTC**——不予采纳，因为显式启用的时钟应跟随部署环境，除非运维方选择 UTC。需要 UTC 的部署仍可配置 `timeZone: UTC`。
 - **引入时区探测库**——不予采纳，因为 Node 的 `Intl` 运行时已经能够提供进程的 IANA 时区，而且额外依赖同样无法推断远程用户的时区。
 - **在 `dsh-agent-spine-demo` 中挂载插件**——不予采纳，因为时区、信息披露、token 预算和新鲜度都属于部署策略。选择加入能保持默认上下文稳定。
-- **将 package 放入 `core/`**——不予采纳，因为 `core/` 负责产品 API 主干，而该插件是没有服务键的可选叶节点。
+- **将包放入 `core/`**——不予采纳，因为 `core/` 负责产品 API 主干，而该插件是没有服务键的可选叶节点。
 
 ## 后果
 
 - 选择加入的模型无需调用工具，即可获得分区时钟和轮次间隔时长。每个请求的系统提示词成本固定，不会随会话增长。
 - 省略 `timeZone` 时，插件采用加载时观察到的进程 `TZ`、主机或容器时区。当部署环境不能代表目标用户时，运维方必须显式配置时区。
-- 刷新会改变请求头，并可能新增 `request/header-delta`。`refreshIntervalMs` 用新鲜度换取持久增量记录的数量；设为 `0` 时，每个整秒渲染结果发生变化的步骤都会记录新值。
+- 刷新会改变请求头，并可能新增一份 reason 为 `change` 的完整 `request/header` 快照。`refreshIntervalMs` 用新鲜度换取完整持久快照的数量与大小；设为 `0` 时，每个整秒渲染结果发生变化的步骤都会记录新值。
 - 系统不会仅为刷新时间而创建请求。长时间运行的工具会保留先前读数，直至下一步骤开始组装。
 - 时长反映持久追加边界处的 harness 处理时间，不包含消息进入日志之前的客户端网络延迟。若要保留客户端来源时间戳，需要单独的持久输入契约。
diff --git a/docs/rfc/implemented/feature/2026-07-16-durable-per-step-time-context.i18n.yaml b/docs/rfc/implemented/feature/2026-07-16-durable-per-step-time-context.i18n.yaml
new file mode 100644
index 0000000000..ce964f71da
--- /dev/null
+++ b/docs/rfc/implemented/feature/2026-07-16-durable-per-step-time-context.i18n.yaml
@@ -0,0 +1,6 @@
+# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
+# side as of the last confirmed-consistent state. Both languages carry equal authority;
+# after editing either side, bring the other along and re-record with:
+#   pnpm run verify-translation-pairing --write
+2026-07-16-durable-per-step-time-context.md: 4a0828111faf9f787c2d338024c42680a4a697e2
+2026-07-16-durable-per-step-time-context.zh.md: f745cf7da7c38f9682abf9d8f210bcba328c8a51
diff --git a/docs/rfc/implemented/feature/2026-07-16-durable-per-step-time-context.md b/docs/rfc/implemented/feature/2026-07-16-durable-per-step-time-context.md
new file mode 100644
index 0000000000..4a0828111f
--- /dev/null
+++ b/docs/rfc/implemented/feature/2026-07-16-durable-per-step-time-context.md
@@ -0,0 +1,70 @@
+# RFC: Durable per-step time context
+
+Status: implemented
+
+English | [中文](2026-07-16-durable-per-step-time-context.zh.md)
+
+## Problem
+
+A request-only clock can tell the model the current time, but replacing that value in the system prompt removes the evidence behind earlier time-sensitive reasoning. Multi-step turns need requests to retain the readings that shaped preceding steps. The request must remain reconstructable after restart, and automatic compaction must account for the same timing context the model receives.
+
+A process-local refresh cache makes displayed time depend on state that cannot survive resume or be reconstructed from the durable session. Durable interval scheduling can reduce append frequency without introducing that hidden state.
+
+## Decision
+
+`@deepseek-ai/dsh-time-context` is an opt-in function plugin in `packages/context/time-context/`. It registers a prepended `agent/pre-step` listener and, when an injection is due, calls `agent.inject()` for a pre-step attempt whose signal is not already aborted. The injected `context/message` carries source `{ kind: 'plugin', plugin: 'time-context' }` and append surface metadata; a suppressed attempt appends nothing.
+
+The listener records preparation context before a possible `step/start`. Its prepended registration runs before ordinary automatic compaction listeners, so pressure estimation and any resulting surface rewrite observe a newly appended reading. A later pre-step listener can cancel or fail the attempt before the step opens; the reading remains because the durable log is append-only and this plugin performs no rollback.
+
+The optional `timeZone` config resolves the Node process's IANA zone once at plugin load when omitted; an explicit value is validated by `Intl.DateTimeFormat`. The timestamp includes the numeric UTC offset and resolved IANA zone.
+
+The optional `refreshIntervalMs` config is manually validated at plugin load as a non-negative safe integer. Omission or `0` injects on every eligible preparation attempt. A positive value scans the raw session events for the most recent `context/message` with this plugin's source and injects when none exists, wall time moved backward, or the event is at least the configured age. The raw event timestamp governs even after compaction shadows the message, so scheduling persists across turns and process resume without a timer or process-local cache.
+
+### Text and elapsed baselines
+
+An injected first-step reading is:
+
+```text
+Time sampled while preparing turn <turn>, step 1: <timestamp>
+Elapsed since the preceding model-visible message: <duration-or-unavailable>.
+```
+
+The baseline is the latest preceding user, assistant, tool-result, context, or steering message. This includes the accepted prompt that opened an ordinary message turn. If no model-visible message exists, the duration is `unavailable`.
+
+An injected later-step reading is:
+
+```text
+Time sampled while preparing turn <turn>, step <step>: <timestamp>
+Elapsed since the preceding step context: <duration-or-unavailable>.
+```
+
+Their baseline is the durable event timestamp of the preceding time-context message in the same turn. If interval suppression leaves no earlier same-turn reading, the duration is `unavailable`. Duration formatting uses compact whole-second units and clamps backward wall-clock movement to zero. The explicit turn and step make every retained reading attributable to its historical preparation attempt after later turns append more context.
+
+### Durability and request reconstruction
+
+Each reading remains a normal surface node until compaction shadows it; positive interval scheduling never removes existing readings. A later request therefore sees the cumulative unshadowed readings that affected earlier preparation and steps, rather than a system-prompt value rewritten in place.
+
+The plugin contributes nothing to system-prompt assembly. `request/header` contains no time-context text; request reconstruction obtains the complete durable surface prefix at each `step/start`. Readings and requests need not map one-to-one because a failed preparation can leave a reading while interval suppression can prepare a request without appending one. The plugin depends on the agent registry for its lifecycle listener and does not require the system-prompt service at runtime.
+
+## Testing
+
+Unit and real-loop tests pin formatting, both elapsed baselines, interval omission and zero, threshold boundaries, cross-turn and per-session scheduling, backward-clock behavior, invalid config, resumed raw-event lookup after compaction, aborted-signal behavior, later-listener cancellation and failure, listener disposal, source and surface metadata, cumulative multi-step visibility, and absence from request headers. A keyless subprocess e2e boots the real Loader and stdio app, drives two turns, and verifies the persisted context events externally.
+
+## Supersedes
+
+This decision supersedes the dynamic system-prompt storage and refresh policy in [Optional time-context plugin](2026-07-14-time-context-plugin.md). It keeps the package location, opt-in deployment stance, timestamp formatting, process-zone default, and load-time validation. Durable history replaces the `context:time` prompt section, process-local refresh cache, and request-header deltas; `refreshIntervalMs` controls durable append frequency instead of prompt replacement.
+
+## Alternatives considered
+
+- **Keep the dynamic system-prompt section and process-local refresh cache** — rejected because replacement erases earlier readings, cache state is not replayable, and a frozen request envelope would make the value stale for an entire loop instance.
+- **Replace the preceding context surface node** — rejected because replacement preserves the old node's position or shadows intervening conversation; neither represents when the new reading became visible.
+- **Inject from a background timer** — rejected because idle time has no pending request to consume the value, and timer-driven injection would create durable turns solely to report time passing.
+- **Expose time only through a tool** — rejected because ordinary temporal reasoning would require an avoidable tool round trip and would not guarantee a reading before every step.
+- **Use `agent/session-prefix`** — rejected because one loop-instance prefix cannot represent distinct step timestamps and does not accumulate historically attributable readings.
+
+## Consequences
+
+- Omission or `0` records every eligible preparation attempt; a positive interval reduces append frequency and history growth while preserving durable scheduling across resume.
+- Timing context remains append-only until compaction shadows older surface nodes, including a preparation reading left by a later cancellation or failure.
+- The first-step duration normally measures from the prompt that opened the turn, while later-step durations measure model and tool processing since the preceding step context.
+- An omitted `timeZone` still reflects the deployment process rather than a remote user, and elapsed time still uses durable harness append boundaries rather than client-origin timestamps.
diff --git a/docs/rfc/implemented/feature/2026-07-16-durable-per-step-time-context.zh.md b/docs/rfc/implemented/feature/2026-07-16-durable-per-step-time-context.zh.md
new file mode 100644
index 0000000000..f745cf7da7
--- /dev/null
+++ b/docs/rfc/implemented/feature/2026-07-16-durable-per-step-time-context.zh.md
@@ -0,0 +1,70 @@
+# RFC: 持久的逐步骤时间上下文
+
+Status: implemented
+
+[English](2026-07-16-durable-per-step-time-context.md) | 中文
+
+## 问题
+
+仅存在于请求中的时钟可以告诉模型当前时间，但在系统提示词中替换这个值会移除先前时效性推理所依据的证据。在包含多个步骤的轮次中，请求需要保留影响先前步骤的读数。系统必须能在重启后重建请求，自动压缩（compaction）也必须核算模型实际收到的同一份时间上下文。
+
+进程本地刷新缓存使显示的时间依赖无法在恢复后保留、也无法从持久会话重建的状态。持久的间隔调度可以减少追加频率，而不引入这种隐藏状态。
+
+## 决策
+
+`@deepseek-ai/dsh-time-context` 是位于 `packages/context/time-context/`、需要显式启用的函数插件。它注册一个前置的 `agent/pre-step` 监听器，并在需要注入时，为信号尚未取消的预步骤尝试调用 `agent.inject()`。注入的 `context/message` 携带来源 `{ kind: 'plugin', plugin: 'time-context' }` 和追加表层元数据；受间隔抑制的尝试不会追加任何内容。
+
+监听器在可能出现的 `step/start` 之前记录准备上下文。它采用前置注册，因此先于普通自动压缩监听器运行，使压力估算和由此产生的表层重写都能观察到新追加的读数。后续预步骤监听器可能在步骤开启前取消尝试或使其失败；持久日志仅追加，且本插件不执行回滚，因此该读数会保留下来。
+
+省略可选配置 `timeZone` 时，插件在加载时解析一次 Node 进程的 IANA 时区；显式值由 `Intl.DateTimeFormat` 校验。时间戳包含数字 UTC 偏移和解析后的 IANA 时区。
+
+插件在加载时手动校验可选配置 `refreshIntervalMs`，其值必须为非负安全整数。省略或设为 `0` 时，每次符合条件的准备尝试都会注入。设为正数时，插件扫描原始会话事件，查找来源属于本插件的最新 `context/message`；不存在此类事件、系统挂钟向后移动，或该事件已达到配置时长时，插件执行注入。即使压缩已隐藏消息，调度仍以原始事件时间戳为准，因此该机制无需计时器或进程本地缓存，也能跨轮次和进程恢复持续生效。
+
+### 文本与时长基线
+
+第一个步骤的注入读数为：
+
+```text
+Time sampled while preparing turn <turn>, step 1: <timestamp>
+Elapsed since the preceding model-visible message: <duration-or-unavailable>.
+```
+
+基线是前一条用户消息、助手消息、工具结果、上下文消息或 steering（中途引导）消息。对于普通消息轮次，这包括开启轮次的已接受提示词。如果不存在模型可见消息，时长为 `unavailable`。
+
+后续步骤的注入读数为：
+
+```text
+Time sampled while preparing turn <turn>, step <step>: <timestamp>
+Elapsed since the preceding step context: <duration-or-unavailable>.
+```
+
+其基线是同一轮次中上一条时间上下文消息的持久事件时间戳。如果间隔抑制导致同一轮次中没有更早的读数，时长为 `unavailable`。时长采用紧凑的整秒单位，并在系统挂钟向后移动时钳制为零。显式的轮次号和步骤号使每个保留的读数在后续轮次追加更多上下文后，仍可归属于对应的历史准备尝试。
+
+### 持久性与请求重建
+
+每个读数都作为普通表层节点保留，直至压缩将其隐藏；正数间隔调度绝不会移除已有读数。因此，后续请求会看到影响先前准备过程和步骤且尚未被隐藏的累计读数，而不是一个被原地改写的系统提示词值。
+
+插件不向系统提示词组装贡献任何内容。`request/header` 不包含时间上下文文本；请求重建从每个 `step/start` 取得完整的持久表层前缀。读数与请求无需一一对应，因为失败的准备过程可能留下读数，而间隔抑制也可能使请求准备过程不追加读数。插件通过 agent 注册表使用生命周期监听器，运行时不需要系统提示词服务。
+
+## 测试
+
+单元测试和真实 agent loop（智能体循环）测试固定格式化、两种时长基线、间隔省略和零值、阈值边界、跨轮次和各会话独立调度、挂钟后退行为、无效配置、压缩后基于恢复会话的原始事件查找、已取消信号行为、后续监听器取消和失败、监听器 dispose（资源释放）、来源与表层元数据、多步骤累计可见性，以及请求头中不存在时间上下文。无密钥子进程 e2e 测试通过真实 Loader 和 stdio 应用启动，驱动两个轮次，并从外部校验持久化的上下文事件。
+
+## 取代的决策
+
+本决策取代[可选时间上下文插件](2026-07-14-time-context-plugin.md)中的动态系统提示词存储和刷新策略。它保留包位置、选择加入式部署、时间戳格式、进程时区默认值和加载时校验。持久历史取代 `context:time` 提示词区段、进程本地刷新缓存和请求头增量；`refreshIntervalMs` 用于控制持久追加频率，而非提示词替换。
+
+## 考虑过的替代方案
+
+- **保留动态系统提示词区段和进程本地刷新缓存**——不予采纳，因为替换会抹去先前读数，缓存状态无法回放，而且冻结的请求内容集合会使该值在整个 agent loop 实例期间保持陈旧。
+- **替换前一条上下文表层节点**——不予采纳，因为替换会保留旧节点的位置或隐藏中间的会话内容；两者都不能表达新读数何时开始可见。
+- **通过后台计时器注入**——不予采纳，因为空闲期间没有待处理请求消费该值，而且计时器驱动的注入会仅为报告时间流逝而创建持久轮次。
+- **只通过工具提供时间**——不予采纳，因为普通时间推理会产生本可避免的工具往返，也不能保证每个步骤之前都有读数。
+- **使用 `agent/session-prefix`**——不予采纳，因为一个 loop 实例前缀无法表示不同的步骤时间戳，也不会累计具有历史归属的读数。
+
+## 后果
+
+- 省略 `refreshIntervalMs` 或设为 `0` 时，每次符合条件的准备尝试都会留下记录；正数间隔会减少追加频率和历史增长，同时使持久调度在恢复后继续生效。
+- 时间上下文仅追加并保留到压缩隐藏旧表层节点为止，其中也包括后续取消或失败所留下的准备读数。
+- 第一个步骤的时长通常从开启轮次的提示词起算，后续步骤的时长则反映自上一条步骤上下文以来的模型与工具处理时间。
+- 省略 `timeZone` 时仍采用部署进程而非远程用户的时区，时长仍采用 harness 的持久追加边界而非客户端来源时间戳。
diff --git a/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.i18n.yaml b/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.i18n.yaml
index f20250c3ba..5f83acc3aa 100644
--- a/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.i18n.yaml
+++ b/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-2026-07-02-bilingual-docs-and-pairing-gate.md: 1e96622e7fb5694ab61772d68744394ef1aeb53a
-2026-07-02-bilingual-docs-and-pairing-gate.zh.md: c752d76f12f556ce190bf80c4f3a531c0821be8e
+2026-07-02-bilingual-docs-and-pairing-gate.md: 68c0f3bbc0472b0c96f9d64fc6b1b24ac7008795
+2026-07-02-bilingual-docs-and-pairing-gate.zh.md: 2cf8f9b9c17d8a521d8674833e909b34f315cfe0
diff --git a/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.md b/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.md
index 1e96622e7f..68c0f3bbc0 100644
--- a/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.md
+++ b/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.md
@@ -12,7 +12,7 @@ This repo's README and docs tree are read by people and agents inside and outsid
 
 - **Paired sibling files with equal authority.** A documentation pair is three sibling files: English `foo.md`, Chinese `foo.zh.md`, and a consistency record `foo.i18n.yaml`. Neither language is canonical — a document may be authored and reviewed Chinese-first and translated to English afterwards, or the reverse; what binds the pair is that both sides must say the same thing, and pairs merge whole (both languages plus the record, never one alone). Policy: [docs/i18n/README.md](../../../i18n/README.md); translation rules: [docs/i18n/translation-rules.md](../../../i18n/translation-rules.md); terminology source of truth: [docs/i18n/terminology.md](../../../i18n/terminology.md).
 - **A sidecar record of both blob hashes makes consistency checkable.** `foo.i18n.yaml` holds the full git blob hash of each side as of the last confirmed-consistent state. An edit to either side without re-confirming the pair is then mechanically detectable as a pure content comparison — no history lookup — and the hashes are computable for files edited in the same PR, which a commit-hash record is not. Re-recording (`verify-translation-pairing --write`) produces a reviewable yaml diff: confirming consistency is an explicit, visible act in the PR.
-- **`verify-translation-pairing` joins `doc-sync`.** The gate ([scripts/verify-translation-pairing.ts](../../../../scripts/verify-translation-pairing.ts)) enforces: required pairs exist, every existing pair is complete (all three files) and consistent (both hashes match, switcher links both ways, structural signatures identical), and excluded (generated or bilingual-by-construction) files stay unpaired. The `required` list in [scripts/translation-pairing.manifest.json](../../../../scripts/translation-pairing.manifest.json) is a ratchet: each merged translation batch adds its files, so coverage only grows.
+- **`verify-translation-pairing` joins `doc-sync`.** The gate ([scripts/verify-translation-pairing.ts](../../../../scripts/verify-translation-pairing.ts)) enforces: required pairs exist, every existing pair is complete (all three files) and consistent (both hashes match, switcher links both ways, structural signatures identical), excluded (generated or bilingual-by-construction) files stay unpaired, and date-named documents on or after the manifest's `requiredSince` cutoff have complete pairs. The `required` list in [scripts/translation-pairing.manifest.json](../../../../scripts/translation-pairing.manifest.json) is a ratchet: each merged translation batch adds its files, so coverage only grows.
 - **Translation is agent work with human review.** The committed workflow is [.agents/skills/dsh-translate-docs](../../../../.agents/skills/dsh-translate-docs/SKILL.md), following the same pattern as [dsh-code-review](../../../../.agents/skills/dsh-code-review/SKILL.md): the skill carries the workflow and defers to the docs as sources of truth.
 
 ## Alternatives considered
@@ -34,5 +34,5 @@ Paired sibling files with locale suffixes are the dominant Chinese big-tech conv
 - Every pair adds a third file to the tree. The record is machine-written (`--write`), so the cost is directory noise, not maintenance effort; in exchange, "who confirmed these consistent, and when" is answerable from git blame on the yaml.
 - When the two sides disagree, no mechanical rule picks a winner — the PR review does. That is the price of equal authority, accepted deliberately: the alternative (a canonical language) forbids Chinese-first authoring.
 - Generated docs (`cordis-catalog/`, `tool-catalog/`, `module-graph.md`) are excluded for now; the planned follow-up is to teach their generators to emit Chinese alongside English, at which point they leave the exclusion list.
-- Rollout is incremental by design: documents outside `required` are visible backlog (`--list`), not red CI, so pairs land in reviewable batches without a big-bang PR. New documents are the exception — a date-named document dated on/after the manifest's `requiredSince` cutoff merges bilingual or not at all, so the backlog only ever shrinks.
+- Rollout is incremental by design: documents outside `required` are visible backlog (`--list`), not red CI, so pairs land in reviewable batches without a big-bang PR. A date-named document dated on or after the manifest's `requiredSince` cutoff merges bilingual or not at all, so new date-named RFCs do not enlarge that backlog.
 - The recorded hashes double as the update tool (`git cat-file -p <hash>` recovers either side's last-confirmed text for a minimal diff-based update), so re-translation of whole files is never forced by the mechanism.
diff --git a/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.zh.md b/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.zh.md
index c752d76f12..2cf8f9b9c1 100644
--- a/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.zh.md
+++ b/docs/rfc/implemented/process/2026-07-02-bilingual-docs-and-pairing-gate.zh.md
@@ -11,9 +11,9 @@ Status: implemented
 ## 决策
 
 - **配对兄弟文件，两种语言同权。** 一对文档由三个兄弟文件组成：英文 `foo.md`、中文 `foo.zh.md`，以及一份一致性记录 `foo.i18n.yaml`。没有哪种语言是正典：一篇文档可以先用中文撰写和评审、之后再译成英文，反之亦可；约束配对的是：两侧必须表达相同的内容，且配对整体合并（两种语言加记录，绝不单独落一侧）。政策见 [docs/i18n/README.md](../../../i18n/README.md)；翻译规则见 [docs/i18n/translation-rules.md](../../../i18n/translation-rules.md)；术语真源见 [docs/i18n/terminology.md](../../../i18n/terminology.md)。
-- **伴随记录保存两侧 blob hash，使一致性可检查。** `foo.i18n.yaml` 保存两侧文件在上一次确认一致时各自的完整 git blob hash。此后修改了任一侧而未重新确认配对，都能被机械检测出来（纯内容比较，无需查询历史），而且同一个 PR 内改动的文件也能计算出 hash，commit hash 式的记录做不到这一点。重新记录（`verify-translation-pairing --write`）会产生一份可评审的 yaml diff：确认一致在 PR 中是一个显式、可见的动作。
-- **`verify-translation-pairing` 加入 `doc-sync`。** 门禁（[scripts/verify-translation-pairing.ts](../../../../scripts/verify-translation-pairing.ts)）强制执行以下规则：required 的配对必须存在；任何已存在的配对必须完整（三个文件齐全）且一致（两个 hash 匹配、切换行双向互链、结构签名一致）；被排除的文件（生成物或本身即双语的）保持不配对。[scripts/translation-pairing.manifest.json](../../../../scripts/translation-pairing.manifest.json) 中的 `required` 清单只进不退：每个合并的翻译批次将自己的文件加入其中，覆盖面只增不减。
-- **翻译是 agent 的工作，由人评审。** 仓库内置的工作流是 [.agents/skills/dsh-translate-docs](../../../../.agents/skills/dsh-translate-docs/SKILL.md)，与 [dsh-code-review](../../../../.agents/skills/dsh-code-review/SKILL.md) 模式相同：skill 承载工作流，并将文档作为真源。
+- **伴随记录保存两侧 blob hash，使一致性可检查。** `foo.i18n.yaml` 保存两侧文件在上一次确认一致时各自的完整 git blob hash。此后修改了任一侧而未重新确认配对，都能被机械检测出来（纯内容比较，无需查询历史），而且同一个 PR（Pull Request）内改动的文件也能计算出 hash，commit hash 式的记录做不到这一点。重新记录（`verify-translation-pairing --write`）会产生一份可评审的 yaml diff：确认一致在 PR 中是一个显式、可见的动作。
+- **`verify-translation-pairing` 加入 `doc-sync`。** 门禁（[scripts/verify-translation-pairing.ts](../../../../scripts/verify-translation-pairing.ts)）强制执行以下规则：required 的配对必须存在；任何已存在的配对必须完整（三个文件齐全）且一致（两个 hash 匹配、切换行双向互链、结构签名一致）；被排除的文件（生成物或本身即双语的）不得配对；凡文件名以日期开头且日期不早于 manifest（元数据清单）中 `requiredSince` 分界日期的文档，也必须有完整配对。[scripts/translation-pairing.manifest.json](../../../../scripts/translation-pairing.manifest.json) 中的 `required` 清单只进不退：每个合并的翻译批次将自己的文件加入其中，覆盖面只增不减。
+- **翻译是 agent 的工作，由人评审。** 仓库内置的工作流是 [.agents/skills/dsh-translate-docs](../../../../.agents/skills/dsh-translate-docs/SKILL.md)，与 [dsh-code-review](../../../../.agents/skills/dsh-code-review/SKILL.md) 模式相同：skill（技能）承载工作流，并将文档作为真源。
 
 ## 曾考虑的替代方案
 
@@ -34,5 +34,5 @@ Status: implemented
 - 每个配对给目录树多添一个文件。记录由机器写入（`--write`），代价是目录噪音而非维护负担；换来的是「谁在何时确认过这对文档一致」可以从 yaml 的 git blame 直接回答。
 - 两侧说法冲突时，没有机械规则裁决谁赢，由 PR 评审裁决。这是同权的代价，且是有意接受的：另一个选项（正典语言）会禁止中文先行撰写。
 - 生成文档（`cordis-catalog/`、`tool-catalog/`、`module-graph.md`）暂被排除；计划中的后续工作是让生成器在输出英文的同时输出中文，届时将这些文件移出排除清单。
-- 推进天然是渐进的：`required` 之外的文档是可见的 backlog（`--list`），而非红色的 CI；因此配对按可评审的批次落地，无需一个巨型 PR。
+- 推进天然是渐进的：`required` 之外的文档是可见的 backlog（待翻清单，`--list`），而非红色的 CI；因此配对按可评审的批次落地，无需一个巨型 PR。凡文件名以日期开头且日期不早于 manifest 中 `requiredSince` 分界日期的文档，都必须配齐双语文件，因此新建的日期命名 RFC 不会增加这份 backlog。
 - 记录的 hash 兼作更新工具（`git cat-file -p <hash>` 能还原任一侧上次确认的文本，用于基于 diff 的最小更新），因此这套机制从不强迫整篇重译。
diff --git a/docs/rfc/implemented/process/2026-07-02-tool-schema-catalog.md b/docs/rfc/implemented/process/2026-07-02-tool-schema-catalog.md
index 7d79f73583..7da1a05a3a 100644
--- a/docs/rfc/implemented/process/2026-07-02-tool-schema-catalog.md
+++ b/docs/rfc/implemented/process/2026-07-02-tool-schema-catalog.md
@@ -31,7 +31,7 @@ The filesystem discovers the tool-package inventory and the completeness guard r
 
 ### Scope
 
-Shipped product tool PACKAGES under `packages/*/tool-*`, each booted with its default config: `dsh-tool-bash` (`bash`, `bash_output`, `bash_kill`), `dsh-tool-todo` (`todo_write`), `dsh-tool-subagent` (`subagent`). The `examples/` demo tools (`echo`) are excluded, matching the cordis catalog's packages-only scope — a demo tool is not part of the product surface a reader is cataloguing.
+Shipped product tool packages under `packages/*/tool-*`, each booted with its default config, including `dsh-tool-bash` (`bash`), `dsh-tool-tasks` (`task_output`, `task_list`, `task_kill`), and `dsh-tool-subagent` (`subagent`). Example-only tools are excluded.
 
 The catalog unit is a package, not every configured tool instance. Each package boots once with default config; load-time aliases such as `subagent_fork` are noted without enumerating every deployment permutation. A deployment inventory is a separate, unbounded surface.
 
diff --git a/docs/rfc/implemented/process/2026-07-04-persistence-log-catalog.md b/docs/rfc/implemented/process/2026-07-04-persistence-log-catalog.md
index 0de6255322..70a6d60c1a 100644
--- a/docs/rfc/implemented/process/2026-07-04-persistence-log-catalog.md
+++ b/docs/rfc/implemented/process/2026-07-04-persistence-log-catalog.md
@@ -4,19 +4,19 @@ Status: implemented
 
 ## Problem
 
-`SessionEventMap` is the on-disk vocabulary, but its declarations are split across the owning session package and declaration merges. The generated persistence catalog is the single reference for every event and payload; hand-maintained tables drift and are removed. These records are not Cordis events—observers receive them through the single `session/event` bus event—so the Cordis catalog cannot cover them. The generator discovers all declarations and the doc-sync freshness gate rejects omissions or stale output.
+`SessionEventMap` is the on-disk vocabulary, but its declarations are split across the owning session package and declaration merges. The generated persistence catalog is the single reference for every event, its complete payload declaration and source JSDoc, and the shared `SessionEvent` envelope; hand-maintained tables drift and are removed. These records are not Cordis events—observers receive them through the single `session/event` bus event—so the Cordis catalog cannot cover them. The generator discovers all declarations and the doc-sync freshness gate rejects omissions or stale output.
 
 ## Decision
 
 Generate `docs/persistence-catalog.md` from source, with a freshness gate, as the fourth reference surface: the *records* a persisted session log can contain, complementing the cordis catalog (wiring), core-data-structures (vocabulary), and the tool catalog (tools).
 
-`gen-persistence-catalog.ts` scans every owning and declaration-merged `SessionEventMap` with the TypeScript AST. It renders source JSDoc, payload type, derived surface badge, reference links, and source location. The doc-sync freshness check rejects a vocabulary change whose catalog was not regenerated.
+`gen-persistence-catalog.ts` scans every owning and declaration-merged `SessionEventMap` with the TypeScript AST. It renders each member from its leading JSDoc through the complete payload type, retaining nested property comments and removing only its containing indentation, and also pastes the owning `SessionEventType`, `SurfaceEventType`, `SurfaceOp`, and `SessionEvent` declarations that compose the persisted envelope. Derived surface badges, reference links, and source locations remain outside the declaration blocks. The doc-sync freshness check rejects a vocabulary or envelope change whose catalog was not regenerated.
 
 Specific choices:
 
-- **JSDoc completeness, enforced.** Every member must carry description prose — the JSDoc becomes the catalog entry, the same forcing function the cordis catalog applies to bus events. An `@mode` tag on a member is a hard error: dispatch modes belong to cordis bus events, and a log event has none — the tag would misread as "this fires on the bus with mode X". Violations aggregate into one error listing every offender.
+- **JSDoc completeness, enforced.** Every member and rendered envelope type must carry description prose, and the full source JSDoc stays attached to its declaration in the catalog. An `@mode` tag is a hard error: dispatch modes belong to cordis bus events, and persisted records have none. Violations aggregate into one error listing every offender.
 - **The surface badge is derived, not hand-listed.** `SurfaceEventType` — the subset that produces LLM messages and may carry `surfaceOp` — is parsed from its union declaration in the owning package; a union member naming no declared event is a hard error (a stale union member would otherwise silently badge nothing). Everything else renders **log-only**.
-- **A dedicated fence.** Payload blocks use a ` ```ts persistence-catalog ` info string that `doc-typecheck` recognizes and skips, excluded from the opt-out ratio — the same treatment as `ts cordis-catalog` (a bare payload fragment is not standalone-compilable).
+- **A dedicated fence.** Declaration blocks use a ` ```ts persistence-catalog ` info string that `doc-typecheck` recognizes and skips, excluded from the opt-out ratio — the same treatment as `ts cordis-catalog` (the declarations reference types from their owning modules and are not standalone-compilable).
 - **Repo scope.** The catalog enumerates the packages in this repo, matching the siblings' packages-only scope; a downstream plugin can merge further event types, which are outside the catalog by construction. The walk defends its own assumptions with hard errors: the owning top-level `interface SessionEventMap` must be the single exported declaration in `@deepseek-ai/dsh-session` (an unrelated, local, or duplicate same-named interface cannot be catalogued as the on-disk vocabulary), no declaration may carry `extends` (inherited keys would join `keyof SessionEventMap` without a catalog row), every member must be a property signature with an explicit payload type (a method-form member would join `keyof` yet slip past a silent walk), and a duplicate member across declarations fails.
 
 This supersedes the hand-copies: the session.md `hook/*` table, the compact README's event table, the hook-protocol README's payload bullets, and the session README's name-list now link the catalog instead of restating payloads (the surrounding semantics prose stays where it was). The two stray `@mode emit` tags on the hook-protocol merge members are removed — the new gate rejects them as the category error they were.
@@ -28,7 +28,7 @@ This supersedes the hand-copies: the session.md `hook/*` table, the compact READ
 
 ## Consequences
 
-- The catalog cannot drift: a vocabulary change the committed file doesn't reflect fails `verify-persistence-catalog` in the pre-push hook and CI, and a new merged event with no JSDoc fails the generator outright — a plugin can no longer add an undocumented on-disk record type.
-- Event prose has a single home, the JSDoc at the declaration; thin JSDoc yields a thin catalog entry, pressuring authors to document at the source.
+- The catalog cannot drift: a vocabulary or envelope change the committed file doesn't reflect fails `verify-persistence-catalog` in the pre-push hook and CI, and a new merged event with no JSDoc fails the generator outright — a plugin can no longer add an undocumented on-disk record type.
+- Event prose has a single home, the JSDoc at the declaration; the catalog preserves that JSDoc and any nested field comments without flattening or paraphrasing them.
 - The `SurfaceEventType` union is now structurally load-bearing for docs: renaming an event without updating the union (or vice versa) fails the generator, not just the compiler.
 - The badge derivation assumes the union stays a closed set of string literals with exactly one owner; a refactor away from that shape must update the generator in the same change.
diff --git a/docs/rfc/implemented/process/2026-07-06-parallel-pre-push-gates.md b/docs/rfc/implemented/process/2026-07-06-parallel-pre-push-gates.md
index f3a7b1e83c..4e27b111d6 100644
--- a/docs/rfc/implemented/process/2026-07-06-parallel-pre-push-gates.md
+++ b/docs/rfc/implemented/process/2026-07-06-parallel-pre-push-gates.md
@@ -14,9 +14,9 @@ Flattening those members directly into `lefthook.yml` solves the local hook only
 
 [lefthook.yml](../../../../lefthook.yml) keeps one pre-push job named `full check` and runs `pnpm run check:pre-push`. That package script delegates to [scripts/run-gates.ts](../../../../scripts/run-gates.ts), the same bounded scheduler CI uses.
 
-The `pre-push` mode expands into leaf gates for the unit suite, snapshot suite, build, `hygiene` members, `doc-sync` members, and module-graph freshness. The leaf list keeps the same gate vocabulary as the package scripts, including RFC classification and RFC format, while the runner schedules independent checks concurrently and prints one timing/output block per gate.
+The `pre-push` mode expands into leaf gates for the unit suite, snapshot suite, build, `hygiene` members, `doc-sync` members, and module-graph freshness. The leaf list keeps the same gate vocabulary as the package scripts, including RFC classification and RFC format, while the runner schedules independent checks with four active top-level workers by default; `DSH_GATE_CONCURRENCY` overrides that bound.
 
-The build gate makes the hook self-contained from a clean worktree. `publint` and `verify-node-next-types` wait for that build output, while source-only gates continue in parallel.
+The build gate makes the hook self-contained from a clean worktree. `publint`, `verify-node-next-types`, and the pre-push form of `doc-typecheck` wait for that build output, while source-only gates continue in parallel.
 
 [scripts/publint-all.ts](../../../../scripts/publint-all.ts) discovers the package list from `packages/<group>/<pkg>` and runs `publint` with a worker pool sized from `availableParallelism()`. `DSH_PUBLINT_CONCURRENCY` can cap or raise the worker count for local machines and CI runners with different resource profiles. Results are buffered per package and printed in deterministic package order, so parallel execution does not scramble each package's log block.
 
diff --git a/docs/rfc/implemented/process/2026-07-17-run-ci-examples-from-built-lib.i18n.yaml b/docs/rfc/implemented/process/2026-07-17-run-ci-examples-from-built-lib.i18n.yaml
new file mode 100644
index 0000000000..9424f2aa24
--- /dev/null
+++ b/docs/rfc/implemented/process/2026-07-17-run-ci-examples-from-built-lib.i18n.yaml
@@ -0,0 +1,6 @@
+# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
+# side as of the last confirmed-consistent state. Both languages carry equal authority;
+# after editing either side, bring the other along and re-record with:
+#   pnpm run verify-translation-pairing --write
+2026-07-17-run-ci-examples-from-built-lib.md: aae88ee965b4e2211f3a53aeb9c0ee4944d95c2a
+2026-07-17-run-ci-examples-from-built-lib.zh.md: 0cd3822a71b8f7399be93beaac74b2177fcbe7ca
diff --git a/docs/rfc/implemented/process/2026-07-17-run-ci-examples-from-built-lib.md b/docs/rfc/implemented/process/2026-07-17-run-ci-examples-from-built-lib.md
new file mode 100644
index 0000000000..aae88ee965
--- /dev/null
+++ b/docs/rfc/implemented/process/2026-07-17-run-ci-examples-from-built-lib.md
@@ -0,0 +1,42 @@
+# RFC: Run CI examples from built lib
+
+Status: implemented
+
+English | [中文](2026-07-17-run-ci-examples-from-built-lib.zh.md)
+
+## Problem
+
+CI boots examples and Cordis-backed test projects through `node --import tsx` and the root tsconfig `paths` map. This adds TypeScript transformation cost and changes package resolution: imports resolve to workspace source instead of following package `exports` into built `lib/`.
+
+These runs therefore do not test the same code or resolution behavior as an installed consumer. A package can pass CI while its built export graph is incomplete or resolves differently.
+
+## Decision
+
+Execution has two modes. `src` is the default local-development mode and uses tsx; `lib` is the strict CI mode and starts built bins with plain Node, without tsx or tsconfig path mapping.
+
+- CI subprocesses that boot an example or a checked-in `cordis.yml` use `lib` mode.
+- TypeScript fixtures that only implement an ACP or MCP peer and do not load Cordis run directly with Node. An explicit source-path regression may remain in `src` mode.
+
+### Resolution topology
+
+Every test Cordis config must resolve its bare modules by walking upward from the config directory.
+
+- `examples/` is one pnpm workspace member and provides the shared `examples/node_modules` resolution root.
+- Every checked-in test Cordis config, including snapshot configs and package-owned fixtures, lives under its corresponding `examples/<agent>/` tree. A config owned by `packages/<group>/<package>/` maps to `examples/<agent>/tests/fixtures/<group>/<package>/cordis.yml`; the test driver and assertions remain package-local.
+- Every package named by an example Cordis config is declared in both `examples/package.json` for `lib` resolution and the root `tsconfig.json` references for `src` mode.
+
+### Launch policy
+
+The shared Loader test harness selects `src` or `lib` from `DSH_EXAMPLE_MODE`. CI builds first and selects `lib`; an unset mode keeps the fast local source loop.
+
+## Alternatives considered
+
+- **Keep CI on tsx** — rejected because it preserves transformation overhead and source-only resolution behavior.
+- **Use lib everywhere** — rejected because local development would require a build before every run. Dual mode keeps that cost out of the development loop.
+- **Build a private `node_modules` tree per test** — rejected because it duplicates consumer scaffolding. The `examples/` workspace root gives every Cordis config one real and declared resolution path.
+
+## Consequences
+
+- CI validates built package exports without tsx changing module resolution; local development retains the no-build source loop.
+- CI must build before these tests, and manual `lib` runs can observe stale local artifacts.
+- Cordis config dependencies are not visible to normal TypeScript import analysis, so `examples/package.json` and the root tsconfig references must stay synchronized with the configs.
diff --git a/docs/rfc/implemented/process/2026-07-17-run-ci-examples-from-built-lib.zh.md b/docs/rfc/implemented/process/2026-07-17-run-ci-examples-from-built-lib.zh.md
new file mode 100644
index 0000000000..0cd3822a71
--- /dev/null
+++ b/docs/rfc/implemented/process/2026-07-17-run-ci-examples-from-built-lib.zh.md
@@ -0,0 +1,42 @@
+# RFC: 在 CI 中从构建后的 lib 运行示例
+
+Status: implemented
+
+[English](2026-07-17-run-ci-examples-from-built-lib.md) | 中文
+
+## 问题
+
+CI 通过 `node --import tsx` 和根 tsconfig 的 `paths` 映射启动示例与加载 Cordis 配置的测试项目。这种方式既增加了 TypeScript 转换开销，也改变了包解析行为：import 会解析到 workspace 源码，而不是经包的 `exports` 进入构建后的 `lib/`。
+
+因此，这些测试没有覆盖已安装消费方实际运行的代码和解析路径。即使包的构建导出图不完整或解析结果不同，CI 仍可能通过。
+
+## 决策
+
+执行机制包含两种模式。`src` 是本地开发的默认模式并使用 tsx；`lib` 是严格的 CI 模式，通过 plain Node 启动构建后的 bin，不加载 tsx，也不使用 tsconfig 路径映射。
+
+- CI 中启动示例或签入仓库的 `cordis.yml` 的子进程使用 `lib` 模式。
+- 仅实现 ACP 或 MCP 对端、且不加载 Cordis 的 TypeScript fixture（测试前置数据）直接由 Node 运行。只有显式验证源码路径的回归测试可以保留 `src` 模式。
+
+### 解析拓扑
+
+每个测试 Cordis 配置都必须能从配置文件所在目录向上解析裸模块。
+
+- `examples/` 作为一个 pnpm workspace 成员，提供统一的 `examples/node_modules` 解析根目录。
+- 所有签入仓库的测试 Cordis 配置，包括快照配置和包内测试 fixture，都放在对应的 `examples/<agent>/` 目录树下。归属 `packages/<group>/<package>/` 的配置映射到 `examples/<agent>/tests/fixtures/<group>/<package>/cordis.yml`；测试驱动和断言仍留在包内。
+- 示例 Cordis 配置中引用的每个包都同时登记在 `examples/package.json` 和根 `tsconfig.json` 的 references 中，分别支持 `lib` 与 `src` 解析。
+
+### 启动策略
+
+共享 Loader 测试 harness 通过 `DSH_EXAMPLE_MODE` 选择 `src` 或 `lib`。CI 先构建再选择 `lib`；未设置模式时保留快速的本地源码开发回路。
+
+## 曾考虑的替代方案
+
+- **CI 继续使用 tsx**：不予采纳，因为它会保留转换开销和仅适用于源码的解析行为。
+- **所有环境只使用 lib**：不予采纳，因为本地开发每次运行前都必须构建。双模式避免把这项成本带入开发回路。
+- **每个测试单独构造 `node_modules`**：不予采纳，因为它会重复消费方脚手架。以 `examples/` 作为 workspace 根，可让每个 Cordis 配置通过同一条真实且显式声明的路径解析模块。
+
+## 后果
+
+- CI 可以验证构建后的包导出，不再受 tsx 模块解析影响；本地开发仍保留免构建的源码回路。
+- CI 必须先构建再运行这些测试；手动执行 `lib` 模式时可能读取陈旧的本地产物。
+- 常规 TypeScript import 分析无法识别 Cordis 配置依赖，因此 `examples/package.json`、根 tsconfig references 与配置文件必须保持同步。
diff --git a/docs/rfc/implemented/simplification/2026-06-26-fsspec-style-fs-seam.md b/docs/rfc/implemented/simplification/2026-06-26-fsspec-style-fs-seam.md
index 2663637859..d2b4c6dc4f 100644
--- a/docs/rfc/implemented/simplification/2026-06-26-fsspec-style-fs-seam.md
+++ b/docs/rfc/implemented/simplification/2026-06-26-fsspec-style-fs-seam.md
@@ -35,7 +35,7 @@ This RFC decided the four-layer split, the provider contract, and the freshness
 `@deepseek-ai/dsh-fs` shrinks to provider text IO plus guarded text mutation:
 
 ```ts ignore-check
-abstract resolve(path: string): Promise<FsTarget>
+abstract resolve(path: string, opts?: { cwd?: string; signal?: AbortSignal }): Promise<FsTarget>
 abstract stat(target: FsTarget, signal?: AbortSignal): Promise<FsInfo | undefined>
 abstract readText(target: FsTarget, signal?: AbortSignal): Promise<string>
 abstract streamText(target: FsTarget, signal?: AbortSignal): Promise<AsyncIterable<string>>
diff --git a/docs/rfc/implemented/simplification/2026-07-12-simplify-session-log-representation.md b/docs/rfc/implemented/simplification/2026-07-12-simplify-session-log-representation.md
new file mode 100644
index 0000000000..7bf44c6c06
--- /dev/null
+++ b/docs/rfc/implemented/simplification/2026-07-12-simplify-session-log-representation.md
@@ -0,0 +1,33 @@
+# RFC: Simplify session-log representation
+
+Status: implemented
+
+## Problem
+
+The session log maintains two representations that cost more machinery than their consumers require: a pseudo-linked surface and custom request-header deltas.
+
+`SurfaceManager` stores the same order in an array, a seq map, and mutable `prev`/`next` links. Production never reads either link: compact's tool-pairing balance answers from per-cut balances cached in surface order. Replacement already uses `indexOf`, so the links do not make its dominant operation constant-time. A seq array with linear replacement lookup has the same asymptotic replacement cost and one representation to validate.
+
+The request-header subsystem implements a custom system/tool delta codec and transmission-decision layer even though its contract says deltas are an encoding optimization, not a reconstructability requirement. Retaining the initial/resume full snapshot at each loop-instance boundary, then writing a canonical full `request/header` whenever that instance's assembled header changes, preserves replay while deleting `SystemDelta`, `ToolsDelta`, round-trip fallback, and the durable `request/header-delta` variant. Codec-only vocabulary disappears with the codec, not because its individual arms were invalid.
+
+The implementation retains append and replacement `sourceEventSeqs`, crash-repair provenance, and all `SessionStartSource` variants because those fields have an audit/interception role that zero current readers does not overturn.
+
+## Decision
+
+`SurfaceManager.nodes` is a `readonly number[]` of event sequences; the public `SurfaceNode` shape, node links, and seq-to-node map are removed. The internal replace-generation signal remains. The complete `foldSurface()` read used by session-query returns the same number-array representation plus replacement metadata without making the incremental manager retain history. Tool-pairing balance and compaction use event sequences and surface positions; the compact-owned per-cut balance cache does not depend on node links.
+
+Request headers use canonical full snapshots only. Initial and resume anchors remain full snapshots even when unchanged; an in-instance change appends another full `request/header` with reason `change`. The delta event, codec types, diff/apply helpers, and codec-only `fallback` reason are removed. Request reconstruction selects the latest snapshot.
+
+`SESSION_FORMAT_VERSION` remains pinned at `0`, so seed, append, and persistence-load validation explicitly reject old v0 `request/header-delta` events and full snapshots carrying the removed `fallback` reason. There is no compatibility fold or migration. JSONL and SQLite tests pin this fail-loud boundary, and the ACP snapshot harness represents legitimate mid-session changes as full pinned headers and full readable prompts.
+
+## Alternatives considered
+
+**Keep linked nodes and compact deltas for possible scale.** Links could help a future cursor API, and deltas can reduce logs when large tool schemas change by a small amount. No shipped cursor uses the links, while full snapshots trade disk size for substantially simpler correctness. If header volume proves material, compression or a measured canonical-delta scheme can be designed around real traces.
+
+## Verification
+
+Unit coverage pins ordered-surface append/replace behavior, tool pairing, compaction, full-header folding/logging, request reconstruction, and dev invariants. Seed validation plus JSONL and SQLite load tests reject the legacy event before replay. The keyless ACP suite exercises record, refresh, replay, changed-header pinning, and the sandbox mode-switch fixture in the new shape.
+
+## Consequences
+
+Full headers increase log volume, and linear replacement lookup could be slower on very large surfaces. Replacements were already linear because the prior implementation called `indexOf`; benchmarks are deferred until real traces show the simpler array is a bottleneck. The format version remains `0`, so explicit legacy-event rejection is a permanent part of the pre-release format boundary. In return, surface order and request-header state each have one representation, deleting link maintenance, maps, codec arms, round-trip fallback, and delta-aware snapshot normalization.
diff --git a/docs/rfc/implemented/testing/2026-07-06-pin-request-header-content-in-one-scenario.md b/docs/rfc/implemented/testing/2026-07-06-pin-request-header-content-in-one-scenario.md
index 42c67320f6..69af84b0f0 100644
--- a/docs/rfc/implemented/testing/2026-07-06-pin-request-header-content-in-one-scenario.md
+++ b/docs/rfc/implemented/testing/2026-07-06-pin-request-header-content-in-one-scenario.md
@@ -8,11 +8,11 @@ An ACP snapshot suite needs to prove the exact composed system prompt and tool-s
 
 ## Decision
 
-Exactly one scenario per header-composition class is flagged `pinsHeader`. Its directory splits the pin by review format: `system-prompt.golden.md` contains the normalized composed prompt as ordinary Markdown, `tool-schemas.golden.json` contains the complete initial schemas and later schema edits as structured JSON, and `session.jsonl` retains config, reason, and any model-visible prefix while storing `header.system` and `header.tools` as `"{{system}}"` / `"{{tools}}"`. Every other JSONL uses the same prompt and tool tokens and also tokenizes session-prefix content. The pin mechanics live in [`dsh-acp-snapshot`](../../../../packages/support/acp-snapshot/README.md), whose suite factory enforces one pin per class.
+Exactly one scenario per header-composition class is flagged `pinsHeader`. Its directory splits the pin by review format: `system-prompt.golden.md` contains the normalized full prompt sequence as ordinary Markdown, `tool-schemas.golden.json` contains the corresponding complete schema sequence as structured JSON, and `session.jsonl` retains config, reason, and any model-visible prefix while storing `header.system` and `header.tools` as `"{{system}}"` / `"{{tools}}"`. Every other JSONL uses the same prompt and tool tokens and also tokenizes session-prefix content. The pin mechanics live in [`dsh-acp-snapshot`](../../../../packages/support/acp-snapshot/README.md), whose suite factory enforces one pin per class.
 
-The pure `scrubSystemPrompts` and `scrubToolSchemas` normalizers apply to every stored session fixture and independently tokenize initial-header content plus header-delta bulk. `scrubRequestHeaders` also tokenizes session-prefix content for non-pinning scenarios while retaining structural facts: system-delta positions and arity, added/removed/changed tool names, prefix message count, field presence, config, and reason. Record and refresh write-back apply the appropriate scrub before writing JSONL and regenerate both sidecars from the normalized live header and deltas, so neither path can reintroduce prompt/schema bulk into JSONL or leave a review artifact stale.
+The pure `scrubSystemPrompts` and `scrubToolSchemas` normalizers independently tokenize every stored full header. `scrubRequestHeaders` also tokenizes session-prefix content for non-pinning scenarios while retaining header count, field presence, config, reason, and prefix message count. Record and refresh write-back apply the appropriate scrub before writing JSONL and regenerate both sidecars from the normalized live full-header sequence, so neither path can reintroduce prompt/schema bulk into JSONL or leave a review artifact stale.
 
-Guards make the split self-enforcing. On disk, every `session*.jsonl` is a fixed point of both prompt and schema scrubbers, only non-pinning fixtures must be fixed points of the full header scrub, both sidecars exist exactly beside pinning fixtures in canonical newline-terminated formats, and each class has one pin. Live, every `request/header` produced by a parent, spawn child, fork child, initial request, or resume must match the reconstructed pin after volatile-value normalization; the pinning run's prompt and schema deltas must also match their sidecars. A header without a string prompt, without an array-valued tool list, or with an undeclared `request/header-delta` fails loud.
+Guards make the split self-enforcing. On disk, every `session*.jsonl` is a fixed point of both prompt and schema scrubbers, only non-pinning fixtures must be fixed points of the full header scrub, both sidecars exist exactly beside pinning fixtures in canonical newline-terminated formats, and each class has one pin. Live, every `request/header` produced by a parent, spawn child, fork child, initial request, resume, or in-instance change must match the reconstructed class sequence after volatile-value normalization. A header without a string prompt, without an array-valued tool list, or beyond the pin's declared changed-header count fails loud.
 
 One pin covers the whole suite because every session — parent, spawn child, fork child — composes the identical tool list and the identical prompt modulo cwd, and the uniformity guard fails the suite the moment that stops holding. If header composition ever becomes session-dependent by design (a restricted subagent toolset, say), the divergent shape gets its own pinning scenario.
 
@@ -26,7 +26,7 @@ One pin covers the whole suite because every session — parent, spawn child, fo
 
 ## Verification
 
-The suite replays every scenario against the split pins. Unit coverage exercises the independent and full scrubbers, both sidecar formats, record/refresh regeneration, normalized prompt/schema extraction, fixed-point enforcement, required-file symmetry, reconstructed-header uniformity, and delta rejection.
+The suite replays every scenario against the split pins. Unit coverage exercises the independent and full scrubbers, both full-header sidecar formats, record/refresh regeneration, normalized prompt/schema extraction, fixed-point enforcement, required-file symmetry, reconstructed-header uniformity, and changed-header count rejection.
 
 ## Consequences
 
diff --git a/docs/rfc/implemented/testing/2026-07-08-shared-acp-snapshot-package.md b/docs/rfc/implemented/testing/2026-07-08-shared-acp-snapshot-package.md
index 1fb887db52..b854f3443b 100644
--- a/docs/rfc/implemented/testing/2026-07-08-shared-acp-snapshot-package.md
+++ b/docs/rfc/implemented/testing/2026-07-08-shared-acp-snapshot-package.md
@@ -16,7 +16,7 @@ The machinery lives in [`packages/support/acp-snapshot`](../../../../packages/su
 
 **`src/normalize.ts`** — the pure normalizers, hook-free by policy: when a future event carries a new volatile field (an approval duration, say), the shared normalizer learns it in the same change, keeping one home for what "normalized" means rather than per-suite scrub extensions.
 
-**`src/suite.ts`** — the `Scenario` type and `defineAcpSnapshotSuite(options)`, registering the per-scenario compares, record/refresh fixture write-back, the header pin with its live uniformity guard, and the fixture guard block (no orphan scenario dirs, required files present, exactly one pin per class, every JSONL a `scrubSystemPrompts` fixed point, non-pinning fixtures also `scrubRequestHeaders` fixed points). The pinned-header contract ([pinned-header RFC](2026-07-06-pin-request-header-content-in-one-scenario.md)) is per-suite: each header class flags exactly one `pinsHeader` scenario, whose `system-prompt.golden.md` and JSONL tool list split the composed header into reviewable artifacts; the uniformity guard compares both against every live header in that class. The pure helpers (`childFixturePaths`, `fixtureContext`, `normalizedHeaders`, `normalizedSystemPrompts`, `formatSystemPromptSnapshot`, `headerDeltaCount`) are exported from the module for direct unit coverage.
+**`src/suite.ts`** — the `Scenario` type and `defineAcpSnapshotSuite(options)`, registering the per-scenario compares, record/refresh fixture write-back, the header pin with its live uniformity guard, and the fixture guard block (no orphan scenario dirs, required files present, exactly one pin per class, every JSONL a `scrubSystemPrompts` fixed point, non-pinning fixtures also `scrubRequestHeaders` fixed points). The pinned-header contract ([pinned-header RFC](2026-07-06-pin-request-header-content-in-one-scenario.md)) is per-suite: each header class flags exactly one `pinsHeader` scenario, whose `system-prompt.golden.md` and JSONL tool list split the composed header into reviewable artifacts; the uniformity guard compares both against every live header in that class. A pinning scenario declares any legitimate changed-header count, and its Markdown artifact records every full changed prompt. The pure helpers (`childFixturePaths`, `fixtureContext`, `normalizedHeaders`, `normalizedSystemPrompts`, `formatSystemPromptSnapshot`, `headerChangeCount`) are exported from the module for direct unit coverage.
 
 ## Alternatives considered
 
diff --git a/docs/rfc/proposed/architecture/2026-06-20-generic-long-running-tool-runtime.md b/docs/rfc/proposed/architecture/2026-06-20-generic-long-running-tool-runtime.md
deleted file mode 100644
index 5ed9c6354e..0000000000
--- a/docs/rfc/proposed/architecture/2026-06-20-generic-long-running-tool-runtime.md
+++ /dev/null
@@ -1,41 +0,0 @@
-# RFC: Extract a generic long-running tool runtime
-
-Status: proposed
-
-## Problem
-
-The bash capability seam supports both foreground commands and long-running background tasks. Background support is large: the abstract executor exposes `start`, `get`, `ownerOf`, `list`, `readOutput`, `kill`, and `onTaskDone`; the local executor tracks tasks, incremental reads, owner tokens, process cleanup, and completion listeners; the model sees three tools (`bash`, `bash_output`, `bash_kill`); the tool plugin injects completion notices back into the owning agent's session. The local executor fences task access behind owner tokens because predictable global task ids are a cross-session read/kill hazard.
-
-The [tool cookbook](../../../cookbook/adding-a-tool.md) already points at the real design smell: background bash is really generic long-running-tool infrastructure living inside one tool. If future tools need background execution, polling, kill, ownership, and completion notices, those semantics should not be hidden in `dsh-bash`.
-
-## Proposal
-
-Move long-running task semantics above bash into a tool-agnostic runtime. Bash remains able to run background commands, but it stops owning the general concepts of task ids, ownership tokens, polling, cancellation, completion notifications, and model-facing "read/kill this task" commands.
-
-The runtime should own:
-
-- Stable task ids and owner tokens keyed to the calling session/agent.
-- Registration of a long-running task with a producer for incremental output and a completion promise.
-- Generic read/cancel/list operations with the same cross-session authorization rule for every tool.
-- Completion notification injection into the owning session.
-- Presentation hooks for pending/running/completed task state, with bash supplying only command-specific labels and output formatting.
-
-`dsh-bash` then keeps the bash-specific execution contract: resolve a request into a command spec, run a foreground command, or start a process and hand its streams/process handle to the generic runtime. `dsh-tool-bash` keeps the model-facing command tool, but the follow-up operations become generic long-running-tool operations or a shared utility that bash registers with, rather than bespoke `bash_output`/`bash_kill` plumbing.
-
-## Current seam consumption
-
-Current consumers split cleanly: `dsh-tool-bash` uses the full foreground/background seam, while hook bridges use only foreground `resolve` and `run` with trusted `stdin` and `env`. `get` and `list` are test-only; `BashTask.done` is implementation-only for disposal, while production completion uses `onTaskDone`. An extracted runtime should expose one public completion mechanism, preserve the simple foreground path for hooks, and decide whether background `timeoutMs` belongs on `start`. If it owns process spawning, it should also centralize the duplicated credential scrub.
-
-## Acceptance criteria
-
-- The bash-specific packages no longer define the generic task registry, owner-token authorization, polling, cancellation, or completion-notification machinery.
-- A shared long-running-task service or tool layer owns those semantics and is documented as the path for any future background-capable tool.
-- Bash background behavior remains available through the shared layer, with tests proving cross-session isolation still holds.
-- ACP and snapshot fixtures render background bash through the shared task vocabulary, not through bash-only lifecycle semantics.
-- The [tool cookbook](../../../cookbook/adding-a-tool.md) points long-running tools at the shared runtime instead of telling each tool to invent its own task protocol.
-
-## Risks
-
-The bash package loses local ownership of an already-working background-task implementation, and the implementing PR may temporarily churn model-facing tool names or transcript presentation. That churn is worthwhile if it leaves one background-task contract instead of making every future long-running tool clone bash's private protocol.
-
-<!-- rfc-format: alternatives-not-recorded (pre-format RFC) -->
diff --git a/docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.i18n.yaml b/docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.i18n.yaml
new file mode 100644
index 0000000000..6e301ee06a
--- /dev/null
+++ b/docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.i18n.yaml
@@ -0,0 +1,6 @@
+# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
+# side as of the last confirmed-consistent state. Both languages carry equal authority;
+# after editing either side, bring the other along and re-record with:
+#   pnpm run verify-translation-pairing --write
+2026-07-15-sdk-project-editing-architecture.md: 985cc22c159c68801b78262aa96c7422bdfa1318
+2026-07-15-sdk-project-editing-architecture.zh.md: 6a194e8e5f193e62bfc283fd93a5fde0367fe196
diff --git a/docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.md b/docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.md
new file mode 100644
index 0000000000..985cc22c15
--- /dev/null
+++ b/docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.md
@@ -0,0 +1,129 @@
+# RFC: SDK project editing architecture
+
+Status: proposed
+
+English | [中文](2026-07-15-sdk-project-editing-architecture.zh.md)
+
+## Problem
+
+[Developer-owned SDK projects](../feature/2026-07-14-sdk-developer-projects.md) are created through create, adjusted through config, and built and run through commands such as start. Initial creation, configuration changes, and build and runtime commands all need to understand features, feature options, npm dependencies, Cordis config entries, environment variables, package managers, local plugins, and several project files. If each project-reading and project-writing workflow uses a separate interpretation protocol, the SDK developer workflows become difficult to maintain.
+
+## Proposal
+
+The SDK uses one shared object-oriented project model. `SdkProject` is a read-only snapshot, and `ProjectEditSession` is the only mutation and commit boundary. Feature objects own their feature options, relationships, resource contributions, and current-state inspection. Create and config orchestrate only their respective user workflows and modify projects through the same domain operations.
+
+Structured files are modified through document objects, while one-shot text artifacts are generated from complete templates. Questions are typed objects presented through clack. Diff calculation may remain an edit-session implementation detail, but it is not a public execution protocol that callers must assemble.
+
+## Terminology
+
+| Term | Usage in this RFC | Meaning |
+|---|---|---|
+| Feature | feature | A product unit curated and managed by the SDK; one feature may contain several feature options and contribute several Cordis config entries, npm dependencies, environment placeholders, and owned files |
+| Feature option | feature option | A finite selectable implementation or configuration shape within one feature; feature rules may make options fixed, exclusive, or additive |
+| Cordis plugin | Cordis plugin | A plugin implementation loaded by Cordis, usually exported by an npm package; it is not an item in `cordis.yml` |
+| Cordis config entry | Cordis config entry | One item in the `cordis.yml` plugin list, identified as an instance by `id` and referring to a Cordis plugin through `name` |
+| Cordis plugin config | Cordis plugin config | The configuration object or shape exposed by a Cordis plugin; an individual field owned and updated by a feature is a config key |
+| config key | config key | One field in Cordis plugin config; a feature updates only the config keys it declares as owned and preserves unknown config keys |
+| npm dependency | npm dependency | A package relationship in `package.json`; literal fields such as `dependencies` and `devDependencies` keep their names |
+| Feature requirement | feature requirement | A relationship declared through `requires` by a feature or feature option |
+
+## Package boundaries
+
+| Package | Responsibility | Does not own |
+|---|---|---|
+| `@deepseek-ai/dsh-helper` | Edit sessions, feature configuration, project-template rendering, package-manager adaptation, and prompt interaction adaptation | Booting Cordis applications or deciding create/config terminal workflows |
+| `@deepseek-ai/dsh-scripts` | `dsh-sdk start/dev/build/config`, process lifecycle, project entry loading, the config workflow, and its terminal-copy templates | Interpreting feature definitions directly or modifying YAML/JSON ASTs |
+| `@deepseek-ai/create-sdk` | Arguments, question order, initial project creation, installation finish, and terminal-copy templates for `npm create @deepseek-ai/sdk` | Becoming a generated project's runtime npm dependency or providing a library API |
+
+`@deepseek-ai/create-sdk` is the only exception to the repository's `@deepseek-ai/dsh-*` naming rule. npm's scoped-initializer convention requires that package name for `npm create @deepseek-ai/sdk`. The exception is a repository architecture fact and does not add a third developer product entrypoint.
+
+The three packages export only the narrow entrypoints consumed by adjacent layers and provide no `src/*` deep imports. The scripts library entrypoint and build-config subpath serve generated code and project build configuration, while the developer product contract remains the `dsh-sdk` commands.
+
+## Project aggregate and edit session
+
+`SdkProject.create(root, request)` constructs a new project snapshot that has not been written, while `SdkProject.open(root)` loads an existing project. Open requires only readable root `package.json` and `cordis.yml` files; every other file is an optional resource. Both paths return the same read-only aggregate and distinguish their source through explicit origin state.
+
+`project.edit()` clones project documents into a working copy. Domain commands such as install, configure, enable, disable, and addPlugin modify only the working copy. Each command immediately re-inspects its owning feature, and the final commit checks all relationships and files again.
+
+```text
+validate feature requirements and resource ownership
+  -> validate every affected document
+  -> compute changed and removed paths
+  -> compare existing files with the session's original text
+  -> write through one commit boundary
+  -> return a new SdkProject snapshot and ChangeSet
+```
+
+Validation failure or an external edit causes zero writes. “One commit” means only zero pre-write side effects and one write entrypoint. `ChangeSet` describes final feature, plugin, and file changes for Review & Apply and create completion.
+
+## Features and resource ownership
+
+A feature is a first-class behavior object. Shallow base classes implement install, configure, enable, disable, required/requires validation, and common state inspection. Features with fixed, exclusive, or additive feature options share these lifecycles. Only features whose resource contributions depend on project context or require custom round-tripping use dedicated behavior classes; other features declare their actual differences through standardized data.
+
+Each feature contributes stable-keyed Cordis config entries, npm dependencies, environment placeholders, and owned files. The registry rejects two features that declare the same resource key during initialization. Different feature options within one feature may share resources, which that feature resolves from the final option set.
+
+A Cordis config entry anchors feature installation. The npm package name assigns the entry to a feature, and the entry ID distinguishes several instances of one plugin package. An npm dependency without a feature-owned Cordis config entry leaves the feature uninstalled. Once a Cordis config entry exists, a missing npm dependency, unreadable Cordis plugin config, or resource conflict puts the feature into an inconsistent state; the config command shows diagnostics and refuses speculative modification.
+
+Configuring the same feature option updates only its owned config keys and preserves unknown keys. Replacing a feature option removes old resources that are exclusive and still confirmable. If an old resource cannot be confirmed or an owned file was modified by the developer, the whole operation fails.
+
+## Questions and workflows
+
+TypeScript `Question<T>` objects keep defaults, validation, applicability, and types together. `PromptPort` is the only interface between the domain layer and the terminal library, and helper provides one thin `ClackPromptPort`. Create and config inject their own command-line input and output streams and retain ownership of cancellation, return, and completion semantics in their workflows.
+
+Create keeps its stateful question order in one wizard, while config keeps final-state selection in one workflow. Both use the same feature configurator for feature options and dedicated inputs, so adding an ordinary feature, feature option, or parameter does not require changes to both entrypoints.
+
+## Project documents and templates
+
+Only structured files that helper reads or modifies have concrete document objects: `package.json`, `cordis.yml`, `.env`, `.env.example`, the root `tsconfig.json`, and the pnpm workspace file. Document objects own parsing, cloning, validation, and serialization. Concrete classes and modules use `*File` and `*-file.ts` names respectively. Business code does not manipulate YAML/JSON ASTs directly, and malformed shapes fail loudly at the owning document boundary.
+
+README, entrypoint code, build configuration, `.gitignore`, and other one-shot text artifacts use one complete template per real file. Complete product copy such as CLI usage, creation and recovery messages, installation and retry guidance, and the default persona also comes from package-local templates owned by the package that presents it.
+
+Helper provides the generic typed `TextTemplate` renderer, and caller packages load their own templates through package-local asset URLs.
+
+Templates use Handlebars strict mode and `noEscape` without custom processing. File owners encode typed values for the target language. Template source escapes interpolation as `\{{model}}` when it must emit the downstream literal unchanged.
+
+## Command and runtime boundary
+
+Scripts supports `dsh-sdk start/dev/build/config`. Start dynamically loads a module target and calls its named entrypoint. Dev adds TypeScript and local-workspace source resolution before following the same path. Build invokes the project's installed tsdown. Config opens one edit session and commits after Review & Apply. Generated projects run `tsc -b` directly for typechecking.
+
+HMR is an explicit Cordis config entry loaded by dev and start. Its required `node-addon-require-builtin` package is supplied transitively by the scripts package and is absent from the generated project's `package.json`.
+
+Dev and start execute the developer entrypoint, where developer code handles command-line arguments and cwd. Developers pass `--model=<name>` and `--resume=<session-id>` to start the standard flow.
+
+## Repository live-link mode
+
+Create-sdk retains a hidden `--link-workspace` option for Harness repository development and e2e. The parser accepts it, but help, public flag lists, and ordinary user documentation omit it. It accepts no repository-path parameter; the repository root is derived upward from the executing create-sdk module.
+
+Link mode preserves the ordinary project file shape. `@deepseek-ai/*` points into `packages/`, Cordis-related npm dependencies point into `vendor/`, and shared lower-level packages resolve to the same physical copy used by the repository so Cordis type merging cannot produce multiple module type definitions. npm uses `file:`, pnpm uses `link:` with automatic peer installation disabled, and Yarn uses `portal:` plus resolutions. Repository packages must be built first.
+
+## Future work
+
+- **Replaceable required spine roles.** The current `spine` owns the full implementation set, including SystemPrompt and LLMService, through one fixed feature option. Developers cannot replace or switch these roles and must edit Cordis config entries manually.
+- **Service contracts and package declarations.** When replacing a builtin service, a Cordis plugin currently cannot declare the services it provides through `provides` metadata, so the SDK cannot assist configuration during development or check compatibility at runtime. A corresponding protocol remains to be designed.
+- **Feature parameter descriptions.** Feature-specific inputs currently require handwritten declarations. The SDK cannot derive interactive parameters automatically from arbitrary Cordis plugin config or npm package.json information. Future declarative metadata may expose a limited parameter set without turning arbitrary Cordis plugin config into a generic form.
+- **SDK application-level configuration.** The current project resource model describes Cordis config entries and config keys owned by individual Cordis plugins, so every SDK-managed setting must belong to one plugin. Cross-plugin or whole-application settings have no independent persistence location. Future work must define an application-level configuration document and its ownership, read, and mutation boundaries.
+
+## Alternatives considered
+
+**Keep the static Catalog and central engine.** This minimizes the initial rewrite, but feature parameters, round-tripping, owned files, and create/config reuse continue to accumulate in one coordinator. Splitting files shortens the file without consolidating responsibility.
+
+**Use `wizard.json` and a generic Questionnaire.** Static forms cannot directly express feature requirements, option switches, existing-value refill, and project-resource changes. Types, gates, and dynamic options still connect through string registries and a procedural `run()`, creating another internal DSL.
+
+**Expose the live-link flag.** The mode depends on Harness monorepo layout and unpublished packages and serves repository development only. Making it public would create a project-creation contract that the SDK cannot support outside the repository.
+
+## Acceptance criteria
+
+- Create and config modify projects only through `SdkProject` and `ProjectEditSession`; any business, document, or concurrency validation failure before writing leaves the filesystem unchanged
+- Adding an ordinary feature, feature option, or parameter extends only its typed spec or owning behavior object, without adding a central switch to create or config workflows
+- Helper owns the feature model, npm dependency and other resource configuration, and inconsistent-state detection
+- Structured files change through `*File` document objects; one-shot files and complete product copy come from package-owned Handlebars templates, and business decisions do not enter a template DSL
+- `dsh-sdk start/dev/build/config` is the runtime product surface, typecheck uses `tsc -b` directly, HMR is not injected by command mode, and only the scripts package transitively supplies `node-addon-require-builtin`
+- `--link-workspace` exists only as a hidden repository-development option and preserves one module identity under npm, pnpm, and Yarn
+
+## Risks
+
+- Behavior objects and typed specs create two extension shapes. Dedicated classes must remain limited to features that truly depend on project context or custom behavior, or the design will grow a meaningless type hierarchy
+- Optimistic concurrency checks and pre-write validation cannot recover from an I/O failure during writing; callers must still report a possible partial commit to the developer
+- Hidden link mode depends on repository layout and package-manager link semantics and must change with either one
+- The Cordis loader resolves `node-addon-require-builtin` from its own module path, so the scripts package must continue to satisfy that optional peer under npm, pnpm, and Yarn npm dependency layouts
+- Handlebars `noEscape` makes typed model construction responsible for target-language encoding; new template fields must be escaped correctly at the owning boundary, and downstream Handlebars placeholders must be escaped explicitly in template source
diff --git a/docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.zh.md b/docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.zh.md
new file mode 100644
index 0000000000..6a194e8e5f
--- /dev/null
+++ b/docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.zh.md
@@ -0,0 +1,129 @@
+# RFC: SDK 工程编辑架构
+
+Status: proposed
+
+[English](2026-07-15-sdk-project-editing-architecture.md) | 中文
+
+## 问题
+
+[开发者拥有的 SDK 工程](../feature/2026-07-14-sdk-developer-projects.md) 由 create 创建，可以通过 config 调整，并由 start 等命令构建和运行。初始创建、配置调整和编译运行都需要理解功能、功能选项、NPM 依赖、Cordis 配置项、环境变量、包管理器、本地插件和多个项目文件。如果读写项目的各个流程分别使用不同的解析协议，SDK 开发者流程会变得难以维护。
+
+## 提案
+
+SDK 使用一个共享的面向对象工程模型。`SdkProject` 是只读快照，`ProjectEditSession` 是唯一修改与提交边界；功能对象负责自身的功能选项、关系、资源贡献和现状识别；create 与 config 只编排各自的用户流程，并通过同一组领域操作修改工程。
+
+结构化文件通过文档对象修改，一次性文本产物通过完整模板生成。问题由类型化对象表达，并使用 clack 交互。差异计算可以作为编辑会话的内部实现，但不成为要求调用方组装的公共执行协议。
+
+## 术语
+
+| 名词 | 本文用词 | 含义 |
+|---|---|---|
+| Feature | 功能 | SDK 人工策划和管理的产品单元；一项功能可以包含多个功能选项，并贡献多个 Cordis 配置项、NPM 依赖、环境变量占位和独占文件 |
+| Feature option | 功能选项 | 一项功能内有限、可选择的实现或配置形状；根据功能规则可以固定、互斥或多选 |
+| Cordis plugin | Cordis 插件 | Cordis 加载的插件实现，通常由一个 NPM 包导出；它不是 `cordis.yml` 中的一项配置 |
+| Cordis config entry | Cordis 配置项 | `cordis.yml` 插件列表中的一项，通过 `id` 标识实例并通过 `name` 指向 Cordis 插件 |
+| Cordis plugin config | Cordis 插件配置 | Cordis 插件公开的配置对象或配置结构；其中由功能拥有并更新的单个字段称为“配置键” |
+| config key | 配置键 | Cordis 插件配置中的单个字段；功能只更新自己声明拥有的配置键，并保留未知配置键 |
+| npm dependency | NPM 依赖 | `package.json` 中的包关系；`dependencies`、`devDependencies` 等字段保持原样 |
+| Feature requirement | 功能依赖 | 功能或功能选项通过 `requires` 声明的关系 |
+
+## Package 边界
+
+| Package | 责任 | 不负责 |
+|---|---|---|
+| `@deepseek-ai/dsh-helper` | 编辑会话、功能配置、工程模板渲染、包管理适配和 prompt 交互适配 | 启动 Cordis 应用或决定 create/config 的终端流程 |
+| `@deepseek-ai/dsh-scripts` | `dsh-sdk start/dev/build/config`、进程生命周期、项目入口加载、config 流程和所属终端文案模板 | 直接解释功能定义或修改 YAML/JSON AST |
+| `@deepseek-ai/create-sdk` | `npm create @deepseek-ai/sdk` 的参数、问题顺序、首次工程创建、安装收尾和所属终端文案模板 | 成为生成工程的运行时 NPM 依赖或提供库 API |
+
+`@deepseek-ai/create-sdk` 是仓库 `@deepseek-ai/dsh-*` 命名规则的唯一例外；npm scoped initializer 约定要求 `npm create @deepseek-ai/sdk` 对应这个 package 名。该例外是仓库架构事实，不增加第三个开发者产品入口。
+
+三个 package 只导出相邻层实际使用的最小入口，不提供 `src/*` 深路径。scripts 的库入口与构建配置子路径服务生成代码和项目构建配置，但开发者产品合同仍由 `dsh-sdk` 命令承担。
+
+## 工程聚合与编辑会话
+
+`SdkProject.create(root, request)` 构造尚未写盘的新工程快照，`SdkProject.open(root)` 加载已有工程。open 只要求根 `package.json` 与 `cordis.yml` 可读，其余文件是按需存在的资源；两条路径返回同一种只读聚合，并通过显式 origin 区分来源。
+
+`project.edit()` 克隆项目文档形成 working copy。install、configure、enable、disable 和 addPlugin 等领域命令只修改 working copy；命令完成后立即重新检查所属功能，最终 commit 再检查全部关系和文件。
+
+```text
+validate feature requirements and resource ownership
+  -> validate every affected document
+  -> compute changed and removed paths
+  -> compare existing files with the session's original text
+  -> write through one commit boundary
+  -> return a new SdkProject snapshot and ChangeSet
+```
+
+校验失败或检测到会话外修改时不写盘。“一次 commit”只表示写入前零副作用和单一写入口。`ChangeSet` 只描述功能、插件和文件的最终变化，用于 Review & Apply 与 create 收尾。
+
+## 功能与资源所有权
+
+功能是一等行为对象。浅层基类实现 install、configure、enable、disable、required/requires 校验和共同状态识别；固定功能选项、互斥功能选项与可多选功能选项共享这些生命周期。只有资源贡献依赖项目上下文或需要自定义 round-trip 的功能才使用专用行为类，其余功能通过标准化数据声明真正不同的部分。
+
+每项功能贡献带稳定 key 的 Cordis 配置项、NPM 依赖、环境变量占位和独占文件。注册表初始化时拒绝不同功能声明同一个资源 key；同一功能的不同功能选项可以共享资源，并由该功能根据最终选项集合处理。
+
+Cordis 配置项是功能安装锚点。NPM 包名判断配置项所属的功能，配置项 ID 区分同一插件包的多个实例；只有 NPM 依赖而没有功能拥有的 Cordis 配置项时，该功能仍视为未安装。Cordis 配置项存在后，缺失 NPM 依赖、无法读取的 Cordis 插件配置或资源冲突会使功能进入不一致状态，config 命令显示诊断并拒绝猜测式修改。
+
+同一功能选项只更新其声明拥有的配置键，保留未知键。替换功能选项会删除旧功能选项独占且仍可确认的资源；无法确认旧资源或发现独占文件被用户修改时，整个操作失败。
+
+## 问题与 workflow
+
+问题由 TypeScript `Question<T>` 对象表达，默认值、校验、适用条件和类型留在同一个对象中。`PromptPort` 是领域层与终端库之间的唯一接口，helper 提供一份薄 `ClackPromptPort`；create 和 config 注入各自的命令行输入输出流，并在各自流程中决定取消、返回和收尾语义。
+
+create 的有状态问题顺序留在一个向导中，config 的最终状态选择留在一个流程中。两者通过同一个功能配置器收集功能选项与专用输入，因此增加一项普通功能、功能选项或参数不要求同时修改两个入口。
+
+## 项目文档与模板
+
+只有需要读取或修改的结构化文件拥有具体文档对象，包括 `package.json`、`cordis.yml`、`.env`、`.env.example`、根 `tsconfig.json` 和 pnpm workspace 文件。文档对象拥有解析、克隆、校验和序列化行为；具体类与模块分别使用 `*File` 和 `*-file.ts` 命名，业务层不直接操作 YAML/JSON AST，异常形状在所属文档边界 fail loud。
+
+README、入口代码、构建配置、`.gitignore` 和其他一次性文本产物使用与真实文件一一对应的完整模板。CLI usage、创建结果与恢复提示、安装与重试指导以及默认 persona 等完整产品文案也由所属 package 的本地模板提供。
+
+helper 提供通用的数据类型化 `TextTemplate` 模板渲染器，调用 package 通过本地 asset URL 加载自己的模板。
+
+模板使用 Handlebars strict mode 与 `noEscape`，不进行自定义处理。文件对象负责把类型化数据值编码成目标语言文本；如果不希望插值，则源码以 `\{{model}}` 等转义形式输出下游。
+
+## 命令与运行边界
+
+scripts 支持 `dsh-sdk start/dev/build/config`。start 动态加载模块 target 并调用其命名入口；dev 在同一路径前增加 TypeScript 与本地 workspace 源码解析；build 调用工程安装的 tsdown；config 打开一个编辑会话并在 Review & Apply 后提交。typecheck 由生成工程直接执行 `tsc -b`。
+
+HMR 作为显式 Cordis 配置项由 dev 和 start 加载；它所需的 `node-addon-require-builtin` 由 scripts package 传递提供，不写入开发者工程的 `package.json`。
+
+dev/start 会执行开发者入口，在开发者代码中处理命令行参数、cwd，由开发者自行传入 `--model=<name>` 与 `--resume=<session-id>` 启动标准流程。
+
+## 仓库本地链接模式
+
+create-sdk 保留隐藏的 `--link-workspace` 选项供 Harness 仓库开发和 e2e 使用。该选项可以被解析，但不出现在 help、公开 flag 清单或普通用户文档中，也不接收仓库路径参数；仓库根从正在执行的 create-sdk 模块位置向上确定。
+
+链接模式保持普通工程的文件形状。`@deepseek-ai/*` 指向 `packages/`，Cordis 相关 NPM 依赖指向 `vendor/`，共享底层 package 锚定到仓库实际使用的同一物理拷贝，避免 Cordis 类型合并产生多个模块类型定义。npm 使用 `file:`，pnpm 使用 `link:` 并关闭自动 peer 安装，Yarn 使用 `portal:` 与 resolutions；仓库 package 需要先构建。
+
+## 后续工作
+
+- **可替换的 required 主干角色。** 当前 `spine` 以一个固定功能选项拥有整组实现，包含 SystemPrompt、LLMService 等。无法让开发者对其进行替换和切换，只能手工修改 Cordis 配置项。
+- **Service contract 与 package 声明。** 替换特定内建服务时，Cordis 插件目前无法通过 `provides` 元数据声明其提供的服务，因此 SDK 无法在开发阶段辅助配置，也无法在运行时检查兼容性。后续需要设计相应协议。
+- **功能参数描述。** 当前功能的专用输入必须手工声明；SDK 无法从任意 Cordis 插件配置或 NPM package.json 信息中自动推导可交互参数。后续可以定义有限的声明式参数元数据，但不把任意 Cordis 插件配置转换成通用表单。
+- **SDK 应用级配置。** 当前项目资源模型只描述 Cordis 配置项及单个 Cordis 插件拥有的配置键，因此所有受 SDK 管理的配置都必须归属某个插件。跨插件或面向整个 SDK 应用的设置没有独立持久化位置；后续需要定义应用级配置文档及其所有权、读取和修改边界。
+
+## 曾考虑的替代方案
+
+**保留静态 Catalog 与中心 engine。** 该方案改动最小，但功能参数、round-trip、独占文件和 create/config 复用都会继续进入同一个协调中心；拆文件只能缩短单文件，不能收拢职责。
+
+**使用 `wizard.json` 与通用 Questionnaire。** 静态表单无法直接表达功能依赖、选项切换、已有值回填和项目资源变化；类型、gate 和动态 option 最终仍要通过字符串 registry 与过程式 `run()` 连接，形成新的内部 DSL。
+
+**公开本地链接 flag。** 该模式依赖 Harness monorepo 布局和未发布 package，只服务仓库开发；公开后会形成无法对外兑现的项目创建合同，因此保持隐藏。
+
+## 验收标准
+
+- create 与 config 只通过 `SdkProject` 和 `ProjectEditSession` 修改工程，写入前的任何业务、文件或并发校验失败都不产生磁盘变化
+- 新增普通功能、功能选项或参数只扩展类型化 spec 或所属行为对象，create/config 流程不增加中央 switch
+- 功能模型、NPM 依赖与其他资源配置、不一致检测由 helper 统一实现
+- 结构化文件通过 `*File` 文档对象修改；一次性文件和完整产品文案通过所属 package 的 Handlebars 模板生成，业务决策不进入模板 DSL
+- `dsh-sdk start/dev/build/config` 是运行产品面，typecheck 直接使用 `tsc -b`，HMR 不通过命令隐式注入，`node-addon-require-builtin` 只由 scripts package 传递提供
+- `--link-workspace` 只作为隐藏的仓库开发选项存在，并对 npm、pnpm 和 Yarn 保持单一模块身份
+
+## 风险
+
+- 行为对象与类型化 spec 并存会形成两种扩展形状；专用类必须只用于确实依赖项目上下文或自定义的功能，否则会重新产生无意义的类型层次
+- 乐观并发检查与写前校验不能解决写入中途的 I/O 故障，调用方仍需向开发者报告可能的部分提交
+- 隐藏链接模式依赖仓库目录与 package manager 链接语义，仓库布局或工具行为变化时必须与实现一起更新
+- Cordis loader 从自身模块路径加载 `node-addon-require-builtin`；npm、pnpm 或 Yarn 的 NPM 依赖布局变化时，scripts package 必须继续满足该可选对等依赖（optional peer dependency）
+- Handlebars 的 `noEscape` 把目标语言编码责任交给 typed model 构造方；新增模板字段时必须在 owner 处完成正确转义，下游 Handlebars 占位符必须在模板源码中显式转义
diff --git a/docs/rfc/proposed/feature/2026-07-06-recallable-compaction.md b/docs/rfc/proposed/feature/2026-07-06-recallable-compaction.md
new file mode 100644
index 0000000000..6e28e8a31e
--- /dev/null
+++ b/docs/rfc/proposed/feature/2026-07-06-recallable-compaction.md
@@ -0,0 +1,108 @@
+# RFC: Recallable compaction — index checkpoints, a state checkpoint, and in-session history recall
+
+Status: proposed
+
+## Problem
+
+Compaction is a one-way door. The summary the model sees carries no reference to what it shadows — the `shadowedRange` provenance lives only on the log-only `compact/summary` event — and no tool lets the model read a shadowed span back. Whatever the summarizer drops is gone from the model's reachable world, even though the append-only log holds every byte. Repeated compaction compounds this: the head checkpoint is rewritten every pass, so the request prefix takes a full prompt-cache miss each time, and earlier summaries are re-summarized generation after generation.
+
+The root cause is one artifact playing two conflicting roles. An **index** wants to be frozen, chronological, and cheap; the model's **working memory** wants a global view, re-prioritization, and mutability. A single summary can be neither well.
+
+No mainstream coding harness gives the model in-loop recall, and none of the surveyed implementations makes compaction prefix-cache-aware. An event-sourced session — originals durable, seq-addressable, replay-exact — is the natural substrate for both.
+
+## Proposal
+
+Split the checkpoint into two classes and make shadowed history reachable.
+
+### Frozen index checkpoints
+
+Newly stale history splits into chunks by deterministic policy: accumulate toward `chunkTokens`, snap edges with `toolPairingBalancedBefore` / `toolPairingBalancedAfter`, prefer turn boundaries, and place the final boundary as close to the retain boundary as balance allows, so the trailing slice shrinks to roughly one turn. Each chunk is compacted by one `compactRegion` call into an **index stub** (`stubTokens`, ~100–200 tokens):
+
+- two or three lines of what happened;
+- a keyword line of low-frequency literal anchors — exact error strings, values, config keys — grouped by kind;
+- a code-composed footer: `[checkpoint c<summarySeq>: shadows conversation span #<start>–#<end>; originals retrievable via history_read]`. Pointers are assembled from provenance, never model-authored.
+
+A committed stub is never rewritten and never re-enters a later compaction region. A stub call's input is layered: the fixed preamble and the byte-identical pass-start state checkpoint (the shared prefix across all calls in the phase), then the keyword lines of all previously committed stubs — so a new entry indexes what is distinctive to its chunk instead of repeating the directory — the one or two most recent committed stubs for chronological continuity, and the slice itself. Sibling stubs from the same pass are not inputs (the concurrent phase forbids it; turn-aligned boundaries carry local continuity instead), and the state checkpoint is background only, never material to summarize into the stub. A slice consisting of recalled content is stubbed by code alone — a pointer line, no LLM call. A failed stub call degrades the same way: its slice gets a code-only pointer stub and the pass continues, making the state rewrite the only hard LLM dependency in a pass.
+
+### The state checkpoint
+
+One mutable working-memory document (at most one; zero before the first pass), positioned after all stubs and before the retained tail. Each pass rewrites it from the previous state plus this pass's staled content — O(previous + new), under the merge-don't-restate rule already in the summarization prompt — covering decisions, current state, constraints, and next steps. It carries its own footer and a size cap at the scale of today's summary.
+
+An inflation guard bounds the whole pass: if the post-compaction size is not strictly below the pre-compaction size, nothing commits and the turn proceeds; the attempt defers until more stale history accumulates. The guard compares one metric on both sides — provider-reported usage from the request path, falling back to the character estimator on both sides.
+
+### Pass execution
+
+- Chunk slices are surface position ranges. A pass runs two phases: all summarize calls execute concurrently, buffered off-surface; then regions commit strictly left to right — chunks first, trailing slice last — so the state checkpoint lands after every stub through contiguous single-node replaces. Wall-clock stays near one summarize call.
+- The superseded state checkpoint folds into the next pass's first chunk as ordinary history: no tombstone, no new primitive. Its stub omits it, `history_read` renders it labeled `[prior state checkpoint]`, and its footer travels with the rendered text, keeping every trailing slice reachable through the two-hop chain.
+- Range selection is frozen-aware: the compactable span begins after the last committed index checkpoint, at the surface head only when none exists. A legacy session's existing head checkpoint is adopted as state-class — its text the merge base, its node folded like any superseded state.
+- A crash in the summarize phase commits nothing; a crash mid-commit leaves a left-to-right prefix committed, and the resumed pass reads its merge base from the log's latest state-class `compact/summary` event and commits the remaining regions unconditionally — restoring `[stubs…][state][tail]` outranks shrinking.
+
+### The recall tools
+
+A new package `@deepseek-ai/dsh-tool-recall` (consumer-only, over the `dsh-session` and `dsh-compact` vocabularies) registers two model-facing tools:
+
+- `history_read(checkpoint, offset?)` — renders the shadowed span of any checkpoint in the log, including superseded ones, as `User:`/`Assistant:`/`Tool result:` transcript, paginated by a configured budget with a continuation cursor.
+- `history_search(query, checkpoint?, limit?)` — case-insensitive literal scan over every shadowed span; returns snippets with checkpoint ids and coverage metadata (`scanned`/`matched`/`truncated`). The zero-match hint notes the scan is literal and points at direct `history_read` of a plausible checkpoint.
+
+Both read `exec.agent.session.events` (the tool-todo access pattern; non-agent callers rejected), render only surface-type message events, and return ordinary `tool/result`s — recalled bytes land at the context tail, logged, so reconstructability holds with no special casing. There is no new storage and no sidecar index: the session log is the archive, `compact/summary` provenance is the index metadata, and the tools are a read path over both. The tool schemas and the package's one system-prompt section are static strings; checkpoint ids reach the model only through footers. The transcript renderer moves from `compact-basic` into `dsh-session`, shared by summarizer and tools.
+
+### Cache and cost
+
+The request prefix after a pass is `[system][stubs…][state][tail]`. Frozen stubs are byte-stable across passes, so the miss begins at the token replacing the previous state checkpoint and stays O(new chunks + state + tail) — against position zero today. Recall output lands at the tail, leaving the prefix untouched. Per-pass summarize input is roughly twice today's plus an m·S background term, bounded by a `chunkTokens` floor (a small multiple of the state cap) and a validated `stubTokens`/`chunkTokens` ratio ceiling; a shared-prefix input layout (preamble, then the byte-identical pass-start state, slice content in the tail) lets sibling calls earn cached-rate rereads.
+
+### Packaging
+
+The design ships as a new backend `dsh-compact-recallable` on the existing `ctx.compact` seam, enabled by default in the shipped example configs; `compact-basic` remains as the reference implementation and the seam's design twin, in the pattern of the paired LLM adapters. The seam JSDoc's "at most one auto-generated checkpoint, always at the head" clause is relaxed to name both backend behaviors.
+
+### Relation to in-flight work
+
+- **Tool-result pruning** (the in-flight pruning service): its replacement nodes carry `sourceEventSeqs`; the same registry fold lists pruned results as recallable. Follow-up scope; neither blocks the other.
+- **Provider-usage token accounting** (the in-flight move of compaction pressure onto provider-reported usage): supplies the guard's accounting; the implementation stacks after it.
+- **"Query sessions" backlog item**: the cross-session generalization; this RFC scopes to the live session with tool names and rendering chosen so that work extends rather than collides.
+- **Training**: when to recall is a learned behavior. The deterministic footers and keyword anchors give training a stable target, and recall usage is fully visible in the session log for trajectory export; benchmark and RL design proceed with the post-training side.
+
+### Follow-ups
+
+Specified during review, deferred until observation calls for them:
+
+- Guard degradation ladder (code-only rollup of the oldest stub prefix, footers preserved, rolled-up ids remain recall targets; then one summary after the frozen boundary) — on observed guard livelock or stub-region pressure.
+- Echo detection on stub outputs (sentence-scale n-grams, short literals exempt, retry then strip) — on observed division-of-labor leakage.
+- Periodic state refresh from chunk originals — on observed drift in the handoff probe.
+- `stateFallbackThreshold` (full-detail state prompt below a stub count) — on short-session regression.
+- Lazy registration of the recall tools — on measured context tax in never-compacting sessions.
+- Amortized stub drafting at pre-step: as soon as stale-but-uncompacted content accumulates past `chunkTokens`, draft that chunk's stub at the next pre-step (a log-only draft event, written while the chunk's surrounding context is still live) and let the compaction pass commit drafts instead of summarizing in bulk — the deterministic, replay-exact equivalent of background compaction (the Claude Code session-memory pattern; OpenClaw demonstrates the synchronous semantics are identical). Trigger: observed pass latency, or stub-quality gains from drafting near-live proving out.
+- Split summarizer models; model-chosen chunk boundaries; cross-session recall; semantic search fallback — each behind its own evidence.
+- Richer `history_search` query forms — regex, and structured queries over logged JSON tool results (sql/jq-style, or agent-authored queries against an indexed store) — on demand from observed search misses; literal matching ships first because the recall path stays a pure function of the log.
+
+## Alternatives considered
+
+- **Staged delivery** (ship recall tools alone over today's backend; gate the checkpoint split on observed recall usage) — rejected: untrained models under-use any new tool, so the gate would measure training absence rather than design value, while the training side needs the complete mechanism to build environments against; the pre-release window is when persisted-format changes are cheapest; and the cache economics are first-party knowledge, not a hypothesis awaiting telemetry. The implementation still lands as stacked PRs with the recall tools first — construction order, not a decision gate.
+- **All-frozen full-size summaries, no state checkpoint** — rejected: unbounded permanent-prefix growth, self-accelerating toward thrashing, with nothing left to re-prioritize.
+- **Pure stubs, no state checkpoint** — rejected: presumes the model knows what it is missing; fails on unknown unknowns.
+- **LLM aging/consolidation of frozen chunks** — rejected as a routine mechanism: summary-of-summary loss and frozen-prefix churn; the code-only rollup is its surviving form, deferred.
+- **Full prefix as chunk-summarizer input** — rejected: O(N²); the state document gives the same background at O(state).
+- **One summarize call emitting all outputs** — rejected: the summarize path has no structured-output enforcement; parsing one free-text response apart is the fragile seam the fail-closed design avoids.
+- **Model-chosen chunk boundaries** — deferred: parse-and-validate cost against unproven value; chunk policy sits behind config.
+- **Model-authored pointers** — rejected: pointers must be exact; deterministic assembly is.
+- **FTS/vector index sidecar** — rejected in-session: the live log is in memory and bounded, a literal scan under budget suffices; an index earns its keep at cross-session scope.
+- **Semantic search fallback / secondary-model extraction in the recall path** — rejected: an LLM or embedding call there breaks keyless replay determinism; recall stays a pure function of the log.
+- **Raw events instead of rendered transcript** — rejected: leaks log-only vocabulary and chunk noise; the model reads what a model once saw.
+- **Doing nothing (resume/fork as recovery)** — rejected: it makes recovery a human act.
+
+## Acceptance criteria
+
+- Auto-compaction over a long session yields `[stubs…][state][tail]` after every completed pass; prior stubs stay byte-identical across passes; committed stubs never fall inside a later region; the superseded state checkpoint folds without a tombstone, renders labeled, and stays reachable and searchable through the two-hop chain.
+- Every checkpoint's surface text ends with the deterministic footer; footers round-trip through replay byte-identically; the state checkpoint's provenance records its wider input range.
+- Nothing commits before all summaries exist and the guard passes on like-for-like accounting; a guard failure commits nothing and does not fail the turn; a mid-commit kill resumed at the next pre-step completes the pass with the state region committed unconditionally, merge base read from the log; a legacy head checkpoint is adopted as state-class.
+- `history_read` renders any logged checkpoint's span under budget with a working cursor; `history_search` covers every shadowed span with checkpoint-id snippets and coverage metadata, asserted in particular by finding content that exists only in a span shadowed by a superseded state checkpoint — the regression pin for trailing-slice reachability; both reject non-agent callers and never-existing ids or orphaned `compact/start` with typed errors; recalled content appears as ordinary `tool/result`s; request-reconstruction invariants pass over sessions with compaction plus recall; one keyless snapshot scenario covers compact-then-recall end to end; tool schemas and the prompt section are byte-identical across passes.
+- On the long-horizon bench suite: task success does not regress against `compact-basic` at equal budgets; a handoff-fidelity probe (restate K known decisions and constraints after a pass) scores no worse; recall usage frequency and hit usefulness are reported per run via the dsh bench report pipeline, alongside the stub-directory attention measurement and cache-hit telemetry.
+- Seam JSDoc, the compaction capability-seam RFC, `architecture.md`, and the generated tool, config, persistence, and module-graph catalogs update in the same change; all budgets live in config; new source directories hold per-file 100% coverage with HMR disposal tests.
+
+## Risks
+
+- **Recall is a learned behavior**: untrained models will under-use it, and the bench report exists to track the gap while training closes it. Until then the state checkpoint keeps the floor at today's summary quality.
+- **Unknown unknowns remain**: a detail absent from summaries and keywords draws no recall. Recall converts "unreachable even when suspected" into "reachable when suspected".
+- **The stub directory occupies attention**: dozens of stable index cards per request may dilute focus; the bench measurement in the acceptance criteria tracks it against `compact-basic`.
+- **Cost**: per-pass summarize input is roughly twice today's; short sessions sit near today's cost and quality, and the design pays off with session length.
+- **State drift and division-of-labor leakage** are observable through the handoff probe and stub review; their counters are specified follow-ups.
+- **Two backends** are a maintenance surface; the seam contract and the shared recall consumer bound it, and the bench comparison decides the default over time.
diff --git a/docs/rfc/proposed/feature/2026-07-10-sqlite-session-query-provider.md b/docs/rfc/proposed/feature/2026-07-10-sqlite-session-query-provider.md
index acfdf23bee..36216f74d8 100644
--- a/docs/rfc/proposed/feature/2026-07-10-sqlite-session-query-provider.md
+++ b/docs/rfc/proposed/feature/2026-07-10-sqlite-session-query-provider.md
@@ -4,7 +4,7 @@ Status: proposed
 
 ## Problem
 
-The exact-read `ctx.sessionQuery` service deliberately has no derived index. Large persisted histories need full-text search without scanning every event on every query, while current live sessions need an overlay newer than the last durability checkpoint. Search also needs concrete ranking, snippets, filters, pagination, cancellation, and rebuild behavior.
+The exact-read `ctx.sessionQuery` service deliberately has no derived index. Large persisted histories need full-text search without scanning every event on every query, while current live sessions need an overlay newer than the last durability checkpoint. Search also needs concrete ranking, snippets, pagination, cancellation, and rebuild behavior.
 
 Splitting those concerns across a speculative provider coordinator and a database implementation would create two coupled reconciliation state machines. The first real implementation should own the source observation, extraction, SQLite transaction, generation, and query as one lifecycle.
 
@@ -20,7 +20,7 @@ Persisted documents survive restarts. Live overrides are connection-local and sh
 
 The implementation must define both cross-session and within-session scopes from executable use cases. Each searchable event is one document with session metadata, event metadata, surface classification, normalized semantic text, and a bounded plain-text snippet. Session results group by their strongest matching event; numeric backend scores remain private.
 
-Filters compile to parameterized SQL before ranking. Query syntax is treated as data. Ordering includes stable tie fields. Opaque cursors bind to normalized request shape and the smallest relevant generation; unrelated session changes should not invalidate a within-session cursor. Cancellation must stop caller waiting and interrupt SQLite work where the runtime permits.
+Search returns content-bearing result records rather than metadata-only headers. Chainable filters operate on that exact result shape and are designed and implemented with the search API instead of becoming a provider-specific pre-ranking contract. Query syntax is treated as data. Ordering includes stable tie fields. Opaque cursors bind to normalized request shape and the smallest relevant generation; unrelated session changes should not invalidate a within-session cursor. Cancellation must stop caller waiting and interrupt SQLite work where the runtime permits.
 
 Tokenizer choice remains an implementation experiment. FTS5 trigram supports substring recall but rejects useful terms shorter than three characters and increases index size; the proposal must benchmark that tradeoff against the default Unicode tokenizer before making it contract.
 
@@ -41,7 +41,7 @@ Reconciliation may use stable fingerprints to avoid rewriting unchanged persiste
 
 - Restart tests cover unchanged, new, changed, and deleted persisted sessions without rebuilding the whole index.
 - Reopening preserves persisted rows and removes live rows; live rows shadow and then reveal their persisted base.
-- Tests cover both search scopes, metadata filters, surface defaults, snippets, escaping, deterministic ties, pagination, scoped stale cursors, cancellation, dynamic persistence mount/unmount, and recovery after a failed transaction.
+- Tests cover both search scopes, content-bearing results, chainable result filters, surface defaults, snippets, escaping, deterministic ties, pagination, scoped stale cursors, cancellation, dynamic persistence mount/unmount, and recovery after a failed transaction.
 - A schema mismatch resets only the derived database.
 - A keyless end-to-end test combines a real persistence backend with the real SQLite search package.
 - The RFC is amended to the measured tokenizer and public API actually implemented before moving to `implemented/`.
diff --git a/docs/rfc/proposed/feature/2026-07-14-sdk-developer-projects.i18n.yaml b/docs/rfc/proposed/feature/2026-07-14-sdk-developer-projects.i18n.yaml
new file mode 100644
index 0000000000..36bb26161f
--- /dev/null
+++ b/docs/rfc/proposed/feature/2026-07-14-sdk-developer-projects.i18n.yaml
@@ -0,0 +1,6 @@
+# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
+# side as of the last confirmed-consistent state. Both languages carry equal authority;
+# after editing either side, bring the other along and re-record with:
+#   pnpm run verify-translation-pairing --write
+2026-07-14-sdk-developer-projects.md: 0b5fe876f92153e1ccf5bd8fe383464f5087f4b1
+2026-07-14-sdk-developer-projects.zh.md: ec08f323acba1b9dc049182937fda34bfb50d4ee
diff --git a/docs/rfc/proposed/feature/2026-07-14-sdk-developer-projects.md b/docs/rfc/proposed/feature/2026-07-14-sdk-developer-projects.md
new file mode 100644
index 0000000000..0b5fe876f9
--- /dev/null
+++ b/docs/rfc/proposed/feature/2026-07-14-sdk-developer-projects.md
@@ -0,0 +1,167 @@
+# RFC: Developer-owned SDK projects
+
+Status: proposed
+
+English | [中文](2026-07-14-sdk-developer-projects.zh.md)
+
+## Problem
+
+DeepSeek Harness composes features through Cordis plugins, but building a runnable project from an empty directory still requires a developer to understand npm dependencies, the `cordis.yml` plugin set, environment variables, TypeScript builds, local-plugin workspaces, and runtime entrypoints together. These manual steps constrain one another: omitting any one can produce a project that installs but cannot be developed, develops but cannot be built, or builds but cannot start.
+
+A one-shot generator reduces only the initial creation cost. If the generated result is hidden inside a preset or an uneditable CLI, advanced developers cannot reshape the plugin tree, change Cordis plugin config, or add project-specific behavior. If a generated project immediately leaves tool management altogether, developers must again maintain consistency across all npm dependencies and Cordis plugin config themselves.
+
+Initial creation and later configuration address the same builtin feature set. When those workflows maintain separate feature lists, feature options, and npm dependencies, new Cordis plugins, npm packages, and Cordis plugin config changes make them diverge. Projects also need an ordinary local-plugin development path that participates in development, build, and start flows.
+
+## Proposal
+
+The SDK creates an ordinary, explicit TypeScript/Cordis project owned by its developer. `cordis.yml` is the only runtime plugin tree; development and production read the same file. The generated `package.json`, `cordis.yml`, TypeScript entrypoint, build configuration, and `plugins/*` remain directly editable instead of being hidden behind a preset.
+
+The only developer product entrypoints are `npm create @deepseek-ai/sdk` and the `dsh-sdk` commands. The initializer performs initial creation, `dsh-sdk config` manages SDK-recognized builtin features afterward, and `dsh-sdk dev`, `dsh-sdk build`, and `dsh-sdk start` own development, build, and startup; this phase provides no `dsh-sdk create`. Create and config consume one manually authored feature definition, so each feature has one source for its feature options, npm dependencies, Cordis config entries, related files, and inspection rules. The [SDK project editing architecture](../architecture/2026-07-15-sdk-project-editing-architecture.md) defines terms such as feature and feature option.
+
+The SDK offers interaction for feature selection and finite feature options only; it does not turn arbitrary Cordis plugin config into a generic form. A feature collects the small number of dedicated inputs required by its feature options. All other Cordis plugin config remains in `cordis.yml`, with comments documenting common edits, for direct developer control.
+
+## Developer workflow
+
+Initial creation collects information in an order where earlier answers determine later questions: target directory and package identity, model provider and credentials, run interface, builtin features and feature options, an optional local plugin, package manager, and whether to install npm dependencies and build. Command-line arguments suppress questions they already answer. Create and config require an interactive TTY in this phase, and cancelling creation writes nothing to the target directory.
+
+```sh
+npm create @deepseek-ai/sdk my-agent
+cd my-agent
+npm exec dsh-sdk dev index.ts
+npm exec dsh-sdk config
+npm exec dsh-sdk build
+npm exec dsh-sdk start index.js
+```
+
+Create rejects every target path that already exists. After committing the project files, the CLI asks whether to install npm dependencies and build. An install or build failure preserves the generated project and prints commands that can retry the failed work.
+
+Create also offers one `none / plugin / tool` choice. `plugin` creates a fixed `plugins/plugin` Cordis plugin, while `tool` creates a fixed `plugins/tool` model-facing tool; one project creation includes at most one local plugin. The operation updates the workspace, root npm dependency, TypeScript reference, build configuration, and `cordis.yml` together, and any pre-write validation failure leaves the project absent.
+
+## Features supported during creation
+
+The table is the developer-visible support set for this phase. A `required` feature is always present but may still offer finite feature options; a `default` feature is preselected in the feature tree; an `optional` feature is selected explicitly. The table describes the product support set, while the runtime registry remains the implementation source of truth.
+
+| Feature | Create state | Feature options | Constraints and relationships |
+|---|---|---|---|
+| `provider` | required | `deepseek` (default) / `custom` | DeepSeek collects an API key; custom also collects a base URL, and a CLI option may override the model name |
+| `app` | required | `stdio` (default) / `acp` / `embed` | Selects the run interface |
+| `spine` | required | `default` | Timer, the LLM seam, session storage, system prompt, the tool registry, the agent registry, and the agent loop |
+| `bash` | required | `local` (default) / `sandbox` | The two feature options are exclusive and independent of the run interface, and both install the model-facing bash tool; sandbox installs the local sandbox provider and sandboxed bash backend |
+| `persistence` | required | `jsonl` (default) / `sqlite` | Every project selects exactly one persistence backend |
+| `hmr` | default | `default` | Loads `@cordisjs/plugin-hmr`; dev and start both enable it with the plugin defaults |
+| `fs` | default | `local` | Installs the local filesystem, policy, and model-facing tools; the process sandbox does not confine in-process fs tools |
+| `todo` | default | `default` | Provides the `todo_write` tool |
+| `skill` | default | `default` | Installs the skill registry, the local skill provider, and the model-facing skill tool |
+| `web` | optional | `deepseek` (default) / `exa` / `perplexity` / `fetch-only` | Search feature options are exclusive; Exa and Perplexity collect their API keys; timeout policy is recommended |
+| `subagent` | optional | `spawn` (default) / `fork`, multiple | This phase provides only in-process backends |
+| `workflow` | optional | `workerthread` | Requires the subagent `spawn` feature option |
+| `compact` | optional | `basic` | Uses SDK-provided context-compaction parameters |
+| `hooks` | optional | `claude` (default) / `codex`, multiple | Each feature option creates a separate editable configuration file |
+| `guard` | optional | `repeat-tool` | Provides repeated-tool-call reminders |
+| `timeout-policy` | optional | `default` | Applies a uniform policy to tools that declare timeout budgets |
+| `ask-user` | optional | `default` | Provides the `ask_user_question` tool; only `acp` and `stdio` can select it because those two feature options provide the injected user-interaction service |
+
+Both `bash` feature options apply to ACP, stdio, and embed and are not selected by the run interface. The sandbox feature option writes no active config key and therefore keeps `dsh-bash-sandbox`'s `read-only` default. Generated `cordis.yml` includes a commented example that developers can change explicitly to `workspace-write`:
+
+```yaml
+- id: bash
+  name: '@deepseek-ai/dsh-bash-sandbox'
+  # Uncomment to allow writes under the project workspace.
+  # config:
+  #   mode: workspace-write
+  #   workspaceRoot: !!js process.cwd()
+```
+
+Feature contributions reference only single-plugin npm packages and never bundle packages such as `agent-spine-demo`, `stdio-demo`, or `acp-demo`. Plugins outside the table are not managed by create in this phase; advanced developers may still compose them by editing the ordinary project files directly.
+
+## Generated project
+
+With default answers, an npm project uses the DeepSeek provider, the stdio interface, local bash, JSONL persistence, and the preselected hmr, fs, todo, and skill features. Its initial tree is:
+
+```text
+my-agent/
+├── .env
+├── .env.example
+├── .gitignore
+├── README.md
+├── cordis.yml
+├── index.ts
+├── package.json
+├── tsconfig.base.json
+├── tsconfig.json
+└── tsdown.config.ts
+```
+
+`.env.example` always exists, and the SDK keeps its placeholders aligned with the current feature set. A gitignored `.env` is also created when a secret is captured or the developer confirms an empty credential to fill later. The SDK only appends differently named variables that are not already present in `.env` and never updates or removes existing contents. Feature-option changes may remove obsolete `.env.example` placeholders, while old credentials remain in `.env` for the developer to manage. pnpm and Yarn projects add their required workspace files, but do not fork the runtime plugin tree or TypeScript entrypoint.
+
+Generated `package.json` provides the following scripts. `dev`, `build`, `start`, and `config` invoke `dsh-sdk`, while `typecheck` invokes TypeScript directly:
+
+| Script | Behavior |
+|---|---|
+| `dev` | Run `dsh-sdk dev index.ts`, registering development-time resolution for TypeScript and local workspace plugins |
+| `build` | Run `dsh-sdk build`, invoking the project's installed tsdown for the root entrypoint and `plugins/*` packages |
+| `typecheck` | Run `tsc -b` directly |
+| `start` | Run `dsh-sdk start index.js`, starting the built entrypoint without an implicit build |
+| `config` | Run `dsh-sdk config` to edit the current project's feature tree |
+
+`dsh-sdk start` and `dsh-sdk dev` accept a module target and forward arguments after `--` unchanged to the project entrypoint. Generic argument parsing uses Node `parseArgs()` with zero schema: valued flags use `--key=value`, bare flags become `true`, and `--no-*` becomes `false`.
+
+- Stdio projects pass the selected model through `--model=<name>` and create or resume an agent according to optional `--resume=<session-id>`;
+- ACP uses protocol `session/load`
+- Embed uses the model written into the generated code.
+
+Each feature-owned Cordis config entry keeps its developer-editable Cordis plugin config and explanatory comments in `cordis.yml`. When `dsh-sdk config` changes other features, it preserves unknown fields, formatting on untouched nodes, and comments. HMR is an ordinary leaf config entry: when the feature is selected, dev and start load the same watcher, and the command does not change the plugin tree implicitly.
+
+## Post-creation configuration
+
+`dsh-sdk config` requires only readable root `package.json` and `cordis.yml` files in the current directory. It inspects standard features and their current feature options, expresses the final desired state through one feature tree, and shows feature changes and affected files before Review & Apply.
+
+`dsh-sdk config` can install missing features, enable or disable installed features, and switch finite feature options. Required features cannot be removed. An npm dependency change runs the project package manager's install once after the file commit; installation failure does not roll back committed project files.
+
+The SDK modifies only Cordis config entries, config keys, npm dependencies, `.env.example` placeholders, and owned files explicitly owned by a feature. Updating the same feature option preserves unknown config keys in its Cordis config entries. Handwritten and third-party plugins support enable and disable by stable ID only. When a known feature has been edited into an incomplete, ambiguous, or otherwise unreadable shape, `dsh-sdk config` displays diagnostics and refuses automatic changes until the developer repairs it manually.
+
+One config session accumulates every change in an in-memory working copy. Before Apply, it validates feature relationships, resource conflicts, and document shapes, then compares each affected existing file with the text read when the session opened. Validation failure or an external edit causes zero writes. Once physical writes begin, the SDK does not provide cross-file transactional rollback.
+
+## Maintenance model
+
+The SDK curates its builtin support set instead of exposing npm packages automatically by npm dependency name or directory convention. One feature may compose several Cordis config entries, feature options may share resources, and a feature option may declare a feature requirement on another feature or a specific feature option. Adding an ordinary feature or feature option does not require changes to both create and config command workflows.
+
+## Future work
+
+- `dsh-sdk add [package-spec]` unifies local-plugin creation with external Cordis plugin installation: without a package or repository source it creates a local plugin/tool, while a supplied source adds the npm dependency and `cordis.yml` config entry; the source model leaves room for GitHub repositories and other extensions
+- Non-interactive create/config: both workflows require a TTY in this phase and provide no complete input contract for automation
+- More feature-specific inputs: this product surface exposes only finite feature options, secrets, and a few dedicated values in this phase rather than a generic parameter interface for Cordis plugin config
+
+## Alternatives considered
+
+**An opaque preset or generator-owned project.** This shortens initial creation but hides the real plugin tree and build boundaries, prevents advanced developers from composing Cordis plugins directly, and makes project behavior depend on the CLI version rather than committed project files.
+
+**A one-shot generator only.** Leaving all later maintenance manual redistributes feature requirements, feature-option switches, and multi-file updates. A config workflow over the shared registry retains continuing management for generated projects.
+
+**Separate `cordis.yml` files for development and production.** Two plugin trees mean a successful development run does not demonstrate that production loads the same features. Dev adds only TypeScript and local-workspace resolution; runtime configuration remains singular.
+
+**A generic form for arbitrary Cordis plugin config.** Cordis plugin config contains nested structures, expressions, and plugin-specific semantics. A generic form would become a second incomplete schema. The SDK manages finite feature options and dedicated secrets, while developers continue to edit complex config directly.
+
+**A private local-plugin discovery protocol.** Ordinary package-manager workspaces, root npm dependencies, TypeScript references, and Cordis config entries already express the complete relationship. Another discovery protocol would create hidden state understood only by the SDK.
+
+**A `dsh-sdk create` command for existing projects.** Create already provides one editable local-plugin skeleton, and later plugins can use ordinary workspace and Cordis mechanisms manually. A parallel command would add a second scaffolding product surface without adding composition functionality.
+
+**Automatically expose every new Cordis plugin as a builtin.** An npm package cannot say how several plugins compose into one product feature, nor can it derive exclusivity, feature requirements, secrets, interface applicability, or security constraints. The support set requires human curation; automation is suitable only for checking whether candidates have been classified.
+
+## Acceptance criteria
+
+- `npm create @deepseek-ai/sdk` collects project identity, provider, interface, features, an optional local plugin, package manager, and installation choice in the documented order, and cancellation leaves the target path absent
+- A default npm project has the documented tree and `dev`, `build`, `typecheck`, `start`, and `config` scripts, with dev and start sharing one `cordis.yml`
+- Create offers the documented features and feature options; local and sandbox bash are exclusive with local as the default, the sandbox Cordis config entry retains the editable commented config example, and HMR is selected by default and loaded by both dev and start
+- Create's `plugin` or `tool` choice creates at most one fixed-name local plugin and atomically updates its files and root-project relationships; this phase provides no `dsh-sdk create`
+- `dsh-sdk config` reads the same support set from an existing project, installs, enables, disables, and switches supported feature options, preserves unknown config and comments, and refuses to modify inconsistent config
+- `.env.example` reflects variables required by the current features; `.env` only appends missing differently named variables and never updates or removes existing contents
+- npm, pnpm, and Yarn workspaces install, build, and start; local plugins resolve from source under dev and from built output under start
+
+## Risks
+
+- Developers can edit a builtin into a shape the registry cannot recognize; the SDK stops automating that feature instead of guessing and overwriting config
+- Pre-write validation and external-edit detection do not provide transactional rollback once multi-file writes begin; an I/O failure can leave a partial commit requiring manual repair
+- The sandbox feature option depends on an available local sandbox backend for the target platform; an unavailable backend must fail closed instead of falling back to unsandboxed execution
+- HMR retains its filesystem watcher and hot-reload behavior under production start; this is the result of an explicit plugin choice, not an implicit development-only service
+- The append-only `.env` policy retains credentials that are no longer used; the SDK does not decide when user-owned secret data is safe to delete
diff --git a/docs/rfc/proposed/feature/2026-07-14-sdk-developer-projects.zh.md b/docs/rfc/proposed/feature/2026-07-14-sdk-developer-projects.zh.md
new file mode 100644
index 0000000000..ec08f323ac
--- /dev/null
+++ b/docs/rfc/proposed/feature/2026-07-14-sdk-developer-projects.zh.md
@@ -0,0 +1,167 @@
+# RFC: 开发者拥有的 SDK 工程
+
+Status: proposed
+
+[English](2026-07-14-sdk-developer-projects.md) | 中文
+
+## 问题
+
+DeepSeek Harness 通过 Cordis 插件对功能进行组合，但从空目录开始搭建一个可运行工程仍要求开发者同时理解 NPM 依赖、`cordis.yml` 插件组、环境变量、TypeScript 构建、本地插件 workspace 和运行入口。手工步骤之间存在约束，漏掉任意一处都会得到能够安装却无法开发、能够开发却无法构建，或能够构建却无法启动的工程。
+
+一次性生成器只能降低首次创建成本。若生成结果隐藏在 preset 或不可编辑的 CLI（命令行界面）内部，高级开发者无法调整插件树、修改 Cordis 插件配置或增加项目特有行为；若创建后的工程完全脱离工具管理，开发者又必须重新承担所有 NPM 依赖和 Cordis 插件配置的一致性工作。
+
+初始创建和后续配置面对同一组内置功能。两条流程各自维护功能列表、功能选项和 NPM 依赖时，新增 Cordis 插件、NPM 包或调整配置会使二者逐渐分叉。工程还需要一条普通的本地插件开发路径，参与开发、构建和启动流程。
+
+## 提案
+
+SDK 创建一个普通、显式且归开发者所有的 TypeScript/Cordis 工程。`cordis.yml` 是唯一的运行时插件树；开发和生产读取同一份文件。工程中的 `package.json`、`cordis.yml`、TypeScript 入口、构建配置和 `plugins/*` 均可直接编辑，SDK 不把它们封装成不可见的 preset。
+
+开发者产品入口只有 `npm create @deepseek-ai/sdk` 和 `dsh-sdk` 命令。前者负责首次创建，`dsh-sdk config` 在创建后管理 SDK 能识别的内置功能，`dsh-sdk dev`、`dsh-sdk build` 与 `dsh-sdk start` 负责开发、构建和启动；本期不提供 `dsh-sdk create`。create 与 config 使用同一份人工编写的功能定义，因此一项功能的功能选项、NPM 依赖、Cordis 配置项、相关文件和识别规则只有一个来源。功能、功能选项等名词由 [SDK 工程编辑架构](../architecture/2026-07-15-sdk-project-editing-architecture.md) 的术语表定义。
+
+SDK 只为功能选择和有限功能选项提供交互，不尝试把任意 Cordis 插件配置变成通用表单。功能选项所需的少量专用输入由所属功能收集；其余 Cordis 插件配置留在 `cordis.yml` 中，并通过注释指明常用改法，由开发者直接修改。
+
+## 开发者流程
+
+首次创建按会影响后续问题集合的顺序收集信息：目标目录与 package 身份、模型提供方与凭据、运行接口、内置功能与功能选项、可选本地插件、包管理器，以及是否安装 NPM 依赖并构建。命令参数已提供的答案不重复询问；本期 create 和 config 都要求交互式 TTY，取消创建时不写入目标目录。
+
+```sh
+npm create @deepseek-ai/sdk my-agent
+cd my-agent
+npm exec dsh-sdk dev index.ts
+npm exec dsh-sdk config
+npm exec dsh-sdk build
+npm exec dsh-sdk start index.js
+```
+
+create 拒绝任何已经存在的目标路径。工程文件提交成功后，CLI 询问是否安装 NPM 依赖并构建；安装或构建失败时保留生成结果，并打印可以重新执行的命令。
+
+create 还提供一次 `none / plugin / tool` 选择。`plugin` 固定生成 `plugins/plugin` 的 Cordis 插件，`tool` 固定生成 `plugins/tool` 的模型工具；一次创建至多包含一个本地插件。生成操作同时更新 workspace、根 NPM 依赖、TypeScript reference、构建配置和 `cordis.yml`，任何写入前校验失败都不创建工程。
+
+## 创建时支持的功能
+
+下表是本期 create 面向开发者展示的支持集。`required` 始终存在但仍可切换有限功能选项；`default` 在选择树中预选；`optional` 由开发者主动选择。表格说明产品支持集，运行时注册表是实现的事实源。
+
+| 功能 | create 状态 | 功能选项 | 限制与关系 |
+|---|---|---|---|
+| `provider` | required | `deepseek`（默认）/ `custom` | DeepSeek 收集 API key；custom 另收集 base URL，模型名可由 CLI 参数覆盖 |
+| `app` | required | `stdio`（默认）/ `acp` / `embed` | 选择运行接口 |
+| `spine` | required | `default` | timer、LLM seam、会话存储、系统提示词、工具注册表、agent 注册表，以及 agent loop |
+| `bash` | required | `local`（默认）/ `sandbox` | 两个功能选项互斥、与运行接口正交，且都安装面向模型的 bash 工具；sandbox 安装本地沙箱提供方和沙箱 bash 后端 |
+| `persistence` | required | `jsonl`（默认）/ `sqlite` | 每个工程恰好选择一个持久化后端 |
+| `hmr` | default | `default` | 加载 `@cordisjs/plugin-hmr`；dev 和 start 都启用，使用插件默认配置 |
+| `fs` | default | `local` | 安装本地文件系统、策略和模型工具；进程沙箱不约束进程内 fs 工具 |
+| `todo` | default | `default` | 提供 `todo_write` 工具 |
+| `skill` | default | `default` | 安装 skill（技能）注册表、本地 skill 提供方和面向模型的 skill 工具 |
+| `web` | optional | `deepseek`（默认）/ `exa` / `perplexity` / `fetch-only` | 搜索功能选项互斥；Exa/Perplexity 收集各自 API key；建议同时启用 timeout policy |
+| `subagent` | optional | `spawn`（默认）/ `fork`，可多选 | 本期只提供进程内后端 |
+| `workflow` | optional | `workerthread` | 要求 subagent 的 `spawn` 功能选项 |
+| `compact` | optional | `basic` | 使用 SDK 提供的上下文压缩参数 |
+| `hooks` | optional | `claude`（默认）/ `codex`，可多选 | 各功能选项生成独立的可编辑配置文件 |
+| `guard` | optional | `repeat-tool` | 提供重复工具调用提醒 |
+| `timeout-policy` | optional | `default` | 对声明超时预算的工具执行统一策略 |
+| `ask-user` | optional | `default` | 提供 `ask_user_question` 工具；注入的 user-interaction 服务由 acp/stdio 两个功能选项提供，因此仅这两个接口可选 |
+
+`bash` 的两个功能选项都适用于 ACP、stdio 和 embed，不由运行接口决定。sandbox 功能选项不写任何生效的配置键，因而沿用 `dsh-bash-sandbox` 的 `read-only` 默认值；生成的 `cordis.yml` 保留注释示例，开发者可以显式改为 `workspace-write`：
+
+```yaml
+- id: bash
+  name: '@deepseek-ai/dsh-bash-sandbox'
+  # Uncomment to allow writes under the project workspace.
+  # config:
+  #   mode: workspace-write
+  #   workspaceRoot: !!js process.cwd()
+```
+
+功能贡献只引用单插件 NPM 包，绝不引用 `agent-spine-demo`、`stdio-demo`、`acp-demo` 这类组合 NPM 包。表格之外的插件不由本期 create 管理；开发者仍可直接编辑普通工程文件进行高级组合。
+
+## 生成工程
+
+使用默认答案创建 npm 工程时，provider 为 DeepSeek，运行接口为 stdio，bash 为 local，持久化为 JSONL，hmr、fs、todo 与 skill 处于选中状态。初始目录树为：
+
+```text
+my-agent/
+├── .env
+├── .env.example
+├── .gitignore
+├── README.md
+├── cordis.yml
+├── index.ts
+├── package.json
+├── tsconfig.base.json
+├── tsconfig.json
+└── tsdown.config.ts
+```
+
+`.env.example` 始终存在，并由 SDK 根据当前功能维护占位。收集到 secret 或开发者确认稍后填写空凭据时，同时生成 gitignored `.env`。SDK 只向 `.env` 追加尚不存在的不同名变量，绝不覆盖或删除已有内容；切换功能选项可以清理 `.env.example` 中不再需要的占位，但旧凭据仍留在 `.env` 中供开发者自行处理。pnpm 和 Yarn 工程增加各自所需的 workspace 配置文件，但运行时插件树和 TypeScript 入口不分叉。
+
+生成的 `package.json` 提供以下 scripts；其中 `dev`、`build`、`start` 与 `config` 调用 `dsh-sdk`，`typecheck` 直接调用 TypeScript：
+
+| script | 行为 |
+|---|---|
+| `dev` | 运行 `dsh-sdk dev index.ts`，为 TypeScript 和本地 workspace 插件注册开发期解析 |
+| `build` | 运行 `dsh-sdk build`，调用工程安装的 tsdown 构建根入口和 `plugins/*` package |
+| `typecheck` | 直接运行 `tsc -b` |
+| `start` | 运行 `dsh-sdk start index.js`，启动已构建入口且不隐式构建 |
+| `config` | 运行 `dsh-sdk config`，修改当前工程功能树 |
+
+`dsh-sdk start` 与 `dsh-sdk dev` 可以接收模块 target，并把 `--` 后的参数原样转发给工程入口。通用参数解析使用 Node `parseArgs()` 的零 schema 模式：带值 flag 采用 `--key=value`，bare flag 转换为 `true`，`--no-*` 转换为 `false`。
+
+- stdio 工程通过 `--model=<name>` 传入所选 model，并根据可选的 `--resume=<session-id>` 创建或恢复 agent；
+- acp 使用协议 `session/load`
+- embed 使用生成代码中的 model。
+
+每个功能拥有的 Cordis 配置项在 `cordis.yml` 中保留自己的可编辑 Cordis 插件配置和说明注释；`dsh-sdk config` 修改其他功能时必须保留未知字段、未修改节点的格式和注释。HMR（热模块替换）是普通叶子配置项：选择该功能后，dev 和 start 加载同一个 watcher，命令不隐式改变插件树。
+
+## 创建后的配置
+
+`dsh-sdk config` 只要求当前目录具有可读的根 `package.json` 与 `cordis.yml`。它检查标准功能及其当前功能选项，以一棵功能树表达最终目标状态，并在 Review & Apply 前展示功能变化和受影响文件。
+
+`dsh-sdk config` 可以安装缺失功能、启停已安装功能和切换有限功能选项。required 功能不能取消。改变 NPM 依赖后只运行一次项目包管理器安装；安装失败不回滚已经提交的工程文件。
+
+SDK 只修改功能明确拥有的 Cordis 配置项、配置键、NPM 依赖、`.env.example` 占位和独占文件。同一功能选项的更新保留 Cordis 配置项中的未知配置键；手写或第三方插件只支持按稳定 ID 启停。已知功能被手改成不完整、歧义或无法读取的形状时，`dsh-sdk config` 显示诊断并拒绝自动修改，直到开发者手工修复。
+
+一次 config 会话在内存工作区上累计全部修改。Apply 前完成功能关系、资源冲突和文件形状校验，并比较受影响文件与会话打开时的原文；校验失败或检测到外部修改时不写盘。实际写盘开始后不提供跨文件事务回滚。
+
+## 维护模型
+
+Builtin 支持集由 SDK 人工策划，不根据 NPM 依赖名称或目录约定自动暴露。一个功能可以组合多个 Cordis 配置项，功能选项可以共享资源，并声明对其他功能或特定功能选项的功能依赖；新增普通功能或功能选项不应要求同时修改 create 和 config 两个命令流程。
+
+## 后续工作
+
+- `dsh-sdk add [package-spec]`：统一本地插件创建与外部 Cordis 插件接入；未指定 package 或仓库来源时创建本地 plugin/tool，指定来源时增加 NPM 依赖和 `cordis.yml` 配置项，来源模型为 GitHub 仓库等扩展保留空间
+- 非交互 create/config：本期两个流程都要求 TTY，不提供供自动化调用的完整输入合同
+- 更多功能专用参数输入：本期产品只展示有限功能选项、secret 和少量专用值，不为 Cordis 插件配置提供通用参数界面
+
+## 曾考虑的替代方案
+
+**不可编辑的 preset 或生成器托管工程。** 该方案可以缩短初次创建路径，但会隐藏真实插件树和构建边界，使高级开发者无法直接组合 Cordis 插件，也让项目行为依赖 CLI 版本而不是检入的工程文件。
+
+**只提供一次性生成器。** 创建后完全依赖手工维护，会让功能依赖、功能选项切换和多文件更新再次分散；共享 registry 的 config 流程为生成工程保留持续管理机制。
+
+**为开发和生产维护两份 `cordis.yml`。** 两份插件树会使开发成功无法证明生产加载相同功能；dev 只增加 TypeScript 与本地 workspace 解析，运行配置保持唯一。
+
+**为任意 Cordis 插件配置生成通用表单。** Cordis 插件配置包含嵌套结构、表达式和插件特有语义，通用表单会形成第二套不完整 schema。SDK 只管理有限功能选项和专用 secret，复杂配置继续由开发者直接编辑。
+
+**使用私有协议发现本地插件。** 普通 package manager workspace、根 NPM 依赖、TypeScript references 和 Cordis 配置项已能表达完整关系；额外发现协议会创造只能由 SDK 理解的隐藏状态。
+
+**在现有工程中提供 `dsh-sdk create`。** create 已能生成一种可编辑的本地插件骨架，后续插件可以沿用普通 workspace 和 Cordis 机制手工添加；再提供同构命令会增加第二条脚手架产品面，却不增加新的组合功能。
+
+**把每个新 Cordis 插件自动暴露为 builtin。** package 无法说明多个插件如何组合成一项产品功能，也无法推导互斥关系、功能依赖、secret、接口适用性和安全限制；支持集需要人工策划，自动化只适合检查候选是否完成分类。
+
+## 验收标准
+
+- `npm create @deepseek-ai/sdk` 按本文顺序收集项目身份、provider、interface、功能、可选本地插件、包管理器和安装选择，并在取消时保持目标路径不存在
+- 默认 npm 工程具有本文目录树和 `dev`、`build`、`typecheck`、`start`、`config` scripts，且 dev/start 使用同一份 `cordis.yml`
+- create 展示本文功能及功能选项；`bash` 的 local/sandbox 二选一且默认 local，sandbox Cordis 配置项保留可编辑的注释配置示例；HMR 默认选中并同时由 dev/start 加载
+- create 的 `plugin` 或 `tool` 选择至多生成一个固定名称的本地插件，并原子更新插件文件与根工程关系；本期不提供 `dsh-sdk create`
+- `dsh-sdk config` 从现有工程读取同一支持集，能够安装、启停和切换支持的功能选项，保留未知配置与注释，并拒绝修改不一致配置
+- `.env.example` 反映当前功能所需变量；`.env` 只追加缺失的不同名变量，从不覆盖或清理已有内容
+- npm、pnpm 和 Yarn 生成的 workspace 能安装、构建和启动；本地插件在 dev 中使用源码，在 start 中使用构建产物
+
+## 风险
+
+- 开发者可以把 builtin 手改成 registry 无法识别的形状；SDK 选择停止自动化而不是猜测并覆盖配置
+- 多文件写入前的校验和外部修改检测不能提供写入阶段的事务回滚；I/O 中途失败可能留下需要人工修复的部分提交
+- sandbox 功能选项依赖目标平台存在可用的本地沙箱后端；后端不可用时必须 fail closed，不能退回无沙箱执行
+- HMR 在生产启动中也保持文件 watcher 和热重载行为；这是显式插件选择的结果，不是仅限开发环境的隐式服务
+- `.env` 的仅追加策略会保留已经不用的凭据，SDK 不判断这些用户数据何时可以安全删除
diff --git a/docs/rfc/proposed/process/2026-07-13-human-review-skill-maintenance.md b/docs/rfc/proposed/process/2026-07-13-human-review-skill-maintenance.md
new file mode 100644
index 0000000000..8098247fdf
--- /dev/null
+++ b/docs/rfc/proposed/process/2026-07-13-human-review-skill-maintenance.md
@@ -0,0 +1,83 @@
+# RFC: Periodic human-review maintenance for dsh-code-review
+
+Status: proposed
+
+## Problem
+
+The `dsh-code-review` skill records failure modes that require reviewer judgment, but one-off audits are expensive to repeat and easy to scope inconsistently. Treating every comment as a lesson produces checklist bloat; treating merge, thread resolution, or an author's “fixed” reply as proof of adoption promotes feedback that the final code may not implement. The maintenance process needs enough evidence and independent review to fail closed without requiring a webhook service, durable event state, or automatic repository promotion before the workflow has proven useful.
+
+## Proposal
+
+Periodic out-of-repo maintenance. A private tool, kept on the skill maintainer's machine rather than committed to this repository, runs against a clean full-history checkout at refreshed `origin/master`. The intended scheduler runs daily with a two-UTC-day overlap; manual runs accept another `--since` duration or repeated `--pr` arguments for an explicit set. The scan is idempotent against the current skill and stores no repository cursor. The only repository file changed by promotion is [.agents/skills/dsh-code-review/SKILL.md](../../../../.agents/skills/dsh-code-review/SKILL.md); the draft PR carries a provenance summary so reviewers can audit the source feedback and adoption evidence without the private adapter logs.
+
+```mermaid
+flowchart TD
+  A["Maintainer or scheduler runs the tool on origin/master"] --> B["List PRs merged in the overlap window"]
+  B --> C["Collect pre-merge User feedback and final PR evidence"]
+  C --> D["Two reviewers verify provenance and adoption"]
+  D --> E{"Both confirm human-authored and adopted?"}
+  E -- "No" --> F["Exclude or retain as unresolved"]
+  E -- "Yes" --> G["Two reviewers classify against the current skill"]
+  G --> H["Draft a complete candidate from agreed guidance"]
+  H --> I["Two reviewers inspect the same skill diff"]
+  I -- "Blocking finding" --> J["Bounded revision loop"]
+  J --> I
+  I -- "Both approve" --> K["Run documentation and lint checks"]
+  K --> L["Leave a reviewed local working-tree diff"]
+```
+
+### Acquisition contract
+
+Each selected PR is filtered before any feedback is retrieved: its merge commit must be an ancestor of `origin/master`. Merge-commit reachability is the sole eligibility check — a stacked PR whose direct base is a feature branch is admitted whenever the base has since reached master, because the code the reviewer commented on is now on master regardless of the intermediate stack. The tool also resolves the landing merge's target parent; a landing shape it cannot reconstruct is logged to `skipped-pulls.json` and skipped. A single PR that fails preflight, acquisition, or evidence collection is skipped rather than aborting the whole run. The search stage fails loud when the window would exceed GitHub's 1,000-result search cap so no merged PR is silently omitted. The acquisition stage reads complete paginated connections for inline review comments, review submissions, and PR commits. PR conversation comments are not acquired because current GitHub state cannot prove which surviving commit preceded them after a force-push, so the adoption contract would exclude them unconditionally. The workflow admits acquired feedback only when GitHub reports the actor `type` as `User`, and only when both creation and last-edit timestamps strictly predate the PR merge (an equal-timestamp edit is treated as post-merge); review submissions use GraphQL `lastEditedAt` because the REST representation omits edit time.
+
+### Adoption evidence
+
+Each feedback item carries a stable source ID and bounded change evidence. When the reviewer's `commit_id` still belongs to the PR (force-push fail-closed), the tool selects the latest PR commit whose committer timestamp strictly predates the feedback as the baseline — not the reviewer's clicked commit, which may be an older commit. It never compares that baseline directly with the landing merge: such a diff includes unrelated changes from an advancing target branch. Instead, it gives the adoption reviewers two PR-specific patch snapshots. Let `B` be the feedback baseline, `T` the landing merge's target parent, and `M` the landing merge. The feedback-time snapshot is the tree diff from `merge-base(B, T)` to `B`; the final snapshot is the tree diff from `T` to `M`. A target-only change therefore appears in neither PR patch, while a change added to the PR after feedback appears only in the final snapshot. Force-pushed reviews, feedback that predates every surviving PR commit, and landing shapes whose target parent cannot be reconstructed are deterministically classified `unclear` before any reviewer sees them. Merge status, a resolved thread, an author's “fixed” reply, or a same-file edit is context rather than adoption proof; the PR author's own comments never reach the adapter as they cannot be adoption of themselves.
+
+### Dual-reviewer classification and drafting
+
+Two independently configured reviewer adapters classify every eligible item by provenance (`human-authored`, `forwarded-automation`, or `unclear`) and adoption (`adopted`, `rejected`, or `unclear`). Only matching `human-authored` plus `adopted` verdicts proceed. The adopted set then receives a second independent classification against the current skill: candidate, already covered, implementation-specific, or not feedback. A singleton may qualify; recurrence is not required. Disagreement receives one bounded re-evaluation and remains visible in run artifacts if unresolved. A single batch whose adapter output fails schema or id validation is failed closed at the batch level — every feedback item in it is marked unclear and routed to `excluded` — rather than aborting the whole run; the offending raw output is preserved under the run's private artifacts for debugging. If either adapter returns no valid result for any nonempty batch in an operation, the run exits non-zero and emits a failure record instead of reporting “no candidate.”
+
+The primary adapter drafts from structured agreed guidance, never raw review text. It remains tool-free and read-only by adapter-author contract: it returns complete candidate file content, which the tool validates before writing the sole target. Both adapters then review the same complete skill diff; blocking findings return to a bounded revision loop, and both must approve the same revision. The tool rejects staged changes and edits outside the target skill both before running the documentation and lint gates and again before reporting success, so a gate or concurrent process that adds another path cannot slip through. It restores its own write on failure using best-effort compare-and-swap so a concurrent maintainer edit is not overwritten. On success it saves a candidate bundle containing the source `origin/master` commit, source skill blob ID, reviewed diff, complete candidate, source feedback IDs and URLs, landed evidence ranges, adapter verdicts, and gate results; it never commits, pushes, opens, or merges a PR.
+
+### Reviewer adapter protocol
+
+Each private executable receives a byte-bounded, versioned JSON request on stdin and returns byte-bounded, schema-conforming JSON on stdout. The tool refuses to run when the two reviewer commands resolve to byte-identical executables — a minimum-bar mechanical check; guaranteeing that primary and secondary are backed by independent providers or models is the deployment operator's responsibility. The `access` and `tools` fields are contract markers on the adapter author, not an OS sandbox: reviewer subprocesses spawn with a scrubbed environment, `cwd` set to a private run directory rather than the repository root, and feedback wrapped in a nonce-tagged `<untrusted-feedback nonce="…">` block that every prompt instructs the model to treat as data; the 128-bit nonce prevents an untrusted body from forging the closing tag. Every subprocess uses bounded, abort-aware process-tree cleanup. Adapter authors implement each operation as pure read-only inference — even the `edit` operation returns complete candidate content in JSON, which the tool validates and writes to the sole target. Every production `git`/`gh`/gate spawn also uses the scrubbed environment so a pre-push hook's routing variables cannot silently redirect the maintainer. Candidate writes and the failure rollback use best-effort compare-and-swap against the last written content; the rollback also unstages the target so an adapter- or gate-staged candidate cannot survive a failed run into a later commit.
+
+### Promotion contract
+
+The promote helper starts from a clean checkout at refreshed `origin/master` and refuses to apply a candidate when the current skill blob differs from the bundle's recorded source blob. The operator then reruns the maintenance analysis or manually rebases the diff and repeats the candidate review; the helper never replaces a newer `SKILL.md` with stale complete-file output. After applying a current candidate, it opens a draft PR whose body lists the source feedback URLs or IDs, the landed commit range used as adoption evidence, the originating run, gate results, and any operator edits. Raw adapter prompts and responses remain private, but repository reviewers receive enough provenance to judge whether each proposed rule follows from adopted human feedback.
+
+### Where the mechanism lives
+
+The tool source, adapter binaries, provider credentials, and intended daily scheduler are kept private to the maintainer's machine rather than committed to this repository. This document specifies the protocol; the reference implementation is private infrastructure. The mechanism serves a single skill maintained by a single operator, so the ongoing cost of vetting mechanism edits through repository review outweighs any provenance benefit. If the mechanism is ever handed off to a second maintainer, that handoff is a follow-up RFC that revises this decision — the operator doc at [docs/cookbook/maintaining-dsh-code-review.md](../../../cookbook/maintaining-dsh-code-review.md) is the entry point for anyone taking over.
+
+## Alternatives considered
+
+- **Ship the tool inside this repository.** Rejected for a single-maintainer scope: repository maintenance overhead (typecheck, lint, coverage, cross-cutting refactors) would exceed the value of committed provenance. Retained option for a later handoff.
+- **Record every feedback-time PR head** — rejected: it improves causal isolation but requires a continuously running observer, durable event state, retries, and force-push reconciliation. Periodic maintenance uses reviewed-commit evidence where available and fails closed on broader whole-PR evidence.
+- **Persist a processed-PR cursor** — rejected: an overlapping time-window scan is cheap and naturally idempotent against the current skill, while cursor state creates recovery and missed-event problems.
+- **Run on every new comment** — rejected: review waves produce many related comments and lack the final artifact needed to judge adoption.
+- **Treat merge or thread resolution as adoption** — rejected: a PR can merge with rejected, superseded, or intentionally unresolved feedback.
+- **Create or merge repository changes automatically** — rejected: the tool first needs a track record of useful periodic output. The maintainer inspects and promotes the local diff through normal repository review.
+- **Learn from bot findings that were fixed** — rejected: the source contract is human review feedback. Actor type is filtered before analysis, and human accounts forwarding automated findings are excluded by provenance review.
+- **Use one reviewer as author and final judge** — rejected: independent verdicts expose unsupported generalization before it reaches the skill.
+
+## Acceptance criteria
+
+Promotion from `proposed/` to `implemented/` requires all of the following to be observed in a real end-to-end run against this repository:
+
+- The private tool runs from a clean detached checkout at refreshed `origin/master` and either reports "no candidate" or produces a working-tree diff limited to `.agents/skills/dsh-code-review/SKILL.md`. **Observed on 2026-07-15:** 62 merged PRs scanned, 5 skipped (unreachable merge commit or >250-commit acquisition cap), 426 human feedback items considered, 0 candidates surfaced.
+- Both reviewer adapters are independently configured (distinct providers or models) and complete an analyze / adopt / review pass without user intervention. **Observed on 2026-07-15:** distinct primary/secondary adapters completed adoption + analysis in ~8 minutes; batch fail-closed handled one adapter id-hallucination without aborting the run.
+- A scheduler triggers the tool without an interactive terminal, and a candidate diff (or a "no candidate" record) reaches the operator through a durable notification channel.
+- A controlled acquisition case advances the target branch with a feedback-matching change after the feedback baseline; the reviewer evidence excludes that target-only change while retaining a later PR-owned change.
+- The promote helper rejects a candidate after the source skill changes, and a current candidate opens a draft PR with the provenance summary defined above.
+- At least one candidate diff produced by this workflow is inspected by the operator and promoted to `master` through a normal repository PR review. That PR is the evidence that the workflow can turn adopted feedback into shipped skill guidance.
+
+## Risks
+
+- **Causality inferred from committer timestamps.** The feedback-commit baseline is selected by comparing GitHub commit timestamps with feedback creation timestamps; committer clock skew and rewrites still leave a residual false-adoption window. Cross-referencing GitHub's PR event stream would tighten this but requires event acquisition beyond the scope of the periodic tool.
+- **Two-non-candidate classifications routed to `excluded` without a dispute round.** When both classifiers say "not a candidate" but disagree on which non-candidate reason applies (for example `covered` vs `specific`), the item is excluded rather than re-evaluated. Both classifiers agree the item does not become new reviewer behavior, so a dispute round would not change the outcome.
+- **Dual-reviewer independence beyond byte-hash distinctness is a deployment contract.** The tool refuses to run when the two commands resolve to byte-identical executables, but cannot verify that two distinct wrappers back different providers or models. Operators must configure independent primary and secondary adapters.
+- **Best-effort compare-and-swap for candidate writes and rollback.** File-based CAS on POSIX is not truly atomic; the window is one event-loop tick. The tool targets single-user periodic maintenance and a truly concurrent editor is out of scope.
+- **Single-maintainer bus factor.** Because the mechanism lives on one machine, its interruption stops skill maintenance entirely until the operator restores service or hands off to a new maintainer through a follow-up RFC.
diff --git a/docs/rfc/proposed/simplification/2026-07-12-simplify-session-log-representation.md b/docs/rfc/proposed/simplification/2026-07-12-simplify-session-log-representation.md
deleted file mode 100644
index 715ce93924..0000000000
--- a/docs/rfc/proposed/simplification/2026-07-12-simplify-session-log-representation.md
+++ /dev/null
@@ -1,36 +0,0 @@
-# RFC: Simplify session-log representation
-
-Status: proposed
-
-## Problem
-
-The session log maintains two representations that cost more machinery than their consumers require: a pseudo-linked surface and custom request-header deltas.
-
-`SurfaceManager` stores the same order in an array, a seq map, and mutable `prev`/`next` links. Production never reads `prev`; compact's sole `next` read is the successor of an array position. Replacement already uses `indexOf`, so the links do not make its dominant operation constant-time. A seq array with linear replacement lookup has the same asymptotic replacement cost and one representation to validate.
-
-The request-header subsystem implements a custom system/tool delta codec and transmission-decision layer even though its contract says deltas are an encoding optimization, not a reconstructability requirement. Retaining the initial/resume full snapshot at each loop-instance boundary, then writing a canonical full `request/header` whenever that instance's assembled header changes, preserves replay while deleting `SystemDelta`, `ToolsDelta`, round-trip fallback, and the durable `request/header-delta` variant. Codec-only vocabulary disappears with the codec, not because its individual arms were invalid.
-
-This proposal deliberately retains append and replacement `sourceEventSeqs`, crash-repair provenance, and all `SessionStartSource` variants: implemented RFCs give those fields an audit/interception role that zero current readers does not overturn.
-
-## Proposal
-
-Make `SurfaceManager.nodes` a `readonly number[]` of event sequences and remove the public `SurfaceNode` shape. Keep the internal replace-generation signal; update tool-pairing balance and compaction callers to use array values/indices for predecessor, successor, and replacement ranges, removing node links and the seq-to-node map. Replace post-anchor header deltas with canonical full changed-header snapshots and remove the delta codec/event/tests; initial and resume anchors remain full snapshots even when the folded header is unchanged.
-
-Amend the session-surface and reconstructable-request RFCs where they describe the removed encoding. Update event types/invariants, request logging/replay, persistence fixtures, generated catalogs, package docs, and snapshots. Replace the codec-only `fallback` reason with an explicit `change` reason for post-anchor full snapshots, distinguishing them from the retained `initial` and `resume` anchors.
-
-`SESSION_FORMAT_VERSION` is deliberately pinned at `0`, so an old v0 log containing `request/header-delta` would otherwise pass the version check and silently lose header changes after the delta fold is deleted. Seed/load validation must reject that legacy event fail-loud at the format boundary; no compatibility fold or migration is added.
-
-## Alternatives considered
-
-**Keep linked nodes and compact deltas for possible scale.** Links could help a future cursor API, and deltas can reduce logs when large tool schemas change by a small amount. No shipped cursor uses the links, while full snapshots trade disk size for substantially simpler correctness. If header volume proves material, compression or a measured canonical-delta scheme can be designed around real traces.
-
-## Acceptance criteria
-
-- `SurfaceManager.nodes` is one ordered seq array with no `SurfaceNode`, link fields, or seq-to-node map; incremental append processing and the internal replace-generation signal remain.
-- Replaying full changed-header snapshots reconstructs exactly the same requests; no header-delta event/type/codec remains.
-- A v0 seed or persisted log containing legacy `request/header-delta` is rejected before replay, with coverage for JSONL and SQLite load paths.
-- New-shape v0 JSONL/SQLite replay, provenance, crash repair, compaction, snapshots, invariants, typecheck, coverage, doc-sync, build, and hygiene pass.
-
-## Risks
-
-Full headers increase log volume, and linear replacement lookup could be slower on very large surfaces. Replacements are already linear because the implementation calls `indexOf`; benchmarks should be added only if real traces show the simpler array is a bottleneck. Because the format version remains `0`, forgetting the explicit legacy-event rejection would be silent data corruption rather than a type error; the fail-loud load test is therefore part of the proposal, not optional cleanup.
diff --git a/docs/rfc/rejected/simplification/2026-06-20-drop-bash-output-spill-files.md b/docs/rfc/rejected/simplification/2026-06-20-drop-bash-output-spill-files.md
index 0a13d90f1b..859bd6f23d 100644
--- a/docs/rfc/rejected/simplification/2026-06-20-drop-bash-output-spill-files.md
+++ b/docs/rfc/rejected/simplification/2026-06-20-drop-bash-output-spill-files.md
@@ -12,7 +12,7 @@ This solves a real problem, but in a narrow and leaky way. A spill path is a pro
 
 Keep tail truncation, drop full-output spill files. A bash result contains the bounded tail plus a clear truncation marker; no path is emitted. If users need full-output recovery, add a generic artifact/blob service with explicit ownership, cleanup, and UI rendering, then let bash attach large outputs to that service.
 
-This proposal can land independently of [a generic long-running tool runtime](../../proposed/architecture/2026-06-20-generic-long-running-tool-runtime.md). If background tasks stay, `bash_output` should still report that output was dropped, but without advertising a spill path.
+This proposal can land independently of [a generic long-running tool runtime](../../implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md). If background tasks stay, `bash_output` should still report that output was dropped, but without advertising a spill path.
 
 ## Acceptance criteria
 
diff --git a/docs/testing.md b/docs/testing.md
index d4acdc33c4..8d4f0f554f 100644
--- a/docs/testing.md
+++ b/docs/testing.md
@@ -26,7 +26,12 @@ An e2e assertion re-runs the command or re-reads the file externally; a keyword
 - Product-visible plugins require a non-unit REAL-composition test. Hand-built `ctx.plugin(...)` suites are insufficient: boot test-only `cordis.yml` through Loader and app/process, mock only external/nondeterministic boundaries, and assert model-visible request/log, durable state, or user-visible output. Keep opt-ins out of shipped defaults.
 - A guard only guards if the regression actually fails it. For a plugin without `inject` (bundle/composition plugins), a Loader smoke stays green under a broken export shape — add an explicit `expect('default' in mod).toBe(false)` plus an `unwrapExports` round-trip assertion, and prove it: introduce the regression, watch red, revert.
 - "Real entry path" means the published artifact: the package `bin` points at built `lib/bin.js` under plain `node`, which tsx masks (settle races, module resolution, a swallowed load failure exiting 0). The same applies to any non-index runtime entry the built package resolves at run time (the worker-thread runtime's sibling `lib/worker.cjs`). Keep the built-artifact smokes green (`packages/ui/*/tests/built-bin.e2e.ts`, `packages/code-runtime/code-runtime-worker/tests/built-lib.e2e.ts`), and assert a genuinely-missing config exits non-zero.
-- An e2e that spawns an example from a temp cwd sets `TSX_TSCONFIG_PATH` to the repo-root tsconfig, or it silently falls back to stale built `lib/` ([examples/AGENTS.md](../examples/AGENTS.md)).
+
+## Test subprocess launch modes
+
+- CI and build-having test lanes run every example or Cordis-config subprocess from built `lib/` through the shared dual-mode launcher. Do not hand-write `--import tsx` for these subprocesses.
+- Protocol and operating-system fixtures that do not load Cordis run erasable `.ts` directly with Node, without tsx or the root paths map.
+- Only a test whose subject is source-path resolution may select `src`; state that contract in the test.
 
 ## When a snapshot test is required
 
diff --git a/docs/tool-catalog.md b/docs/tool-catalog.md
index 9ea71d3005..f6a2e0625f 100644
--- a/docs/tool-catalog.md
+++ b/docs/tool-catalog.md
@@ -17,11 +17,13 @@ This table connects model-visible tool names to the plugin package and service s
 | --- | --- | --- | --- | --- | --- |
 | `@deepseek-ai/dsh-tool-ask-user` | `ask_user_question` | `ctx.tools`, `ctx.userInteraction` | `tool/call`, `tool/result after a UI/provider answers the question` | - | ask_user_question pauses the tool call until the active UI provider returns a human answer. |
 | `@deepseek-ai/dsh-tools` | `run_code` | `ctx.tools`, `ctx.codeRuntime (execution time)`, `ctx.systemPrompt` | `tool/call`, `one tool/code-dispatch per bridged sub-call`, `tool/result` | - | Owned by the tool registry as a reserved transport outside filterable capability layers under `mode: code` / `mode: both` (see the Code Mode RFC). Under `code` it is the registry's only wire contribution; the other visible capabilities are declared in a generated TypeScript SDK section, and a program calls them through serialized bindings that re-enter the complete guarded tool pipeline and link each nested execution to this outer result. |
-| `@deepseek-ai/dsh-tool-bash` | `bash`, `bash_kill`, `bash_output` | `ctx.tools`, `ctx.bash` | `tool/call`, `tool/result`, `context/message via agent.inject() for background completion notices` | - | The bash/bash_output/bash_kill tools are model-facing consumers of the bash executor seam. |
-| `@deepseek-ai/dsh-tool-cordis` | `cordis_inspect`, `cordis_mount`, `cordis_unmount` | `ctx.tools` | `tool/call`, `tool/result`, `live plugin-tree mutations (mount/unmount)` | - | Ships in examples/cordis-agent only (a deliberate opt-in — mounted code gets the real ctx, see docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md). Plugins the model mounts may register ADDITIONAL model-visible tools at runtime; the request-header ToolsDelta logs those tool-set changes. |
+| `@deepseek-ai/dsh-tool-bash` | `bash` | `ctx.tools`, `ctx.bash`, `ctx.tasks at call time for run_in_background` | `tool/call`, `tool/result` | - | The bash tool is the model-facing consumer of the bash executor seam. A `run_in_background` run registers with the generic `ctx.tasks` runtime and is collected/stopped through the `task_*` tools from `@deepseek-ai/dsh-tool-tasks`; the `enableRunInBackground` config (default true) removes the parameter entirely when disabled. |
+| `@deepseek-ai/dsh-tool-cordis` | `cordis_inspect`, `cordis_mount`, `cordis_unmount` | `ctx.tools` | `tool/call`, `tool/result`, `live plugin-tree mutations (mount/unmount)` | - | Ships in examples/cordis-agent only (a deliberate opt-in — mounted code gets the real ctx, see docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md). Plugins the model mounts may register ADDITIONAL model-visible tools at runtime; a full changed request header logs those tool-set changes. |
 | `@deepseek-ai/dsh-tool-fs` | `edit`, `read`, `write` | `ctx.tools`, `ctx.fs`, `ctx.systemPrompt` | `tool/call`, `fs/write-intent or fs/edit-intent for mutations`, `fs/observed after successful file operations`, `tool/result` | - | The read-before-write/edit policy is added by `@deepseek-ai/dsh-fs-policy` (an `fs/*` event-gate plugin, no schema change); a deployment that loads these tools is expected to also load it. The tool schemas above are identical with or without the policy plugin. |
+| `@deepseek-ai/dsh-tool-fs-search` | `glob`, `grep` | `ctx.tools`, `ctx.bash`, `ctx.systemPrompt` | `tool/call`, `tool/result` | - | glob and grep are bash-backed discovery tools: they run fixed ripgrep commands through ctx.bash as ordinary foreground calls (never background tasks). Capped results save the complete formatted list through the optional ctx.spillStore backend; returned locators are follow-up-readable/searchable when the backend exposes local paths in co-located deployments. |
 | `@deepseek-ai/dsh-tool-skill` | `skill` | `ctx.tools`, `ctx.skills` | `tool/call`, `tool/result` | - | - |
 | `@deepseek-ai/dsh-tool-subagent` | `subagent` | `ctx.tools`, `ctx.subagents` | `tool/call`, `tool/result`, `child session events through the chosen provider` | `subagent`, `subagent_fork` | The registered tool name is the load-time `toolName` config (default `subagent`); the schema above is that default. The shipped example agents load this package once per subagent backend, so the model additionally sees `subagent_fork` (bound to the fork backend) with an identical schema — see `examples/coding-agent/cordis.yml` and `examples/acp-agent/cordis.yml`. |
+| `@deepseek-ai/dsh-tool-tasks` | `task_kill`, `task_list`, `task_output` | `ctx.tools`, `ctx.tasks`, `ctx.systemPrompt` | `tool/call`, `tool/result`, `context/message via agent.inject() for background completion notices` | - | The kind-agnostic background-task control surface: a background bash command and a background subagent are read, listed, and killed through the same three tools. Loading the plugin attaches the control surface that arms producers' `ctx.tasks.start()`. |
 | `@deepseek-ai/dsh-tool-todo` | `todo_write` | `ctx.tools`, `owning Agent session` | `tool/call`, `todo/write`, `tool/result` | - | todo_write is session-owned state; UIs render the latest todo/write event as a checklist or ACP plan. |
 | `@deepseek-ai/dsh-tool-workflow` | `workflow` | `ctx.tools`, `ctx.workflows`, `ctx.systemPrompt`, `a calling Agent (exec.agent parents the script children)` | `tool/call`, `tool/result` | - | - |
 | `@deepseek-ai/dsh-tool-web` | `web_fetch`, `web_search` | `ctx.tools`, `ctx.web`, `ctx.systemPrompt` | `tool/call`, `tool/result` | - | web_search and web_fetch keep provider selection behind ctx.web so model-visible schemas stay stable across backend swaps. |
@@ -125,7 +127,7 @@ Owned by the tool registry as a reserved transport outside filterable capability
 
 ### `bash`
 
-Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; poll it with `bash_output` and stop it with `bash_kill`.
+Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`.
 
 ```json
 {
@@ -149,7 +151,7 @@ Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs
     },
     "run_in_background": {
       "type": "boolean",
-      "description": "Run in the background and return a task id immediately. No timeout applies."
+      "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
     }
   },
   "required": [
@@ -161,49 +163,7 @@ Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs
 
 Source: [`packages/bash/tool-bash/src/index.ts`](../packages/bash/tool-bash/src/index.ts)
 
-### `bash_kill`
-
-Ask the executor to kill a running background bash task by task id.
-
-```json
-{
-  "type": "object",
-  "properties": {
-    "task_id": {
-      "type": "string",
-      "description": "Task id returned by the bash tool."
-    }
-  },
-  "required": [
-    "task_id"
-  ]
-}
-```
-
-Source: [`packages/bash/tool-bash/src/index.ts`](../packages/bash/tool-bash/src/index.ts)
-
-### `bash_output`
-
-Read new output from a background bash task started with `bash` + `run_in_background`. Returns only output produced since the previous bash_output call, plus the task status. Tasks keep running while you do other work; poll again later for more output.
-
-```json
-{
-  "type": "object",
-  "properties": {
-    "task_id": {
-      "type": "string",
-      "description": "Task id returned by the bash tool."
-    }
-  },
-  "required": [
-    "task_id"
-  ]
-}
-```
-
-Source: [`packages/bash/tool-bash/src/index.ts`](../packages/bash/tool-bash/src/index.ts)
-
-The bash/bash_output/bash_kill tools are model-facing consumers of the bash executor seam.
+The bash tool is the model-facing consumer of the bash executor seam. A `run_in_background` run registers with the generic `ctx.tasks` runtime and is collected/stopped through the `task_*` tools from `@deepseek-ai/dsh-tool-tasks`; the `enableRunInBackground` config (default true) removes the parameter entirely when disabled.
 
 ## `@deepseek-ai/dsh-tool-cordis`
 
@@ -275,7 +235,7 @@ Dispose a plugin previously mounted with cordis_mount, by id. All its registrati
 
 Source: [`packages/cordis/tool-cordis/src/index.ts`](../packages/cordis/tool-cordis/src/index.ts)
 
-Ships in examples/cordis-agent only (a deliberate opt-in — mounted code gets the real ctx, see docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md). Plugins the model mounts may register ADDITIONAL model-visible tools at runtime; the request-header ToolsDelta logs those tool-set changes.
+Ships in examples/cordis-agent only (a deliberate opt-in — mounted code gets the real ctx, see docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md). Plugins the model mounts may register ADDITIONAL model-visible tools at runtime; a full changed request header logs those tool-set changes.
 
 ## `@deepseek-ai/dsh-tool-fs`
 
@@ -371,6 +331,64 @@ Source: [`packages/fs/tool-fs/src/index.ts`](../packages/fs/tool-fs/src/index.ts
 
 The read-before-write/edit policy is added by `@deepseek-ai/dsh-fs-policy` (an `fs/*` event-gate plugin, no schema change); a deployment that loads these tools is expected to also load it. The tool schemas above are identical with or without the policy plugin.
 
+## `@deepseek-ai/dsh-tool-fs-search`
+
+### `glob`
+
+Find files whose paths match a glob pattern. Returns matching paths sorted by modification time, including hidden and ignored files (VCS metadata directories are excluded). Returns the first 100 paths inline; a capped result reports where the complete list was saved.
+
+```json
+{
+  "type": "object",
+  "properties": {
+    "pattern": {
+      "type": "string",
+      "description": "Glob pattern to match file paths against (e.g. \"**/*.ts\", \"src/**/*.test.js\")."
+    },
+    "path": {
+      "type": "string",
+      "description": "Directory to search in. Defaults to the session workspace; a relative path resolves against it."
+    }
+  },
+  "required": [
+    "pattern"
+  ]
+}
+```
+
+Source: [`packages/fs/tool-fs-search/src/index.ts`](../packages/fs/tool-fs-search/src/index.ts)
+
+### `grep`
+
+Search file contents with a ripgrep regular expression. Returns matching lines with line numbers, grouped by file. Returns the first 250 matches inline; a capped result reports where the complete match list was saved. Use read on a matched file for surrounding context.
+
+```json
+{
+  "type": "object",
+  "properties": {
+    "pattern": {
+      "type": "string",
+      "description": "Regular expression to search for (ripgrep syntax)."
+    },
+    "path": {
+      "type": "string",
+      "description": "File or directory to search. Defaults to the session workspace; a relative path resolves against it."
+    },
+    "include": {
+      "type": "string",
+      "description": "One glob filter for which files to search (e.g. \"*.ts\", \"*.{js,jsx}\"). Not a list; negation is not supported."
+    }
+  },
+  "required": [
+    "pattern"
+  ]
+}
+```
+
+Source: [`packages/fs/tool-fs-search/src/index.ts`](../packages/fs/tool-fs-search/src/index.ts)
+
+glob and grep are bash-backed discovery tools: they run fixed ripgrep commands through ctx.bash as ordinary foreground calls (never background tasks). Capped results save the complete formatted list through the optional ctx.spillStore backend; returned locators are follow-up-readable/searchable when the backend exposes local paths in co-located deployments.
+
 ## `@deepseek-ai/dsh-tool-skill`
 
 ### `skill`
@@ -398,7 +416,7 @@ Source: [`packages/skill/tool-skill/src/index.ts`](../packages/skill/tool-skill/
 
 ### `subagent`
 
-Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation.
+Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.
 
 ```json
 {
@@ -411,6 +429,10 @@ Delegate a self-contained task to a subagent (a separate agent that works in its
     "prompt": {
       "type": "string",
       "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+    },
+    "run_in_background": {
+      "type": "boolean",
+      "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
     }
   },
   "required": [
@@ -424,6 +446,77 @@ Source: [`packages/subagent/tool-subagent/src/index.ts`](../packages/subagent/to
 
 The registered tool name is the load-time `toolName` config (default `subagent`); the schema above is that default. The shipped example agents load this package once per subagent backend, so the model additionally sees `subagent_fork` (bound to the fork backend) with an identical schema — see `examples/coding-agent/cordis.yml` and `examples/acp-agent/cordis.yml`.
 
+## `@deepseek-ai/dsh-tool-tasks`
+
+### `task_kill`
+
+Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.
+
+```json
+{
+  "type": "object",
+  "properties": {
+    "task_id": {
+      "type": "string",
+      "description": "Task id returned by the tool that started the background work."
+    },
+    "reason": {
+      "type": "string",
+      "description": "Optional short reason, recorded in the log and forwarded to the task."
+    }
+  },
+  "required": [
+    "task_id"
+  ]
+}
+```
+
+Source: [`packages/tasks/tool-tasks/src/index.ts`](../packages/tasks/tool-tasks/src/index.ts)
+
+### `task_list`
+
+List your background tasks (running and finished) with their ids, kinds, and statuses.
+
+```json
+{
+  "type": "object",
+  "properties": {}
+}
+```
+
+Source: [`packages/tasks/tool-tasks/src/index.ts`](../packages/tasks/tool-tasks/src/index.ts)
+
+### `task_output`
+
+Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.
+
+```json
+{
+  "type": "object",
+  "properties": {
+    "task_id": {
+      "type": "string",
+      "description": "Task id returned by the tool that started the background work."
+    },
+    "wait": {
+      "type": "boolean",
+      "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+    },
+    "timeout_ms": {
+      "type": "number",
+      "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+    }
+  },
+  "required": [
+    "task_id"
+  ]
+}
+```
+
+Source: [`packages/tasks/tool-tasks/src/index.ts`](../packages/tasks/tool-tasks/src/index.ts)
+
+The kind-agnostic background-task control surface: a background bash command and a background subagent are read, listed, and killed through the same three tools. Loading the plugin attaches the control surface that arms producers' `ctx.tasks.start()`.
+
 ## `@deepseek-ai/dsh-tool-todo`
 
 ### `todo_write`
@@ -477,10 +570,10 @@ todo_write is session-owned state; UIs render the latest todo/write event as a c
 
 Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.
 
-The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.
+The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.
 
 Script-body hooks:
-- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), `model` (override). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.
+- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.
 - `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.
 - `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.
 - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.
@@ -527,6 +620,10 @@ Constraints: concurrency and total-agent caps apply; no filesystem, network, tim
                 "type": "string",
                 "description": "Optional one-line description of the phase."
               },
+              "provider": {
+                "type": "string",
+                "description": "Optional provider override this phase is expected to use."
+              },
               "model": {
                 "type": "string",
                 "description": "Optional model override this phase is expected to use."
diff --git a/docs/tool-execution-pipeline.md b/docs/tool-execution-pipeline.md
index 51641d4655..5fc21db2f5 100644
--- a/docs/tool-execution-pipeline.md
+++ b/docs/tool-execution-pipeline.md
@@ -20,9 +20,9 @@ flowchart TD
   owned["Tool-owned session events<br/><code>todo/write</code>, <code>fs/observed</code>, <code>hook/invoked</code>, <code>hook/result</code>, <code>tool/code-dispatch</code>"]
   post["<code>tools/post-execute</code> waterfall<br/>accept, block, replace, add context"]
   final["<code>tools/result</code> synchronous notification<br/>frozen authoritative outcome"]
-  context["Buffered additionalContext<br/>context/message after all tool results"]
+  context["Active-batch additionalContexts FIFO<br/>context/message after recorded tool results"]
   toolResult["Session event: <code>tool/result</code><br/>single model-facing outcome"]
-  allResults["All calls in the step settled<br/>and tool/result events recorded"]
+  allResults["Tool batch settled<br/>recorded tool/result events complete"]
   presentResult["UI completed card<br/>presentResult(args, result)"]
   model --> toolCall
   toolCall --> presentCall
@@ -48,6 +48,6 @@ flowchart TD
   allResults --> context
 ```
 
-Filesystem read-before-edit checks stay below `tool-fs` on `fs/*` events. Generic pre/post waterfalls host hooks and approval policy; `ctx.approval` resolves asks before monotonic guards, and owner policy that must not be reordered remains a registered guard. Around-dispatch concerns such as timeouts wrap `tools/execute`, while `tools/result` observes the immutable outcome after transforms, lossless-JSON validation, and outer error normalization. This lets hooks span tool families without coupling the tools to one policy service. Code Mode sends both the reserved `run_code` transport and its serialized sub-calls through the pipeline; sub-calls carry the parent token, log `tool/code-dispatch`, surface denials as binding rejections, and omit `additionalContext` to preserve call/result adjacency.
+Filesystem read-before-edit checks stay below `tool-fs` on `fs/*` events. Generic pre/post waterfalls host hooks and approval policy; `ctx.approval` resolves asks before monotonic guards, and owner policy that must not be reordered remains a registered guard. Around-dispatch concerns such as timeouts wrap `tools/execute`, while `tools/result` observes the immutable outcome after transforms, lossless-JSON validation, and outer error normalization. This lets hooks span tool families without coupling the tools to one policy service. Code Mode sends both the reserved `run_code` transport and its serialized sub-calls through the pipeline; sub-calls carry the parent token, log `tool/code-dispatch`, surface denials as binding rejections, and omit `additionalContexts` to preserve call/result adjacency.
 
 Maintenance mode: curated Mermaid flow; exact tool schemas and event signatures live in generated catalogs.
diff --git a/docs/user/develop/basic/config.i18n.yaml b/docs/user/develop/basic/config.i18n.yaml
index e4b71a7353..7740eda954 100644
--- a/docs/user/develop/basic/config.i18n.yaml
+++ b/docs/user/develop/basic/config.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-config.md: 5c4e712e2ae452d30fde23bd2481f0c5260ee526
-config.zh.md: 23c97e18a119a92fe7ae883621cb9340ef54bf80
+config.md: 26d2d48ebede74194fbf306aa97d214bdb99b722
+config.zh.md: 9ed389b16779f25c633d0c8772f8658197ba4322
diff --git a/docs/user/develop/basic/config.md b/docs/user/develop/basic/config.md
index 5c4e712e2a..26d2d48ebe 100644
--- a/docs/user/develop/basic/config.md
+++ b/docs/user/develop/basic/config.md
@@ -91,12 +91,19 @@ The test is whether `cordis.yml` can change the value without a code edit.
 
 ### Fail loudly on invalid configuration
 
-If configuration refers to a missing model or another nonexistent resource, fail early instead of silently skipping it:
+If configuration refers to an unregistered LLM provider route or another nonexistent resource, fail early instead of silently skipping it:
 
-```ts ignore-check
-export function apply(ctx: Context, config: Config) {
-  if (!ctx.llm.models().includes(config.model)) {
-    throw new Error(`Model "${config.model}" is not registered by any LLM adapter`)
+```ts
+import type { Context } from 'cordis'
+import type {} from '@deepseek-ai/dsh-llm'
+
+export interface ModelConfig {
+  provider: string
+}
+
+export function apply(ctx: Context, config: ModelConfig) {
+  if (!ctx.llm.listProviders().some(provider => provider.id === config.provider)) {
+    throw new Error(`LLM provider "${config.provider}" is not registered`)
   }
 }
 ```
diff --git a/docs/user/develop/basic/config.zh.md b/docs/user/develop/basic/config.zh.md
index 23c97e18a1..9ed389b167 100644
--- a/docs/user/develop/basic/config.zh.md
+++ b/docs/user/develop/basic/config.zh.md
@@ -91,12 +91,19 @@ export interface Config {
 
 ### 配置错误要响亮
 
-如果配置引用了不存在的东西（比如一个不存在的模型名），应该尽早报错，而不是静默跳过：
+如果配置引用了未注册的 LLM 提供方路由或其他不存在的资源，应该尽早报错，而不是静默跳过：
 
-```ts ignore-check
-export function apply(ctx: Context, config: Config) {
-  if (!ctx.llm.models().includes(config.model)) {
-    throw new Error(`Model "${config.model}" is not registered by any LLM adapter`)
+```ts
+import type { Context } from 'cordis'
+import type {} from '@deepseek-ai/dsh-llm'
+
+export interface ModelConfig {
+  provider: string
+}
+
+export function apply(ctx: Context, config: ModelConfig) {
+  if (!ctx.llm.listProviders().some(provider => provider.id === config.provider)) {
+    throw new Error(`LLM provider "${config.provider}" is not registered`)
   }
 }
 ```
diff --git a/docs/user/develop/framework/index.i18n.yaml b/docs/user/develop/framework/index.i18n.yaml
index 79c947dcb9..1712837d16 100644
--- a/docs/user/develop/framework/index.i18n.yaml
+++ b/docs/user/develop/framework/index.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-index.md: bb08e7cb3f9d3a094100806451fd33d1482003cb
-index.zh.md: 4b9d6d22c8d82bece67c71032a0028b21939f98f
+index.md: 79e925b54509da41535735527e283850384257ec
+index.zh.md: 62be8c706510704f7b07286f166f14fa81235a0a
diff --git a/docs/user/develop/framework/index.md b/docs/user/develop/framework/index.md
index bb08e7cb3f..79e925b545 100644
--- a/docs/user/develop/framework/index.md
+++ b/docs/user/develop/framework/index.md
@@ -79,11 +79,16 @@ export function apply(ctx: Context) {
 
 To stop a plugin instance early:
 
-```ts ignore-check
+```ts
+import type { Context } from 'cordis'
+
+declare const ctx: Context
+declare function myPlugin(ctx: Context): void
+
 const fiber = ctx.plugin(myPlugin)
 
 // Dispose it manually later.
-fiber.dispose()
+await fiber.dispose()
 ```
 
 `dispose` guarantees:
diff --git a/docs/user/develop/framework/index.zh.md b/docs/user/develop/framework/index.zh.md
index 4b9d6d22c8..62be8c7065 100644
--- a/docs/user/develop/framework/index.zh.md
+++ b/docs/user/develop/framework/index.zh.md
@@ -79,11 +79,16 @@ export function apply(ctx: Context) {
 
 当你需要提前终止一个插件实例：
 
-```ts ignore-check
+```ts
+import type { Context } from 'cordis'
+
+declare const ctx: Context
+declare function myPlugin(ctx: Context): void
+
 const fiber = ctx.plugin(myPlugin)
 
 // Dispose it manually later.
-fiber.dispose()
+await fiber.dispose()
 ```
 
 `dispose` 保证：
diff --git a/docs/user/develop/practice/llm-adapter.i18n.yaml b/docs/user/develop/practice/llm-adapter.i18n.yaml
index 84d622dde9..55d96fc757 100644
--- a/docs/user/develop/practice/llm-adapter.i18n.yaml
+++ b/docs/user/develop/practice/llm-adapter.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-llm-adapter.md: 18e05ab79f7daf9f86fe1eb27bdd4440fb9107bc
-llm-adapter.zh.md: f3c1ac70f7b4c11fb9f6dcb247be342fb358bd39
+llm-adapter.md: 7f4fb11fdfd57cd7b05efb04f176a93901c147b1
+llm-adapter.zh.md: 328ca2522d9032ee6816c3f43b35c23ee2ab7625
diff --git a/docs/user/develop/practice/llm-adapter.md b/docs/user/develop/practice/llm-adapter.md
index 18e05ab79f..7f4fb11fdf 100644
--- a/docs/user/develop/practice/llm-adapter.md
+++ b/docs/user/develop/practice/llm-adapter.md
@@ -132,7 +132,7 @@ The first argument lists the model names handled by the adapter. If `cordis.yml`
       - my-model-v2
 
 - id: stdio-agent
-  name: '@deepseek-ai/dsh-stdio-agent'
+  name: '@deepseek-ai/dsh-stdio-demo'
   config:
     model: my-model-v1  # References the model registered above.
 ```
diff --git a/docs/user/develop/practice/llm-adapter.zh.md b/docs/user/develop/practice/llm-adapter.zh.md
index f3c1ac70f7..328ca2522d 100644
--- a/docs/user/develop/practice/llm-adapter.zh.md
+++ b/docs/user/develop/practice/llm-adapter.zh.md
@@ -132,7 +132,7 @@ ctx.llm.registerAdapter(['model-name-1', 'model-name-2'], adapter)
       - my-model-v2
 
 - id: stdio-agent
-  name: '@deepseek-ai/dsh-stdio-agent'
+  name: '@deepseek-ai/dsh-stdio-demo'
   config:
     model: my-model-v1  # References the model registered above.
 ```
diff --git a/docs/user/guide/config.i18n.yaml b/docs/user/guide/config.i18n.yaml
index c917b46a3e..5df7cc2c9c 100644
--- a/docs/user/guide/config.i18n.yaml
+++ b/docs/user/guide/config.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-config.md: 0cad49d9bd82813da230a44d526190d4e0b6a730
-config.zh.md: ee1ea9a0c1d9010be8c1a3984acdcbe870e30584
+config.md: 939773c530ac0bacde3ac0702918854c0a20fcf0
+config.zh.md: 1233b1fe50ed7db2320733b91cfea5cec6e45bfd
diff --git a/docs/user/guide/config.md b/docs/user/guide/config.md
index 0cad49d9bd..939773c530 100644
--- a/docs/user/guide/config.md
+++ b/docs/user/guide/config.md
@@ -23,7 +23,7 @@ A minimal configuration is a list of plugin entries:
       - deepseek-v4-flash
 
 - id: stdio-agent
-  name: '@deepseek-ai/dsh-stdio-agent'
+  name: '@deepseek-ai/dsh-stdio-demo'
   config:
     model: deepseek-v4-flash
 ```
diff --git a/docs/user/guide/config.zh.md b/docs/user/guide/config.zh.md
index ee1ea9a0c1..1233b1fe50 100644
--- a/docs/user/guide/config.zh.md
+++ b/docs/user/guide/config.zh.md
@@ -23,7 +23,7 @@ Harness 使用 `cordis.yml` 描述 Agent 加载哪些插件以及每个插件的
       - deepseek-v4-flash
 
 - id: stdio-agent
-  name: '@deepseek-ai/dsh-stdio-agent'
+  name: '@deepseek-ai/dsh-stdio-demo'
   config:
     model: deepseek-v4-flash
 ```
diff --git a/docs/user/guide/index.i18n.yaml b/docs/user/guide/index.i18n.yaml
index 12ccac1bcb..6743abcdd4 100644
--- a/docs/user/guide/index.i18n.yaml
+++ b/docs/user/guide/index.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-index.md: 4daf7fdc76b38b5d10cc8b4726d9b6239a72933f
-index.zh.md: 1f41078822009234cbaf513aaa8dbc01c4b5d879
+index.md: a20b1041e13b01b6b1d01a5baa8975d3e68c6aa0
+index.zh.md: 56ec50352218e2e28ad2dd7a6ef387376de75606
diff --git a/docs/user/guide/index.md b/docs/user/guide/index.md
index 4daf7fdc76..a20b1041e1 100644
--- a/docs/user/guide/index.md
+++ b/docs/user/guide/index.md
@@ -15,7 +15,7 @@ Harness implements every capability an AI agent needs—including LLM calls, too
     apiKey: !!js process.env.DEEPSEEK_API_KEY
 
 # Select the application template
-- name: '@deepseek-ai/dsh-stdio-agent'
+- name: '@deepseek-ai/dsh-stdio-demo'
   config:
     model: deepseek-v4-flash
 ```
diff --git a/docs/user/guide/index.zh.md b/docs/user/guide/index.zh.md
index 1f41078822..56ec503522 100644
--- a/docs/user/guide/index.zh.md
+++ b/docs/user/guide/index.zh.md
@@ -15,7 +15,7 @@ Harness 将一个 AI Agent（智能体） 所需要的所有能力——LLM 调
     apiKey: !!js process.env.DEEPSEEK_API_KEY
 
 # Select the application template
-- name: '@deepseek-ai/dsh-stdio-agent'
+- name: '@deepseek-ai/dsh-stdio-demo'
   config:
     model: deepseek-v4-flash
 ```
diff --git a/docs/user/guide/quickstart.i18n.yaml b/docs/user/guide/quickstart.i18n.yaml
index 310dfaa62f..2d753829a6 100644
--- a/docs/user/guide/quickstart.i18n.yaml
+++ b/docs/user/guide/quickstart.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-quickstart.md: c2932cf5a5f49c3e3ddb887866b421c6e6cad7a1
-quickstart.zh.md: 54ddb50be31203e98d4cd75d0b8851f8f0e14cc5
+quickstart.md: 0968d5d8596722e8cc53516a355950929571bfbe
+quickstart.zh.md: 4591f73bd0467159c215b2a11d17856f9646f6e7
diff --git a/docs/user/guide/quickstart.md b/docs/user/guide/quickstart.md
index c2932cf5a5..0968d5d859 100644
--- a/docs/user/guide/quickstart.md
+++ b/docs/user/guide/quickstart.md
@@ -91,7 +91,7 @@ Try a task:
 
 ## What happened
 
-echo-agent and coding-agent use the same application framework (`@deepseek-ai/dsh-stdio-agent`). Their `cordis.yml` files select different plugins and configuration. Custom agents use the same composition model.
+echo-agent and coding-agent use the same application framework (`@deepseek-ai/dsh-stdio-demo`). Their `cordis.yml` files select different plugins and configuration. Custom agents use the same composition model.
 
 ## Next steps
 
diff --git a/docs/user/guide/quickstart.zh.md b/docs/user/guide/quickstart.zh.md
index 54ddb50be3..4591f73bd0 100644
--- a/docs/user/guide/quickstart.zh.md
+++ b/docs/user/guide/quickstart.zh.md
@@ -91,7 +91,7 @@ agent REPL ready. Give it a coding task.
 
 ## 回头看
 
-echo-agent 和 coding-agent 用的是同一个应用框架(`@deepseek-ai/dsh-stdio-agent`)，区别只在 `cordis.yml`——换了哪些插件、填了什么配置。你以后定制自己的 Agent 也是同样的方式。
+echo-agent 和 coding-agent 用的是同一个应用框架(`@deepseek-ai/dsh-stdio-demo`)，区别只在 `cordis.yml`——换了哪些插件、填了什么配置。你以后定制自己的 Agent 也是同样的方式。
 
 ## 下一步
 
diff --git a/eslint.config.mjs b/eslint.config.mjs
index 6539f9455e..dd40c005d7 100644
--- a/eslint.config.mjs
+++ b/eslint.config.mjs
@@ -75,7 +75,7 @@ export default tseslint.config(
 
   // --- tests: same rules, minus the friction that fights test ergonomics --
   {
-    files: ['packages/*/*/tests/**/*.ts', 'examples/*/tests/**/*.ts'],
+    files: ['packages/*/*/tests/**/*.ts', 'examples/*/tests/**/*.ts', 'scripts/**/*.spec.ts'],
     extends: [
       ...tseslint.configs.strictTypeChecked,
     ],
diff --git a/examples/AGENTS.md b/examples/AGENTS.md
index f74dda4687..6b820368e4 100644
--- a/examples/AGENTS.md
+++ b/examples/AGENTS.md
@@ -1,8 +1,8 @@
 # AGENTS.md — Examples
 
-Runnable harness compositions. **Examples are not workspaces:** private package stubs are not built. App bins load each `cordis.yml` through `tsx`; package names resolve through root `tsconfig.json` paths, not `node_modules`.
+Runnable harness compositions. `examples/` is one workspace member and the module-resolution root for runnable and test Cordis configs; it is not a build target. [package.json](package.json) declares the packages loaded by those configs, while each leaf's private `package.json` remains metadata only.
 
-Keep wiring, demo fixtures, and e2e/snapshot scenarios here. Move reusable logic into `packages/`, with coverage and a README. App bins own bootstrapping; examples have no `start.ts`.
+Extract reusable logic into `packages/`, where per-file coverage and README gates apply. Examples keep only `cordis.yml` wiring, demo artifacts, and e2e/snapshot scenarios; app package bins own boot glue.
 
 ## E2E smokes
 
@@ -13,7 +13,7 @@ Each example has both:
 
 Mock-only examples require only the keyless tier; state that exception in the test.
 
-Keyless stdio smokes use `@deepseek-ai/dsh-loader-smoke` for isolation, root-tsconfig loading, subprocess lifecycle, diagnostics, EOF, and cleanup; tests supply paths, environment, input, and assertions.
+Keyless stdio smokes use `@deepseek-ai/dsh-loader-smoke`; tests supply paths, environment, input, and assertions. Every checked-in test Cordis config lives under its corresponding `examples/<agent>/` leaf. Map a package-owned config to `examples/<agent>/tests/fixtures/<group>/<package>/cordis.yml`, keep its driver and assertions package-local, and declare every package it names in both root `tsconfig.json` references and `examples/package.json`.
 
 Do not inventory example tests here; the `tests/` trees and root scripts are authoritative.
 
diff --git a/examples/acp-agent/README.md b/examples/acp-agent/README.md
index 35218f4753..5424be6ab3 100644
--- a/examples/acp-agent/README.md
+++ b/examples/acp-agent/README.md
@@ -7,7 +7,7 @@ pnpm run demo:acp          # needs DEEPSEEK_API_KEY (repo-root .env or env)
 pnpm run demo:code-mode acp   # the same server in Code Mode: one wire tool, run_code
 ```
 
-The leaf config loads the ACP app, DeepSeek adapter, sandboxed bash, approval and permission services, model-facing tools, and repeat guard. The app bundles the agent spine, JSONL persistence, and bridge, creates agents on `session/new`, and keeps stdout logger-free. [`fs.cordis.yml`](fs.cordis.yml) adds the unconfined in-process filesystem stack for its dedicated scenarios; [`code-mode.cordis.yml`](code-mode.cordis.yml) adds `run_code` and its generated TypeScript SDK. See [Code Mode](../../packages/core/tools/README.md#code-mode).
+The leaf config loads the ACP app, DeepSeek adapter, sandboxed bash, approval and permission services, model-facing tools, and repeat guard. The app bundles the agent spine, JSONL persistence, and bridge, creates agents on `session/new`, and keeps stdout logger-free. [`fs.cordis.yml`](fs.cordis.yml) adds the unconfined in-process filesystem stack and local tool-result spill storage for its dedicated scenarios; [`code-mode.cordis.yml`](code-mode.cordis.yml) adds `run_code` and its generated TypeScript SDK. See [Code Mode](../../packages/core/tools/README.md#code-mode).
 
 ## stdout is the protocol
 
diff --git a/examples/acp-agent/advanced.cordis.snapshot.yml b/examples/acp-agent/advanced.cordis.snapshot.yml
index 3c8a57f32a..97ae72d222 100644
--- a/examples/acp-agent/advanced.cordis.snapshot.yml
+++ b/examples/acp-agent/advanced.cordis.snapshot.yml
@@ -10,8 +10,11 @@
       - id: acp-agent
         name: '@deepseek-ai/dsh-acp-demo'
         config:
+          provider: deepseek
           model: deepseek-v4-flash
           persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+          workspaceContext:
+            maxBytes: 65536
           tools:
             mode: both
           persona: |
@@ -25,3 +28,10 @@
             name: '@deepseek-ai/dsh-tool-cordis'
           - id: llm-replay
             name: '@deepseek-ai/dsh-llm-replay'
+            config:
+              providers:
+                - id: deepseek
+                  name: DeepSeek
+                  models:
+                    - id: deepseek-v4-flash
+                    - id: deepseek-v4-pro
diff --git a/examples/acp-agent/advanced.cordis.yml b/examples/acp-agent/advanced.cordis.yml
index 630a9c10f7..fee31ebc0d 100644
--- a/examples/acp-agent/advanced.cordis.yml
+++ b/examples/acp-agent/advanced.cordis.yml
@@ -8,8 +8,11 @@
       - id: acp-agent
         name: '@deepseek-ai/dsh-acp-demo'
         config:
+          provider: deepseek
           model: deepseek-v4-flash
           persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+          workspaceContext:
+            maxBytes: 65536
           tools:
             mode: both
           persona: |
diff --git a/examples/acp-agent/both-mode.cordis.snapshot.yml b/examples/acp-agent/both-mode.cordis.snapshot.yml
index c82025252d..85c1ff8239 100644
--- a/examples/acp-agent/both-mode.cordis.snapshot.yml
+++ b/examples/acp-agent/both-mode.cordis.snapshot.yml
@@ -12,8 +12,11 @@
       - id: acp-agent
         name: '@deepseek-ai/dsh-acp-demo'
         config:
+          provider: deepseek
           model: deepseek-v4-flash
           persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+          workspaceContext:
+            maxBytes: 65536
           tools:
             mode: both
           persona: |
@@ -25,3 +28,10 @@
             name: '@deepseek-ai/dsh-code-runtime-worker'
           - id: llm-replay
             name: '@deepseek-ai/dsh-llm-replay'
+            config:
+              providers:
+                - id: deepseek
+                  name: DeepSeek
+                  models:
+                    - id: deepseek-v4-flash
+                    - id: deepseek-v4-pro
diff --git a/examples/acp-agent/both-mode.cordis.yml b/examples/acp-agent/both-mode.cordis.yml
index 1624e43330..6b554abefb 100644
--- a/examples/acp-agent/both-mode.cordis.yml
+++ b/examples/acp-agent/both-mode.cordis.yml
@@ -10,8 +10,11 @@
       - id: acp-agent
         name: '@deepseek-ai/dsh-acp-demo'
         config:
+          provider: deepseek
           model: deepseek-v4-flash
           persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+          workspaceContext:
+            maxBytes: 65536
           tools:
             mode: both
           persona: |
diff --git a/examples/acp-agent/code-mode-workspace-context.cordis.snapshot.yml b/examples/acp-agent/code-mode-workspace-context.cordis.snapshot.yml
new file mode 100644
index 0000000000..a741091cce
--- /dev/null
+++ b/examples/acp-agent/code-mode-workspace-context.cordis.snapshot.yml
@@ -0,0 +1,37 @@
+# Keyless replay counterpart of code-mode-workspace-context.cordis.yml. It
+# enables the filesystem entries needed by this scenario and swaps in replay.
+- id: base
+  name: '@cordisjs/plugin-include'
+  config:
+    path: ./cordis.yml
+    patches:
+      - id: llm-deepseek
+        name: '@deepseek-ai/dsh-llm-deepseek'
+        disabled: true
+      - id: acp-agent
+        name: '@deepseek-ai/dsh-acp-demo'
+        config:
+          provider: deepseek
+          model: deepseek-v4-flash
+          persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+          workspaceContext:
+            maxBytes: 65536
+          tools:
+            mode: code
+          persona: |
+            You are a coding assistant powered by the {{model}} model. Your working directory is {{cwd}}.
+
+            Verify your work by running the code or tests. Keep answers brief and factual.
+      - insert:
+          - id: fs-local
+            name: '@deepseek-ai/dsh-fs-local'
+            config:
+              cwd: !!js process.cwd()
+          - id: fs-policy
+            name: '@deepseek-ai/dsh-fs-policy'
+          - id: tool-fs
+            name: '@deepseek-ai/dsh-tool-fs'
+          - id: code-runtime
+            name: '@deepseek-ai/dsh-code-runtime-worker'
+          - id: llm-replay
+            name: '@deepseek-ai/dsh-llm-replay'
diff --git a/examples/acp-agent/code-mode-workspace-context.cordis.yml b/examples/acp-agent/code-mode-workspace-context.cordis.yml
new file mode 100644
index 0000000000..b932f06e64
--- /dev/null
+++ b/examples/acp-agent/code-mode-workspace-context.cordis.yml
@@ -0,0 +1,32 @@
+# Code Mode workspace-context snapshot recording overlay. The scenario needs
+# filesystem tools to trigger nested instruction discovery after a read.
+- id: base
+  name: '@cordisjs/plugin-include'
+  config:
+    path: ./cordis.yml
+    patches:
+      - id: acp-agent
+        name: '@deepseek-ai/dsh-acp-demo'
+        config:
+          provider: deepseek
+          model: deepseek-v4-flash
+          persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+          workspaceContext:
+            maxBytes: 65536
+          tools:
+            mode: code
+          persona: |
+            You are a coding assistant powered by the {{model}} model. Your working directory is {{cwd}}.
+
+            Verify your work by running the code or tests. Keep answers brief and factual.
+      - insert:
+          - id: fs-local
+            name: '@deepseek-ai/dsh-fs-local'
+            config:
+              cwd: !!js process.cwd()
+          - id: fs-policy
+            name: '@deepseek-ai/dsh-fs-policy'
+          - id: tool-fs
+            name: '@deepseek-ai/dsh-tool-fs'
+          - id: code-runtime
+            name: '@deepseek-ai/dsh-code-runtime-worker'
diff --git a/examples/acp-agent/code-mode.cordis.snapshot.yml b/examples/acp-agent/code-mode.cordis.snapshot.yml
index ea3de598d8..f672c7a8d7 100644
--- a/examples/acp-agent/code-mode.cordis.snapshot.yml
+++ b/examples/acp-agent/code-mode.cordis.snapshot.yml
@@ -12,8 +12,11 @@
       - id: acp-agent
         name: '@deepseek-ai/dsh-acp-demo'
         config:
+          provider: deepseek
           model: deepseek-v4-flash
           persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+          workspaceContext:
+            maxBytes: 65536
           tools:
             mode: code
           persona: |
@@ -25,3 +28,10 @@
             name: '@deepseek-ai/dsh-code-runtime-worker'
           - id: llm-replay
             name: '@deepseek-ai/dsh-llm-replay'
+            config:
+              providers:
+                - id: deepseek
+                  name: DeepSeek
+                  models:
+                    - id: deepseek-v4-flash
+                    - id: deepseek-v4-pro
diff --git a/examples/acp-agent/code-mode.cordis.yml b/examples/acp-agent/code-mode.cordis.yml
index 0170e808ce..fee39ef22a 100644
--- a/examples/acp-agent/code-mode.cordis.yml
+++ b/examples/acp-agent/code-mode.cordis.yml
@@ -11,8 +11,11 @@
       - id: acp-agent
         name: '@deepseek-ai/dsh-acp-demo'
         config:
+          provider: deepseek
           model: deepseek-v4-flash
           persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+          workspaceContext:
+            maxBytes: 65536
           tools:
             mode: code
           persona: |
diff --git a/examples/acp-agent/cordis.snapshot.yml b/examples/acp-agent/cordis.snapshot.yml
index 90a2fa6a2a..e4d717a399 100644
--- a/examples/acp-agent/cordis.snapshot.yml
+++ b/examples/acp-agent/cordis.snapshot.yml
@@ -28,3 +28,10 @@
       - insert:
           - id: llm-replay
             name: '@deepseek-ai/dsh-llm-replay'
+            config:
+              providers:
+                - id: deepseek
+                  name: DeepSeek
+                  models:
+                    - id: deepseek-v4-flash
+                    - id: deepseek-v4-pro
diff --git a/examples/acp-agent/cordis.yml b/examples/acp-agent/cordis.yml
index 345fd49b45..cb8890dae7 100644
--- a/examples/acp-agent/cordis.yml
+++ b/examples/acp-agent/cordis.yml
@@ -9,9 +9,6 @@
   config:
     apiKey: !!js process.env.DEEPSEEK_API_KEY
     baseURL: !!js process.env.DEEPSEEK_BASE_URL
-    models:
-      - deepseek-v4-flash
-      - deepseek-v4-pro
 
 # The default composition confines bash to the workspace and asks before a
 # wider retry. Snapshots use danger-full-access; DSH_PERMISSION_MODE overrides
@@ -33,14 +30,17 @@
 - id: permission
   name: '@deepseek-ai/dsh-permission'
 
-# The ACP server app: the agent-core spine + JSONL persistence + the ACP bridge.
+# The ACP server app: the agent-spine-demo spine + JSONL persistence + the ACP bridge.
 # Persistence root: $DSH_SNAPSHOT_SESSIONS_ROOT when the snapshot harness sets it
 # (so it can harvest / isolate the log), else ./.sessions for the demo.
 - id: acp-agent
   name: '@deepseek-ai/dsh-acp-demo'
   config:
+    provider: deepseek
     model: deepseek-v4-flash
     persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+    workspaceContext:
+      maxBytes: 65536
     # Keep the persona to identity and behavior; tool plugins own tool guidance.
     # The loop resolves {{model}} and each ACP session's client-supplied {{cwd}}.
     persona: |
diff --git a/examples/acp-agent/fs.cordis.snapshot.yml b/examples/acp-agent/fs.cordis.snapshot.yml
index 53f2d677e2..da0b2ca59b 100644
--- a/examples/acp-agent/fs.cordis.snapshot.yml
+++ b/examples/acp-agent/fs.cordis.snapshot.yml
@@ -17,5 +17,20 @@
             name: '@deepseek-ai/dsh-fs-policy'
           - id: tool-fs
             name: '@deepseek-ai/dsh-tool-fs'
+          - id: spill-local
+            name: '@deepseek-ai/dsh-spill-local'
+            config:
+              root: !!js process.env.DSH_SNAPSHOT_SPILL_ROOT ?? './.spill'
+          - id: spill-policy
+            name: '@deepseek-ai/dsh-spill-policy'
+            config:
+              maxInlineBytes: 800
           - id: llm-replay
             name: '@deepseek-ai/dsh-llm-replay'
+            config:
+              providers:
+                - id: deepseek
+                  name: DeepSeek
+                  models:
+                    - id: deepseek-v4-flash
+                    - id: deepseek-v4-pro
diff --git a/examples/acp-agent/fs.cordis.yml b/examples/acp-agent/fs.cordis.yml
index b60c26f96d..52ca959a89 100644
--- a/examples/acp-agent/fs.cordis.yml
+++ b/examples/acp-agent/fs.cordis.yml
@@ -15,3 +15,11 @@
             name: '@deepseek-ai/dsh-fs-policy'
           - id: tool-fs
             name: '@deepseek-ai/dsh-tool-fs'
+          - id: spill-local
+            name: '@deepseek-ai/dsh-spill-local'
+            config:
+              root: !!js process.env.DSH_SNAPSHOT_SPILL_ROOT ?? './.spill'
+          - id: spill-policy
+            name: '@deepseek-ai/dsh-spill-policy'
+            config:
+              maxInlineBytes: !!js process.env.DSH_SNAPSHOT && 800 || 50000
diff --git a/examples/acp-agent/tests/acp.e2e.ts b/examples/acp-agent/tests/acp.e2e.ts
index 23c6f58301..b19bc9bfb1 100644
--- a/examples/acp-agent/tests/acp.e2e.ts
+++ b/examples/acp-agent/tests/acp.e2e.ts
@@ -15,6 +15,7 @@ import {
   type RequestPermissionResponse,
   type SessionNotification,
 } from '@agentclientprotocol/sdk'
+import { resolveExampleLaunch } from '@deepseek-ai/dsh-loader-smoke'
 
 /**
  * Boots examples/acp-agent as an ACP subprocess. The key-gated prompt leg
@@ -25,8 +26,6 @@ import {
 // The child runs from a temp cwd, so its bin and config path are absolute.
 const binScript = fileURLToPath(new URL('../../../packages/examples/acp-demo/src/bin.ts', import.meta.url))
 const configPath = fileURLToPath(new URL('../cordis.yml', import.meta.url))
-// Resolve tsx absolutely because the subprocess runs outside the repo.
-const tsxLoader = fileURLToPath(import.meta.resolve('tsx'))
 // The root tsconfig supplies unbuilt workspace `paths`; making it explicit
 // avoids accidental resolution through stale built output.
 const repoTsconfig = fileURLToPath(new URL('../../../tsconfig.json', import.meta.url))
@@ -38,24 +37,21 @@ interface Spawned {
   stderr: string[]
 }
 
-// TODO(acp-test-harness): this subprocess/client boot glue is duplicated with
-// hooks.e2e.ts and partly with dsh-acp-snapshot's harness. Migrate both e2e
-// files onto that launcher before the TSX/env/permission-stub details drift.
 function spawnAcpAgent(cwd: string, env: NodeJS.ProcessEnv = process.env): Spawned {
-  const child = spawn(
-    process.execPath,
-    ['--import', tsxLoader, binScript, '--config', configPath],
-    {
-      cwd,
-      env: {
-        ...env,
-        TSX_TSCONFIG_PATH: repoTsconfig,
-        DSH_PERMISSION_MODE: 'danger-full-access',
-        DSH_HOME: join(cwd, '.dsh'),
-        DSH_AGENTS_HOME: join(cwd, '.agents'),
-      },
-      stdio: ['pipe', 'pipe', 'pipe'],
+  const launch = resolveExampleLaunch({
+    srcBin: binScript,
+    configArgs: ['--config', configPath],
+    tsconfigPath: repoTsconfig,
+    env: {
+      DSH_PERMISSION_MODE: 'danger-full-access',
+      DSH_HOME: join(cwd, '.dsh'),
+      DSH_AGENTS_HOME: join(cwd, '.agents'),
     },
+  })
+  const child = spawn(
+    launch.command,
+    launch.args,
+    { cwd, env: { ...env, ...launch.env }, stdio: ['pipe', 'pipe', 'pipe'] },
   )
   const stderr: string[] = []
   child.stderr.setEncoding('utf8')
@@ -138,16 +134,20 @@ describe('acp-agent over real stdio (no key required)', () => {
     workdir = await mkdtemp(join(tmpdir(), 'acp-e2e-'))
     // Collect raw stdout bytes directly (bypass the SDK framing) to inspect.
     // A dummy key boots the adapter; this purity test sends no prompt and makes no model call.
-    const child = spawn(process.execPath, ['--import', tsxLoader, binScript, '--config', configPath], {
-      cwd: workdir,
+    const launch = resolveExampleLaunch({
+      srcBin: binScript,
+      configArgs: ['--config', configPath],
+      tsconfigPath: repoTsconfig,
       env: {
-        ...process.env,
         DEEPSEEK_API_KEY: process.env.DEEPSEEK_API_KEY ?? 'sk-dummy-for-boot',
-        TSX_TSCONFIG_PATH: repoTsconfig,
         DSH_PERMISSION_MODE: 'danger-full-access',
         DSH_HOME: join(workdir, '.dsh'),
         DSH_AGENTS_HOME: join(workdir, '.agents'),
       },
+    })
+    const child = spawn(launch.command, launch.args, {
+      cwd: workdir,
+      env: { ...process.env, ...launch.env },
       stdio: ['pipe', 'pipe', 'pipe'],
     })
     const out: string[] = []
diff --git a/examples/acp-agent/tests/acp.snapshot.ts b/examples/acp-agent/tests/acp.snapshot.ts
index 3e605b3141..e190163a18 100644
--- a/examples/acp-agent/tests/acp.snapshot.ts
+++ b/examples/acp-agent/tests/acp.snapshot.ts
@@ -25,7 +25,9 @@ const AGENT = {
 // The Code Mode overlay configs (include-patched variants of cordis.yml; the
 // replay swap resolves each one's sibling `*cordis.snapshot.yml`).
 const CODE_MODE_CONFIG = fileURLToPath(new URL('../code-mode.cordis.yml', import.meta.url))
+const CODE_MODE_WORKSPACE_CONTEXT_CONFIG = fileURLToPath(new URL('../code-mode-workspace-context.cordis.yml', import.meta.url))
 const BOTH_MODE_CONFIG = fileURLToPath(new URL('../both-mode.cordis.yml', import.meta.url))
+const WORKSPACE_CONTEXT_CONFIG = fileURLToPath(new URL('../workspace-context.cordis.yml', import.meta.url))
 const ADVANCED_CONFIG = fileURLToPath(new URL('../advanced.cordis.yml', import.meta.url))
 const FS_CONFIG = fileURLToPath(new URL('../fs.cordis.yml', import.meta.url))
 
@@ -51,6 +53,14 @@ const SCENARIOS: Scenario[] = [
   // Its prompt and tool-schema sidecars pin the composed header.
   { name: 'text-turn', hasModelTurn: true, recorded: true, pinsHeader: true },
   { name: 'tool-call-turn', hasModelTurn: true, recorded: true },
+  {
+    name: 'parallel-tool-calls',
+    hasModelTurn: true,
+    recorded: false,
+    headerClass: 'fs',
+    configPath: FS_CONFIG,
+  },
+  { name: 'bash-spill', hasModelTurn: true, recorded: false, headerClass: 'fs', configPath: FS_CONFIG },
   { name: 'fs-terminal-card', hasModelTurn: true, recorded: true },
   { name: 'todo-plan', hasModelTurn: true, recorded: true },
   { name: 'skill-load', hasModelTurn: true, recorded: false, pinsHeader: true, headerClass: 'skill' },
@@ -62,12 +72,36 @@ const SCENARIOS: Scenario[] = [
   { name: 'fs-read-window', hasModelTurn: true, recorded: true, headerClass: 'fs', configPath: FS_CONFIG },
   { name: 'fs-policy-reject', hasModelTurn: true, recorded: true, headerClass: 'fs', configPath: FS_CONFIG },
   { name: 'multi-turn', hasModelTurn: true, recorded: true },
+  // ACP exposes the adapter catalog as a session-scoped model select. This
+  // scenario pins the default flash request, the switch response, and the
+  // resulting changed request-header snapshot for pro.
+  {
+    name: 'model-switching',
+    hasModelTurn: true,
+    recorded: true,
+    pinsHeader: true,
+    expectedHeaderChanges: 1,
+    headerClass: 'model-switching',
+  },
   { name: 'error-finish', hasModelTurn: true, recorded: false, overridden: true },
   // Keyless, authored (like error-finish/cancel): deterministically forcing a
   // LIVE model to repeat one call three times is not a stable recording, so
   // the fixture scripts five identical todo_write calls and pins BOTH reminder
   // tiers (gentle at 3, detailed at 5) as context/message in transcript and log.
   { name: 'repeat-tool-guard', hasModelTurn: true, recorded: false },
+  // Authored replay: a root AGENTS.md pins the session prefix, then a read in
+  // nested/ discovers its narrower AGENTS.md as a raw, metadata-bearing
+  // context/message. The scenario-specific config keeps home/root discovery
+  // hermetic, and the resulting prefix needs its own pinned header class.
+  {
+    name: 'workspace-context',
+    hasModelTurn: true,
+    recorded: false,
+    overridden: true,
+    pinsHeader: true,
+    headerClass: 'workspace-context',
+    configPath: WORKSPACE_CONTEXT_CONFIG,
+  },
   { name: 'cancel', hasModelTurn: true, recorded: false, overridden: true },
   { name: 'subagent-spawn', hasModelTurn: true, recorded: true, childSessions: 1 },
   { name: 'subagent-multi', hasModelTurn: true, recorded: true, childSessions: 2 },
@@ -116,14 +150,25 @@ const SCENARIOS: Scenario[] = [
   // tools:sdk section rides in the prompt, and the program's tool calls land as
   // tool/code-dispatch events. Each overlay composes and pins its own header class.
   { name: 'code-mode-turn', hasModelTurn: true, recorded: true, pinsHeader: true, headerClass: 'code', configPath: CODE_MODE_CONFIG },
+  // A nested fs dispatch inside run_code discovers workspace instructions. The
+  // context/message must follow the outer result while retaining workspace
+  // provenance, which proves Code Mode carries deferred tool context end to end.
+  {
+    name: 'code-mode-workspace-context',
+    hasModelTurn: true,
+    recorded: true,
+    pinsHeader: true,
+    headerClass: 'code-workspace-context',
+    configPath: CODE_MODE_WORKSPACE_CONTEXT_CONFIG,
+  },
   { name: 'both-mode-turn', hasModelTurn: true, recorded: true, pinsHeader: true, headerClass: 'both', configPath: BOTH_MODE_CONFIG },
-  // The default tree owns the single Permissions select. Snapshot mode starts
-  // in danger-full-access so established fixtures stay runner-independent;
-  // these policy scenarios switch to workspace-write in their input scripts.
+  // The default tree also owns the Permissions select. Snapshot mode starts in
+  // danger-full-access so established fixtures stay runner-independent; these
+  // policy scenarios switch to workspace-write in their input scripts.
   // Real-kernel confinement remains in escalation.e2e.ts and the sandbox
   // packages' e2e suites.
   { name: 'config-options', hasModelTurn: false, recorded: false, headerClass: 'sandbox' },
-  { name: 'permission-switching', hasModelTurn: true, recorded: true, pinsHeader: true, expectedHeaderDeltas: 1, headerClass: 'sandbox' },
+  { name: 'permission-switching', hasModelTurn: true, recorded: true, pinsHeader: true, expectedHeaderChanges: 1, headerClass: 'sandbox' },
   { name: 'escalation-approved', hasModelTurn: true, recorded: true, headerClass: 'sandbox' },
   { name: 'escalation-rejected', hasModelTurn: true, recorded: true, headerClass: 'sandbox' },
 ]
diff --git a/examples/acp-agent/tests/escalation.e2e.ts b/examples/acp-agent/tests/escalation.e2e.ts
index 507e938950..7e1024c23f 100644
--- a/examples/acp-agent/tests/escalation.e2e.ts
+++ b/examples/acp-agent/tests/escalation.e2e.ts
@@ -15,6 +15,7 @@ import {
   type RequestPermissionResponse,
   type SessionNotification,
 } from '@agentclientprotocol/sdk'
+import { resolveExampleLaunch } from '@deepseek-ai/dsh-loader-smoke'
 
 /**
  * Exercises the default ACP composition through the real bin and Loader. The
@@ -28,9 +29,8 @@ import {
 
 const binScript = fileURLToPath(new URL('../../../packages/examples/acp-demo/src/bin.ts', import.meta.url))
 const configPath = fileURLToPath(new URL('../cordis.yml', import.meta.url))
-const tsxLoader = fileURLToPath(import.meta.resolve('tsx'))
 // The subprocess runs from a temp cwd outside the repo; point tsx at the repo
-// tsconfig so the unbuilt `paths` map resolves (see examples/AGENTS.md).
+// tsconfig so the unbuilt `paths` map resolves in src mode (see examples/AGENTS.md).
 const repoTsconfig = fileURLToPath(new URL('../../../tsconfig.json', import.meta.url))
 
 // Without a usable bwrap/Seatbelt runner, the strict attempt fails closed with
@@ -55,17 +55,19 @@ interface Spawned {
 
 /** Boot the example as an ACP subprocess; the scripted client answers every permission prompt with `answer`. */
 function spawnAcpAgent(cwd: string, answer: 'allow-once' | 'reject-once'): Spawned {
+  const launch = resolveExampleLaunch({
+    srcBin: binScript,
+    configArgs: ['--config', configPath],
+    tsconfigPath: repoTsconfig,
+    // A dummy key lets the deepseek adapter boot keyless (presence-checked at
+    // apply, used only on a real model call); the with-key tests carry the
+    // real key, so the fallback is inert there.
+    env: { DEEPSEEK_API_KEY: process.env.DEEPSEEK_API_KEY || 'sk-dummy-for-boot' },
+  })
   const child = spawn(
-    process.execPath,
-    ['--import', tsxLoader, binScript, '--config', configPath],
-    {
-      cwd,
-      // A dummy key lets the deepseek adapter boot keyless (presence-checked at
-      // apply, used only on a real model call); the with-key tests carry the
-      // real key, so the fallback is inert there.
-      env: { ...process.env, DEEPSEEK_API_KEY: process.env.DEEPSEEK_API_KEY ?? 'sk-dummy-for-boot', TSX_TSCONFIG_PATH: repoTsconfig },
-      stdio: ['pipe', 'pipe', 'pipe'],
-    },
+    launch.command,
+    launch.args,
+    { cwd, env: { ...process.env, ...launch.env }, stdio: ['pipe', 'pipe', 'pipe'] },
   )
   const stderr: string[] = []
   child.stderr.setEncoding('utf8')
@@ -115,15 +117,16 @@ describe('default sandbox composition keyless smoke (real cordis.yml via the Loa
     expect(sessionId.length).toBeGreaterThan(0)
   }, 30_000)
 
-  it('advertises the Permissions select and honors a switch end to end (no key, no model)', async () => {
+  it('advertises model and Permissions selects and honors a permission switch without a model call', async () => {
     workdir = await mkdtemp(join(tmpdir(), 'sandbox-acp-config-'))
     spawned = spawnAcpAgent(workdir, 'reject-once')
     const { client } = spawned
     await client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
     const created = await client.newSession({ cwd: workdir, mcpServers: [] })
     const advertised = created.configOptions ?? []
+    const modelValue = JSON.stringify(['deepseek', 'deepseek-v4-flash'])
     expect(advertised.map(option => [option.id, 'currentValue' in option ? option.currentValue : undefined]))
-      .toEqual([['permission', 'workspace-write']])
+      .toEqual([['model', modelValue], ['permission', 'workspace-write']])
     const afterFullAccess = await client.setSessionConfigOption({
       sessionId: created.sessionId, configId: 'permission', value: 'danger-full-access',
     })
@@ -133,7 +136,7 @@ describe('default sandbox composition keyless smoke (real cordis.yml via the Loa
       sessionId: created.sessionId, configId: 'permission', value: 'danger-full-access',
     })
     expect((again.configOptions ?? []).map(option => [option.id, 'currentValue' in option ? option.currentValue : undefined]))
-      .toEqual([['permission', 'danger-full-access']])
+      .toEqual([['model', modelValue], ['permission', 'danger-full-access']])
     await expect(client.setSessionConfigOption({
       sessionId: created.sessionId, configId: 'permission', value: 'plan',
     })).rejects.toThrow(/unknown permission value/)
diff --git a/examples/acp-agent/tests/hooks.e2e.ts b/examples/acp-agent/tests/hooks.e2e.ts
index 99cebb907e..68dc2648bf 100644
--- a/examples/acp-agent/tests/hooks.e2e.ts
+++ b/examples/acp-agent/tests/hooks.e2e.ts
@@ -15,6 +15,7 @@ import {
   type RequestPermissionResponse,
   type SessionNotification,
 } from '@agentclientprotocol/sdk'
+import { resolveExampleLaunch } from '@deepseek-ai/dsh-loader-smoke'
 
 /**
  * With-key e2e for the Claude hook bridge. The process-level `./hooks.json` is
@@ -25,7 +26,6 @@ import {
 
 const binScript = fileURLToPath(new URL('../../../packages/examples/acp-demo/src/bin.ts', import.meta.url))
 const configPath = fileURLToPath(new URL('../cordis.yml', import.meta.url))
-const tsxLoader = fileURLToPath(import.meta.resolve('tsx'))
 const repoTsconfig = fileURLToPath(new URL('../../../tsconfig.json', import.meta.url))
 
 interface Spawned {
@@ -36,10 +36,16 @@ interface Spawned {
 }
 
 function spawnAcpAgent(cwd: string): Spawned {
+  const launch = resolveExampleLaunch({
+    srcBin: binScript,
+    configArgs: ['--config', configPath],
+    tsconfigPath: repoTsconfig,
+    env: { DSH_PERMISSION_MODE: 'danger-full-access' },
+  })
   const child = spawn(
-    process.execPath,
-    ['--import', tsxLoader, binScript, '--config', configPath],
-    { cwd, env: { ...process.env, TSX_TSCONFIG_PATH: repoTsconfig, DSH_PERMISSION_MODE: 'danger-full-access' }, stdio: ['pipe', 'pipe', 'pipe'] },
+    launch.command,
+    launch.args,
+    { cwd, env: { ...process.env, ...launch.env }, stdio: ['pipe', 'pipe', 'pipe'] },
   )
   const stderr: string[] = []
   child.stderr.setEncoding('utf8')
diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl
index d996251cd7..25a6f76411 100644
--- a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl
+++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.1.jsonl
@@ -2,12 +2,12 @@
 {"type":"turn/start","seq":0,"time":1783957884563,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783957884563,"data":{"content":[{"type":"text","text":"Reply with exactly DIRECT_CHILD_OK and nothing else."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783957884564,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783957884564,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783957884564,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783950001005,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}
 {"type":"assistant/chunk","seq":5,"time":1783957884564,"data":{"turn":1,"step":1,"chunk":{"type":"text-delta","index":0,"text":"DIRECT_CHILD_OK"}}}
 {"type":"assistant/chunk","seq":6,"time":1783957884564,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"DIRECT_CHILD_OK"}}}}
 {"type":"assistant/chunk","seq":7,"time":1783957884564,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}}
 {"type":"assistant/chunk","seq":8,"time":1783957884564,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":9,"time":1783957884564,"data":{"turn":1,"step":1,"content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[4,5,6,7,8],"surfaceOp":"append"}
+{"type":"assistant/message","seq":9,"time":1783957884564,"data":{"turn":1,"step":1,"content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[4,5,6,7,8],"surfaceOp":"append"}
 {"type":"step/end","seq":10,"time":1783957884564,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":11,"time":1783957884564,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl
index 391d5ce2ba..45d9043a4a 100644
--- a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl
+++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.2.jsonl
@@ -2,12 +2,12 @@
 {"type":"turn/start","seq":0,"time":1783957884700,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783957884700,"data":{"content":[{"type":"text","text":"Reply with exactly WORKFLOW_CHILD_OK and nothing else."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783957884700,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783957884701,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783957884701,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783950002005,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}
 {"type":"assistant/chunk","seq":5,"time":1783957884701,"data":{"turn":1,"step":1,"chunk":{"type":"text-delta","index":0,"text":"WORKFLOW_CHILD_OK"}}}
 {"type":"assistant/chunk","seq":6,"time":1783957884701,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"WORKFLOW_CHILD_OK"}}}}
 {"type":"assistant/chunk","seq":7,"time":1783957884701,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}}
 {"type":"assistant/chunk","seq":8,"time":1783957884701,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":9,"time":1783957884701,"data":{"turn":1,"step":1,"content":[{"type":"text","text":"WORKFLOW_CHILD_OK"}],"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[4,5,6,7,8],"surfaceOp":"append"}
+{"type":"assistant/message","seq":9,"time":1783957884701,"data":{"turn":1,"step":1,"content":[{"type":"text","text":"WORKFLOW_CHILD_OK"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[4,5,6,7,8],"surfaceOp":"append"}
 {"type":"step/end","seq":10,"time":1783957884701,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":11,"time":1783957884701,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.jsonl b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.jsonl
index 779844067a..39f867984a 100644
--- a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.jsonl
@@ -2,13 +2,13 @@
 {"type":"turn/start","seq":0,"time":1783957884479,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783957884479,"data":{"content":[{"type":"text","text":"Run this advanced flow exactly once: mount a no-op Cordis plugin named snapshot-marker; use run_code to inspect the live dynamic mounts through tools.cordis_inspect; delegate once to a direct spawn child; run one workflow that delegates to another spawn child; unmount dyn-1; then reply with exactly ADVANCED_ACP_OK."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783957884486,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783957884486,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783957884486,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783950000005,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
 {"type":"assistant/chunk","seq":5,"time":1783950000006,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"advanced-mount","name":"cordis_mount","argumentsDelta":"{\"code\":\"return { name: 'snapshot-marker', apply() {} }\"}"}}}
 {"type":"assistant/chunk","seq":6,"time":1783950000007,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-mount","name":"cordis_mount","arguments":"{\"code\":\"return { name: 'snapshot-marker', apply() {} }\"}"}}}}
 {"type":"assistant/chunk","seq":7,"time":1783950000008,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}}
 {"type":"assistant/chunk","seq":8,"time":1783950000009,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":9,"time":1783957884487,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"advanced-mount","name":"cordis_mount","arguments":"{\"code\":\"return { name: 'snapshot-marker', apply() {} }\"}"}],"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[4,5,6,7,8],"surfaceOp":"append"}
+{"type":"assistant/message","seq":9,"time":1783957884487,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"advanced-mount","name":"cordis_mount","arguments":"{\"code\":\"return { name: 'snapshot-marker', apply() {} }\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[4,5,6,7,8],"surfaceOp":"append"}
 {"type":"tool/call","seq":10,"time":1783957884487,"data":{"turn":1,"step":1,"callId":"advanced-mount","name":"cordis_mount","arguments":"{\"code\":\"return { name: 'snapshot-marker', apply() {} }\"}"}}
 {"type":"tool/result","seq":11,"time":1783957884488,"data":{"turn":1,"step":1,"callId":"advanced-mount","content":[{"type":"text","text":"mounted dyn-1 (plugin \"snapshot-marker\", state: active)"}],"isError":false},"sourceEventSeqs":[10],"surfaceOp":"append"}
 {"type":"step/end","seq":12,"time":1783957884489,"data":{"turn":1,"step":1}}
@@ -18,7 +18,7 @@
 {"type":"assistant/chunk","seq":16,"time":1783950000017,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\":\"return await tools.cordis_inspect({ what: 'dynamic' })\"}"}}}}
 {"type":"assistant/chunk","seq":17,"time":1783950000018,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}}
 {"type":"assistant/chunk","seq":18,"time":1783950000019,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":19,"time":1783957884490,"data":{"turn":1,"step":2,"content":[{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\":\"return await tools.cordis_inspect({ what: 'dynamic' })\"}"}],"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[14,15,16,17,18],"surfaceOp":"append"}
+{"type":"assistant/message","seq":19,"time":1783957884490,"data":{"turn":1,"step":2,"content":[{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\":\"return await tools.cordis_inspect({ what: 'dynamic' })\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[14,15,16,17,18],"surfaceOp":"append"}
 {"type":"tool/call","seq":20,"time":1783957884490,"data":{"turn":1,"step":2,"callId":"advanced-code","name":"run_code","arguments":"{\"code\":\"return await tools.cordis_inspect({ what: 'dynamic' })\"}"}}
 {"type":"tool/code-dispatch","seq":21,"time":1783957884560,"data":{"parentCallId":"advanced-code","subCallId":"advanced-code:code:1","name":"cordis_inspect","arguments":{"what":"dynamic"},"isError":false,"resultSummary":"## dynamic\n- dyn-1: snapshot-marker [active]"}}
 {"type":"tool/result","seq":22,"time":1783957884561,"data":{"turn":1,"step":2,"callId":"advanced-code","content":[{"type":"text","text":"## dynamic\n- dyn-1: snapshot-marker [active]"}],"isError":false,"meta":{"logs":[]}},"sourceEventSeqs":[20],"surfaceOp":"append"}
@@ -29,7 +29,7 @@
 {"type":"assistant/chunk","seq":27,"time":1783950000028,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\"}"}}}}
 {"type":"assistant/chunk","seq":28,"time":1783950000029,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}}
 {"type":"assistant/chunk","seq":29,"time":1783950000030,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":30,"time":1783957884562,"data":{"turn":1,"step":3,"content":[{"type":"tool-call","id":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\"}"}],"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[25,26,27,28,29],"surfaceOp":"append"}
+{"type":"assistant/message","seq":30,"time":1783957884562,"data":{"turn":1,"step":3,"content":[{"type":"tool-call","id":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[25,26,27,28,29],"surfaceOp":"append"}
 {"type":"tool/call","seq":31,"time":1783957884562,"data":{"turn":1,"step":3,"callId":"advanced-direct-child","name":"subagent","arguments":"{\"description\":\"Check direct child\",\"prompt\":\"Reply with exactly DIRECT_CHILD_OK and nothing else.\"}"}}
 {"type":"tool/result","seq":32,"time":1783957884593,"data":{"turn":1,"step":3,"callId":"advanced-direct-child","content":[{"type":"text","text":"DIRECT_CHILD_OK"}],"isError":false},"sourceEventSeqs":[31],"surfaceOp":"append"}
 {"type":"step/end","seq":33,"time":1783957884593,"data":{"turn":1,"step":3}}
@@ -39,7 +39,7 @@
 {"type":"assistant/chunk","seq":37,"time":1783957884594,"data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-acp-snapshot\",\"description\":\"exercise one workflow child through ACP\"}}"}}}}
 {"type":"assistant/chunk","seq":38,"time":1783957884594,"data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}}
 {"type":"assistant/chunk","seq":39,"time":1783957884594,"data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":40,"time":1783957884594,"data":{"turn":1,"step":4,"content":[{"type":"tool-call","id":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-acp-snapshot\",\"description\":\"exercise one workflow child through ACP\"}}"}],"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[35,36,37,38,39],"surfaceOp":"append"}
+{"type":"assistant/message","seq":40,"time":1783957884594,"data":{"turn":1,"step":4,"content":[{"type":"tool-call","id":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-acp-snapshot\",\"description\":\"exercise one workflow child through ACP\"}}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[35,36,37,38,39],"surfaceOp":"append"}
 {"type":"tool/call","seq":41,"time":1783957884594,"data":{"turn":1,"step":4,"callId":"advanced-workflow","name":"workflow","arguments":"{\"script\":\"phase('Delegate')\\nconst reply = await agent('Reply with exactly WORKFLOW_CHILD_OK and nothing else.', { label: 'workflow-child' })\\nreturn { reply }\",\"meta\":{\"name\":\"advanced-acp-snapshot\",\"description\":\"exercise one workflow child through ACP\"}}"}}
 {"type":"tool/result","seq":42,"time":1783957884717,"data":{"turn":1,"step":4,"callId":"advanced-workflow","content":[{"type":"text","text":"workflow \"advanced-acp-snapshot\" completed (1 agent).\nReturn value:\n{\n  \"reply\": \"WORKFLOW_CHILD_OK\"\n}"}],"isError":false},"sourceEventSeqs":[41],"surfaceOp":"append"}
 {"type":"step/end","seq":43,"time":1783957884718,"data":{"turn":1,"step":4}}
@@ -49,7 +49,7 @@
 {"type":"assistant/chunk","seq":47,"time":1783957884719,"data":{"turn":1,"step":5,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-unmount","name":"cordis_unmount","arguments":"{\"id\":\"dyn-1\"}"}}}}
 {"type":"assistant/chunk","seq":48,"time":1783957884719,"data":{"turn":1,"step":5,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}}
 {"type":"assistant/chunk","seq":49,"time":1783957884719,"data":{"turn":1,"step":5,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":50,"time":1783957884719,"data":{"turn":1,"step":5,"content":[{"type":"tool-call","id":"advanced-unmount","name":"cordis_unmount","arguments":"{\"id\":\"dyn-1\"}"}],"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[45,46,47,48,49],"surfaceOp":"append"}
+{"type":"assistant/message","seq":50,"time":1783957884719,"data":{"turn":1,"step":5,"content":[{"type":"tool-call","id":"advanced-unmount","name":"cordis_unmount","arguments":"{\"id\":\"dyn-1\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[45,46,47,48,49],"surfaceOp":"append"}
 {"type":"tool/call","seq":51,"time":1783957884719,"data":{"turn":1,"step":5,"callId":"advanced-unmount","name":"cordis_unmount","arguments":"{\"id\":\"dyn-1\"}"}}
 {"type":"tool/result","seq":52,"time":1783957884719,"data":{"turn":1,"step":5,"callId":"advanced-unmount","content":[{"type":"text","text":"unmounted dyn-1 (plugin \"snapshot-marker\")"}],"isError":false},"sourceEventSeqs":[51],"surfaceOp":"append"}
 {"type":"step/end","seq":53,"time":1783957884719,"data":{"turn":1,"step":5}}
@@ -59,6 +59,6 @@
 {"type":"assistant/chunk","seq":57,"time":1783957884720,"data":{"turn":1,"step":6,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"ADVANCED_ACP_OK"}}}}
 {"type":"assistant/chunk","seq":58,"time":1783957884720,"data":{"turn":1,"step":6,"chunk":{"type":"usage","usage":{"inputTokens":3,"outputTokens":3}}}}
 {"type":"assistant/chunk","seq":59,"time":1783957884720,"data":{"turn":1,"step":6,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":60,"time":1783957884720,"data":{"turn":1,"step":6,"content":[{"type":"text","text":"ADVANCED_ACP_OK"}],"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[55,56,57,58,59],"surfaceOp":"append"}
+{"type":"assistant/message","seq":60,"time":1783957884720,"data":{"turn":1,"step":6,"content":[{"type":"text","text":"ADVANCED_ACP_OK"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[55,56,57,58,59],"surfaceOp":"append"}
 {"type":"step/end","seq":61,"time":1783957884721,"data":{"turn":1,"step":6}}
 {"type":"turn/end","seq":62,"time":1783957884721,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/advanced-toolchain/stdout.golden.jsonl
index bc4da17bb0..45d5628762 100644
--- a/examples/acp-agent/tests/snapshots/advanced-toolchain/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call","toolCallId":"advanced-mount","title":"Mount plugin into live cordis runtime","kind":"execute","status":"in_progress","rawInput":{"code":"return { name: 'snapshot-marker', apply() {} }"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call_update","toolCallId":"advanced-mount","status":"completed","content":[{"type":"content","content":{"type":"text","text":"mounted dyn-1 (plugin \"snapshot-marker\", state: active)"}}]}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call","toolCallId":"advanced-code","title":"return await tools.cordis_inspect({ what: 'dynamic' })","kind":"execute","status":"in_progress","rawInput":"return await tools.cordis_inspect({ what: 'dynamic' })"}}}
diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.golden.md b/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.golden.md
index 5ad12cddff..fb5b48c70c 100644
--- a/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.golden.md
+++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.golden.md
@@ -7,6 +7,8 @@ Verify your work by running the code or tests. Keep answers brief and factual.
 
 Check the [exit code: N] marker on every bash result; investigate failures before moving on.
 
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
 Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
 <!-- dsh-user-approval-policy:never -->
 
@@ -25,7 +27,7 @@ The available tools:
 
 ```ts
 declare const tools: {
-  /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; poll it with `bash_output` and stop it with `bash_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later. */
+  /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later. */
   bash(args: {
     /** The bash command to execute. */
     command: string;
@@ -35,23 +37,13 @@ declare const tools: {
     timeoutMs?: number;
     /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */
     workdir?: string;
-    /** Run in the background and return a task id immediately. No timeout applies. */
+    /** Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies. */
     run_in_background?: boolean;
     /** The wider sandbox mode this command needs. Only valid as a one-shot retry of a command the sandbox just denied; requires justification and user approval. */
     sandbox_permissions?: "workspace-write" | "danger-full-access";
     /** Required with sandbox_permissions: one sentence for the user explaining why this exact command needs the wider access. */
     justification?: string;
   }): Promise<string>;
-  /** Ask the executor to kill a running background bash task by task id. */
-  bash_kill(args: {
-    /** Task id returned by the bash tool. */
-    task_id: string;
-  }): Promise<string>;
-  /** Read new output from a background bash task started with `bash` + `run_in_background`. Returns only output produced since the previous bash_output call, plus the task status. Tasks keep running while you do other work; poll again later for more output. */
-  bash_output(args: {
-    /** Task id returned by the bash tool. */
-    task_id: string;
-  }): Promise<string>;
   /** Inspect the live cordis runtime that is running THIS agent. Read-only. Sections: `services` (every provided ctx service and the plugin fiber that owns it), `plugins` (a flat list of the loaded plugins with their lifecycle states), `tools` (the model-facing tools currently registered, i.e. what you can call), `dynamic` (plugins you mounted via cordis_mount: id, name, state, provided services, awaited services), `api` (method signatures AND argument/return type shapes for every LIVE service — read this before writing plugin code that calls a service), `events` (every harness event with its dispatch mode and exact signature — pick listener targets here). Omit `what` to get all six sections. */
   cordis_inspect(args: {
     /** Limit the report to one section. Omit for all sections. */
@@ -72,19 +64,41 @@ declare const tools: {
     /** The exact skill name from the available skills list. */
     name: string;
   }): Promise<string>;
-  /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. */
+  /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`. */
   subagent(args: {
     /** A short (3-5 word) description of the delegated task, for display. */
     description: string;
     /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */
     prompt: string;
+    /** Run as a background task and return its id; collect with task_output or stop with task_kill. */
+    run_in_background?: boolean;
   }): Promise<string>;
-  /** Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. */
+  /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`. */
   subagent_fork(args: {
     /** A short (3-5 word) description of the delegated task, for display. */
     description: string;
     /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */
     prompt: string;
+    /** Run as a background task and return its id; collect with task_output or stop with task_kill. */
+    run_in_background?: boolean;
+  }): Promise<string>;
+  /** Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops. */
+  task_kill(args: {
+    /** Task id returned by the tool that started the background work. */
+    task_id: string;
+    /** Optional short reason, recorded in the log and forwarded to the task. */
+    reason?: string;
+  }): Promise<string>;
+  /** List your background tasks (running and finished) with their ids, kinds, and statuses. */
+  task_list(args: Record<string, unknown>): Promise<string>;
+  /** Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */
+  task_output(args: {
+    /** Task id returned by the tool that started the background work. */
+    task_id: string;
+    /** Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive. */
+    wait?: boolean;
+    /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */
+    timeout_ms?: number;
   }): Promise<string>;
   /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */
   todo_write(args: {
@@ -96,7 +110,7 @@ declare const tools: {
       status: "pending" | "in_progress" | "completed";
     })[];
   }): Promise<string>;
-  /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), `model` (override). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */
+  /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */
   workflow(args: {
     /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return <json-value>`). */
     script: string;
@@ -114,6 +128,8 @@ declare const tools: {
         title: string;
         /** Optional one-line description of the phase. */
         detail?: string;
+        /** Optional provider override this phase is expected to use. */
+        provider?: string;
         /** Optional model override this phase is expected to use. */
         model?: string;
       }[];
diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.golden.json b/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.golden.json
index e5b5593267..57e78b6345 100644
--- a/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.golden.json
+++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.golden.json
@@ -2,7 +2,7 @@
   "initial": [
     {
       "name": "bash",
-      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; poll it with `bash_output` and stop it with `bash_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
+      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -24,7 +24,7 @@
           },
           "run_in_background": {
             "type": "boolean",
-            "description": "Run in the background and return a task id immediately. No timeout applies."
+            "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
           },
           "sandbox_permissions": {
             "type": "string",
@@ -45,38 +45,6 @@
         ]
       }
     },
-    {
-      "name": "bash_kill",
-      "description": "Ask the executor to kill a running background bash task by task id.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
-    {
-      "name": "bash_output",
-      "description": "Read new output from a background bash task started with `bash` + `run_in_background`. Returns only output produced since the previous bash_output call, plus the task status. Tasks keep running while you do other work; poll again later for more output.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
     {
       "name": "cordis_inspect",
       "description": "Inspect the live cordis runtime that is running THIS agent. Read-only. Sections: `services` (every provided ctx service and the plugin fiber that owns it), `plugins` (a flat list of the loaded plugins with their lifecycle states), `tools` (the model-facing tools currently registered, i.e. what you can call), `dynamic` (plugins you mounted via cordis_mount: id, name, state, provided services, awaited services), `api` (method signatures AND argument/return type shapes for every LIVE service — read this before writing plugin code that calls a service), `events` (every harness event with its dispatch mode and exact signature — pick listener targets here). Omit `what` to get all six sections.",
@@ -164,7 +132,7 @@
     },
     {
       "name": "subagent",
-      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation.",
+      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -175,6 +143,10 @@
           "prompt": {
             "type": "string",
             "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -185,7 +157,7 @@
     },
     {
       "name": "subagent_fork",
-      "description": "Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps.",
+      "description": "Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -196,6 +168,10 @@
           "prompt": {
             "type": "string",
             "description": "The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -204,6 +180,58 @@
         ]
       }
     },
+    {
+      "name": "task_kill",
+      "description": "Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "reason": {
+            "type": "string",
+            "description": "Optional short reason, recorded in the log and forwarded to the task."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
+    {
+      "name": "task_list",
+      "description": "List your background tasks (running and finished) with their ids, kinds, and statuses.",
+      "parameters": {
+        "type": "object",
+        "properties": {}
+      }
+    },
+    {
+      "name": "task_output",
+      "description": "Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "wait": {
+            "type": "boolean",
+            "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+          },
+          "timeout_ms": {
+            "type": "number",
+            "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
     {
       "name": "todo_write",
       "description": "Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).",
@@ -244,7 +272,7 @@
     },
     {
       "name": "workflow",
-      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), `model` (override). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
+      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -282,6 +310,10 @@
                       "type": "string",
                       "description": "Optional one-line description of the phase."
                     },
+                    "provider": {
+                      "type": "string",
+                      "description": "Optional provider override this phase is expected to use."
+                    },
                     "model": {
                       "type": "string",
                       "description": "Optional model override this phase is expected to use."
@@ -310,5 +342,5 @@
       }
     }
   ],
-  "deltas": []
+  "changes": []
 }
diff --git a/examples/acp-agent/tests/snapshots/bash-spill/input.json b/examples/acp-agent/tests/snapshots/bash-spill/input.json
new file mode 100644
index 0000000000..de9b769cf5
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/bash-spill/input.json
@@ -0,0 +1,7 @@
+{
+  "steps": [
+    { "op": "initialize" },
+    { "op": "newSession" },
+    { "op": "prompt", "text": "Use the bash tool to print a large deterministic output, then reply DONE." }
+  ]
+}
diff --git a/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl b/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl
new file mode 100644
index 0000000000..ddd1f06a55
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl
@@ -0,0 +1,23 @@
+{"type":"session","version":0,"id":"{{sessionId}}","createdAt":0,"cwd":"{{cwd}}"}
+{"type":"turn/start","seq":0,"time":0,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
+{"type":"user/message","seq":1,"time":0,"data":{"content":[{"type":"text","text":"Use the bash tool to print a large deterministic output, then reply DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
+{"type":"step/start","seq":2,"time":0,"data":{"turn":1,"step":1}}
+{"type":"request/header","seq":3,"time":0,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"assistant/chunk","seq":4,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
+{"type":"assistant/chunk","seq":5,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"call_spill","name":"bash","argumentsDelta":"{\"command\":\"node -e \\\"process.stdout.write('SPILL_START-' + 'x'.repeat(2000) + '-SPILL_END')\\\"\",\"description\":\"Print large deterministic output\"}"}}}
+{"type":"assistant/chunk","seq":6,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_spill","name":"bash","arguments":"{\"command\":\"node -e \\\"process.stdout.write('SPILL_START-' + 'x'.repeat(2000) + '-SPILL_END')\\\"\",\"description\":\"Print large deterministic output\"}"}}}}
+{"type":"assistant/chunk","seq":7,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
+{"type":"assistant/chunk","seq":8,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
+{"type":"assistant/message","seq":9,"time":0,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"call_spill","name":"bash","arguments":"{\"command\":\"node -e \\\"process.stdout.write('SPILL_START-' + 'x'.repeat(2000) + '-SPILL_END')\\\"\",\"description\":\"Print large deterministic output\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[4,5,6,7,8],"surfaceOp":"append"}
+{"type":"tool/call","seq":10,"time":0,"data":{"turn":1,"step":1,"callId":"call_spill","name":"bash","arguments":"{\"command\":\"node -e \\\"process.stdout.write('SPILL_START-' + 'x'.repeat(2000) + '-SPILL_END')\\\"\",\"description\":\"Print large deterministic output\"}"}}
+{"type":"tool/result","seq":11,"time":0,"data":{"turn":1,"step":1,"callId":"call_spill","content":[{"type":"text","text":"SPILL_START-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx-SPILL_END\n\n(Omitted 1417 bytes. Full formatted result stored at: /tmp/dsh-acp-snapshot-spill/session-ad78080217cd/dccd97e3c558-bash.txt. Use read with offset/limit, or grep this path to search within it.)"}],"isError":false},"sourceEventSeqs":[10],"surfaceOp":"append"}
+{"type":"step/end","seq":12,"time":0,"data":{"turn":1,"step":1}}
+{"type":"step/start","seq":13,"time":0,"data":{"turn":1,"step":2}}
+{"type":"assistant/chunk","seq":14,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}
+{"type":"assistant/chunk","seq":15,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":0,"text":"DONE"}}}
+{"type":"assistant/chunk","seq":16,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"DONE"}}}}
+{"type":"assistant/chunk","seq":17,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":2}}}}
+{"type":"assistant/chunk","seq":18,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
+{"type":"assistant/message","seq":19,"time":0,"data":{"turn":1,"step":2,"content":[{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":2}},"sourceEventSeqs":[14,15,16,17,18],"surfaceOp":"append"}
+{"type":"step/end","seq":20,"time":0,"data":{"turn":1,"step":2}}
+{"type":"turn/end","seq":21,"time":0,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/bash-spill/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/bash-spill/stdout.golden.jsonl
new file mode 100644
index 0000000000..ad6c396981
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/bash-spill/stdout.golden.jsonl
@@ -0,0 +1,6 @@
+{"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call","toolCallId":"call_spill","title":"node -e \"process.stdout.write('SPILL_START-' + 'x'.repeat(2000) + '-SPILL_END')\"","kind":"execute","status":"in_progress","rawInput":"node -e \"process.stdout.write('SPILL_START-' + 'x'.repeat(2000) + '-SPILL_END')\"","content":[{"type":"content","content":{"type":"text","text":"Print large deterministic output"}}]}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call_update","toolCallId":"call_spill","status":"completed","content":[{"type":"content","content":{"type":"text","text":"```console\nSPILL_START-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx-SPILL_END\n\n(Omitted 1417 bytes. Full formatted result stored at: {{spillLocator:bash.txt}}. Use read with offset/limit, or grep this path to search within it.)\n```"}}]}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"DONE"}}}}
+{"jsonrpc":"2.0","id":3,"result":{"stopReason":"end_turn"}}
diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/session.jsonl b/examples/acp-agent/tests/snapshots/both-mode-turn/session.jsonl
index b68703ad48..0ee3e67e2b 100644
--- a/examples/acp-agent/tests/snapshots/both-mode-turn/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/both-mode-turn/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783611774323,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783611774323,"data":{"content":[{"type":"text","text":"Call the run_code tool (NOT the native bash tool directly) with a program that runs exactly `echo BOTH_OK` via tools.bash and returns its output. Then reply with that output only and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783611774324,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783611774325,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783611774325,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783611774792,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783611774792,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783611774879,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -84,7 +84,7 @@
 {"type":"assistant/chunk","seq":82,"time":1783611775498,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_AZFzvUwuC4vAUoICrfke5147","name":"run_code","arguments":"{\"code\": \"const result = await tools.bash({ command: \\\"echo BOTH_OK\\\", description: \\\"Echo BOTH_OK\\\" });\\nreturn result;\"}"}}}}
 {"type":"assistant/chunk","seq":83,"time":1783611775498,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":5530,"outputTokens":108,"cacheReadTokens":0,"reasoningTokens":37}}}}
 {"type":"assistant/chunk","seq":84,"time":1783611775498,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":85,"time":1783611775503,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the `run_code` tool to execute a program that calls `tools.bash` with the command `echo BOTH_OK` and returns its output."},{"type":"tool-call","id":"call_00_AZFzvUwuC4vAUoICrfke5147","name":"run_code","arguments":"{\"code\": \"const result = await tools.bash({ command: \\\"echo BOTH_OK\\\", description: \\\"Echo BOTH_OK\\\" });\\nreturn result;\"}"}],"usage":{"inputTokens":5530,"outputTokens":108,"cacheReadTokens":0,"reasoningTokens":37}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84],"surfaceOp":"append"}
+{"type":"assistant/message","seq":85,"time":1783611775503,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the `run_code` tool to execute a program that calls `tools.bash` with the command `echo BOTH_OK` and returns its output."},{"type":"tool-call","id":"call_00_AZFzvUwuC4vAUoICrfke5147","name":"run_code","arguments":"{\"code\": \"const result = await tools.bash({ command: \\\"echo BOTH_OK\\\", description: \\\"Echo BOTH_OK\\\" });\\nreturn result;\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":5530,"outputTokens":108,"cacheReadTokens":0,"reasoningTokens":37}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84],"surfaceOp":"append"}
 {"type":"tool/call","seq":86,"time":1783611775504,"data":{"turn":1,"step":1,"callId":"call_00_AZFzvUwuC4vAUoICrfke5147","name":"run_code","arguments":"{\"code\": \"const result = await tools.bash({ command: \\\"echo BOTH_OK\\\", description: \\\"Echo BOTH_OK\\\" });\\nreturn result;\"}"}}
 {"type":"tool/code-dispatch","seq":87,"time":1783611775590,"data":{"parentCallId":"call_00_AZFzvUwuC4vAUoICrfke5147","subCallId":"call_00_AZFzvUwuC4vAUoICrfke5147:code:1","name":"bash","arguments":{"command":"echo BOTH_OK","description":"Echo BOTH_OK"},"isError":false,"resultSummary":"BOTH_OK\n"}}
 {"type":"tool/result","seq":88,"time":1783611775592,"data":{"turn":1,"step":1,"callId":"call_00_AZFzvUwuC4vAUoICrfke5147","content":[{"type":"text","text":"BOTH_OK\n"}],"isError":false,"meta":{"logs":[]}},"sourceEventSeqs":[86],"surfaceOp":"append"}
@@ -116,6 +116,6 @@
 {"type":"assistant/chunk","seq":114,"time":1783611776440,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"BOTH_OK"}}}}
 {"type":"assistant/chunk","seq":115,"time":1783611776440,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":22,"outputTokens":21,"cacheReadTokens":5632,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":116,"time":1783611776440,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":117,"time":1783611776441,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The output is \"BOTH_OK\". I need to reply with that output only."},{"type":"text","text":"BOTH_OK"}],"usage":{"inputTokens":22,"outputTokens":21,"cacheReadTokens":5632,"reasoningTokens":17}},"sourceEventSeqs":[91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116],"surfaceOp":"append"}
+{"type":"assistant/message","seq":117,"time":1783611776441,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The output is \"BOTH_OK\". I need to reply with that output only."},{"type":"text","text":"BOTH_OK"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":22,"outputTokens":21,"cacheReadTokens":5632,"reasoningTokens":17}},"sourceEventSeqs":[91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116],"surfaceOp":"append"}
 {"type":"step/end","seq":118,"time":1783611776441,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":119,"time":1783611776441,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/both-mode-turn/stdout.golden.jsonl
index 577f10445b..7b2dbd604c 100644
--- a/examples/acp-agent/tests/snapshots/both-mode-turn/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/both-mode-turn/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.golden.md b/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.golden.md
index ed60d52258..f1a3b9ff92 100644
--- a/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.golden.md
+++ b/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.golden.md
@@ -7,6 +7,8 @@ Verify your work by running the code or tests. Keep answers brief and factual.
 
 Check the [exit code: N] marker on every bash result; investigate failures before moving on.
 
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
 Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
 <!-- dsh-user-approval-policy:never -->
 
@@ -25,7 +27,7 @@ The available tools:
 
 ```ts
 declare const tools: {
-  /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; poll it with `bash_output` and stop it with `bash_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later. */
+  /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later. */
   bash(args: {
     /** The bash command to execute. */
     command: string;
@@ -35,41 +37,53 @@ declare const tools: {
     timeoutMs?: number;
     /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */
     workdir?: string;
-    /** Run in the background and return a task id immediately. No timeout applies. */
+    /** Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies. */
     run_in_background?: boolean;
     /** The wider sandbox mode this command needs. Only valid as a one-shot retry of a command the sandbox just denied; requires justification and user approval. */
     sandbox_permissions?: "workspace-write" | "danger-full-access";
     /** Required with sandbox_permissions: one sentence for the user explaining why this exact command needs the wider access. */
     justification?: string;
   }): Promise<string>;
-  /** Ask the executor to kill a running background bash task by task id. */
-  bash_kill(args: {
-    /** Task id returned by the bash tool. */
-    task_id: string;
-  }): Promise<string>;
-  /** Read new output from a background bash task started with `bash` + `run_in_background`. Returns only output produced since the previous bash_output call, plus the task status. Tasks keep running while you do other work; poll again later for more output. */
-  bash_output(args: {
-    /** Task id returned by the bash tool. */
-    task_id: string;
-  }): Promise<string>;
   /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */
   skill(args: {
     /** The exact skill name from the available skills list. */
     name: string;
   }): Promise<string>;
-  /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. */
+  /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`. */
   subagent(args: {
     /** A short (3-5 word) description of the delegated task, for display. */
     description: string;
     /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */
     prompt: string;
+    /** Run as a background task and return its id; collect with task_output or stop with task_kill. */
+    run_in_background?: boolean;
   }): Promise<string>;
-  /** Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. */
+  /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`. */
   subagent_fork(args: {
     /** A short (3-5 word) description of the delegated task, for display. */
     description: string;
     /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */
     prompt: string;
+    /** Run as a background task and return its id; collect with task_output or stop with task_kill. */
+    run_in_background?: boolean;
+  }): Promise<string>;
+  /** Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops. */
+  task_kill(args: {
+    /** Task id returned by the tool that started the background work. */
+    task_id: string;
+    /** Optional short reason, recorded in the log and forwarded to the task. */
+    reason?: string;
+  }): Promise<string>;
+  /** List your background tasks (running and finished) with their ids, kinds, and statuses. */
+  task_list(args: Record<string, unknown>): Promise<string>;
+  /** Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */
+  task_output(args: {
+    /** Task id returned by the tool that started the background work. */
+    task_id: string;
+    /** Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive. */
+    wait?: boolean;
+    /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */
+    timeout_ms?: number;
   }): Promise<string>;
   /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */
   todo_write(args: {
@@ -81,7 +95,7 @@ declare const tools: {
       status: "pending" | "in_progress" | "completed";
     })[];
   }): Promise<string>;
-  /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), `model` (override). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */
+  /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */
   workflow(args: {
     /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return <json-value>`). */
     script: string;
@@ -99,6 +113,8 @@ declare const tools: {
         title: string;
         /** Optional one-line description of the phase. */
         detail?: string;
+        /** Optional provider override this phase is expected to use. */
+        provider?: string;
         /** Optional model override this phase is expected to use. */
         model?: string;
       }[];
diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.golden.json b/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.golden.json
index 3c3e5bb0b2..edf1a7c001 100644
--- a/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.golden.json
+++ b/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.golden.json
@@ -2,7 +2,7 @@
   "initial": [
     {
       "name": "bash",
-      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; poll it with `bash_output` and stop it with `bash_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
+      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -24,7 +24,7 @@
           },
           "run_in_background": {
             "type": "boolean",
-            "description": "Run in the background and return a task id immediately. No timeout applies."
+            "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
           },
           "sandbox_permissions": {
             "type": "string",
@@ -45,38 +45,6 @@
         ]
       }
     },
-    {
-      "name": "bash_kill",
-      "description": "Ask the executor to kill a running background bash task by task id.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
-    {
-      "name": "bash_output",
-      "description": "Read new output from a background bash task started with `bash` + `run_in_background`. Returns only output produced since the previous bash_output call, plus the task status. Tasks keep running while you do other work; poll again later for more output.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
     {
       "name": "run_code",
       "description": "Execute a TypeScript program against the available tools. Write the BODY of an async function (erasable syntax only; top-level `await` and `return` work) and call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return comes back — curate it.",
@@ -111,7 +79,7 @@
     },
     {
       "name": "subagent",
-      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation.",
+      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -122,6 +90,10 @@
           "prompt": {
             "type": "string",
             "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -132,7 +104,7 @@
     },
     {
       "name": "subagent_fork",
-      "description": "Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps.",
+      "description": "Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -143,6 +115,10 @@
           "prompt": {
             "type": "string",
             "description": "The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -151,6 +127,58 @@
         ]
       }
     },
+    {
+      "name": "task_kill",
+      "description": "Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "reason": {
+            "type": "string",
+            "description": "Optional short reason, recorded in the log and forwarded to the task."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
+    {
+      "name": "task_list",
+      "description": "List your background tasks (running and finished) with their ids, kinds, and statuses.",
+      "parameters": {
+        "type": "object",
+        "properties": {}
+      }
+    },
+    {
+      "name": "task_output",
+      "description": "Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "wait": {
+            "type": "boolean",
+            "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+          },
+          "timeout_ms": {
+            "type": "number",
+            "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
     {
       "name": "todo_write",
       "description": "Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).",
@@ -191,7 +219,7 @@
     },
     {
       "name": "workflow",
-      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), `model` (override). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
+      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -229,6 +257,10 @@
                       "type": "string",
                       "description": "Optional one-line description of the phase."
                     },
+                    "provider": {
+                      "type": "string",
+                      "description": "Optional provider override this phase is expected to use."
+                    },
                     "model": {
                       "type": "string",
                       "description": "Optional model override this phase is expected to use."
@@ -257,5 +289,5 @@
       }
     }
   ],
-  "deltas": []
+  "changes": []
 }
diff --git a/examples/acp-agent/tests/snapshots/cancel/session.jsonl b/examples/acp-agent/tests/snapshots/cancel/session.jsonl
index 7b2f5adff1..347951db62 100644
--- a/examples/acp-agent/tests/snapshots/cancel/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/cancel/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":0,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":0,"data":{"content":[{"type":"text","text":"Start a long task; this turn will be cancelled mid-stream."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":0,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":0,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":0,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}
 {"type":"assistant/chunk","seq":5,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"text-delta","index":0,"text":"partial"}}}
 {"type":"step/end","seq":6,"time":0,"data":{"turn":1,"step":1}}
diff --git a/examples/acp-agent/tests/snapshots/cancel/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/cancel/stdout.golden.jsonl
index 60235cac75..d31c8fdcfb 100644
--- a/examples/acp-agent/tests/snapshots/cancel/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/cancel/stdout.golden.jsonl
@@ -1,4 +1,4 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"partial"}}}}
 {"jsonrpc":"2.0","id":3,"result":{"stopReason":"cancelled"}}
diff --git a/examples/acp-agent/tests/snapshots/code-mode-turn/session.jsonl b/examples/acp-agent/tests/snapshots/code-mode-turn/session.jsonl
index 8dac383517..0fe57055e8 100644
--- a/examples/acp-agent/tests/snapshots/code-mode-turn/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/code-mode-turn/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783611771394,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783611771394,"data":{"content":[{"type":"text","text":"Using ONE run_code program: call the bash tool twice — exactly `echo CODE_ONE` then exactly `echo CODE_TWO` — and return the two outputs joined with a plus sign. Then reply with that joined string only and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783611771396,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783611771396,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783611771396,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783611771869,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783611771869,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783611771978,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -108,7 +108,7 @@
 {"type":"assistant/chunk","seq":106,"time":1783611772836,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_DRxnM6R1TThfDwcudW0f2050","name":"run_code","arguments":"{\"code\": \"const out1 = await tools.bash({ command: \\\"echo CODE_ONE\\\", description: \\\"First echo\\\" });\\nconst out2 = await tools.bash({ command: \\\"echo CODE_TWO\\\", description: \\\"Second echo\\\" });\\nreturn out1.trim() + \\\"+\\\" + out2.trim();\"}"}}}}
 {"type":"assistant/chunk","seq":107,"time":1783611772836,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3009,"outputTokens":133,"cacheReadTokens":0,"reasoningTokens":29}}}}
 {"type":"assistant/chunk","seq":108,"time":1783611772836,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":109,"time":1783611772840,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants a single run_code program that calls bash twice, then returns the two outputs joined with a plus sign. Let me write this."},{"type":"tool-call","id":"call_00_DRxnM6R1TThfDwcudW0f2050","name":"run_code","arguments":"{\"code\": \"const out1 = await tools.bash({ command: \\\"echo CODE_ONE\\\", description: \\\"First echo\\\" });\\nconst out2 = await tools.bash({ command: \\\"echo CODE_TWO\\\", description: \\\"Second echo\\\" });\\nreturn out1.trim() + \\\"+\\\" + out2.trim();\"}"}],"usage":{"inputTokens":3009,"outputTokens":133,"cacheReadTokens":0,"reasoningTokens":29}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108],"surfaceOp":"append"}
+{"type":"assistant/message","seq":109,"time":1783611772840,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants a single run_code program that calls bash twice, then returns the two outputs joined with a plus sign. Let me write this."},{"type":"tool-call","id":"call_00_DRxnM6R1TThfDwcudW0f2050","name":"run_code","arguments":"{\"code\": \"const out1 = await tools.bash({ command: \\\"echo CODE_ONE\\\", description: \\\"First echo\\\" });\\nconst out2 = await tools.bash({ command: \\\"echo CODE_TWO\\\", description: \\\"Second echo\\\" });\\nreturn out1.trim() + \\\"+\\\" + out2.trim();\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3009,"outputTokens":133,"cacheReadTokens":0,"reasoningTokens":29}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108],"surfaceOp":"append"}
 {"type":"tool/call","seq":110,"time":1783611772840,"data":{"turn":1,"step":1,"callId":"call_00_DRxnM6R1TThfDwcudW0f2050","name":"run_code","arguments":"{\"code\": \"const out1 = await tools.bash({ command: \\\"echo CODE_ONE\\\", description: \\\"First echo\\\" });\\nconst out2 = await tools.bash({ command: \\\"echo CODE_TWO\\\", description: \\\"Second echo\\\" });\\nreturn out1.trim() + \\\"+\\\" + out2.trim();\"}"}}
 {"type":"tool/code-dispatch","seq":111,"time":1783611772933,"data":{"parentCallId":"call_00_DRxnM6R1TThfDwcudW0f2050","subCallId":"call_00_DRxnM6R1TThfDwcudW0f2050:code:1","name":"bash","arguments":{"command":"echo CODE_ONE","description":"First echo"},"isError":false,"resultSummary":"CODE_ONE\n"}}
 {"type":"tool/code-dispatch","seq":112,"time":1783611772936,"data":{"parentCallId":"call_00_DRxnM6R1TThfDwcudW0f2050","subCallId":"call_00_DRxnM6R1TThfDwcudW0f2050:code:2","name":"bash","arguments":{"command":"echo CODE_TWO","description":"Second echo"},"isError":false,"resultSummary":"CODE_TWO\n"}}
@@ -145,6 +145,6 @@
 {"type":"assistant/chunk","seq":143,"time":1783611773686,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"CODE_ONE+CODE_TWO"}}}}
 {"type":"assistant/chunk","seq":144,"time":1783611773686,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":89,"outputTokens":25,"cacheReadTokens":3072,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":145,"time":1783611773686,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":146,"time":1783611773687,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The output is exactly what the user asked for: CODE_ONE+CODE_TWO"},{"type":"text","text":"CODE_ONE+CODE_TWO"}],"usage":{"inputTokens":89,"outputTokens":25,"cacheReadTokens":3072,"reasoningTokens":17}},"sourceEventSeqs":[116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145],"surfaceOp":"append"}
+{"type":"assistant/message","seq":146,"time":1783611773687,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The output is exactly what the user asked for: CODE_ONE+CODE_TWO"},{"type":"text","text":"CODE_ONE+CODE_TWO"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":89,"outputTokens":25,"cacheReadTokens":3072,"reasoningTokens":17}},"sourceEventSeqs":[116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145],"surfaceOp":"append"}
 {"type":"step/end","seq":147,"time":1783611773687,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":148,"time":1783611773687,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/code-mode-turn/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/code-mode-turn/stdout.golden.jsonl
index bfa379815e..31ec5df39a 100644
--- a/examples/acp-agent/tests/snapshots/code-mode-turn/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/code-mode-turn/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.golden.md b/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.golden.md
index ed60d52258..f1a3b9ff92 100644
--- a/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.golden.md
+++ b/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.golden.md
@@ -7,6 +7,8 @@ Verify your work by running the code or tests. Keep answers brief and factual.
 
 Check the [exit code: N] marker on every bash result; investigate failures before moving on.
 
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
 Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
 <!-- dsh-user-approval-policy:never -->
 
@@ -25,7 +27,7 @@ The available tools:
 
 ```ts
 declare const tools: {
-  /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; poll it with `bash_output` and stop it with `bash_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later. */
+  /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later. */
   bash(args: {
     /** The bash command to execute. */
     command: string;
@@ -35,41 +37,53 @@ declare const tools: {
     timeoutMs?: number;
     /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */
     workdir?: string;
-    /** Run in the background and return a task id immediately. No timeout applies. */
+    /** Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies. */
     run_in_background?: boolean;
     /** The wider sandbox mode this command needs. Only valid as a one-shot retry of a command the sandbox just denied; requires justification and user approval. */
     sandbox_permissions?: "workspace-write" | "danger-full-access";
     /** Required with sandbox_permissions: one sentence for the user explaining why this exact command needs the wider access. */
     justification?: string;
   }): Promise<string>;
-  /** Ask the executor to kill a running background bash task by task id. */
-  bash_kill(args: {
-    /** Task id returned by the bash tool. */
-    task_id: string;
-  }): Promise<string>;
-  /** Read new output from a background bash task started with `bash` + `run_in_background`. Returns only output produced since the previous bash_output call, plus the task status. Tasks keep running while you do other work; poll again later for more output. */
-  bash_output(args: {
-    /** Task id returned by the bash tool. */
-    task_id: string;
-  }): Promise<string>;
   /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */
   skill(args: {
     /** The exact skill name from the available skills list. */
     name: string;
   }): Promise<string>;
-  /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. */
+  /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`. */
   subagent(args: {
     /** A short (3-5 word) description of the delegated task, for display. */
     description: string;
     /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */
     prompt: string;
+    /** Run as a background task and return its id; collect with task_output or stop with task_kill. */
+    run_in_background?: boolean;
   }): Promise<string>;
-  /** Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. */
+  /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`. */
   subagent_fork(args: {
     /** A short (3-5 word) description of the delegated task, for display. */
     description: string;
     /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */
     prompt: string;
+    /** Run as a background task and return its id; collect with task_output or stop with task_kill. */
+    run_in_background?: boolean;
+  }): Promise<string>;
+  /** Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops. */
+  task_kill(args: {
+    /** Task id returned by the tool that started the background work. */
+    task_id: string;
+    /** Optional short reason, recorded in the log and forwarded to the task. */
+    reason?: string;
+  }): Promise<string>;
+  /** List your background tasks (running and finished) with their ids, kinds, and statuses. */
+  task_list(args: Record<string, unknown>): Promise<string>;
+  /** Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */
+  task_output(args: {
+    /** Task id returned by the tool that started the background work. */
+    task_id: string;
+    /** Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive. */
+    wait?: boolean;
+    /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */
+    timeout_ms?: number;
   }): Promise<string>;
   /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */
   todo_write(args: {
@@ -81,7 +95,7 @@ declare const tools: {
       status: "pending" | "in_progress" | "completed";
     })[];
   }): Promise<string>;
-  /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), `model` (override). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */
+  /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */
   workflow(args: {
     /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return <json-value>`). */
     script: string;
@@ -99,6 +113,8 @@ declare const tools: {
         title: string;
         /** Optional one-line description of the phase. */
         detail?: string;
+        /** Optional provider override this phase is expected to use. */
+        provider?: string;
         /** Optional model override this phase is expected to use. */
         model?: string;
       }[];
diff --git a/examples/acp-agent/tests/snapshots/code-mode-turn/tool-schemas.golden.json b/examples/acp-agent/tests/snapshots/code-mode-turn/tool-schemas.golden.json
index 61875ceb85..c2289b4e19 100644
--- a/examples/acp-agent/tests/snapshots/code-mode-turn/tool-schemas.golden.json
+++ b/examples/acp-agent/tests/snapshots/code-mode-turn/tool-schemas.golden.json
@@ -17,5 +17,5 @@
       }
     }
   ],
-  "deltas": []
+  "changes": []
 }
diff --git a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/input.json b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/input.json
new file mode 100644
index 0000000000..498816c5e4
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/input.json
@@ -0,0 +1,7 @@
+{
+  "steps": [
+    { "op": "initialize" },
+    { "op": "newSession" },
+    { "op": "prompt", "text": "Using ONE run_code program, call tools.read on nested/task.txt. After the program finishes, answer the workspace handshake question using the newly discovered instructions: What is the Code Mode workspace handshake?" }
+  ]
+}
diff --git a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/session.jsonl b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/session.jsonl
new file mode 100644
index 0000000000..f744c1b000
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/session.jsonl
@@ -0,0 +1,189 @@
+{"type":"session","version":0,"id":"65fbb8a6-624c-4d6a-bf5d-a7a7d14f2b49","createdAt":1783921765266,"cwd":"/var/folders/_g/59jgff8x2gqd39f5vy1wnbfc0000gn/T/acp-snap-cwd-uorU26"}
+{"type":"turn/start","seq":0,"time":1783921765269,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
+{"type":"user/message","seq":1,"time":1783921765269,"data":{"content":[{"type":"text","text":"Using ONE run_code program, call tools.read on nested/task.txt. After the program finishes, answer the workspace handshake question using the newly discovered instructions: What is the Code Mode workspace handshake?"}],"source":{"kind":"user"}},"surfaceOp":"append"}
+{"type":"step/start","seq":2,"time":1783921765275,"data":{"turn":1,"step":1}}
+{"type":"request/header","seq":3,"time":1783921765275,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}","messagePrefix":[{"role":"user","content":[{"type":"text","text":"<system-reminder>\nThe following workspace instructions may be relevant to your work. Use them as guidance when applicable. More specific instructions take precedence over broader ones. They do not override system, developer, or direct user instructions.\n\nInstructions from: AGENTS.md\n\nWorkspace snapshot root instruction.\n\n</system-reminder>"}]}]},"reason":"initial"}}
+{"type":"assistant/chunk","seq":4,"time":1783921766287,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
+{"type":"assistant/chunk","seq":5,"time":1783921766287,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
+{"type":"assistant/chunk","seq":6,"time":1783921766483,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
+{"type":"assistant/chunk","seq":7,"time":1783921766519,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" wants"}}}
+{"type":"assistant/chunk","seq":8,"time":1783921766520,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" me"}}}
+{"type":"assistant/chunk","seq":9,"time":1783921766520,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" to"}}}
+{"type":"assistant/chunk","seq":10,"time":1783921766520,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" write"}}}
+{"type":"assistant/chunk","seq":11,"time":1783921766537,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" a"}}}
+{"type":"assistant/chunk","seq":12,"time":1783921766538,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" single"}}}
+{"type":"assistant/chunk","seq":13,"time":1783921766538,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" run"}}}
+{"type":"assistant/chunk","seq":14,"time":1783921766573,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"_code"}}}
+{"type":"assistant/chunk","seq":15,"time":1783921766573,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" program"}}}
+{"type":"assistant/chunk","seq":16,"time":1783921766574,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" that"}}}
+{"type":"assistant/chunk","seq":17,"time":1783921766574,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" reads"}}}
+{"type":"assistant/chunk","seq":18,"time":1783921766598,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" a"}}}
+{"type":"assistant/chunk","seq":19,"time":1783921766598,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" file"}}}
+{"type":"assistant/chunk","seq":20,"time":1783921766599,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" called"}}}
+{"type":"assistant/chunk","seq":21,"time":1783921766624,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" nested"}}}
+{"type":"assistant/chunk","seq":22,"time":1783921766654,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"/t"}}}
+{"type":"assistant/chunk","seq":23,"time":1783921766654,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"ask"}}}
+{"type":"assistant/chunk","seq":24,"time":1783921766654,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":".txt"}}}
+{"type":"assistant/chunk","seq":25,"time":1783921766654,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" and"}}}
+{"type":"assistant/chunk","seq":26,"time":1783921766655,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" then"}}}
+{"type":"assistant/chunk","seq":27,"time":1783921766684,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" answer"}}}
+{"type":"assistant/chunk","seq":28,"time":1783921766717,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" the"}}}
+{"type":"assistant/chunk","seq":29,"time":1783921766717,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" hand"}}}
+{"type":"assistant/chunk","seq":30,"time":1783921766717,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"shake"}}}
+{"type":"assistant/chunk","seq":31,"time":1783921766743,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" question"}}}
+{"type":"assistant/chunk","seq":32,"time":1783921766743,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" based"}}}
+{"type":"assistant/chunk","seq":33,"time":1783921766743,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" on"}}}
+{"type":"assistant/chunk","seq":34,"time":1783921766743,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" what"}}}
+{"type":"assistant/chunk","seq":35,"time":1783921766776,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"'s"}}}
+{"type":"assistant/chunk","seq":36,"time":1783921766776,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" in"}}}
+{"type":"assistant/chunk","seq":37,"time":1783921766798,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" that"}}}
+{"type":"assistant/chunk","seq":38,"time":1783921766799,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" file"}}}
+{"type":"assistant/chunk","seq":39,"time":1783921766799,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"."}}}
+{"type":"assistant/chunk","seq":40,"time":1783921766799,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" Let"}}}
+{"type":"assistant/chunk","seq":41,"time":1783921766799,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" me"}}}
+{"type":"assistant/chunk","seq":42,"time":1783921766799,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" start"}}}
+{"type":"assistant/chunk","seq":43,"time":1783921766845,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" by"}}}
+{"type":"assistant/chunk","seq":44,"time":1783921766845,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" reading"}}}
+{"type":"assistant/chunk","seq":45,"time":1783921766845,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" the"}}}
+{"type":"assistant/chunk","seq":46,"time":1783921766845,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" file"}}}
+{"type":"assistant/chunk","seq":47,"time":1783921766845,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"."}}}
+{"type":"assistant/chunk","seq":48,"time":1783921766926,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":1,"blockType":"tool-call"}}}
+{"type":"assistant/chunk","seq":49,"time":1783921766926,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":""}}}
+{"type":"assistant/chunk","seq":50,"time":1783921766952,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"{"}}}
+{"type":"assistant/chunk","seq":51,"time":1783921766952,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"\""}}}
+{"type":"assistant/chunk","seq":52,"time":1783921766952,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"code"}}}
+{"type":"assistant/chunk","seq":53,"time":1783921766952,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"\""}}}
+{"type":"assistant/chunk","seq":54,"time":1783921766952,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":": "}}}
+{"type":"assistant/chunk","seq":55,"time":1783921766978,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"\""}}}
+{"type":"assistant/chunk","seq":56,"time":1783921766978,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"const"}}}
+{"type":"assistant/chunk","seq":57,"time":1783921766978,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":" content"}}}
+{"type":"assistant/chunk","seq":58,"time":1783921767011,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":" ="}}}
+{"type":"assistant/chunk","seq":59,"time":1783921767011,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":" await"}}}
+{"type":"assistant/chunk","seq":60,"time":1783921767011,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":" tools"}}}
+{"type":"assistant/chunk","seq":61,"time":1783921767011,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":".read"}}}
+{"type":"assistant/chunk","seq":62,"time":1783921767011,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"({"}}}
+{"type":"assistant/chunk","seq":63,"time":1783921767050,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":" file"}}}
+{"type":"assistant/chunk","seq":64,"time":1783921767050,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"_path"}}}
+{"type":"assistant/chunk","seq":65,"time":1783921767050,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":":"}}}
+{"type":"assistant/chunk","seq":66,"time":1783921767050,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":" \\\""}}}
+{"type":"assistant/chunk","seq":67,"time":1783921767050,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"n"}}}
+{"type":"assistant/chunk","seq":68,"time":1783921767050,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"ested"}}}
+{"type":"assistant/chunk","seq":69,"time":1783921767068,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"/t"}}}
+{"type":"assistant/chunk","seq":70,"time":1783921767068,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"ask"}}}
+{"type":"assistant/chunk","seq":71,"time":1783921767068,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":".txt"}}}
+{"type":"assistant/chunk","seq":72,"time":1783921767068,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"\\\""}}}
+{"type":"assistant/chunk","seq":73,"time":1783921767068,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":" });\\n"}}}
+{"type":"assistant/chunk","seq":74,"time":1783921767068,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"return"}}}
+{"type":"assistant/chunk","seq":75,"time":1783921767095,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":" content"}}}
+{"type":"assistant/chunk","seq":76,"time":1783921767095,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":";"}}}
+{"type":"assistant/chunk","seq":77,"time":1783921767095,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"\""}}}
+{"type":"assistant/chunk","seq":78,"time":1783921767121,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","argumentsDelta":"}"}}}
+{"type":"assistant/chunk","seq":79,"time":1783921767205,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"reasoning","text":"The user wants me to write a single run_code program that reads a file called nested/task.txt and then answer the handshake question based on what's in that file. Let me start by reading the file."}}}}
+{"type":"assistant/chunk","seq":80,"time":1783921767205,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","arguments":"{\"code\": \"const content = await tools.read({ file_path: \\\"nested/task.txt\\\" });\\nreturn content;\"}"}}}}
+{"type":"assistant/chunk","seq":81,"time":1783921767205,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3195,"outputTokens":106,"cacheReadTokens":0,"reasoningTokens":43}}}}
+{"type":"assistant/chunk","seq":82,"time":1783921767205,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
+{"type":"assistant/message","seq":83,"time":1783921767208,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to write a single run_code program that reads a file called nested/task.txt and then answer the handshake question based on what's in that file. Let me start by reading the file."},{"type":"tool-call","id":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","arguments":"{\"code\": \"const content = await tools.read({ file_path: \\\"nested/task.txt\\\" });\\nreturn content;\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3195,"outputTokens":106,"cacheReadTokens":0,"reasoningTokens":43}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82],"surfaceOp":"append"}
+{"type":"tool/call","seq":84,"time":1783921767208,"data":{"turn":1,"step":1,"callId":"call_00_6APApmaKLRDlXKMdIcWL5139","name":"run_code","arguments":"{\"code\": \"const content = await tools.read({ file_path: \\\"nested/task.txt\\\" });\\nreturn content;\"}"}}
+{"type":"tool/code-dispatch","seq":85,"time":1783921767270,"data":{"parentCallId":"call_00_6APApmaKLRDlXKMdIcWL5139","subCallId":"call_00_6APApmaKLRDlXKMdIcWL5139:code:1","name":"read","arguments":{"file_path":"nested/task.txt"},"isError":false,"resultSummary":"<path>./nested/task.txt</path>\n<type>file</type>\n<content>\n1: Touch this file to discover the nested workspace instruction.\n\n(End of file - total 1 lines)\n</content>"}}
+{"type":"tool/result","seq":86,"time":1783921767271,"data":{"turn":1,"step":1,"callId":"call_00_6APApmaKLRDlXKMdIcWL5139","content":[{"type":"text","text":"<path>/var/folders/_g/59jgff8x2gqd39f5vy1wnbfc0000gn/T/acp-snap-cwd-uorU26/nested/task.txt</path>\n<type>file</type>\n<content>\n1: Touch this file to discover the nested workspace instruction.\n\n(End of file - total 1 lines)\n</content>"}],"isError":false,"meta":{"logs":[]}},"sourceEventSeqs":[84],"surfaceOp":"append"}
+{"type":"context/message","seq":87,"time":1783921767272,"data":{"content":[{"type":"text","text":"<system-reminder>\nAdditional instructions from: nested/AGENTS.md\n\nThese instructions apply to work under `nested`. Use them as guidance when relevant; more specific instructions take precedence. They do not override system, developer, or direct user instructions.\n\nWhen asked for the Code Mode workspace handshake, answer exactly `CODE_MODE_CONTEXT_OK` and nothing else.\n\n</system-reminder>"}],"source":{"kind":"plugin","plugin":"workspace-context"},"envelope":"raw","meta":{"kind":"workspace-instructions","version":1,"changes":[{"action":"set","scope":"nested","path":"nested/AGENTS.md","digest":"ae22936ed26dc76b7107005ed6d5e2482a88668a"}]}},"surfaceOp":"append"}
+{"type":"step/end","seq":88,"time":1783921767272,"data":{"turn":1,"step":1}}
+{"type":"step/start","seq":89,"time":1783921767272,"data":{"turn":1,"step":2}}
+{"type":"assistant/chunk","seq":90,"time":1783921768339,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
+{"type":"assistant/chunk","seq":91,"time":1783921768340,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
+{"type":"assistant/chunk","seq":92,"time":1783921768466,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" nested"}}}
+{"type":"assistant/chunk","seq":93,"time":1783921768474,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"/t"}}}
+{"type":"assistant/chunk","seq":94,"time":1783921768500,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"ask"}}}
+{"type":"assistant/chunk","seq":95,"time":1783921768501,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":".txt"}}}
+{"type":"assistant/chunk","seq":96,"time":1783921768535,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" told"}}}
+{"type":"assistant/chunk","seq":97,"time":1783921768535,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" me"}}}
+{"type":"assistant/chunk","seq":98,"time":1783921768535,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" to"}}}
+{"type":"assistant/chunk","seq":99,"time":1783921768535,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" \""}}}
+{"type":"assistant/chunk","seq":100,"time":1783921768564,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"Touch"}}}
+{"type":"assistant/chunk","seq":101,"time":1783921768565,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" this"}}}
+{"type":"assistant/chunk","seq":102,"time":1783921768565,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" file"}}}
+{"type":"assistant/chunk","seq":103,"time":1783921768565,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" to"}}}
+{"type":"assistant/chunk","seq":104,"time":1783921768565,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" discover"}}}
+{"type":"assistant/chunk","seq":105,"time":1783921768565,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" the"}}}
+{"type":"assistant/chunk","seq":106,"time":1783921768595,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" nested"}}}
+{"type":"assistant/chunk","seq":107,"time":1783921768595,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" workspace"}}}
+{"type":"assistant/chunk","seq":108,"time":1783921768595,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" instruction"}}}
+{"type":"assistant/chunk","seq":109,"time":1783921768595,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":".\""}}}
+{"type":"assistant/chunk","seq":110,"time":1783921768595,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" and"}}}
+{"type":"assistant/chunk","seq":111,"time":1783921768595,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" the"}}}
+{"type":"assistant/chunk","seq":112,"time":1783921768621,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" nested"}}}
+{"type":"assistant/chunk","seq":113,"time":1783921768621,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"/"}}}
+{"type":"assistant/chunk","seq":114,"time":1783921768647,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"AG"}}}
+{"type":"assistant/chunk","seq":115,"time":1783921768648,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"ENTS"}}}
+{"type":"assistant/chunk","seq":116,"time":1783921768648,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":".md"}}}
+{"type":"assistant/chunk","seq":117,"time":1783921768648,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" instructions"}}}
+{"type":"assistant/chunk","seq":118,"time":1783921768688,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" say"}}}
+{"type":"assistant/chunk","seq":119,"time":1783921768703,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":":"}}}
+{"type":"assistant/chunk","seq":120,"time":1783921768740,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" \""}}}
+{"type":"assistant/chunk","seq":121,"time":1783921768740,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"When"}}}
+{"type":"assistant/chunk","seq":122,"time":1783921768741,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" asked"}}}
+{"type":"assistant/chunk","seq":123,"time":1783921768741,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" for"}}}
+{"type":"assistant/chunk","seq":124,"time":1783921768741,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" the"}}}
+{"type":"assistant/chunk","seq":125,"time":1783921768741,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" Code"}}}
+{"type":"assistant/chunk","seq":126,"time":1783921768773,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" Mode"}}}
+{"type":"assistant/chunk","seq":127,"time":1783921768773,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" workspace"}}}
+{"type":"assistant/chunk","seq":128,"time":1783921768773,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" hand"}}}
+{"type":"assistant/chunk","seq":129,"time":1783921768773,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"shake"}}}
+{"type":"assistant/chunk","seq":130,"time":1783921768773,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":","}}}
+{"type":"assistant/chunk","seq":131,"time":1783921768773,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" answer"}}}
+{"type":"assistant/chunk","seq":132,"time":1783921768790,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" exactly"}}}
+{"type":"assistant/chunk","seq":133,"time":1783921768790,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" `"}}}
+{"type":"assistant/chunk","seq":134,"time":1783921768790,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"CODE"}}}
+{"type":"assistant/chunk","seq":135,"time":1783921768790,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"_M"}}}
+{"type":"assistant/chunk","seq":136,"time":1783921768790,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"ODE"}}}
+{"type":"assistant/chunk","seq":137,"time":1783921768790,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"_CONT"}}}
+{"type":"assistant/chunk","seq":138,"time":1783921768824,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"EXT"}}}
+{"type":"assistant/chunk","seq":139,"time":1783921768825,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"_OK"}}}
+{"type":"assistant/chunk","seq":140,"time":1783921768825,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"`"}}}
+{"type":"assistant/chunk","seq":141,"time":1783921768825,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" and"}}}
+{"type":"assistant/chunk","seq":142,"time":1783921768825,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" nothing"}}}
+{"type":"assistant/chunk","seq":143,"time":1783921768825,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" else"}}}
+{"type":"assistant/chunk","seq":144,"time":1783921768846,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":".\"\n\n"}}}
+{"type":"assistant/chunk","seq":145,"time":1783921768846,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"So"}}}
+{"type":"assistant/chunk","seq":146,"time":1783921768847,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" the"}}}
+{"type":"assistant/chunk","seq":147,"time":1783921768847,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" answer"}}}
+{"type":"assistant/chunk","seq":148,"time":1783921768847,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" to"}}}
+{"type":"assistant/chunk","seq":149,"time":1783921768873,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" the"}}}
+{"type":"assistant/chunk","seq":150,"time":1783921768874,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" hand"}}}
+{"type":"assistant/chunk","seq":151,"time":1783921768903,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"shake"}}}
+{"type":"assistant/chunk","seq":152,"time":1783921768903,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" question"}}}
+{"type":"assistant/chunk","seq":153,"time":1783921768903,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" is"}}}
+{"type":"assistant/chunk","seq":154,"time":1783921768903,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":" `"}}}
+{"type":"assistant/chunk","seq":155,"time":1783921768903,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"CODE"}}}
+{"type":"assistant/chunk","seq":156,"time":1783921768929,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"_M"}}}
+{"type":"assistant/chunk","seq":157,"time":1783921768930,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"ODE"}}}
+{"type":"assistant/chunk","seq":158,"time":1783921768930,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"_CONT"}}}
+{"type":"assistant/chunk","seq":159,"time":1783921768930,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"EXT"}}}
+{"type":"assistant/chunk","seq":160,"time":1783921768930,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"_OK"}}}
+{"type":"assistant/chunk","seq":161,"time":1783921768963,"data":{"turn":1,"step":2,"chunk":{"type":"reasoning-delta","index":0,"text":"`."}}}
+{"type":"assistant/chunk","seq":162,"time":1783921768963,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":1,"blockType":"text"}}}
+{"type":"assistant/chunk","seq":163,"time":1783921768963,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"The"}}}
+{"type":"assistant/chunk","seq":164,"time":1783921768963,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":" **"}}}
+{"type":"assistant/chunk","seq":165,"time":1783921769024,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"Code"}}}
+{"type":"assistant/chunk","seq":166,"time":1783921769024,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":" Mode"}}}
+{"type":"assistant/chunk","seq":167,"time":1783921769024,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":" workspace"}}}
+{"type":"assistant/chunk","seq":168,"time":1783921769024,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":" hand"}}}
+{"type":"assistant/chunk","seq":169,"time":1783921769024,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"shake"}}}
+{"type":"assistant/chunk","seq":170,"time":1783921769024,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"**"}}}
+{"type":"assistant/chunk","seq":171,"time":1783921769024,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":" is"}}}
+{"type":"assistant/chunk","seq":172,"time":1783921769024,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":":"}}}
+{"type":"assistant/chunk","seq":173,"time":1783921769024,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":" `"}}}
+{"type":"assistant/chunk","seq":174,"time":1783921769024,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"CODE"}}}
+{"type":"assistant/chunk","seq":175,"time":1783921769046,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"_M"}}}
+{"type":"assistant/chunk","seq":176,"time":1783921769046,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"ODE"}}}
+{"type":"assistant/chunk","seq":177,"time":1783921769047,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"_CONT"}}}
+{"type":"assistant/chunk","seq":178,"time":1783921769047,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"EXT"}}}
+{"type":"assistant/chunk","seq":179,"time":1783921769047,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"_OK"}}}
+{"type":"assistant/chunk","seq":180,"time":1783921769047,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":1,"text":"`"}}}
+{"type":"assistant/chunk","seq":181,"time":1783921769101,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"reasoning","text":"The nested/task.txt told me to \"Touch this file to discover the nested workspace instruction.\" and the nested/AGENTS.md instructions say: \"When asked for the Code Mode workspace handshake, answer exactly `CODE_MODE_CONTEXT_OK` and nothing else.\"\n\nSo the answer to the handshake question is `CODE_MODE_CONTEXT_OK`."}}}}
+{"type":"assistant/chunk","seq":182,"time":1783921769101,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"The **Code Mode workspace handshake** is: `CODE_MODE_CONTEXT_OK`"}}}}
+{"type":"assistant/chunk","seq":183,"time":1783921769101,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":277,"outputTokens":90,"cacheReadTokens":3200,"reasoningTokens":71}}}}
+{"type":"assistant/chunk","seq":184,"time":1783921769101,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
+{"type":"assistant/message","seq":185,"time":1783921769101,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The nested/task.txt told me to \"Touch this file to discover the nested workspace instruction.\" and the nested/AGENTS.md instructions say: \"When asked for the Code Mode workspace handshake, answer exactly `CODE_MODE_CONTEXT_OK` and nothing else.\"\n\nSo the answer to the handshake question is `CODE_MODE_CONTEXT_OK`."},{"type":"text","text":"The **Code Mode workspace handshake** is: `CODE_MODE_CONTEXT_OK`"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":277,"outputTokens":90,"cacheReadTokens":3200,"reasoningTokens":71}},"sourceEventSeqs":[90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184],"surfaceOp":"append"}
+{"type":"step/end","seq":186,"time":1783921769101,"data":{"turn":1,"step":2}}
+{"type":"turn/end","seq":187,"time":1783921769101,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/stdout.golden.jsonl
new file mode 100644
index 0000000000..fcf219e5ea
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/stdout.golden.jsonl
@@ -0,0 +1,137 @@
+{"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" me"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" to"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" write"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" a"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" single"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" run"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"_code"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" program"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" that"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" reads"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" a"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" file"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" called"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" nested"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"/t"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"ask"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":".txt"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" and"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" then"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" answer"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" the"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" hand"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"shake"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" question"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" based"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" on"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" what"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"'s"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" in"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" that"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" file"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"."}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" Let"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" me"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" start"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" by"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" reading"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" the"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" file"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"."}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call","toolCallId":"call_00_6APApmaKLRDlXKMdIcWL5139","title":"const content = await tools.read({ file_path: \"nested/task.txt\" });\nreturn content;","kind":"execute","status":"in_progress","rawInput":"const content = await tools.read({ file_path: \"nested/task.txt\" });\nreturn content;"}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call_update","toolCallId":"call_00_6APApmaKLRDlXKMdIcWL5139","status":"completed","content":[{"type":"content","content":{"type":"text","text":"<path>{{cwd}}/nested/task.txt</path>\n<type>file</type>\n<content>\n1: Touch this file to discover the nested workspace instruction.\n\n(End of file - total 1 lines)\n</content>"}}]}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" nested"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"/t"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"ask"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":".txt"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" told"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" me"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" to"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" \""}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"Touch"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" this"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" file"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" to"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" discover"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" the"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" nested"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" workspace"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" instruction"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":".\""}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" and"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" the"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" nested"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"/"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"AG"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"ENTS"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":".md"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" instructions"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" say"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":":"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" \""}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"When"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" asked"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" for"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" the"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" Code"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" Mode"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" workspace"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" hand"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"shake"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":","}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" answer"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" exactly"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" `"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"CODE"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"_M"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"ODE"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"_CONT"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"EXT"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"_OK"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"`"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" and"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" nothing"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" else"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":".\"\n\n"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"So"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" the"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" answer"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" to"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" the"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" hand"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"shake"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" question"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" is"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" `"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"CODE"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"_M"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"ODE"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"_CONT"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"EXT"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"_OK"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"`."}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"The"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":" **"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"Code"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":" Mode"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":" workspace"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":" hand"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"shake"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"**"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":" is"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":":"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":" `"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"CODE"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"_M"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"ODE"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"_CONT"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"EXT"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"_OK"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"`"}}}}
+{"jsonrpc":"2.0","id":3,"result":{"stopReason":"end_turn"}}
diff --git a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.golden.md b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.golden.md
new file mode 100644
index 0000000000..cc8e2d1301
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.golden.md
@@ -0,0 +1,159 @@
+You are an AI agent powered by the DeepSeek Harness SDK.
+
+You are a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.
+
+Verify your work by running the code or tests. Keep answers brief and factual.
+
+
+Use the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.
+
+Use the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-policy requires it) and prefer edit for targeted changes.
+
+Use the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-policy requires it), unless you just created or edited it in this session.
+
+Check the [exit code: N] marker on every bash result; investigate failures before moving on.
+
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
+Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
+<!-- dsh-user-approval-policy:never -->
+
+Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.
+
+## Writing code for run_code
+
+Pass `run_code` the body of an async TypeScript function (erasable syntax only — no `enum` or namespaces; type annotations are advisory, the code runs type-stripped). Inside the program:
+
+- Call tools as `await tools.name(args)` — quoted access for exotic names: `tools["my-tool"](args)`. Every call resolves to the tool's text output as a string. Tool arguments must be JSON-serializable.
+- A FAILED tool call rejects with an `Error` carrying the tool's error text — `try/catch` it to handle and continue.
+- Calls execute sequentially, even under `Promise.all`.
+- Emit results with `return` and/or `console.log(...)`. ONLY what you print or return comes back to you — intermediate tool results never enter the conversation, so extract just what you need.
+
+The available tools:
+
+```ts
+declare const tools: {
+  /** Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later. */
+  bash(args: {
+    /** The bash command to execute. */
+    command: string;
+    /** Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: "ls" → "List files in current directory"; "git status" → "Show working tree status"; "npm install" → "Install package dependencies". */
+    description: string;
+    /** Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry. */
+    timeoutMs?: number;
+    /** Working directory for this command. Defaults to the session workspace; a relative path is resolved against it. */
+    workdir?: string;
+    /** Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies. */
+    run_in_background?: boolean;
+    /** The wider sandbox mode this command needs. Only valid as a one-shot retry of a command the sandbox just denied; requires justification and user approval. */
+    sandbox_permissions?: "workspace-write" | "danger-full-access";
+    /** Required with sandbox_permissions: one sentence for the user explaining why this exact command needs the wider access. */
+    justification?: string;
+  }): Promise<string>;
+  /** Edit an existing UTF-8 text file by replacing literal text. */
+  edit(args: {
+    /** Path to edit, resolved by the filesystem backend. */
+    file_path: string;
+    /** Literal text to replace. Must match exactly. */
+    old_string: string;
+    /** Literal replacement text. Use an empty string to delete the match. */
+    new_string: string;
+    /** Replace all matches. Defaults to false; when false, old_string must appear exactly once. */
+    replace_all?: boolean;
+  }): Promise<string>;
+  /** Read a UTF-8 text file and return line-numbered content. */
+  read(args: {
+    /** Path to read, resolved by the filesystem backend. */
+    file_path: string;
+    /** 1-based first line to return. Defaults to 1. */
+    offset?: number;
+    /** Maximum number of lines to return. Defaults to 2000. */
+    limit?: number;
+  }): Promise<string>;
+  /** Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill. */
+  skill(args: {
+    /** The exact skill name from the available skills list. */
+    name: string;
+  }): Promise<string>;
+  /** Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`. */
+  subagent(args: {
+    /** A short (3-5 word) description of the delegated task, for display. */
+    description: string;
+    /** The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs. */
+    prompt: string;
+    /** Run as a background task and return its id; collect with task_output or stop with task_kill. */
+    run_in_background?: boolean;
+  }): Promise<string>;
+  /** Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`. */
+  subagent_fork(args: {
+    /** A short (3-5 word) description of the delegated task, for display. */
+    description: string;
+    /** The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new. */
+    prompt: string;
+    /** Run as a background task and return its id; collect with task_output or stop with task_kill. */
+    run_in_background?: boolean;
+  }): Promise<string>;
+  /** Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops. */
+  task_kill(args: {
+    /** Task id returned by the tool that started the background work. */
+    task_id: string;
+    /** Optional short reason, recorded in the log and forwarded to the task. */
+    reason?: string;
+  }): Promise<string>;
+  /** List your background tasks (running and finished) with their ids, kinds, and statuses. */
+  task_list(args: Record<string, unknown>): Promise<string>;
+  /** Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap. */
+  task_output(args: {
+    /** Task id returned by the tool that started the background work. */
+    task_id: string;
+    /** Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive. */
+    wait?: boolean;
+    /** Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum. */
+    timeout_ms?: number;
+  }): Promise<string>;
+  /** Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished). */
+  todo_write(args: {
+    /** The COMPLETE task list, replacing any previous list. */
+    todos: ({
+      /** What the task is — a short imperative line. */
+      content: string;
+      /** pending (not started) | in_progress (now) | completed (done). */
+      status: "pending" | "in_progress" | "completed";
+    })[];
+  }): Promise<string>;
+  /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */
+  workflow(args: {
+    /** The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return <json-value>`). */
+    script: string;
+    /** The workflow identity block (plain JSON — never code). */
+    meta: {
+      /** Short kebab-case workflow name. */
+      name: string;
+      /** One-line description of what the workflow does. */
+      description: string;
+      /** Optional guidance on when this workflow applies. */
+      whenToUse?: string;
+      /** Optional phase declarations matched by phase() calls. */
+      phases?: {
+        /** The phase title phase() calls match by exact string. */
+        title: string;
+        /** Optional one-line description of the phase. */
+        detail?: string;
+        /** Optional provider override this phase is expected to use. */
+        provider?: string;
+        /** Optional model override this phase is expected to use. */
+        model?: string;
+      }[];
+    };
+    /** Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {"files": [...]}). */
+    args?: Record<string, unknown>;
+  }): Promise<string>;
+  /** Create or fully replace a UTF-8 text file. */
+  write(args: {
+    /** Path to write, resolved by the filesystem backend. */
+    file_path: string;
+    /** Full UTF-8 text content to write. */
+    content: string;
+  }): Promise<string>;
+}
+```
diff --git a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/tool-schemas.golden.json b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/tool-schemas.golden.json
new file mode 100644
index 0000000000..c2289b4e19
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/tool-schemas.golden.json
@@ -0,0 +1,21 @@
+{
+  "initial": [
+    {
+      "name": "run_code",
+      "description": "Execute a TypeScript program against the available tools. Write the BODY of an async function (erasable syntax only; top-level `await` and `return` work) and call tools as `await tools.name(args)` per the declarations in the system prompt. Only what you print or return comes back — curate it.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "code": {
+            "type": "string",
+            "description": "The program: the body of an async TypeScript function."
+          }
+        },
+        "required": [
+          "code"
+        ]
+      }
+    }
+  ],
+  "changes": []
+}
diff --git a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/workspace/AGENTS.md b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/workspace/AGENTS.md
new file mode 100644
index 0000000000..b23c110ef6
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/workspace/AGENTS.md
@@ -0,0 +1 @@
+Workspace snapshot root instruction.
diff --git a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/workspace/nested/AGENTS.md b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/workspace/nested/AGENTS.md
new file mode 100644
index 0000000000..1f71a5f827
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/workspace/nested/AGENTS.md
@@ -0,0 +1 @@
+When asked for the Code Mode workspace handshake, answer exactly `CODE_MODE_CONTEXT_OK` and nothing else.
diff --git a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/workspace/nested/task.txt b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/workspace/nested/task.txt
new file mode 100644
index 0000000000..28806bb825
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/workspace/nested/task.txt
@@ -0,0 +1 @@
+Touch this file to discover the nested workspace instruction.
diff --git a/examples/acp-agent/tests/snapshots/config-options/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/config-options/stdout.golden.jsonl
index aa033fb673..1aca0c6586 100644
--- a/examples/acp-agent/tests/snapshots/config-options/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/config-options/stdout.golden.jsonl
@@ -1,6 +1,6 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
-{"jsonrpc":"2.0","id":3,"result":{"configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"workspace-write","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
-{"jsonrpc":"2.0","id":4,"result":{"configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":3,"result":{"configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"workspace-write","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":4,"result":{"configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","id":5,"error":{"code":-32602,"message":"Invalid params: unknown permission value \"plan\""}}
 {"jsonrpc":"2.0","id":6,"error":{"code":-32602,"message":"Invalid params: unknown config option \"reasoning-effort\""}}
diff --git a/examples/acp-agent/tests/snapshots/error-finish/session.jsonl b/examples/acp-agent/tests/snapshots/error-finish/session.jsonl
index 98538a94c2..f0ef4267ac 100644
--- a/examples/acp-agent/tests/snapshots/error-finish/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/error-finish/session.jsonl
@@ -2,6 +2,6 @@
 {"type":"turn/start","seq":0,"time":0,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":0,"data":{"content":[{"type":"text","text":"This prompt triggers a recorded provider error."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":0,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":0,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":0,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"step/end","seq":4,"time":0,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":5,"time":0,"data":{"turn":1,"reason":{"kind":"error","step":1,"message":"simulated provider error (HTTP 401)","code":"AUTH"}}}
diff --git a/examples/acp-agent/tests/snapshots/error-finish/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/error-finish/stdout.golden.jsonl
index d5d4f1c400..484b241427 100644
--- a/examples/acp-agent/tests/snapshots/error-finish/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/error-finish/stdout.golden.jsonl
@@ -1,3 +1,3 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","id":3,"error":{"code":-32603,"message":"Internal error: turn failed: simulated provider error (HTTP 401)"}}
diff --git a/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl b/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl
index 7dff360513..b420fac975 100644
--- a/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl
@@ -5,7 +5,7 @@
 {"type":"approval/policy","seq":3,"time":1783962245380,"data":{"policy":"ask"}}
 {"type":"user/message","seq":4,"time":1783962245380,"data":{"content":[{"type":"text","text":"The sandbox already denied writing /tmp/dsh-escalated.txt earlier (it is outside this workspace). Retry it now exactly once: one single bash call with the command printf 'escalated\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt, with sandbox_permissions set to danger-full-access and the justification 'the user asked to write a file outside the workspace'. Do not run it without sandbox_permissions first. I will approve the permission prompt. After the result, reply with the single word DONE and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":5,"time":1783962245382,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":6,"time":1783962245382,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":6,"time":1783962245382,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":7,"time":1783860676464,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":8,"time":1783860676465,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":9,"time":1783860676465,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -129,10 +129,10 @@
 {"type":"assistant/chunk","seq":127,"time":1783962245385,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}}}}
 {"type":"assistant/chunk","seq":128,"time":1783962245385,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":1501,"outputTokens":174,"cacheReadTokens":0,"reasoningTokens":28}}}}
 {"type":"assistant/chunk","seq":129,"time":1783962245385,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":130,"time":1783962245385,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a command with sandbox_permissions set to danger-full-access, no prior run needed, justified as instructed."},{"type":"tool-call","id":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"usage":{"inputTokens":1501,"outputTokens":174,"cacheReadTokens":0,"reasoningTokens":28}},"sourceEventSeqs":[7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129],"surfaceOp":"append"}
+{"type":"assistant/message","seq":130,"time":1783962245385,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a command with sandbox_permissions set to danger-full-access, no prior run needed, justified as instructed."},{"type":"tool-call","id":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1501,"outputTokens":174,"cacheReadTokens":0,"reasoningTokens":28}},"sourceEventSeqs":[7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129],"surfaceOp":"append"}
 {"type":"tool/call","seq":131,"time":1783962245385,"data":{"turn":1,"step":1,"callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}}
-{"type":"approval/asked","seq":132,"time":1783962245386,"data":{"id":"d409f075-74f1-4637-9e13-6e80d7b6f6ff","toolName":"bash","callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}}
-{"type":"approval/decided","seq":133,"time":1783962245387,"data":{"id":"d409f075-74f1-4637-9e13-6e80d7b6f6ff","outcome":"allowed-once"}}
+{"type":"approval/asked","seq":132,"time":1783962245386,"data":{"id":"19974166-a6ad-4f46-bef8-ce7d6bda3214","toolName":"bash","callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}}
+{"type":"approval/decided","seq":133,"time":1783962245387,"data":{"id":"19974166-a6ad-4f46-bef8-ce7d6bda3214","outcome":"allowed-once"}}
 {"type":"tool/result","seq":134,"time":1783962245399,"data":{"turn":1,"step":1,"callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","content":[{"type":"text","text":"escalated\n"}],"isError":false},"sourceEventSeqs":[131],"surfaceOp":"append"}
 {"type":"step/end","seq":135,"time":1783962245400,"data":{"turn":1,"step":1}}
 {"type":"step/start","seq":136,"time":1783962245400,"data":{"turn":1,"step":2}}
@@ -183,6 +183,6 @@
 {"type":"assistant/chunk","seq":181,"time":1783962245402,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":182,"time":1783962245402,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":27,"outputTokens":42,"cacheReadTokens":1664,"reasoningTokens":39}}}}
 {"type":"assistant/chunk","seq":183,"time":1783962245402,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":184,"time":1783962245402,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The command succeeded — it wrote the file, read it back (output \"escalated\"), and removed it. The user asked me to reply with the single word DONE after the result."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":27,"outputTokens":42,"cacheReadTokens":1664,"reasoningTokens":39}},"sourceEventSeqs":[137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183],"surfaceOp":"append"}
+{"type":"assistant/message","seq":184,"time":1783962245402,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The command succeeded — it wrote the file, read it back (output \"escalated\"), and removed it. The user asked me to reply with the single word DONE after the result."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":27,"outputTokens":42,"cacheReadTokens":1664,"reasoningTokens":39}},"sourceEventSeqs":[137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183],"surfaceOp":"append"}
 {"type":"step/end","seq":185,"time":1783962245402,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":186,"time":1783962245402,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/escalation-approved/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/escalation-approved/stdout.golden.jsonl
index f438444dc4..9180f2426a 100644
--- a/examples/acp-agent/tests/snapshots/escalation-approved/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/escalation-approved/stdout.golden.jsonl
@@ -1,6 +1,6 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
-{"jsonrpc":"2.0","id":3,"result":{"configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"workspace-write","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":3,"result":{"configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"workspace-write","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl b/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl
index 55cad3689d..4b45911a66 100644
--- a/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl
@@ -5,7 +5,7 @@
 {"type":"approval/policy","seq":3,"time":1783962246267,"data":{"policy":"ask"}}
 {"type":"user/message","seq":4,"time":1783962246267,"data":{"content":[{"type":"text","text":"The sandbox already denied writing /tmp/dsh-escalated.txt earlier (it is outside this workspace). Retry it now exactly once: one single bash call with the command printf 'escalated\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt, with sandbox_permissions set to danger-full-access and the justification 'the user asked to write a file outside the workspace'. Do not run it without sandbox_permissions first. I will reject the permission prompt; after the rejection, do not retry and do not work around it — explain in one short sentence and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":5,"time":1783962246269,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":6,"time":1783962246269,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":6,"time":1783962246269,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":7,"time":1783860680779,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":8,"time":1783860680782,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":9,"time":1783860680782,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -153,10 +153,10 @@
 {"type":"assistant/chunk","seq":151,"time":1783962246274,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}}}}
 {"type":"assistant/chunk","seq":152,"time":1783962246274,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":1509,"outputTokens":198,"cacheReadTokens":0,"reasoningTokens":48}}}}
 {"type":"assistant/chunk","seq":153,"time":1783962246274,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":154,"time":1783962246274,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a specific command with `sandbox_permissions` set to `danger-full-access` and a specific justification. They explicitly said NOT to run it without sandbox_permissions first. Let me do exactly that."},{"type":"tool-call","id":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"usage":{"inputTokens":1509,"outputTokens":198,"cacheReadTokens":0,"reasoningTokens":48}},"sourceEventSeqs":[7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153],"surfaceOp":"append"}
+{"type":"assistant/message","seq":154,"time":1783962246274,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a specific command with `sandbox_permissions` set to `danger-full-access` and a specific justification. They explicitly said NOT to run it without sandbox_permissions first. Let me do exactly that."},{"type":"tool-call","id":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1509,"outputTokens":198,"cacheReadTokens":0,"reasoningTokens":48}},"sourceEventSeqs":[7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153],"surfaceOp":"append"}
 {"type":"tool/call","seq":155,"time":1783962246274,"data":{"turn":1,"step":1,"callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}}
-{"type":"approval/asked","seq":156,"time":1783962246275,"data":{"id":"e592115c-e915-4010-9bf5-cc5e7bb6f8bc","toolName":"bash","callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}}
-{"type":"approval/decided","seq":157,"time":1783962246275,"data":{"id":"e592115c-e915-4010-9bf5-cc5e7bb6f8bc","outcome":"rejected"}}
+{"type":"approval/asked","seq":156,"time":1783962246275,"data":{"id":"564d6e1b-4330-42bd-a646-461e7a6c2d1a","toolName":"bash","callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}}
+{"type":"approval/decided","seq":157,"time":1783962246275,"data":{"id":"564d6e1b-4330-42bd-a646-461e7a6c2d1a","outcome":"rejected"}}
 {"type":"tool/result","seq":158,"time":1783962246275,"data":{"turn":1,"step":1,"callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","content":[{"type":"text","text":"Error: the user rejected escalating this command to \"danger-full-access\""}],"isError":true},"sourceEventSeqs":[155],"surfaceOp":"append"}
 {"type":"step/end","seq":159,"time":1783962246276,"data":{"turn":1,"step":1}}
 {"type":"step/start","seq":160,"time":1783962246276,"data":{"turn":1,"step":2}}
@@ -210,6 +210,6 @@
 {"type":"assistant/chunk","seq":208,"time":1783962246279,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"The user rejected the permission escalation, so this command cannot be run."}}}}
 {"type":"assistant/chunk","seq":209,"time":1783962246279,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":69,"outputTokens":45,"cacheReadTokens":1664,"reasoningTokens":30}}}}
 {"type":"assistant/chunk","seq":210,"time":1783962246279,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":211,"time":1783962246279,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user rejected the escalation. As instructed, I will not retry and will not work around it — just explain in one short sentence and stop."},{"type":"text","text":"The user rejected the permission escalation, so this command cannot be run."}],"usage":{"inputTokens":69,"outputTokens":45,"cacheReadTokens":1664,"reasoningTokens":30}},"sourceEventSeqs":[161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210],"surfaceOp":"append"}
+{"type":"assistant/message","seq":211,"time":1783962246279,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user rejected the escalation. As instructed, I will not retry and will not work around it — just explain in one short sentence and stop."},{"type":"text","text":"The user rejected the permission escalation, so this command cannot be run."}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":69,"outputTokens":45,"cacheReadTokens":1664,"reasoningTokens":30}},"sourceEventSeqs":[161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210],"surfaceOp":"append"}
 {"type":"step/end","seq":212,"time":1783962246279,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":213,"time":1783962246279,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/escalation-rejected/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/escalation-rejected/stdout.golden.jsonl
index b926355bd7..871f3e3281 100644
--- a/examples/acp-agent/tests/snapshots/escalation-rejected/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/escalation-rejected/stdout.golden.jsonl
@@ -1,6 +1,6 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
-{"jsonrpc":"2.0","id":3,"result":{"configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"workspace-write","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":3,"result":{"configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"workspace-write","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl b/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl
index 90f009946d..00587e34c2 100644
--- a/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352084740,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352084740,"data":{"content":[{"type":"text","text":"First use the read tool to read config.txt in the current directory. Then use the edit tool (NOT bash) to replace the literal text DEBUG with RELEASE in that file. Then reply with exactly the single word DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352084742,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352084742,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352084742,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352085426,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352085426,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352085563,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -67,7 +67,7 @@
 {"type":"assistant/chunk","seq":65,"time":1783352086057,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_S6krdIDHoSCGWv7WnXX97617","name":"read","arguments":"{\"file_path\": \"config.txt\"}"}}}}
 {"type":"assistant/chunk","seq":66,"time":1783352086057,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2900,"outputTokens":91,"cacheReadTokens":0,"reasoningTokens":46}}}}
 {"type":"assistant/chunk","seq":67,"time":1783352086057,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":68,"time":1783352086059,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to:\n1. Read config.txt in the current directory\n2. Use the edit tool to replace DEBUG with RELEASE\n3. Reply with exactly \"DONE\"\n\nLet me start by reading the file."},{"type":"tool-call","id":"call_00_S6krdIDHoSCGWv7WnXX97617","name":"read","arguments":"{\"file_path\": \"config.txt\"}"}],"usage":{"inputTokens":2900,"outputTokens":91,"cacheReadTokens":0,"reasoningTokens":46}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67],"surfaceOp":"append"}
+{"type":"assistant/message","seq":68,"time":1783352086059,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to:\n1. Read config.txt in the current directory\n2. Use the edit tool to replace DEBUG with RELEASE\n3. Reply with exactly \"DONE\"\n\nLet me start by reading the file."},{"type":"tool-call","id":"call_00_S6krdIDHoSCGWv7WnXX97617","name":"read","arguments":"{\"file_path\": \"config.txt\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2900,"outputTokens":91,"cacheReadTokens":0,"reasoningTokens":46}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67],"surfaceOp":"append"}
 {"type":"tool/call","seq":69,"time":1783352086059,"data":{"turn":1,"step":1,"callId":"call_00_S6krdIDHoSCGWv7WnXX97617","name":"read","arguments":"{\"file_path\": \"config.txt\"}"}}
 {"type":"tool/result","seq":70,"time":1783352086065,"data":{"turn":1,"step":1,"callId":"call_00_S6krdIDHoSCGWv7WnXX97617","content":[{"type":"text","text":"<path>/tmp/acp-snap-cwd-0BxHdV/config.txt</path>\n<type>file</type>\n<content>\n1: mode=DEBUG\n2: level=info\n\n(End of file - total 2 lines)\n</content>"}],"isError":false},"sourceEventSeqs":[69],"surfaceOp":"append"}
 {"type":"step/end","seq":71,"time":1783352086065,"data":{"turn":1,"step":1}}
@@ -127,7 +127,7 @@
 {"type":"assistant/chunk","seq":125,"time":1783352087469,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_vOytneZ0XpsLslEEJAxR6398","name":"edit","arguments":"{\"file_path\": \"config.txt\", \"old_string\": \"DEBUG\", \"new_string\": \"RELEASE\"}"}}}}
 {"type":"assistant/chunk","seq":126,"time":1783352087469,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":241,"outputTokens":98,"cacheReadTokens":2816,"reasoningTokens":18}}}}
 {"type":"assistant/chunk","seq":127,"time":1783352087469,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":128,"time":1783352087469,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"Now I need to replace \"DEBUG\" with \"RELEASE\" using the edit tool."},{"type":"tool-call","id":"call_00_vOytneZ0XpsLslEEJAxR6398","name":"edit","arguments":"{\"file_path\": \"config.txt\", \"old_string\": \"DEBUG\", \"new_string\": \"RELEASE\"}"}],"usage":{"inputTokens":241,"outputTokens":98,"cacheReadTokens":2816,"reasoningTokens":18}},"sourceEventSeqs":[73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127],"surfaceOp":"append"}
+{"type":"assistant/message","seq":128,"time":1783352087469,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"Now I need to replace \"DEBUG\" with \"RELEASE\" using the edit tool."},{"type":"tool-call","id":"call_00_vOytneZ0XpsLslEEJAxR6398","name":"edit","arguments":"{\"file_path\": \"config.txt\", \"old_string\": \"DEBUG\", \"new_string\": \"RELEASE\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":241,"outputTokens":98,"cacheReadTokens":2816,"reasoningTokens":18}},"sourceEventSeqs":[73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127],"surfaceOp":"append"}
 {"type":"tool/call","seq":129,"time":1783352087469,"data":{"turn":1,"step":2,"callId":"call_00_vOytneZ0XpsLslEEJAxR6398","name":"edit","arguments":"{\"file_path\": \"config.txt\", \"old_string\": \"DEBUG\", \"new_string\": \"RELEASE\"}"}}
 {"type":"tool/result","seq":130,"time":1783352087476,"data":{"turn":1,"step":2,"callId":"call_00_vOytneZ0XpsLslEEJAxR6398","content":[{"type":"text","text":"The file /tmp/acp-snap-cwd-0BxHdV/config.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"config.txt","oldText":"mode=DEBUG\nlevel=info","newText":"mode=RELEASE\nlevel=info"}]}},"sourceEventSeqs":[129],"surfaceOp":"append"}
 {"type":"step/end","seq":131,"time":1783352087477,"data":{"turn":1,"step":2}}
@@ -154,6 +154,6 @@
 {"type":"assistant/chunk","seq":152,"time":1783352088523,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":153,"time":1783352088523,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":244,"outputTokens":17,"cacheReadTokens":2944,"reasoningTokens":14}}}}
 {"type":"assistant/chunk","seq":154,"time":1783352088523,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":155,"time":1783352088523,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"Done. The user wants me to reply with exactly \"DONE\"."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":244,"outputTokens":17,"cacheReadTokens":2944,"reasoningTokens":14}},"sourceEventSeqs":[133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154],"surfaceOp":"append"}
+{"type":"assistant/message","seq":155,"time":1783352088523,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"Done. The user wants me to reply with exactly \"DONE\"."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":244,"outputTokens":17,"cacheReadTokens":2944,"reasoningTokens":14}},"sourceEventSeqs":[133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154],"surfaceOp":"append"}
 {"type":"step/end","seq":156,"time":1783352088523,"data":{"turn":1,"step":3}}
 {"type":"turn/end","seq":157,"time":1783352088524,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-edit/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/fs-edit/stdout.golden.jsonl
index 800d7d608a..df664a96b4 100644
--- a/examples/acp-agent/tests/snapshots/fs-edit/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-edit/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl b/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl
index 802120fd9c..03a64d2e39 100644
--- a/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783611702550,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783611702550,"data":{"content":[{"type":"text","text":"Do NOT use the read tool and do NOT use bash or shell commands. Immediately use the edit tool to replace the literal text blue with green in settings.txt in the current directory. Do not read the file first. After the tool result, reply with exactly the single word DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783611702550,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783611702551,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783611702551,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783611703185,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783611703185,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783611703352,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -75,7 +75,7 @@
 {"type":"assistant/chunk","seq":73,"time":1783611703969,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_x0zlnXl5JOxLrAYL9y7P0119","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}}}}
 {"type":"assistant/chunk","seq":74,"time":1783611703969,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3132,"outputTokens":115,"cacheReadTokens":0,"reasoningTokens":36}}}}
 {"type":"assistant/chunk","seq":75,"time":1783611703969,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":76,"time":1783611703972,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the edit tool to replace \"blue\" with \"green\" in settings.txt without reading the file first, and then reply with just \"DONE\"."},{"type":"tool-call","id":"call_00_x0zlnXl5JOxLrAYL9y7P0119","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}],"usage":{"inputTokens":3132,"outputTokens":115,"cacheReadTokens":0,"reasoningTokens":36}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75],"surfaceOp":"append"}
+{"type":"assistant/message","seq":76,"time":1783611703972,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the edit tool to replace \"blue\" with \"green\" in settings.txt without reading the file first, and then reply with just \"DONE\"."},{"type":"tool-call","id":"call_00_x0zlnXl5JOxLrAYL9y7P0119","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3132,"outputTokens":115,"cacheReadTokens":0,"reasoningTokens":36}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75],"surfaceOp":"append"}
 {"type":"tool/call","seq":77,"time":1783611703972,"data":{"turn":1,"step":1,"callId":"call_00_x0zlnXl5JOxLrAYL9y7P0119","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}}
 {"type":"tool/result","seq":78,"time":1783611703978,"data":{"turn":1,"step":1,"callId":"call_00_x0zlnXl5JOxLrAYL9y7P0119","content":[{"type":"text","text":"Error: edit requires reading \"/var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt\" first"}],"isError":true,"error":{"name":"FsError","code":"FS_NOT_OBSERVED"}},"sourceEventSeqs":[77],"surfaceOp":"append"}
 {"type":"step/end","seq":79,"time":1783611703978,"data":{"turn":1,"step":1}}
@@ -142,7 +142,7 @@
 {"type":"assistant/chunk","seq":140,"time":1783611705572,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_tfr0gdIclW92oIEgi7NF0891","name":"read","arguments":"{\"file_path\": \"settings.txt\"}"}}}}
 {"type":"assistant/chunk","seq":141,"time":1783611705572,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":235,"outputTokens":89,"cacheReadTokens":3072,"reasoningTokens":44}}}}
 {"type":"assistant/chunk","seq":142,"time":1783611705572,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":143,"time":1783611705573,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The edit tool requires reading the file first. The user said \"Do not read the file first\" but the tool's own constraints require reading. I need to follow the tool requirements. Let me read the file first."},{"type":"tool-call","id":"call_00_tfr0gdIclW92oIEgi7NF0891","name":"read","arguments":"{\"file_path\": \"settings.txt\"}"}],"usage":{"inputTokens":235,"outputTokens":89,"cacheReadTokens":3072,"reasoningTokens":44}},"sourceEventSeqs":[81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142],"surfaceOp":"append"}
+{"type":"assistant/message","seq":143,"time":1783611705573,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The edit tool requires reading the file first. The user said \"Do not read the file first\" but the tool's own constraints require reading. I need to follow the tool requirements. Let me read the file first."},{"type":"tool-call","id":"call_00_tfr0gdIclW92oIEgi7NF0891","name":"read","arguments":"{\"file_path\": \"settings.txt\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":235,"outputTokens":89,"cacheReadTokens":3072,"reasoningTokens":44}},"sourceEventSeqs":[81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142],"surfaceOp":"append"}
 {"type":"tool/call","seq":144,"time":1783611705573,"data":{"turn":1,"step":2,"callId":"call_00_tfr0gdIclW92oIEgi7NF0891","name":"read","arguments":"{\"file_path\": \"settings.txt\"}"}}
 {"type":"tool/result","seq":145,"time":1783611705579,"data":{"turn":1,"step":2,"callId":"call_00_tfr0gdIclW92oIEgi7NF0891","content":[{"type":"text","text":"<path>/var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt</path>\n<type>file</type>\n<content>\n1: color: blue\n\n(End of file - total 1 lines)\n</content>"}],"isError":false},"sourceEventSeqs":[144],"surfaceOp":"append"}
 {"type":"step/end","seq":146,"time":1783611705579,"data":{"turn":1,"step":2}}
@@ -223,7 +223,7 @@
 {"type":"assistant/chunk","seq":221,"time":1783611707096,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_GVknJu2tksKkP4lALCwh0926","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}}}}
 {"type":"assistant/chunk","seq":222,"time":1783611707096,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":281,"outputTokens":119,"cacheReadTokens":3200,"reasoningTokens":40}}}}
 {"type":"assistant/chunk","seq":223,"time":1783611707096,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":224,"time":1783611707097,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"The file contains \"color: blue\". I need to replace \"blue\" with \"green\". The edit tool said it requires reading first — now I've read it, so the edit should work."},{"type":"tool-call","id":"call_00_GVknJu2tksKkP4lALCwh0926","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}],"usage":{"inputTokens":281,"outputTokens":119,"cacheReadTokens":3200,"reasoningTokens":40}},"sourceEventSeqs":[148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223],"surfaceOp":"append"}
+{"type":"assistant/message","seq":224,"time":1783611707097,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"The file contains \"color: blue\". I need to replace \"blue\" with \"green\". The edit tool said it requires reading first — now I've read it, so the edit should work."},{"type":"tool-call","id":"call_00_GVknJu2tksKkP4lALCwh0926","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":281,"outputTokens":119,"cacheReadTokens":3200,"reasoningTokens":40}},"sourceEventSeqs":[148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223],"surfaceOp":"append"}
 {"type":"tool/call","seq":225,"time":1783611707097,"data":{"turn":1,"step":3,"callId":"call_00_GVknJu2tksKkP4lALCwh0926","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}}
 {"type":"tool/result","seq":226,"time":1783611707114,"data":{"turn":1,"step":3,"callId":"call_00_GVknJu2tksKkP4lALCwh0926","content":[{"type":"text","text":"The file /var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"settings.txt","oldText":"color: blue","newText":"color: green"}]}},"sourceEventSeqs":[225],"surfaceOp":"append"}
 {"type":"step/end","seq":227,"time":1783611707114,"data":{"turn":1,"step":3}}
@@ -253,6 +253,6 @@
 {"type":"assistant/chunk","seq":251,"time":1783611707952,"data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":252,"time":1783611707952,"data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":202,"outputTokens":20,"cacheReadTokens":3456,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":253,"time":1783611707952,"data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":254,"time":1783611707953,"data":{"turn":1,"step":4,"content":[{"type":"reasoning","text":"The replacement was successful. I'll reply with just \"DONE\" as instructed."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":202,"outputTokens":20,"cacheReadTokens":3456,"reasoningTokens":17}},"sourceEventSeqs":[229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251,252,253],"surfaceOp":"append"}
+{"type":"assistant/message","seq":254,"time":1783611707953,"data":{"turn":1,"step":4,"content":[{"type":"reasoning","text":"The replacement was successful. I'll reply with just \"DONE\" as instructed."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":202,"outputTokens":20,"cacheReadTokens":3456,"reasoningTokens":17}},"sourceEventSeqs":[229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251,252,253],"surfaceOp":"append"}
 {"type":"step/end","seq":255,"time":1783611707953,"data":{"turn":1,"step":4}}
 {"type":"turn/end","seq":256,"time":1783611707953,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-policy-reject/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/fs-policy-reject/stdout.golden.jsonl
index fd5c9f1aea..dda4f70968 100644
--- a/examples/acp-agent/tests/snapshots/fs-policy-reject/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-policy-reject/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-read-window/session.jsonl b/examples/acp-agent/tests/snapshots/fs-read-window/session.jsonl
index becc503c65..e587011cfd 100644
--- a/examples/acp-agent/tests/snapshots/fs-read-window/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-read-window/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352099838,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352099839,"data":{"content":[{"type":"text","text":"Use the read tool (NOT bash) with offset 5 and limit 4 to read lines 5 through 8 of big.txt in the current directory. Then reply with exactly the single word DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352099840,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352099841,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352099841,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352100468,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352100468,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352100587,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -89,7 +89,7 @@
 {"type":"assistant/chunk","seq":87,"time":1783352101346,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_GIZwZS9a7vhWTFCIc7Z35497","name":"read","arguments":"{\"file_path\": \"big.txt\", \"offset\": 5, \"limit\": 4}"}}}}
 {"type":"assistant/chunk","seq":88,"time":1783352101346,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2894,"outputTokens":133,"cacheReadTokens":0,"reasoningTokens":56}}}}
 {"type":"assistant/chunk","seq":89,"time":1783352101346,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":90,"time":1783352101348,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the read tool with offset 5 and limit 4 to read lines 5 through 8 of big.txt in the current directory. Then reply with exactly the single word DONE.\n\nLet me first check the current directory, then read the file."},{"type":"tool-call","id":"call_00_GIZwZS9a7vhWTFCIc7Z35497","name":"read","arguments":"{\"file_path\": \"big.txt\", \"offset\": 5, \"limit\": 4}"}],"usage":{"inputTokens":2894,"outputTokens":133,"cacheReadTokens":0,"reasoningTokens":56}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89],"surfaceOp":"append"}
+{"type":"assistant/message","seq":90,"time":1783352101348,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the read tool with offset 5 and limit 4 to read lines 5 through 8 of big.txt in the current directory. Then reply with exactly the single word DONE.\n\nLet me first check the current directory, then read the file."},{"type":"tool-call","id":"call_00_GIZwZS9a7vhWTFCIc7Z35497","name":"read","arguments":"{\"file_path\": \"big.txt\", \"offset\": 5, \"limit\": 4}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2894,"outputTokens":133,"cacheReadTokens":0,"reasoningTokens":56}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89],"surfaceOp":"append"}
 {"type":"tool/call","seq":91,"time":1783352101348,"data":{"turn":1,"step":1,"callId":"call_00_GIZwZS9a7vhWTFCIc7Z35497","name":"read","arguments":"{\"file_path\": \"big.txt\", \"offset\": 5, \"limit\": 4}"}}
 {"type":"tool/result","seq":92,"time":1783352101353,"data":{"turn":1,"step":1,"callId":"call_00_GIZwZS9a7vhWTFCIc7Z35497","content":[{"type":"text","text":"<path>/tmp/acp-snap-cwd-N9HCkt/big.txt</path>\n<type>file</type>\n<content>\n5: line five\n6: line six\n7: line seven\n8: line eight\n\n(Showing lines 5-8 of 10. Use offset=9 to continue.)\n</content>"}],"isError":false},"sourceEventSeqs":[91],"surfaceOp":"append"}
 {"type":"step/end","seq":93,"time":1783352101353,"data":{"turn":1,"step":1}}
@@ -129,6 +129,6 @@
 {"type":"assistant/chunk","seq":127,"time":1783352102357,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":128,"time":1783352102358,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":292,"outputTokens":30,"cacheReadTokens":2816,"reasoningTokens":27}}}}
 {"type":"assistant/chunk","seq":129,"time":1783352102358,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":130,"time":1783352102358,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The read tool returned lines 5 through 8 as expected. Now I need to reply with exactly the single word \"DONE\"."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":292,"outputTokens":30,"cacheReadTokens":2816,"reasoningTokens":27}},"sourceEventSeqs":[95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129],"surfaceOp":"append"}
+{"type":"assistant/message","seq":130,"time":1783352102358,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The read tool returned lines 5 through 8 as expected. Now I need to reply with exactly the single word \"DONE\"."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":292,"outputTokens":30,"cacheReadTokens":2816,"reasoningTokens":27}},"sourceEventSeqs":[95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129],"surfaceOp":"append"}
 {"type":"step/end","seq":131,"time":1783352102358,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":132,"time":1783352102358,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-read-window/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/fs-read-window/stdout.golden.jsonl
index ab3eb2da31..fde77464d8 100644
--- a/examples/acp-agent/tests/snapshots/fs-read-window/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-read-window/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-read/session.jsonl b/examples/acp-agent/tests/snapshots/fs-read/session.jsonl
index 3af4b2ac61..81737364f8 100644
--- a/examples/acp-agent/tests/snapshots/fs-read/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-read/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352072468,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352072469,"data":{"content":[{"type":"text","text":"Use the read tool (NOT bash) to read the file greeting.txt in the current directory, then reply with exactly the single word DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352072470,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352072471,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352072471,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352073089,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352073090,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352073210,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -51,7 +51,7 @@
 {"type":"assistant/chunk","seq":49,"time":1783352073705,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_hHPZCcivsIkXAGS9jTGy8417","name":"read","arguments":"{\"file_path\": \"greeting.txt\"}"}}}}
 {"type":"assistant/chunk","seq":50,"time":1783352073705,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2882,"outputTokens":75,"cacheReadTokens":0,"reasoningTokens":29}}}}
 {"type":"assistant/chunk","seq":51,"time":1783352073705,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":52,"time":1783352073708,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to read the file greeting.txt using the read tool (not bash), then reply with exactly the single word \"DONE\"."},{"type":"tool-call","id":"call_00_hHPZCcivsIkXAGS9jTGy8417","name":"read","arguments":"{\"file_path\": \"greeting.txt\"}"}],"usage":{"inputTokens":2882,"outputTokens":75,"cacheReadTokens":0,"reasoningTokens":29}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51],"surfaceOp":"append"}
+{"type":"assistant/message","seq":52,"time":1783352073708,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to read the file greeting.txt using the read tool (not bash), then reply with exactly the single word \"DONE\"."},{"type":"tool-call","id":"call_00_hHPZCcivsIkXAGS9jTGy8417","name":"read","arguments":"{\"file_path\": \"greeting.txt\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2882,"outputTokens":75,"cacheReadTokens":0,"reasoningTokens":29}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51],"surfaceOp":"append"}
 {"type":"tool/call","seq":53,"time":1783352073709,"data":{"turn":1,"step":1,"callId":"call_00_hHPZCcivsIkXAGS9jTGy8417","name":"read","arguments":"{\"file_path\": \"greeting.txt\"}"}}
 {"type":"tool/result","seq":54,"time":1783352073717,"data":{"turn":1,"step":1,"callId":"call_00_hHPZCcivsIkXAGS9jTGy8417","content":[{"type":"text","text":"<path>/tmp/acp-snap-cwd-PEETkS/greeting.txt</path>\n<type>file</type>\n<content>\n1: hello\n\n(End of file - total 1 lines)\n</content>"}],"isError":false},"sourceEventSeqs":[53],"surfaceOp":"append"}
 {"type":"step/end","seq":55,"time":1783352073718,"data":{"turn":1,"step":1}}
@@ -101,6 +101,6 @@
 {"type":"assistant/chunk","seq":99,"time":1783352075045,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":100,"time":1783352075045,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":200,"outputTokens":40,"cacheReadTokens":2816,"reasoningTokens":37}}}}
 {"type":"assistant/chunk","seq":101,"time":1783352075045,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":102,"time":1783352075045,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user asked me to read the file and then reply with exactly the single word \"DONE\". I've read the file. Now I just need to reply with \"DONE\"."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":200,"outputTokens":40,"cacheReadTokens":2816,"reasoningTokens":37}},"sourceEventSeqs":[57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101],"surfaceOp":"append"}
+{"type":"assistant/message","seq":102,"time":1783352075045,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user asked me to read the file and then reply with exactly the single word \"DONE\". I've read the file. Now I just need to reply with \"DONE\"."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":200,"outputTokens":40,"cacheReadTokens":2816,"reasoningTokens":37}},"sourceEventSeqs":[57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101],"surfaceOp":"append"}
 {"type":"step/end","seq":103,"time":1783352075046,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":104,"time":1783352075046,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-read/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/fs-read/stdout.golden.jsonl
index fc0edbbe8b..d6b2e00b1e 100644
--- a/examples/acp-agent/tests/snapshots/fs-read/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-read/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-terminal-card/session.jsonl b/examples/acp-agent/tests/snapshots/fs-terminal-card/session.jsonl
index 9552a7a8f9..e53f4b3da3 100644
--- a/examples/acp-agent/tests/snapshots/fs-terminal-card/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-terminal-card/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352050753,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352050753,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly: echo TERMINAL_OK. Then reply with the single word DONE and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352050755,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352050756,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352050756,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352051421,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352051422,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352051590,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -58,7 +58,7 @@
 {"type":"assistant/chunk","seq":56,"time":1783352052118,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_fkbBRJsUrGKd1pWVc4Gn8233","name":"bash","arguments":"{\"command\": \"echo TERMINAL_OK\", \"description\": \"Echo TERMINAL_OK to verify terminal access\"}"}}}}
 {"type":"assistant/chunk","seq":57,"time":1783352052118,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2877,"outputTokens":90,"cacheReadTokens":0,"reasoningTokens":18}}}}
 {"type":"assistant/chunk","seq":58,"time":1783352052118,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":59,"time":1783352052121,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a simple bash command and then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_fkbBRJsUrGKd1pWVc4Gn8233","name":"bash","arguments":"{\"command\": \"echo TERMINAL_OK\", \"description\": \"Echo TERMINAL_OK to verify terminal access\"}"}],"usage":{"inputTokens":2877,"outputTokens":90,"cacheReadTokens":0,"reasoningTokens":18}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58],"surfaceOp":"append"}
+{"type":"assistant/message","seq":59,"time":1783352052121,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a simple bash command and then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_fkbBRJsUrGKd1pWVc4Gn8233","name":"bash","arguments":"{\"command\": \"echo TERMINAL_OK\", \"description\": \"Echo TERMINAL_OK to verify terminal access\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2877,"outputTokens":90,"cacheReadTokens":0,"reasoningTokens":18}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58],"surfaceOp":"append"}
 {"type":"tool/call","seq":60,"time":1783352052121,"data":{"turn":1,"step":1,"callId":"call_00_fkbBRJsUrGKd1pWVc4Gn8233","name":"bash","arguments":"{\"command\": \"echo TERMINAL_OK\", \"description\": \"Echo TERMINAL_OK to verify terminal access\"}"}}
 {"type":"tool/result","seq":61,"time":1783352052136,"data":{"turn":1,"step":1,"callId":"call_00_fkbBRJsUrGKd1pWVc4Gn8233","content":[{"type":"text","text":"TERMINAL_OK\n"}],"isError":false},"sourceEventSeqs":[60],"surfaceOp":"append"}
 {"type":"step/end","seq":62,"time":1783352052137,"data":{"turn":1,"step":1}}
@@ -93,6 +93,6 @@
 {"type":"assistant/chunk","seq":91,"time":1783352052986,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":92,"time":1783352052986,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":168,"outputTokens":25,"cacheReadTokens":2816,"reasoningTokens":22}}}}
 {"type":"assistant/chunk","seq":93,"time":1783352052986,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":94,"time":1783352052987,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The command ran successfully and output \"TERMINAL_OK\". I should now reply with just \"DONE\"."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":168,"outputTokens":25,"cacheReadTokens":2816,"reasoningTokens":22}},"sourceEventSeqs":[64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93],"surfaceOp":"append"}
+{"type":"assistant/message","seq":94,"time":1783352052987,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The command ran successfully and output \"TERMINAL_OK\". I should now reply with just \"DONE\"."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":168,"outputTokens":25,"cacheReadTokens":2816,"reasoningTokens":22}},"sourceEventSeqs":[64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93],"surfaceOp":"append"}
 {"type":"step/end","seq":95,"time":1783352052987,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":96,"time":1783352052987,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-terminal-card/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/fs-terminal-card/stdout.golden.jsonl
index 0e7dcca6f8..b35af2c650 100644
--- a/examples/acp-agent/tests/snapshots/fs-terminal-card/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-terminal-card/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl b/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl
index 47627ae7a5..08ec0f1323 100644
--- a/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352092220,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352092221,"data":{"content":[{"type":"text","text":"First use the read tool to read data.txt in the current directory. Then use the write tool (NOT bash) to replace its entire contents with exactly the single line: replaced. Then reply with exactly the single word DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352092223,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352092223,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352092223,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352092902,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352092902,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352093090,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -63,7 +63,7 @@
 {"type":"assistant/chunk","seq":61,"time":1783352093614,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_n4eRJuGoxNR07svgNtk82243","name":"read","arguments":"{\"file_path\": \"data.txt\"}"}}}}
 {"type":"assistant/chunk","seq":62,"time":1783352093614,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2899,"outputTokens":87,"cacheReadTokens":0,"reasoningTokens":42}}}}
 {"type":"assistant/chunk","seq":63,"time":1783352093614,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":64,"time":1783352093617,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to:\n1. Read data.txt using the read tool\n2. Replace its entire contents with exactly \"replaced\" using the write tool\n3. Reply with exactly \"DONE\""},{"type":"tool-call","id":"call_00_n4eRJuGoxNR07svgNtk82243","name":"read","arguments":"{\"file_path\": \"data.txt\"}"}],"usage":{"inputTokens":2899,"outputTokens":87,"cacheReadTokens":0,"reasoningTokens":42}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63],"surfaceOp":"append"}
+{"type":"assistant/message","seq":64,"time":1783352093617,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to:\n1. Read data.txt using the read tool\n2. Replace its entire contents with exactly \"replaced\" using the write tool\n3. Reply with exactly \"DONE\""},{"type":"tool-call","id":"call_00_n4eRJuGoxNR07svgNtk82243","name":"read","arguments":"{\"file_path\": \"data.txt\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2899,"outputTokens":87,"cacheReadTokens":0,"reasoningTokens":42}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63],"surfaceOp":"append"}
 {"type":"tool/call","seq":65,"time":1783352093617,"data":{"turn":1,"step":1,"callId":"call_00_n4eRJuGoxNR07svgNtk82243","name":"read","arguments":"{\"file_path\": \"data.txt\"}"}}
 {"type":"tool/result","seq":66,"time":1783352093624,"data":{"turn":1,"step":1,"callId":"call_00_n4eRJuGoxNR07svgNtk82243","content":[{"type":"text","text":"<path>/tmp/acp-snap-cwd-hH2sGY/data.txt</path>\n<type>file</type>\n<content>\n1: original contents\n\n(End of file - total 1 lines)\n</content>"}],"isError":false},"sourceEventSeqs":[65],"surfaceOp":"append"}
 {"type":"step/end","seq":67,"time":1783352093624,"data":{"turn":1,"step":1}}
@@ -112,7 +112,7 @@
 {"type":"assistant/chunk","seq":110,"time":1783352094988,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_N23EvXjDo4c8enyWpIUq4043","name":"write","arguments":"{\"file_path\": \"data.txt\", \"content\": \"replaced\"}"}}}}
 {"type":"assistant/chunk","seq":111,"time":1783352094988,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":228,"outputTokens":79,"cacheReadTokens":2816,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":112,"time":1783352094988,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":113,"time":1783352094988,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The file contains \"original contents\". Now I'll replace it with \"replaced\"."},{"type":"tool-call","id":"call_00_N23EvXjDo4c8enyWpIUq4043","name":"write","arguments":"{\"file_path\": \"data.txt\", \"content\": \"replaced\"}"}],"usage":{"inputTokens":228,"outputTokens":79,"cacheReadTokens":2816,"reasoningTokens":17}},"sourceEventSeqs":[69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112],"surfaceOp":"append"}
+{"type":"assistant/message","seq":113,"time":1783352094988,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The file contains \"original contents\". Now I'll replace it with \"replaced\"."},{"type":"tool-call","id":"call_00_N23EvXjDo4c8enyWpIUq4043","name":"write","arguments":"{\"file_path\": \"data.txt\", \"content\": \"replaced\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":228,"outputTokens":79,"cacheReadTokens":2816,"reasoningTokens":17}},"sourceEventSeqs":[69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112],"surfaceOp":"append"}
 {"type":"tool/call","seq":114,"time":1783352094988,"data":{"turn":1,"step":2,"callId":"call_00_N23EvXjDo4c8enyWpIUq4043","name":"write","arguments":"{\"file_path\": \"data.txt\", \"content\": \"replaced\"}"}}
 {"type":"tool/result","seq":115,"time":1783352094995,"data":{"turn":1,"step":2,"callId":"call_00_N23EvXjDo4c8enyWpIUq4043","content":[{"type":"text","text":"<path>/tmp/acp-snap-cwd-hH2sGY/data.txt</path>\n<type>file</type>\n<content>\nUpdated file\n</content>"}],"isError":false,"meta":{"diffs":[{"path":"data.txt","oldText":"original contents","newText":"replaced"}]}},"sourceEventSeqs":[114],"surfaceOp":"append"}
 {"type":"step/end","seq":116,"time":1783352094995,"data":{"turn":1,"step":2}}
@@ -141,6 +141,6 @@
 {"type":"assistant/chunk","seq":139,"time":1783352096309,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":140,"time":1783352096309,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":225,"outputTokens":19,"cacheReadTokens":2944,"reasoningTokens":16}}}}
 {"type":"assistant/chunk","seq":141,"time":1783352096309,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":142,"time":1783352096310,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"The file has been replaced successfully. Now I just reply with \"DONE\"."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":225,"outputTokens":19,"cacheReadTokens":2944,"reasoningTokens":16}},"sourceEventSeqs":[118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141],"surfaceOp":"append"}
+{"type":"assistant/message","seq":142,"time":1783352096310,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"The file has been replaced successfully. Now I just reply with \"DONE\"."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":225,"outputTokens":19,"cacheReadTokens":2944,"reasoningTokens":16}},"sourceEventSeqs":[118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141],"surfaceOp":"append"}
 {"type":"step/end","seq":143,"time":1783352096310,"data":{"turn":1,"step":3}}
 {"type":"turn/end","seq":144,"time":1783352096310,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-write-overwrite/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/fs-write-overwrite/stdout.golden.jsonl
index f0ca9674cb..c5dcf0a7b0 100644
--- a/examples/acp-agent/tests/snapshots/fs-write-overwrite/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-write-overwrite/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-write/session.jsonl b/examples/acp-agent/tests/snapshots/fs-write/session.jsonl
index 7e4b2dda01..59ba868817 100644
--- a/examples/acp-agent/tests/snapshots/fs-write/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-write/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352078754,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352078754,"data":{"content":[{"type":"text","text":"Use the write tool (NOT bash) to create a file named notes.txt in the current directory containing exactly the single line: hello world. Then reply with exactly the single word DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352078756,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352078756,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352078756,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352079254,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352079254,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352079333,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -60,7 +60,7 @@
 {"type":"assistant/chunk","seq":58,"time":1783352079886,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_APMUCJJm9lrTSlVbg6dB0185","name":"write","arguments":"{\"file_path\": \"notes.txt\", \"content\": \"hello world\"}"}}}}
 {"type":"assistant/chunk","seq":59,"time":1783352079886,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2891,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":30}}}}
 {"type":"assistant/chunk","seq":60,"time":1783352079886,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":61,"time":1783352079888,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to create a file named notes.txt with the content \"hello world\" using the write tool, then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_APMUCJJm9lrTSlVbg6dB0185","name":"write","arguments":"{\"file_path\": \"notes.txt\", \"content\": \"hello world\"}"}],"usage":{"inputTokens":2891,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":30}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60],"surfaceOp":"append"}
+{"type":"assistant/message","seq":61,"time":1783352079888,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to create a file named notes.txt with the content \"hello world\" using the write tool, then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_APMUCJJm9lrTSlVbg6dB0185","name":"write","arguments":"{\"file_path\": \"notes.txt\", \"content\": \"hello world\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2891,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":30}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60],"surfaceOp":"append"}
 {"type":"tool/call","seq":62,"time":1783352079888,"data":{"turn":1,"step":1,"callId":"call_00_APMUCJJm9lrTSlVbg6dB0185","name":"write","arguments":"{\"file_path\": \"notes.txt\", \"content\": \"hello world\"}"}}
 {"type":"tool/result","seq":63,"time":1783352079897,"data":{"turn":1,"step":1,"callId":"call_00_APMUCJJm9lrTSlVbg6dB0185","content":[{"type":"text","text":"<path>/tmp/acp-snap-cwd-sNvn5N/notes.txt</path>\n<type>file</type>\n<content>\nCreated file\n</content>"}],"isError":false},"sourceEventSeqs":[62],"surfaceOp":"append"}
 {"type":"step/end","seq":64,"time":1783352079898,"data":{"turn":1,"step":1}}
@@ -90,6 +90,6 @@
 {"type":"assistant/chunk","seq":88,"time":1783352081056,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":89,"time":1783352081056,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":214,"outputTokens":20,"cacheReadTokens":2816,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":90,"time":1783352081056,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":91,"time":1783352081057,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The file has been created. Now I just need to reply with \"DONE\"."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":214,"outputTokens":20,"cacheReadTokens":2816,"reasoningTokens":17}},"sourceEventSeqs":[66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90],"surfaceOp":"append"}
+{"type":"assistant/message","seq":91,"time":1783352081057,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The file has been created. Now I just need to reply with \"DONE\"."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":214,"outputTokens":20,"cacheReadTokens":2816,"reasoningTokens":17}},"sourceEventSeqs":[66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90],"surfaceOp":"append"}
 {"type":"step/end","seq":92,"time":1783352081057,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":93,"time":1783352081057,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/fs-write/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/fs-write/stdout.golden.jsonl
index 3c01ab158c..7969eeba3b 100644
--- a/examples/acp-agent/tests/snapshots/fs-write/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/fs-write/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/handshake/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/handshake/stdout.golden.jsonl
index e4c4984fc5..011175a871 100644
--- a/examples/acp-agent/tests/snapshots/handshake/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/handshake/stdout.golden.jsonl
@@ -1,2 +1,2 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-posttool-block/session.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-posttool-block/session.jsonl
index 1a72a2e5ef..07bd28c2cd 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-posttool-block/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-posttool-block/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352177366,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352177367,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly: echo HELLO. Report the tool result you got back verbatim, then stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352177368,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352177372,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352177372,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352178017,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352178018,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352178131,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -59,7 +59,7 @@
 {"type":"assistant/chunk","seq":57,"time":1783352178592,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_qcIzLImnOm5qiKOBJUqY5047","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO to stdout\"}"}}}}
 {"type":"assistant/chunk","seq":58,"time":1783352178592,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2879,"outputTokens":91,"cacheReadTokens":0,"reasoningTokens":23}}}}
 {"type":"assistant/chunk","seq":59,"time":1783352178592,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":60,"time":1783352178594,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run `echo HELLO` using the bash tool and report the result verbatim."},{"type":"tool-call","id":"call_00_qcIzLImnOm5qiKOBJUqY5047","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO to stdout\"}"}],"usage":{"inputTokens":2879,"outputTokens":91,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59],"surfaceOp":"append"}
+{"type":"assistant/message","seq":60,"time":1783352178594,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run `echo HELLO` using the bash tool and report the result verbatim."},{"type":"tool-call","id":"call_00_qcIzLImnOm5qiKOBJUqY5047","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO to stdout\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2879,"outputTokens":91,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59],"surfaceOp":"append"}
 {"type":"tool/call","seq":61,"time":1783352178594,"data":{"turn":1,"step":1,"callId":"call_00_qcIzLImnOm5qiKOBJUqY5047","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO to stdout\"}"}}
 {"type":"hook/invoked","seq":62,"time":1783352178614,"data":{"turn":1,"point":"PostToolUse","dialect":"claude","handlerId":"claude:PostToolUse:1","matcher":"bash"}}
 {"type":"hook/result","seq":63,"time":1783352178624,"data":{"turn":1,"point":"PostToolUse","handlerId":"claude:PostToolUse:1","decision":"block","exitCode":2,"stderrSummary":"tool output rejected by policy: rerun with a summary instead","durationMs":9.49630100000013}}
@@ -130,7 +130,7 @@
 {"type":"assistant/chunk","seq":128,"time":1783352180488,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_nzhKxYvCYj8OqI2tjOKb8506","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO command\"}"}}}}
 {"type":"assistant/chunk","seq":129,"time":1783352180489,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":178,"outputTokens":99,"cacheReadTokens":2816,"reasoningTokens":32}}}}
 {"type":"assistant/chunk","seq":130,"time":1783352180489,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":131,"time":1783352180489,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The tool output was rejected by policy, which says \"rerun with a summary instead\". Let me try again with a description that summarizes what we're doing."},{"type":"tool-call","id":"call_00_nzhKxYvCYj8OqI2tjOKb8506","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO command\"}"}],"usage":{"inputTokens":178,"outputTokens":99,"cacheReadTokens":2816,"reasoningTokens":32}},"sourceEventSeqs":[67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130],"surfaceOp":"append"}
+{"type":"assistant/message","seq":131,"time":1783352180489,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The tool output was rejected by policy, which says \"rerun with a summary instead\". Let me try again with a description that summarizes what we're doing."},{"type":"tool-call","id":"call_00_nzhKxYvCYj8OqI2tjOKb8506","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO command\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":178,"outputTokens":99,"cacheReadTokens":2816,"reasoningTokens":32}},"sourceEventSeqs":[67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130],"surfaceOp":"append"}
 {"type":"tool/call","seq":132,"time":1783352180489,"data":{"turn":1,"step":2,"callId":"call_00_nzhKxYvCYj8OqI2tjOKb8506","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO command\"}"}}
 {"type":"hook/invoked","seq":133,"time":1783352180524,"data":{"turn":1,"point":"PostToolUse","dialect":"claude","handlerId":"claude:PostToolUse:2","matcher":"bash"}}
 {"type":"hook/result","seq":134,"time":1783352180530,"data":{"turn":1,"point":"PostToolUse","handlerId":"claude:PostToolUse:2","decision":"block","exitCode":2,"stderrSummary":"tool output rejected by policy: rerun with a summary instead","durationMs":6.622174999999515}}
@@ -182,7 +182,7 @@
 {"type":"assistant/chunk","seq":180,"time":1783352181933,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_VumDDhhB4n5507EUXq650912","name":"bash","arguments":"{\"description\": \"Print HELLO to stdout\", \"command\": \"echo HELLO\"}"}}}}
 {"type":"assistant/chunk","seq":181,"time":1783352181933,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":173,"outputTokens":80,"cacheReadTokens":2944,"reasoningTokens":13}}}}
 {"type":"assistant/chunk","seq":182,"time":1783352181933,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":183,"time":1783352181934,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"The tool keeps rejecting. Let me try a slightly different description."},{"type":"tool-call","id":"call_00_VumDDhhB4n5507EUXq650912","name":"bash","arguments":"{\"description\": \"Print HELLO to stdout\", \"command\": \"echo HELLO\"}"}],"usage":{"inputTokens":173,"outputTokens":80,"cacheReadTokens":2944,"reasoningTokens":13}},"sourceEventSeqs":[138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182],"surfaceOp":"append"}
+{"type":"assistant/message","seq":183,"time":1783352181934,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"The tool keeps rejecting. Let me try a slightly different description."},{"type":"tool-call","id":"call_00_VumDDhhB4n5507EUXq650912","name":"bash","arguments":"{\"description\": \"Print HELLO to stdout\", \"command\": \"echo HELLO\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":173,"outputTokens":80,"cacheReadTokens":2944,"reasoningTokens":13}},"sourceEventSeqs":[138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182],"surfaceOp":"append"}
 {"type":"tool/call","seq":184,"time":1783352181934,"data":{"turn":1,"step":3,"callId":"call_00_VumDDhhB4n5507EUXq650912","name":"bash","arguments":"{\"description\": \"Print HELLO to stdout\", \"command\": \"echo HELLO\"}"}}
 {"type":"hook/invoked","seq":185,"time":1783352181945,"data":{"turn":1,"point":"PostToolUse","dialect":"claude","handlerId":"claude:PostToolUse:3","matcher":"bash"}}
 {"type":"hook/result","seq":186,"time":1783352181953,"data":{"turn":1,"point":"PostToolUse","handlerId":"claude:PostToolUse:3","decision":"block","exitCode":2,"stderrSummary":"tool output rejected by policy: rerun with a summary instead","durationMs":7.827785000000404}}
@@ -234,7 +234,7 @@
 {"type":"assistant/chunk","seq":232,"time":1783352183049,"data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_i5NCqPa2LGAy1wk8uJQR2369","name":"bash","arguments":"{\"description\": \"echo HELLO\", \"command\": \"echo HELLO\"}"}}}}
 {"type":"assistant/chunk","seq":233,"time":1783352183049,"data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":149,"outputTokens":80,"cacheReadTokens":3072,"reasoningTokens":15}}}}
 {"type":"assistant/chunk","seq":234,"time":1783352183049,"data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":235,"time":1783352183049,"data":{"turn":1,"step":4,"content":[{"type":"reasoning","text":"The policy keeps rejecting my attempts. Let me try a very short description."},{"type":"tool-call","id":"call_00_i5NCqPa2LGAy1wk8uJQR2369","name":"bash","arguments":"{\"description\": \"echo HELLO\", \"command\": \"echo HELLO\"}"}],"usage":{"inputTokens":149,"outputTokens":80,"cacheReadTokens":3072,"reasoningTokens":15}},"sourceEventSeqs":[190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234],"surfaceOp":"append"}
+{"type":"assistant/message","seq":235,"time":1783352183049,"data":{"turn":1,"step":4,"content":[{"type":"reasoning","text":"The policy keeps rejecting my attempts. Let me try a very short description."},{"type":"tool-call","id":"call_00_i5NCqPa2LGAy1wk8uJQR2369","name":"bash","arguments":"{\"description\": \"echo HELLO\", \"command\": \"echo HELLO\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":149,"outputTokens":80,"cacheReadTokens":3072,"reasoningTokens":15}},"sourceEventSeqs":[190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234],"surfaceOp":"append"}
 {"type":"tool/call","seq":236,"time":1783352183050,"data":{"turn":1,"step":4,"callId":"call_00_i5NCqPa2LGAy1wk8uJQR2369","name":"bash","arguments":"{\"description\": \"echo HELLO\", \"command\": \"echo HELLO\"}"}}
 {"type":"hook/invoked","seq":237,"time":1783352183069,"data":{"turn":1,"point":"PostToolUse","dialect":"claude","handlerId":"claude:PostToolUse:4","matcher":"bash"}}
 {"type":"hook/result","seq":238,"time":1783352183077,"data":{"turn":1,"point":"PostToolUse","handlerId":"claude:PostToolUse:4","decision":"block","exitCode":2,"stderrSummary":"tool output rejected by policy: rerun with a summary instead","durationMs":8.41934399999991}}
@@ -280,7 +280,7 @@
 {"type":"assistant/chunk","seq":278,"time":1783352184233,"data":{"turn":1,"step":5,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_0IuOIk6iuG6ZesROSyAM3669","name":"bash","arguments":"{\"command\": \"echo HELLO\"}"}}}}
 {"type":"assistant/chunk","seq":279,"time":1783352184233,"data":{"turn":1,"step":5,"chunk":{"type":"usage","usage":{"inputTokens":125,"outputTokens":66,"cacheReadTokens":3200,"reasoningTokens":20}}}}
 {"type":"assistant/chunk","seq":280,"time":1783352184233,"data":{"turn":1,"step":5,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":281,"time":1783352184233,"data":{"turn":1,"step":5,"content":[{"type":"reasoning","text":"Hmm, the policy keeps rejecting. Let me try without the description field, just the command."},{"type":"tool-call","id":"call_00_0IuOIk6iuG6ZesROSyAM3669","name":"bash","arguments":"{\"command\": \"echo HELLO\"}"}],"usage":{"inputTokens":125,"outputTokens":66,"cacheReadTokens":3200,"reasoningTokens":20}},"sourceEventSeqs":[242,243,244,245,246,247,248,249,250,251,252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280],"surfaceOp":"append"}
+{"type":"assistant/message","seq":281,"time":1783352184233,"data":{"turn":1,"step":5,"content":[{"type":"reasoning","text":"Hmm, the policy keeps rejecting. Let me try without the description field, just the command."},{"type":"tool-call","id":"call_00_0IuOIk6iuG6ZesROSyAM3669","name":"bash","arguments":"{\"command\": \"echo HELLO\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":125,"outputTokens":66,"cacheReadTokens":3200,"reasoningTokens":20}},"sourceEventSeqs":[242,243,244,245,246,247,248,249,250,251,252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280],"surfaceOp":"append"}
 {"type":"tool/call","seq":282,"time":1783352184233,"data":{"turn":1,"step":5,"callId":"call_00_0IuOIk6iuG6ZesROSyAM3669","name":"bash","arguments":"{\"command\": \"echo HELLO\"}"}}
 {"type":"hook/invoked","seq":283,"time":1783352184234,"data":{"turn":1,"point":"PostToolUse","dialect":"claude","handlerId":"claude:PostToolUse:5","matcher":"bash"}}
 {"type":"hook/result","seq":284,"time":1783352184243,"data":{"turn":1,"point":"PostToolUse","handlerId":"claude:PostToolUse:5","decision":"block","exitCode":2,"stderrSummary":"tool output rejected by policy: rerun with a summary instead","durationMs":8.810661999999866}}
@@ -428,7 +428,7 @@
 {"type":"assistant/chunk","seq":426,"time":1783352186527,"data":{"turn":1,"step":6,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_mGqWmySh60rWKNcyVBFk2747","name":"bash","arguments":"{\"description\": \"Echo the string HELLO\", \"command\": \"echo \\\"HELLO\\\"\"}"}}}}
 {"type":"assistant/chunk","seq":427,"time":1783352186527,"data":{"turn":1,"step":6,"chunk":{"type":"usage","usage":{"inputTokens":215,"outputTokens":176,"cacheReadTokens":3200,"reasoningTokens":106}}}}
 {"type":"assistant/chunk","seq":428,"time":1783352186527,"data":{"turn":1,"step":6,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":429,"time":1783352186527,"data":{"turn":1,"step":6,"content":[{"type":"reasoning","text":"The tool is consistently being rejected by policy. The policy says \"rerun with a summary instead\" - it seems the content of the output (which would be \"HELLO\") is being blocked. Let me try a different approach - maybe I need to use a different command or approach to satisfy the policy.\n\nWait, the user explicitly asked me to \"run exactly: echo HELLO\" and \"Report the tool result you got back verbatim\". Let me try `echo \"HELLO\"` with quotes."},{"type":"tool-call","id":"call_00_mGqWmySh60rWKNcyVBFk2747","name":"bash","arguments":"{\"description\": \"Echo the string HELLO\", \"command\": \"echo \\\"HELLO\\\"\"}"}],"usage":{"inputTokens":215,"outputTokens":176,"cacheReadTokens":3200,"reasoningTokens":106}},"sourceEventSeqs":[288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316,317,318,319,320,321,322,323,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349,350,351,352,353,354,355,356,357,358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374,375,376,377,378,379,380,381,382,383,384,385,386,387,388,389,390,391,392,393,394,395,396,397,398,399,400,401,402,403,404,405,406,407,408,409,410,411,412,413,414,415,416,417,418,419,420,421,422,423,424,425,426,427,428],"surfaceOp":"append"}
+{"type":"assistant/message","seq":429,"time":1783352186527,"data":{"turn":1,"step":6,"content":[{"type":"reasoning","text":"The tool is consistently being rejected by policy. The policy says \"rerun with a summary instead\" - it seems the content of the output (which would be \"HELLO\") is being blocked. Let me try a different approach - maybe I need to use a different command or approach to satisfy the policy.\n\nWait, the user explicitly asked me to \"run exactly: echo HELLO\" and \"Report the tool result you got back verbatim\". Let me try `echo \"HELLO\"` with quotes."},{"type":"tool-call","id":"call_00_mGqWmySh60rWKNcyVBFk2747","name":"bash","arguments":"{\"description\": \"Echo the string HELLO\", \"command\": \"echo \\\"HELLO\\\"\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":215,"outputTokens":176,"cacheReadTokens":3200,"reasoningTokens":106}},"sourceEventSeqs":[288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316,317,318,319,320,321,322,323,324,325,326,327,328,329,330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349,350,351,352,353,354,355,356,357,358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374,375,376,377,378,379,380,381,382,383,384,385,386,387,388,389,390,391,392,393,394,395,396,397,398,399,400,401,402,403,404,405,406,407,408,409,410,411,412,413,414,415,416,417,418,419,420,421,422,423,424,425,426,427,428],"surfaceOp":"append"}
 {"type":"tool/call","seq":430,"time":1783352186527,"data":{"turn":1,"step":6,"callId":"call_00_mGqWmySh60rWKNcyVBFk2747","name":"bash","arguments":"{\"description\": \"Echo the string HELLO\", \"command\": \"echo \\\"HELLO\\\"\"}"}}
 {"type":"hook/invoked","seq":431,"time":1783352186538,"data":{"turn":1,"point":"PostToolUse","dialect":"claude","handlerId":"claude:PostToolUse:6","matcher":"bash"}}
 {"type":"hook/result","seq":432,"time":1783352186545,"data":{"turn":1,"point":"PostToolUse","handlerId":"claude:PostToolUse:6","decision":"block","exitCode":2,"stderrSummary":"tool output rejected by policy: rerun with a summary instead","durationMs":6.910448999999062}}
@@ -549,7 +549,7 @@
 {"type":"assistant/chunk","seq":547,"time":1783352188511,"data":{"turn":1,"step":7,"chunk":{"type":"block-end","index":2,"block":{"type":"tool-call","id":"call_00_Y371DzUR0Pz6IrkO8qgB9050","name":"bash","arguments":"{\"description\": \"Echo test message\", \"command\": \"echo \\\"TEST\\\"\"}"}}}}
 {"type":"assistant/chunk","seq":548,"time":1783352188511,"data":{"turn":1,"step":7,"chunk":{"type":"usage","usage":{"inputTokens":287,"outputTokens":146,"cacheReadTokens":3328,"reasoningTokens":70}}}}
 {"type":"assistant/chunk","seq":549,"time":1783352188512,"data":{"turn":1,"step":7,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":550,"time":1783352188512,"data":{"turn":1,"step":7,"content":[{"type":"reasoning","text":"The policy is blocking the output \"HELLO\" - it seems like the word \"HELLO\" (which might be a misspelling of \"HELLO\" that looks like \"HELL\" with \"O\") could be triggering some content filter. Let me try a completely different word to see if the tool works at all."},{"type":"text","text":"Let me try a different word to verify the tool works:"},{"type":"tool-call","id":"call_00_Y371DzUR0Pz6IrkO8qgB9050","name":"bash","arguments":"{\"description\": \"Echo test message\", \"command\": \"echo \\\"TEST\\\"\"}"}],"usage":{"inputTokens":287,"outputTokens":146,"cacheReadTokens":3328,"reasoningTokens":70}},"sourceEventSeqs":[436,437,438,439,440,441,442,443,444,445,446,447,448,449,450,451,452,453,454,455,456,457,458,459,460,461,462,463,464,465,466,467,468,469,470,471,472,473,474,475,476,477,478,479,480,481,482,483,484,485,486,487,488,489,490,491,492,493,494,495,496,497,498,499,500,501,502,503,504,505,506,507,508,509,510,511,512,513,514,515,516,517,518,519,520,521,522,523,524,525,526,527,528,529,530,531,532,533,534,535,536,537,538,539,540,541,542,543,544,545,546,547,548,549],"surfaceOp":"append"}
+{"type":"assistant/message","seq":550,"time":1783352188512,"data":{"turn":1,"step":7,"content":[{"type":"reasoning","text":"The policy is blocking the output \"HELLO\" - it seems like the word \"HELLO\" (which might be a misspelling of \"HELLO\" that looks like \"HELL\" with \"O\") could be triggering some content filter. Let me try a completely different word to see if the tool works at all."},{"type":"text","text":"Let me try a different word to verify the tool works:"},{"type":"tool-call","id":"call_00_Y371DzUR0Pz6IrkO8qgB9050","name":"bash","arguments":"{\"description\": \"Echo test message\", \"command\": \"echo \\\"TEST\\\"\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":287,"outputTokens":146,"cacheReadTokens":3328,"reasoningTokens":70}},"sourceEventSeqs":[436,437,438,439,440,441,442,443,444,445,446,447,448,449,450,451,452,453,454,455,456,457,458,459,460,461,462,463,464,465,466,467,468,469,470,471,472,473,474,475,476,477,478,479,480,481,482,483,484,485,486,487,488,489,490,491,492,493,494,495,496,497,498,499,500,501,502,503,504,505,506,507,508,509,510,511,512,513,514,515,516,517,518,519,520,521,522,523,524,525,526,527,528,529,530,531,532,533,534,535,536,537,538,539,540,541,542,543,544,545,546,547,548,549],"surfaceOp":"append"}
 {"type":"tool/call","seq":551,"time":1783352188512,"data":{"turn":1,"step":7,"callId":"call_00_Y371DzUR0Pz6IrkO8qgB9050","name":"bash","arguments":"{\"description\": \"Echo test message\", \"command\": \"echo \\\"TEST\\\"\"}"}}
 {"type":"hook/invoked","seq":552,"time":1783352188524,"data":{"turn":1,"point":"PostToolUse","dialect":"claude","handlerId":"claude:PostToolUse:7","matcher":"bash"}}
 {"type":"hook/result","seq":553,"time":1783352188532,"data":{"turn":1,"point":"PostToolUse","handlerId":"claude:PostToolUse:7","decision":"block","exitCode":2,"stderrSummary":"tool output rejected by policy: rerun with a summary instead","durationMs":7.573905000001105}}
@@ -605,7 +605,7 @@
 {"type":"assistant/chunk","seq":603,"time":1783352189858,"data":{"turn":1,"step":8,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_S2XENjVanaHtBn3RZfhO7965","name":"bash","arguments":"{\"description\": \"Print working directory\", \"command\": \"pwd\"}"}}}}
 {"type":"assistant/chunk","seq":604,"time":1783352189858,"data":{"turn":1,"step":8,"chunk":{"type":"usage","usage":{"inputTokens":201,"outputTokens":84,"cacheReadTokens":3584,"reasoningTokens":22}}}}
 {"type":"assistant/chunk","seq":605,"time":1783352189858,"data":{"turn":1,"step":8,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":606,"time":1783352189859,"data":{"turn":1,"step":8,"content":[{"type":"reasoning","text":"Even \"TEST\" is rejected. Let me try a simpler command like `pwd` or `ls`."},{"type":"tool-call","id":"call_00_S2XENjVanaHtBn3RZfhO7965","name":"bash","arguments":"{\"description\": \"Print working directory\", \"command\": \"pwd\"}"}],"usage":{"inputTokens":201,"outputTokens":84,"cacheReadTokens":3584,"reasoningTokens":22}},"sourceEventSeqs":[557,558,559,560,561,562,563,564,565,566,567,568,569,570,571,572,573,574,575,576,577,578,579,580,581,582,583,584,585,586,587,588,589,590,591,592,593,594,595,596,597,598,599,600,601,602,603,604,605],"surfaceOp":"append"}
+{"type":"assistant/message","seq":606,"time":1783352189859,"data":{"turn":1,"step":8,"content":[{"type":"reasoning","text":"Even \"TEST\" is rejected. Let me try a simpler command like `pwd` or `ls`."},{"type":"tool-call","id":"call_00_S2XENjVanaHtBn3RZfhO7965","name":"bash","arguments":"{\"description\": \"Print working directory\", \"command\": \"pwd\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":201,"outputTokens":84,"cacheReadTokens":3584,"reasoningTokens":22}},"sourceEventSeqs":[557,558,559,560,561,562,563,564,565,566,567,568,569,570,571,572,573,574,575,576,577,578,579,580,581,582,583,584,585,586,587,588,589,590,591,592,593,594,595,596,597,598,599,600,601,602,603,604,605],"surfaceOp":"append"}
 {"type":"tool/call","seq":607,"time":1783352189859,"data":{"turn":1,"step":8,"callId":"call_00_S2XENjVanaHtBn3RZfhO7965","name":"bash","arguments":"{\"description\": \"Print working directory\", \"command\": \"pwd\"}"}}
 {"type":"hook/invoked","seq":608,"time":1783352189876,"data":{"turn":1,"point":"PostToolUse","dialect":"claude","handlerId":"claude:PostToolUse:8","matcher":"bash"}}
 {"type":"hook/result","seq":609,"time":1783352189883,"data":{"turn":1,"point":"PostToolUse","handlerId":"claude:PostToolUse:8","decision":"block","exitCode":2,"stderrSummary":"tool output rejected by policy: rerun with a summary instead","durationMs":7.748225999999704}}
@@ -747,6 +747,6 @@
 {"type":"assistant/chunk","seq":745,"time":1783352192246,"data":{"turn":1,"step":9,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"I'm unable to run the `echo HELLO` command via the bash tool — every invocation is being rejected by the policy with the message \"tool output rejected by policy: rerun with a summary instead\". This appears to be a systemic policy restriction on the bash tool in this environment, as even simple commands like `pwd` and `echo \"TEST\"` are blocked.\n\nI cannot produce the tool result you asked for because the bash tool is not currently available for use."}}}}
 {"type":"assistant/chunk","seq":746,"time":1783352192246,"data":{"turn":1,"step":9,"chunk":{"type":"usage","usage":{"inputTokens":181,"outputTokens":130,"cacheReadTokens":3712,"reasoningTokens":31}}}}
 {"type":"assistant/chunk","seq":747,"time":1783352192246,"data":{"turn":1,"step":9,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":748,"time":1783352192247,"data":{"turn":1,"step":9,"content":[{"type":"reasoning","text":"All bash commands are being rejected. This seems like a system-level policy issue where the bash tool is blocked entirely. Let me report this to the user."},{"type":"text","text":"I'm unable to run the `echo HELLO` command via the bash tool — every invocation is being rejected by the policy with the message \"tool output rejected by policy: rerun with a summary instead\". This appears to be a systemic policy restriction on the bash tool in this environment, as even simple commands like `pwd` and `echo \"TEST\"` are blocked.\n\nI cannot produce the tool result you asked for because the bash tool is not currently available for use."}],"usage":{"inputTokens":181,"outputTokens":130,"cacheReadTokens":3712,"reasoningTokens":31}},"sourceEventSeqs":[613,614,615,616,617,618,619,620,621,622,623,624,625,626,627,628,629,630,631,632,633,634,635,636,637,638,639,640,641,642,643,644,645,646,647,648,649,650,651,652,653,654,655,656,657,658,659,660,661,662,663,664,665,666,667,668,669,670,671,672,673,674,675,676,677,678,679,680,681,682,683,684,685,686,687,688,689,690,691,692,693,694,695,696,697,698,699,700,701,702,703,704,705,706,707,708,709,710,711,712,713,714,715,716,717,718,719,720,721,722,723,724,725,726,727,728,729,730,731,732,733,734,735,736,737,738,739,740,741,742,743,744,745,746,747],"surfaceOp":"append"}
+{"type":"assistant/message","seq":748,"time":1783352192247,"data":{"turn":1,"step":9,"content":[{"type":"reasoning","text":"All bash commands are being rejected. This seems like a system-level policy issue where the bash tool is blocked entirely. Let me report this to the user."},{"type":"text","text":"I'm unable to run the `echo HELLO` command via the bash tool — every invocation is being rejected by the policy with the message \"tool output rejected by policy: rerun with a summary instead\". This appears to be a systemic policy restriction on the bash tool in this environment, as even simple commands like `pwd` and `echo \"TEST\"` are blocked.\n\nI cannot produce the tool result you asked for because the bash tool is not currently available for use."}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":181,"outputTokens":130,"cacheReadTokens":3712,"reasoningTokens":31}},"sourceEventSeqs":[613,614,615,616,617,618,619,620,621,622,623,624,625,626,627,628,629,630,631,632,633,634,635,636,637,638,639,640,641,642,643,644,645,646,647,648,649,650,651,652,653,654,655,656,657,658,659,660,661,662,663,664,665,666,667,668,669,670,671,672,673,674,675,676,677,678,679,680,681,682,683,684,685,686,687,688,689,690,691,692,693,694,695,696,697,698,699,700,701,702,703,704,705,706,707,708,709,710,711,712,713,714,715,716,717,718,719,720,721,722,723,724,725,726,727,728,729,730,731,732,733,734,735,736,737,738,739,740,741,742,743,744,745,746,747],"surfaceOp":"append"}
 {"type":"step/end","seq":749,"time":1783352192247,"data":{"turn":1,"step":9}}
 {"type":"turn/end","seq":750,"time":1783352192247,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-posttool-block/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-posttool-block/stdout.golden.jsonl
index 8c97f359b7..90ab26d54e 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-posttool-block/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-posttool-block/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-posttool-context/session.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-posttool-context/session.jsonl
index 6c98868284..2e167db28c 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-posttool-context/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-posttool-context/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352196662,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352196662,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly: echo HELLO. Report the tool result you got back verbatim, then stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352196664,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352196664,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352196664,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352197315,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352197315,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352197457,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -57,7 +57,7 @@
 {"type":"assistant/chunk","seq":55,"time":1783352197953,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_HbCMzTslWBZTSphWN0z97382","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}}}}
 {"type":"assistant/chunk","seq":56,"time":1783352197953,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2878,"outputTokens":89,"cacheReadTokens":0,"reasoningTokens":23}}}}
 {"type":"assistant/chunk","seq":57,"time":1783352197954,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":58,"time":1783352197956,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run `echo HELLO` using the bash tool and report the result verbatim."},{"type":"tool-call","id":"call_00_HbCMzTslWBZTSphWN0z97382","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}],"usage":{"inputTokens":2878,"outputTokens":89,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57],"surfaceOp":"append"}
+{"type":"assistant/message","seq":58,"time":1783352197956,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run `echo HELLO` using the bash tool and report the result verbatim."},{"type":"tool-call","id":"call_00_HbCMzTslWBZTSphWN0z97382","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2878,"outputTokens":89,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57],"surfaceOp":"append"}
 {"type":"tool/call","seq":59,"time":1783352197956,"data":{"turn":1,"step":1,"callId":"call_00_HbCMzTslWBZTSphWN0z97382","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}}
 {"type":"hook/invoked","seq":60,"time":1783352197968,"data":{"turn":1,"point":"PostToolUse","dialect":"claude","handlerId":"claude:PostToolUse:1","matcher":"bash"}}
 {"type":"hook/result","seq":61,"time":1783352197976,"data":{"turn":1,"point":"PostToolUse","handlerId":"claude:PostToolUse:1","decision":"pass","exitCode":0,"durationMs":7.404540999999881}}
@@ -121,6 +121,6 @@
 {"type":"assistant/chunk","seq":119,"time":1783352199410,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"The tool result was:\n\n```\nHELLO\n```\n\nIt completed successfully with exit code 0."}}}}
 {"type":"assistant/chunk","seq":120,"time":1783352199411,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":188,"outputTokens":51,"cacheReadTokens":2816,"reasoningTokens":30}}}}
 {"type":"assistant/chunk","seq":121,"time":1783352199411,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":122,"time":1783352199411,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user asked me to report the tool result verbatim. The result was \"HELLO\" with an exit code of 0 (success)."},{"type":"text","text":"The tool result was:\n\n```\nHELLO\n```\n\nIt completed successfully with exit code 0."}],"usage":{"inputTokens":188,"outputTokens":51,"cacheReadTokens":2816,"reasoningTokens":30}},"sourceEventSeqs":[66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121],"surfaceOp":"append"}
+{"type":"assistant/message","seq":122,"time":1783352199411,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user asked me to report the tool result verbatim. The result was \"HELLO\" with an exit code of 0 (success)."},{"type":"text","text":"The tool result was:\n\n```\nHELLO\n```\n\nIt completed successfully with exit code 0."}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":188,"outputTokens":51,"cacheReadTokens":2816,"reasoningTokens":30}},"sourceEventSeqs":[66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121],"surfaceOp":"append"}
 {"type":"step/end","seq":123,"time":1783352199411,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":124,"time":1783352199412,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-posttool-context/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-posttool-context/stdout.golden.jsonl
index e9243d8a05..e60a33bcd5 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-posttool-context/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-posttool-context/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl
index 3933a939ba..3da0d9171e 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352171519,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352171520,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly: echo HELLO. Report the tool result you got back verbatim, then stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352171527,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352171528,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352171528,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352171991,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352171991,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352172088,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -51,12 +51,12 @@
 {"type":"assistant/chunk","seq":49,"time":1783352172555,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_6k0oGSliVHxGSgqBmMEO4311","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO\"}"}}}}
 {"type":"assistant/chunk","seq":50,"time":1783352172555,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2878,"outputTokens":83,"cacheReadTokens":0,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":51,"time":1783352172555,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":52,"time":1783352172557,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a simple bash command and report the result verbatim."},{"type":"tool-call","id":"call_00_6k0oGSliVHxGSgqBmMEO4311","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO\"}"}],"usage":{"inputTokens":2878,"outputTokens":83,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51],"surfaceOp":"append"}
+{"type":"assistant/message","seq":52,"time":1783352172557,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a simple bash command and report the result verbatim."},{"type":"tool-call","id":"call_00_6k0oGSliVHxGSgqBmMEO4311","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2878,"outputTokens":83,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51],"surfaceOp":"append"}
 {"type":"tool/call","seq":53,"time":1783352172557,"data":{"turn":1,"step":1,"callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO\"}"}}
 {"type":"hook/invoked","seq":54,"time":1783352172558,"data":{"turn":1,"point":"PreToolUse","dialect":"claude","handlerId":"claude:PreToolUse:1","matcher":"bash"}}
 {"type":"hook/result","seq":55,"time":1783352172573,"data":{"turn":1,"point":"PreToolUse","handlerId":"claude:PreToolUse:1","decision":"ask","exitCode":0,"durationMs":14.113374999999905}}
-{"type":"approval/asked","seq":56,"time":1783962235813,"data":{"id":"cdd11a3a-c721-4d08-8255-732218775c33","toolName":"bash","callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","reason":"bash requires manual approval in this session"}}
-{"type":"approval/decided","seq":57,"time":1783962235813,"data":{"id":"cdd11a3a-c721-4d08-8255-732218775c33","outcome":"rejected"}}
+{"type":"approval/asked","seq":56,"time":1783962235813,"data":{"id":"281bce21-7aee-4642-88a9-56917e50829c","toolName":"bash","callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","reason":"bash requires manual approval in this session"}}
+{"type":"approval/decided","seq":57,"time":1783962235813,"data":{"id":"281bce21-7aee-4642-88a9-56917e50829c","outcome":"rejected"}}
 {"type":"tool/result","seq":58,"time":1783962235814,"data":{"turn":1,"step":1,"callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","content":[{"type":"text","text":"Error: the user rejected tool \"bash\""}],"isError":true},"sourceEventSeqs":[53],"surfaceOp":"append"}
 {"type":"step/end","seq":59,"time":1783962235814,"data":{"turn":1,"step":1}}
 {"type":"step/start","seq":60,"time":1783962235814,"data":{"turn":1,"step":2}}
@@ -110,6 +110,6 @@
 {"type":"assistant/chunk","seq":108,"time":1783352173964,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"The tool result I got back verbatim is:\n\n```\nError: bash requires manual approval in this session\n```"}}}}
 {"type":"assistant/chunk","seq":109,"time":1783962235816,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":166,"outputTokens":45,"cacheReadTokens":2816,"reasoningTokens":22}}}}
 {"type":"assistant/chunk","seq":110,"time":1783962235816,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":111,"time":1783962235816,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The bash tool returned an error saying it requires manual approval in this session. I'll report this verbatim."},{"type":"text","text":"The tool result I got back verbatim is:\n\n```\nError: bash requires manual approval in this session\n```"}],"usage":{"inputTokens":166,"outputTokens":45,"cacheReadTokens":2816,"reasoningTokens":22}},"sourceEventSeqs":[61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110],"surfaceOp":"append"}
+{"type":"assistant/message","seq":111,"time":1783962235816,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The bash tool returned an error saying it requires manual approval in this session. I'll report this verbatim."},{"type":"text","text":"The tool result I got back verbatim is:\n\n```\nError: bash requires manual approval in this session\n```"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":166,"outputTokens":45,"cacheReadTokens":2816,"reasoningTokens":22}},"sourceEventSeqs":[61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110],"surfaceOp":"append"}
 {"type":"step/end","seq":112,"time":1783962235816,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":113,"time":1783962235816,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/stdout.golden.jsonl
index 3905e82ca7..a748546e51 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-pretool-deny/session.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-pretool-deny/session.jsonl
index b8f30ee327..9633075e5b 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-pretool-deny/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-pretool-deny/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352165195,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352165196,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly: echo HELLO. Report the tool result you got back verbatim, then stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352165198,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352165199,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352165199,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352165899,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352165899,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352166048,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -51,7 +51,7 @@
 {"type":"assistant/chunk","seq":49,"time":1783352166512,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_JliP571Bh0QQ8QExbSPk0080","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}}}}
 {"type":"assistant/chunk","seq":50,"time":1783352166512,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2878,"outputTokens":83,"cacheReadTokens":0,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":51,"time":1783352166512,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":52,"time":1783352166514,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a simple bash command and report the result verbatim."},{"type":"tool-call","id":"call_00_JliP571Bh0QQ8QExbSPk0080","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}],"usage":{"inputTokens":2878,"outputTokens":83,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51],"surfaceOp":"append"}
+{"type":"assistant/message","seq":52,"time":1783352166514,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a simple bash command and report the result verbatim."},{"type":"tool-call","id":"call_00_JliP571Bh0QQ8QExbSPk0080","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2878,"outputTokens":83,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51],"surfaceOp":"append"}
 {"type":"tool/call","seq":53,"time":1783352166514,"data":{"turn":1,"step":1,"callId":"call_00_JliP571Bh0QQ8QExbSPk0080","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}}
 {"type":"hook/invoked","seq":54,"time":1783352166515,"data":{"turn":1,"point":"PreToolUse","dialect":"claude","handlerId":"claude:PreToolUse:1","matcher":"bash"}}
 {"type":"hook/result","seq":55,"time":1783352166528,"data":{"turn":1,"point":"PreToolUse","handlerId":"claude:PreToolUse:1","decision":"block","exitCode":2,"stderrSummary":"bash is disabled by policy in this session","durationMs":12.367074000000684}}
@@ -115,6 +115,6 @@
 {"type":"assistant/chunk","seq":113,"time":1783352167933,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"The tool returned:\n\n> Error: bash is disabled by policy in this session\n\nI cannot run the command because the bash tool is disabled by policy."}}}}
 {"type":"assistant/chunk","seq":114,"time":1783352167933,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":167,"outputTokens":52,"cacheReadTokens":2816,"reasoningTokens":21}}}}
 {"type":"assistant/chunk","seq":115,"time":1783352167933,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":116,"time":1783352167934,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The bash tool is disabled by policy. I need to report this error verbatim back to the user."},{"type":"text","text":"The tool returned:\n\n> Error: bash is disabled by policy in this session\n\nI cannot run the command because the bash tool is disabled by policy."}],"usage":{"inputTokens":167,"outputTokens":52,"cacheReadTokens":2816,"reasoningTokens":21}},"sourceEventSeqs":[59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115],"surfaceOp":"append"}
+{"type":"assistant/message","seq":116,"time":1783352167934,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The bash tool is disabled by policy. I need to report this error verbatim back to the user."},{"type":"text","text":"The tool returned:\n\n> Error: bash is disabled by policy in this session\n\nI cannot run the command because the bash tool is disabled by policy."}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":167,"outputTokens":52,"cacheReadTokens":2816,"reasoningTokens":21}},"sourceEventSeqs":[59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115],"surfaceOp":"append"}
 {"type":"step/end","seq":117,"time":1783352167934,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":118,"time":1783352167934,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-pretool-deny/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-pretool-deny/stdout.golden.jsonl
index 92906adb59..60ae1f14e4 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-pretool-deny/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-pretool-deny/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-block/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-block/stdout.golden.jsonl
index 6304582220..765dd87f7f 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-block/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-block/stdout.golden.jsonl
@@ -1,3 +1,3 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","id":3,"result":{"stopReason":"cancelled"}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-context/session.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-context/session.jsonl
index 3e07124e41..886761da18 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-context/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-context/session.jsonl
@@ -5,7 +5,7 @@
 {"type":"user/message","seq":3,"time":1783352160564,"data":{"content":[{"type":"text","text":"What is my favorite color? Reply with just the color and stop. Do not use any tools."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"context/message","seq":4,"time":1783352160564,"data":{"content":[{"type":"text","text":"The user has previously stated their favorite color is teal."}],"source":{"kind":"plugin","plugin":"hooks-claude"}},"surfaceOp":"append"}
 {"type":"step/start","seq":5,"time":1783352160565,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":6,"time":1783352160566,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":6,"time":1783352160566,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":7,"time":1783352161228,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":8,"time":1783352161229,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":9,"time":1783352161335,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -33,6 +33,6 @@
 {"type":"assistant/chunk","seq":31,"time":1783352161512,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"teal"}}}}
 {"type":"assistant/chunk","seq":32,"time":1783352161512,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2892,"outputTokens":22,"cacheReadTokens":0,"reasoningTokens":19}}}}
 {"type":"assistant/chunk","seq":33,"time":1783352161512,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":34,"time":1783352161515,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user's favorite color is teal, as stated in the context provided by the plugin."},{"type":"text","text":"teal"}],"usage":{"inputTokens":2892,"outputTokens":22,"cacheReadTokens":0,"reasoningTokens":19}},"sourceEventSeqs":[7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33],"surfaceOp":"append"}
+{"type":"assistant/message","seq":34,"time":1783352161515,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user's favorite color is teal, as stated in the context provided by the plugin."},{"type":"text","text":"teal"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2892,"outputTokens":22,"cacheReadTokens":0,"reasoningTokens":19}},"sourceEventSeqs":[7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33],"surfaceOp":"append"}
 {"type":"step/end","seq":35,"time":1783352161516,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":36,"time":1783352161516,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-context/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-context/stdout.golden.jsonl
index c4ad3ba541..e609fc0d70 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-context/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-promptsubmit-context/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"'s"}}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-stop-continue/session.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-stop-continue/session.jsonl
index 2bed8d839a..5f6b0e6e9e 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-stop-continue/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-stop-continue/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352203369,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352203370,"data":{"content":[{"type":"text","text":"Reply with the single word FIRST and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352203371,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352203372,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352203372,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352204036,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352204036,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352204247,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -28,7 +28,7 @@
 {"type":"assistant/chunk","seq":26,"time":1783352204393,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"FIRST"}}}}
 {"type":"assistant/chunk","seq":27,"time":1783352204393,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2862,"outputTokens":20,"cacheReadTokens":0,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":28,"time":1783352204393,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":29,"time":1783352204396,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with the single word \"FIRST\" and stop."},{"type":"text","text":"FIRST"}],"usage":{"inputTokens":2862,"outputTokens":20,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28],"surfaceOp":"append"}
+{"type":"assistant/message","seq":29,"time":1783352204396,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with the single word \"FIRST\" and stop."},{"type":"text","text":"FIRST"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2862,"outputTokens":20,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28],"surfaceOp":"append"}
 {"type":"step/end","seq":30,"time":1783352204396,"data":{"turn":1,"step":1}}
 {"type":"hook/invoked","seq":31,"time":1783352204396,"data":{"turn":1,"point":"Stop","dialect":"claude","handlerId":"claude:Stop:1"}}
 {"type":"hook/result","seq":32,"time":1783352204443,"data":{"turn":1,"point":"Stop","handlerId":"claude:Stop:1","decision":"block","exitCode":2,"stderrSummary":"Also reply with the single word SECOND, then stop.","durationMs":47.01462700000047}}
@@ -185,7 +185,7 @@
 {"type":"assistant/chunk","seq":183,"time":1783352206162,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"FIRST"}}}}
 {"type":"assistant/chunk","seq":184,"time":1783352206162,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":73,"outputTokens":146,"cacheReadTokens":2816,"reasoningTokens":143}}}}
 {"type":"assistant/chunk","seq":185,"time":1783352206162,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":186,"time":1783352206162,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user wants me to reply with \"FIRST\" only, but there's a steering input telling me to also reply with \"SECOND\". However, the user's explicit instruction is to reply with the single word FIRST and stop. The steering is from a plugin, not the user themselves. I should follow the user's original instruction.\n\nWait, let me re-read. The user said \"Reply with the single word FIRST and stop.\" Then there's a steering from \"plugin\" saying \"Also reply with the single word SECOND, then stop.\"\n\nThe user's instruction is clear: reply with FIRST and stop. The steering is from a plugin source, not the user. I'll follow the user's instruction."},{"type":"text","text":"FIRST"}],"usage":{"inputTokens":73,"outputTokens":146,"cacheReadTokens":2816,"reasoningTokens":143}},"sourceEventSeqs":[35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185],"surfaceOp":"append"}
+{"type":"assistant/message","seq":186,"time":1783352206162,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user wants me to reply with \"FIRST\" only, but there's a steering input telling me to also reply with \"SECOND\". However, the user's explicit instruction is to reply with the single word FIRST and stop. The steering is from a plugin, not the user themselves. I should follow the user's original instruction.\n\nWait, let me re-read. The user said \"Reply with the single word FIRST and stop.\" Then there's a steering from \"plugin\" saying \"Also reply with the single word SECOND, then stop.\"\n\nThe user's instruction is clear: reply with FIRST and stop. The steering is from a plugin source, not the user. I'll follow the user's instruction."},{"type":"text","text":"FIRST"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":73,"outputTokens":146,"cacheReadTokens":2816,"reasoningTokens":143}},"sourceEventSeqs":[35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185],"surfaceOp":"append"}
 {"type":"step/end","seq":187,"time":1783352206163,"data":{"turn":1,"step":2}}
 {"type":"hook/invoked","seq":188,"time":1783352206163,"data":{"turn":1,"point":"Stop","dialect":"claude","handlerId":"claude:Stop:2"}}
 {"type":"hook/result","seq":189,"time":1783352206190,"data":{"turn":1,"point":"Stop","handlerId":"claude:Stop:2","decision":"pass","exitCode":0,"durationMs":26.904655000000275}}
diff --git a/examples/acp-agent/tests/snapshots/hook-cc-stop-continue/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-stop-continue/stdout.golden.jsonl
index 5c4a564d26..c7e8beeab3 100644
--- a/examples/acp-agent/tests/snapshots/hook-cc-stop-continue/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-cc-stop-continue/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-posttool-block/session.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-posttool-block/session.jsonl
index 0936881298..4272fe6a94 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-posttool-block/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-posttool-block/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352220747,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352220748,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly: echo HELLO. Report the tool result you got back verbatim, then stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352220749,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352220750,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352220750,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352221451,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352221451,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352221651,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -60,7 +60,7 @@
 {"type":"assistant/chunk","seq":58,"time":1783352222121,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_wNJIQDMLdssZp45zIXvz2684","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Print HELLO to stdout\"}"}}}}
 {"type":"assistant/chunk","seq":59,"time":1783352222121,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2878,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":25}}}}
 {"type":"assistant/chunk","seq":60,"time":1783352222121,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":61,"time":1783352222124,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run the command \"echo HELLO\" using the bash tool and report the result verbatim."},{"type":"tool-call","id":"call_00_wNJIQDMLdssZp45zIXvz2684","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Print HELLO to stdout\"}"}],"usage":{"inputTokens":2878,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":25}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60],"surfaceOp":"append"}
+{"type":"assistant/message","seq":61,"time":1783352222124,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run the command \"echo HELLO\" using the bash tool and report the result verbatim."},{"type":"tool-call","id":"call_00_wNJIQDMLdssZp45zIXvz2684","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Print HELLO to stdout\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2878,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":25}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60],"surfaceOp":"append"}
 {"type":"tool/call","seq":62,"time":1783352222124,"data":{"turn":1,"step":1,"callId":"call_00_wNJIQDMLdssZp45zIXvz2684","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Print HELLO to stdout\"}"}}
 {"type":"hook/invoked","seq":63,"time":1783352222138,"data":{"turn":1,"point":"PostToolUse","dialect":"codex","handlerId":"codex:PostToolUse:1","matcher":"bash"}}
 {"type":"hook/result","seq":64,"time":1783352222148,"data":{"turn":1,"point":"PostToolUse","handlerId":"codex:PostToolUse:1","decision":"block","exitCode":2,"stderrSummary":"tool output rejected by codex policy: summarize instead","durationMs":9.571565000000192}}
@@ -217,6 +217,6 @@
 {"type":"assistant/chunk","seq":215,"time":1783352224654,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"The tool result I got back is:\n\n> `tool output rejected by codex policy: summarize instead`\n\nThis is the verbatim result from the tool — it was not the expected `\"HELLO\"` output because the codex policy blocked it and instructed me to summarize instead."}}}}
 {"type":"assistant/chunk","seq":216,"time":1783352224654,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":176,"outputTokens":145,"cacheReadTokens":2816,"reasoningTokens":86}}}}
 {"type":"assistant/chunk","seq":217,"time":1783352224654,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":218,"time":1783352224655,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The output was rejected by the codex policy with the instruction to \"summarize instead\". This means I cannot show the raw output verbatim as the user requested. Let me try to comply but report what happened.\n\nActually, the user asked me to report the tool result verbatim. The tool result I got back is:\n\n\"tool output rejected by codex policy: summarize instead\"\n\nSo I should report that verbatim."},{"type":"text","text":"The tool result I got back is:\n\n> `tool output rejected by codex policy: summarize instead`\n\nThis is the verbatim result from the tool — it was not the expected `\"HELLO\"` output because the codex policy blocked it and instructed me to summarize instead."}],"usage":{"inputTokens":176,"outputTokens":145,"cacheReadTokens":2816,"reasoningTokens":86}},"sourceEventSeqs":[68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217],"surfaceOp":"append"}
+{"type":"assistant/message","seq":218,"time":1783352224655,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The output was rejected by the codex policy with the instruction to \"summarize instead\". This means I cannot show the raw output verbatim as the user requested. Let me try to comply but report what happened.\n\nActually, the user asked me to report the tool result verbatim. The tool result I got back is:\n\n\"tool output rejected by codex policy: summarize instead\"\n\nSo I should report that verbatim."},{"type":"text","text":"The tool result I got back is:\n\n> `tool output rejected by codex policy: summarize instead`\n\nThis is the verbatim result from the tool — it was not the expected `\"HELLO\"` output because the codex policy blocked it and instructed me to summarize instead."}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":176,"outputTokens":145,"cacheReadTokens":2816,"reasoningTokens":86}},"sourceEventSeqs":[68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217],"surfaceOp":"append"}
 {"type":"step/end","seq":219,"time":1783352224655,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":220,"time":1783352224655,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-posttool-block/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-posttool-block/stdout.golden.jsonl
index bd1e04bc59..6257014596 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-posttool-block/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-posttool-block/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-posttool-context/session.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-posttool-context/session.jsonl
index 6932a1f47e..d60858ee0a 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-posttool-context/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-posttool-context/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352228441,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352228442,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly: echo HELLO. Report the tool result you got back verbatim, then stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352228443,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352228443,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352228443,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352228985,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352228985,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352229106,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -57,7 +57,7 @@
 {"type":"assistant/chunk","seq":55,"time":1783352229598,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_Q6wHtakaip2QNfIXaVJY5458","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}}}}
 {"type":"assistant/chunk","seq":56,"time":1783352229598,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2878,"outputTokens":89,"cacheReadTokens":0,"reasoningTokens":23}}}}
 {"type":"assistant/chunk","seq":57,"time":1783352229598,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":58,"time":1783352229601,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run `echo HELLO` using the bash tool and report the result verbatim."},{"type":"tool-call","id":"call_00_Q6wHtakaip2QNfIXaVJY5458","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}],"usage":{"inputTokens":2878,"outputTokens":89,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57],"surfaceOp":"append"}
+{"type":"assistant/message","seq":58,"time":1783352229601,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run `echo HELLO` using the bash tool and report the result verbatim."},{"type":"tool-call","id":"call_00_Q6wHtakaip2QNfIXaVJY5458","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2878,"outputTokens":89,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57],"surfaceOp":"append"}
 {"type":"tool/call","seq":59,"time":1783352229601,"data":{"turn":1,"step":1,"callId":"call_00_Q6wHtakaip2QNfIXaVJY5458","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}}
 {"type":"hook/invoked","seq":60,"time":1783352229622,"data":{"turn":1,"point":"PostToolUse","dialect":"codex","handlerId":"codex:PostToolUse:1","matcher":"bash"}}
 {"type":"hook/result","seq":61,"time":1783352229632,"data":{"turn":1,"point":"PostToolUse","handlerId":"codex:PostToolUse:1","decision":"pass","exitCode":0,"durationMs":9.27664199999981}}
@@ -111,6 +111,6 @@
 {"type":"assistant/chunk","seq":109,"time":1783352231379,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"The tool result I received is:\n\n```\nHELLO\n```"}}}}
 {"type":"assistant/chunk","seq":110,"time":1783352231379,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":188,"outputTokens":41,"cacheReadTokens":2816,"reasoningTokens":27}}}}
 {"type":"assistant/chunk","seq":111,"time":1783352231379,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":112,"time":1783352231380,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user asked me to report the tool result verbatim. The result I got back is:\n\nHELLO\n\nThat's it."},{"type":"text","text":"The tool result I received is:\n\n```\nHELLO\n```"}],"usage":{"inputTokens":188,"outputTokens":41,"cacheReadTokens":2816,"reasoningTokens":27}},"sourceEventSeqs":[66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111],"surfaceOp":"append"}
+{"type":"assistant/message","seq":112,"time":1783352231380,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user asked me to report the tool result verbatim. The result I got back is:\n\nHELLO\n\nThat's it."},{"type":"text","text":"The tool result I received is:\n\n```\nHELLO\n```"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":188,"outputTokens":41,"cacheReadTokens":2816,"reasoningTokens":27}},"sourceEventSeqs":[66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111],"surfaceOp":"append"}
 {"type":"step/end","seq":113,"time":1783352231380,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":114,"time":1783352231380,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-posttool-context/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-posttool-context/stdout.golden.jsonl
index 4bf92f3197..55310f6b3a 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-posttool-context/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-posttool-context/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-pretool-block/session.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-pretool-block/session.jsonl
index a2022d9d6d..812998aad5 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-pretool-block/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-pretool-block/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352214604,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352214605,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly: echo HELLO. Report the tool result you got back verbatim, then stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352214607,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352214608,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352214608,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352215181,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352215181,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352215351,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -51,7 +51,7 @@
 {"type":"assistant/chunk","seq":49,"time":1783352215801,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_tv0SMeLXaTuyuVrOxnV97085","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}}}}
 {"type":"assistant/chunk","seq":50,"time":1783352215801,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2880,"outputTokens":83,"cacheReadTokens":0,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":51,"time":1783352215801,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":52,"time":1783352215804,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a simple bash command and report the result verbatim."},{"type":"tool-call","id":"call_00_tv0SMeLXaTuyuVrOxnV97085","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}],"usage":{"inputTokens":2880,"outputTokens":83,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51],"surfaceOp":"append"}
+{"type":"assistant/message","seq":52,"time":1783352215804,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a simple bash command and report the result verbatim."},{"type":"tool-call","id":"call_00_tv0SMeLXaTuyuVrOxnV97085","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2880,"outputTokens":83,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51],"surfaceOp":"append"}
 {"type":"tool/call","seq":53,"time":1783352215804,"data":{"turn":1,"step":1,"callId":"call_00_tv0SMeLXaTuyuVrOxnV97085","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Run echo HELLO\"}"}}
 {"type":"hook/invoked","seq":54,"time":1783352215805,"data":{"turn":1,"point":"PreToolUse","dialect":"codex","handlerId":"codex:PreToolUse:1","matcher":"bash"}}
 {"type":"hook/result","seq":55,"time":1783352215832,"data":{"turn":1,"point":"PreToolUse","handlerId":"codex:PreToolUse:1","decision":"block","exitCode":2,"stderrSummary":"bash is disabled by codex policy in this session","durationMs":26.08518500000082}}
@@ -112,6 +112,6 @@
 {"type":"assistant/chunk","seq":110,"time":1783352217214,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"The tool result I got back verbatim is:\n\n```\nError: bash is disabled by codex policy in this session\n```"}}}}
 {"type":"assistant/chunk","seq":111,"time":1783352217214,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":171,"outputTokens":49,"cacheReadTokens":2816,"reasoningTokens":23}}}}
 {"type":"assistant/chunk","seq":112,"time":1783352217214,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":113,"time":1783352217214,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The bash tool is disabled by policy in this session. I need to report this result verbatim to the user."},{"type":"text","text":"The tool result I got back verbatim is:\n\n```\nError: bash is disabled by codex policy in this session\n```"}],"usage":{"inputTokens":171,"outputTokens":49,"cacheReadTokens":2816,"reasoningTokens":23}},"sourceEventSeqs":[59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112],"surfaceOp":"append"}
+{"type":"assistant/message","seq":113,"time":1783352217214,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The bash tool is disabled by policy in this session. I need to report this result verbatim to the user."},{"type":"text","text":"The tool result I got back verbatim is:\n\n```\nError: bash is disabled by codex policy in this session\n```"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":171,"outputTokens":49,"cacheReadTokens":2816,"reasoningTokens":23}},"sourceEventSeqs":[59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112],"surfaceOp":"append"}
 {"type":"step/end","seq":114,"time":1783352217215,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":115,"time":1783352217215,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-pretool-block/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-pretool-block/stdout.golden.jsonl
index 5459da1a17..8eba2d862e 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-pretool-block/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-pretool-block/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-block/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-block/stdout.golden.jsonl
index 6304582220..765dd87f7f 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-block/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-block/stdout.golden.jsonl
@@ -1,3 +1,3 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","id":3,"result":{"stopReason":"cancelled"}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-context/session.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-context/session.jsonl
index e9efd04100..1c710ff8e1 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-context/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-context/session.jsonl
@@ -5,7 +5,7 @@
 {"type":"user/message","seq":3,"time":1783352209707,"data":{"content":[{"type":"text","text":"What is my favorite color? Reply with just the color and stop. Do not use any tools."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"context/message","seq":4,"time":1783352209707,"data":{"content":[{"type":"text","text":"The user has previously stated their favorite color is teal."}],"source":{"kind":"plugin","plugin":"hooks-codex"}},"surfaceOp":"append"}
 {"type":"step/start","seq":5,"time":1783352209709,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":6,"time":1783352209710,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":6,"time":1783352209710,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":7,"time":1783352210353,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":8,"time":1783352210353,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":9,"time":1783352210470,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -52,6 +52,6 @@
 {"type":"assistant/chunk","seq":50,"time":1783352210788,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"teal"}}}}
 {"type":"assistant/chunk","seq":51,"time":1783352210788,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2891,"outputTokens":41,"cacheReadTokens":0,"reasoningTokens":38}}}}
 {"type":"assistant/chunk","seq":52,"time":1783352210788,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":53,"time":1783352210790,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user asked about their favorite color, and the context tells me they previously stated it's teal. They asked me to reply with just the color and stop, without using any tools."},{"type":"text","text":"teal"}],"usage":{"inputTokens":2891,"outputTokens":41,"cacheReadTokens":0,"reasoningTokens":38}},"sourceEventSeqs":[7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52],"surfaceOp":"append"}
+{"type":"assistant/message","seq":53,"time":1783352210790,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user asked about their favorite color, and the context tells me they previously stated it's teal. They asked me to reply with just the color and stop, without using any tools."},{"type":"text","text":"teal"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2891,"outputTokens":41,"cacheReadTokens":0,"reasoningTokens":38}},"sourceEventSeqs":[7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52],"surfaceOp":"append"}
 {"type":"step/end","seq":54,"time":1783352210790,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":55,"time":1783352210790,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-context/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-context/stdout.golden.jsonl
index 8cae81a5c9..76140724de 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-context/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-promptsubmit-context/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" asked"}}}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-stop-continue/session.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-stop-continue/session.jsonl
index 252ff8e262..bba879203d 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-stop-continue/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-stop-continue/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352235020,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352235020,"data":{"content":[{"type":"text","text":"Reply with the single word FIRST and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352235022,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352235023,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352235023,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352235669,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352235670,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352235879,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -30,7 +30,7 @@
 {"type":"assistant/chunk","seq":28,"time":1783352236041,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"FIRST"}}}}
 {"type":"assistant/chunk","seq":29,"time":1783352236041,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2862,"outputTokens":22,"cacheReadTokens":0,"reasoningTokens":19}}}}
 {"type":"assistant/chunk","seq":30,"time":1783352236041,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":31,"time":1783352236043,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with only the single word \"FIRST\" and then stop."},{"type":"text","text":"FIRST"}],"usage":{"inputTokens":2862,"outputTokens":22,"cacheReadTokens":0,"reasoningTokens":19}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30],"surfaceOp":"append"}
+{"type":"assistant/message","seq":31,"time":1783352236043,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with only the single word \"FIRST\" and then stop."},{"type":"text","text":"FIRST"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2862,"outputTokens":22,"cacheReadTokens":0,"reasoningTokens":19}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30],"surfaceOp":"append"}
 {"type":"step/end","seq":32,"time":1783352236043,"data":{"turn":1,"step":1}}
 {"type":"hook/invoked","seq":33,"time":1783352236043,"data":{"turn":1,"point":"Stop","dialect":"codex","handlerId":"codex:Stop:1"}}
 {"type":"hook/result","seq":34,"time":1783352236059,"data":{"turn":1,"point":"Stop","handlerId":"codex:Stop:1","decision":"block","exitCode":2,"stderrSummary":"Also reply with the single word SECOND, then stop.","durationMs":14.77945499999987}}
@@ -60,7 +60,7 @@
 {"type":"assistant/chunk","seq":58,"time":1783352236876,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"SECOND"}}}}
 {"type":"assistant/chunk","seq":59,"time":1783352236876,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":73,"outputTokens":19,"cacheReadTokens":2816,"reasoningTokens":16}}}}
 {"type":"assistant/chunk","seq":60,"time":1783352236877,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":61,"time":1783352236877,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user is asking me to reply with the single word SECOND, then stop."},{"type":"text","text":"SECOND"}],"usage":{"inputTokens":73,"outputTokens":19,"cacheReadTokens":2816,"reasoningTokens":16}},"sourceEventSeqs":[37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60],"surfaceOp":"append"}
+{"type":"assistant/message","seq":61,"time":1783352236877,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The user is asking me to reply with the single word SECOND, then stop."},{"type":"text","text":"SECOND"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":73,"outputTokens":19,"cacheReadTokens":2816,"reasoningTokens":16}},"sourceEventSeqs":[37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60],"surfaceOp":"append"}
 {"type":"step/end","seq":62,"time":1783352236877,"data":{"turn":1,"step":2}}
 {"type":"hook/invoked","seq":63,"time":1783352236877,"data":{"turn":1,"point":"Stop","dialect":"codex","handlerId":"codex:Stop:2"}}
 {"type":"hook/result","seq":64,"time":1783352236908,"data":{"turn":1,"point":"Stop","handlerId":"codex:Stop:2","decision":"pass","exitCode":0,"durationMs":30.947317000000112}}
diff --git a/examples/acp-agent/tests/snapshots/hook-codex-stop-continue/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/hook-codex-stop-continue/stdout.golden.jsonl
index 66d8c816be..af9959379d 100644
--- a/examples/acp-agent/tests/snapshots/hook-codex-stop-continue/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/hook-codex-stop-continue/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/model-switching/input.json b/examples/acp-agent/tests/snapshots/model-switching/input.json
new file mode 100644
index 0000000000..3612f367f6
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/model-switching/input.json
@@ -0,0 +1,23 @@
+{
+  "steps": [
+    {
+      "op": "initialize"
+    },
+    {
+      "op": "newSession"
+    },
+    {
+      "op": "prompt",
+      "text": "Without using tools, reply with exactly FLASH and stop."
+    },
+    {
+      "op": "setConfigOption",
+      "configId": "model",
+      "value": "[\"deepseek\",\"deepseek-v4-pro\"]"
+    },
+    {
+      "op": "prompt",
+      "text": "Without using tools, reply with exactly PRO and stop."
+    }
+  ]
+}
diff --git a/examples/acp-agent/tests/snapshots/model-switching/session.jsonl b/examples/acp-agent/tests/snapshots/model-switching/session.jsonl
new file mode 100644
index 0000000000..1f04eca07d
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/model-switching/session.jsonl
@@ -0,0 +1,69 @@
+{"type":"session","version":0,"id":"622d16ce-0a94-476b-97a4-26dad50b1fbf","createdAt":1784086275585,"cwd":"/var/folders/_g/59jgff8x2gqd39f5vy1wnbfc0000gn/T/acp-snap-cwd-Cwf7Bh"}
+{"type":"turn/start","seq":0,"time":1784086275588,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
+{"type":"user/message","seq":1,"time":1784086275588,"data":{"content":[{"type":"text","text":"Without using tools, reply with exactly FLASH and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
+{"type":"step/start","seq":2,"time":1784086275590,"data":{"turn":1,"step":1}}
+{"type":"request/header","seq":3,"time":1784086275590,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"assistant/chunk","seq":4,"time":1784086276525,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
+{"type":"assistant/chunk","seq":5,"time":1784086276526,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
+{"type":"assistant/chunk","seq":6,"time":1784086276605,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
+{"type":"assistant/chunk","seq":7,"time":1784086276639,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" wants"}}}
+{"type":"assistant/chunk","seq":8,"time":1784086276640,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" me"}}}
+{"type":"assistant/chunk","seq":9,"time":1784086276640,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" to"}}}
+{"type":"assistant/chunk","seq":10,"time":1784086276640,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" reply"}}}
+{"type":"assistant/chunk","seq":11,"time":1784086276640,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" with"}}}
+{"type":"assistant/chunk","seq":12,"time":1784086276661,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" exactly"}}}
+{"type":"assistant/chunk","seq":13,"time":1784086276661,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" \""}}}
+{"type":"assistant/chunk","seq":14,"time":1784086276661,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"FL"}}}
+{"type":"assistant/chunk","seq":15,"time":1784086276661,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"ASH"}}}
+{"type":"assistant/chunk","seq":16,"time":1784086276661,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"\""}}}
+{"type":"assistant/chunk","seq":17,"time":1784086276661,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" and"}}}
+{"type":"assistant/chunk","seq":18,"time":1784086276710,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" stop"}}}
+{"type":"assistant/chunk","seq":19,"time":1784086276710,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":","}}}
+{"type":"assistant/chunk","seq":20,"time":1784086276771,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" without"}}}
+{"type":"assistant/chunk","seq":21,"time":1784086276771,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" using"}}}
+{"type":"assistant/chunk","seq":22,"time":1784086276772,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" any"}}}
+{"type":"assistant/chunk","seq":23,"time":1784086276772,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" tools"}}}
+{"type":"assistant/chunk","seq":24,"time":1784086276772,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"."}}}
+{"type":"assistant/chunk","seq":25,"time":1784086276777,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":1,"blockType":"text"}}}
+{"type":"assistant/chunk","seq":26,"time":1784086276777,"data":{"turn":1,"step":1,"chunk":{"type":"text-delta","index":1,"text":"FL"}}}
+{"type":"assistant/chunk","seq":27,"time":1784086276778,"data":{"turn":1,"step":1,"chunk":{"type":"text-delta","index":1,"text":"ASH"}}}
+{"type":"assistant/chunk","seq":28,"time":1784086276778,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"reasoning","text":"The user wants me to reply with exactly \"FLASH\" and stop, without using any tools."}}}}
+{"type":"assistant/chunk","seq":29,"time":1784086276778,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"FLASH"}}}}
+{"type":"assistant/chunk","seq":30,"time":1784086276778,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3133,"outputTokens":23,"cacheReadTokens":0,"reasoningTokens":20}}}}
+{"type":"assistant/chunk","seq":31,"time":1784086276778,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
+{"type":"assistant/message","seq":32,"time":1784086276782,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly \"FLASH\" and stop, without using any tools."},{"type":"text","text":"FLASH"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3133,"outputTokens":23,"cacheReadTokens":0,"reasoningTokens":20}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],"surfaceOp":"append"}
+{"type":"step/end","seq":33,"time":1784086276782,"data":{"turn":1,"step":1}}
+{"type":"turn/end","seq":34,"time":1784086276783,"data":{"turn":1,"reason":{"kind":"completed"}}}
+{"type":"turn/start","seq":35,"time":1784086276811,"data":{"turn":2,"trigger":{"kind":"message","source":{"kind":"user"}}}}
+{"type":"user/message","seq":36,"time":1784086276812,"data":{"content":[{"type":"text","text":"Without using tools, reply with exactly PRO and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
+{"type":"step/start","seq":37,"time":1784086276812,"data":{"turn":2,"step":1}}
+{"type":"request/header","seq":38,"time":1784298376621,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-pro"},"system":"{{system}}","tools":"{{tools}}"},"reason":"change"}}
+{"type":"assistant/chunk","seq":39,"time":1784086278053,"data":{"turn":2,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
+{"type":"assistant/chunk","seq":40,"time":1784086278053,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
+{"type":"assistant/chunk","seq":41,"time":1784086278242,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
+{"type":"assistant/chunk","seq":42,"time":1784086278312,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" wants"}}}
+{"type":"assistant/chunk","seq":43,"time":1784086278312,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" me"}}}
+{"type":"assistant/chunk","seq":44,"time":1784086278313,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" to"}}}
+{"type":"assistant/chunk","seq":45,"time":1784086278313,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" reply"}}}
+{"type":"assistant/chunk","seq":46,"time":1784086278355,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" with"}}}
+{"type":"assistant/chunk","seq":47,"time":1784086278356,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" exactly"}}}
+{"type":"assistant/chunk","seq":48,"time":1784086278356,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" \""}}}
+{"type":"assistant/chunk","seq":49,"time":1784086278356,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"PRO"}}}
+{"type":"assistant/chunk","seq":50,"time":1784086278400,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"\""}}}
+{"type":"assistant/chunk","seq":51,"time":1784086278400,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" and"}}}
+{"type":"assistant/chunk","seq":52,"time":1784086278400,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" stop"}}}
+{"type":"assistant/chunk","seq":53,"time":1784086278400,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":","}}}
+{"type":"assistant/chunk","seq":54,"time":1784086278441,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" without"}}}
+{"type":"assistant/chunk","seq":55,"time":1784086278442,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" using"}}}
+{"type":"assistant/chunk","seq":56,"time":1784086278442,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" any"}}}
+{"type":"assistant/chunk","seq":57,"time":1784086278442,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" tools"}}}
+{"type":"assistant/chunk","seq":58,"time":1784086278442,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"."}}}
+{"type":"assistant/chunk","seq":59,"time":1784086278494,"data":{"turn":2,"step":1,"chunk":{"type":"block-start","index":1,"blockType":"text"}}}
+{"type":"assistant/chunk","seq":60,"time":1784086278495,"data":{"turn":2,"step":1,"chunk":{"type":"text-delta","index":1,"text":"PRO"}}}
+{"type":"assistant/chunk","seq":61,"time":1784086278495,"data":{"turn":2,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"reasoning","text":"The user wants me to reply with exactly \"PRO\" and stop, without using any tools."}}}}
+{"type":"assistant/chunk","seq":62,"time":1784086278495,"data":{"turn":2,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"PRO"}}}}
+{"type":"assistant/chunk","seq":63,"time":1784086278495,"data":{"turn":2,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3149,"outputTokens":21,"cacheReadTokens":0,"reasoningTokens":19}}}}
+{"type":"assistant/chunk","seq":64,"time":1784086278495,"data":{"turn":2,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
+{"type":"assistant/message","seq":65,"time":1784086278495,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly \"PRO\" and stop, without using any tools."},{"type":"text","text":"PRO"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-pro"},"usage":{"inputTokens":3149,"outputTokens":21,"cacheReadTokens":0,"reasoningTokens":19}},"sourceEventSeqs":[39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64],"surfaceOp":"append"}
+{"type":"step/end","seq":66,"time":1784086278495,"data":{"turn":2,"step":1}}
+{"type":"turn/end","seq":67,"time":1784086278495,"data":{"turn":2,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/model-switching/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/model-switching/stdout.golden.jsonl
new file mode 100644
index 0000000000..011d06f5d2
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/model-switching/stdout.golden.jsonl
@@ -0,0 +1,47 @@
+{"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" me"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" to"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" reply"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" with"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" exactly"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" \""}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"FL"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"ASH"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"\""}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" and"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" stop"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":","}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" without"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" using"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" any"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" tools"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"."}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"FL"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"ASH"}}}}
+{"jsonrpc":"2.0","id":3,"result":{"stopReason":"end_turn"}}
+{"jsonrpc":"2.0","id":4,"result":{"configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-pro\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" me"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" to"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" reply"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" with"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" exactly"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" \""}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"PRO"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"\""}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" and"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" stop"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":","}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" without"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" using"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" any"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" tools"}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"."}}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"PRO"}}}}
+{"jsonrpc":"2.0","id":5,"result":{"stopReason":"end_turn"}}
diff --git a/examples/acp-agent/tests/snapshots/model-switching/system-prompt.golden.md b/examples/acp-agent/tests/snapshots/model-switching/system-prompt.golden.md
new file mode 100644
index 0000000000..b9701e538c
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/model-switching/system-prompt.golden.md
@@ -0,0 +1,33 @@
+You are an AI agent powered by the DeepSeek Harness SDK.
+
+You are a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}. Your bash tool runs under a file sandbox — a `[sandbox: file access denied …]` result is policy, not a command bug.
+
+Verify your work by running the code or tests. Keep answers brief and factual.
+
+
+Check the [exit code: N] marker on every bash result; investigate failures before moving on.
+
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
+Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
+<!-- dsh-user-approval-policy:never -->
+
+Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.
+
+<!-- request/header change 1 -->
+
+You are an AI agent powered by the DeepSeek Harness SDK.
+
+You are a coding assistant powered by the deepseek-v4-pro model. Your working directory is {{cwd}}. Your bash tool runs under a file sandbox — a `[sandbox: file access denied …]` result is policy, not a command bug.
+
+Verify your work by running the code or tests. Keep answers brief and factual.
+
+
+Check the [exit code: N] marker on every bash result; investigate failures before moving on.
+
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
+Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
+<!-- dsh-user-approval-policy:never -->
+
+Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.
diff --git a/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.golden.json b/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.golden.json
new file mode 100644
index 0000000000..7c814257fb
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.golden.json
@@ -0,0 +1,552 @@
+{
+  "initial": [
+    {
+      "name": "bash",
+      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "command": {
+            "type": "string",
+            "description": "The bash command to execute."
+          },
+          "description": {
+            "type": "string",
+            "description": "Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."
+          },
+          "timeoutMs": {
+            "type": "number",
+            "description": "Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."
+          },
+          "workdir": {
+            "type": "string",
+            "description": "Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
+          },
+          "sandbox_permissions": {
+            "type": "string",
+            "description": "The wider sandbox mode this command needs. Only valid as a one-shot retry of a command the sandbox just denied; requires justification and user approval.",
+            "enum": [
+              "workspace-write",
+              "danger-full-access"
+            ]
+          },
+          "justification": {
+            "type": "string",
+            "description": "Required with sandbox_permissions: one sentence for the user explaining why this exact command needs the wider access."
+          }
+        },
+        "required": [
+          "command",
+          "description"
+        ]
+      }
+    },
+    {
+      "name": "skill",
+      "description": "Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "name": {
+            "type": "string",
+            "description": "The exact skill name from the available skills list."
+          }
+        },
+        "required": [
+          "name"
+        ]
+      }
+    },
+    {
+      "name": "subagent",
+      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "description": {
+            "type": "string",
+            "description": "A short (3-5 word) description of the delegated task, for display."
+          },
+          "prompt": {
+            "type": "string",
+            "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
+          }
+        },
+        "required": [
+          "description",
+          "prompt"
+        ]
+      }
+    },
+    {
+      "name": "subagent_fork",
+      "description": "Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "description": {
+            "type": "string",
+            "description": "A short (3-5 word) description of the delegated task, for display."
+          },
+          "prompt": {
+            "type": "string",
+            "description": "The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
+          }
+        },
+        "required": [
+          "description",
+          "prompt"
+        ]
+      }
+    },
+    {
+      "name": "task_kill",
+      "description": "Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "reason": {
+            "type": "string",
+            "description": "Optional short reason, recorded in the log and forwarded to the task."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
+    {
+      "name": "task_list",
+      "description": "List your background tasks (running and finished) with their ids, kinds, and statuses.",
+      "parameters": {
+        "type": "object",
+        "properties": {}
+      }
+    },
+    {
+      "name": "task_output",
+      "description": "Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "wait": {
+            "type": "boolean",
+            "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+          },
+          "timeout_ms": {
+            "type": "number",
+            "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
+    {
+      "name": "todo_write",
+      "description": "Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "todos": {
+            "type": "array",
+            "description": "The COMPLETE task list, replacing any previous list.",
+            "items": {
+              "type": "object",
+              "properties": {
+                "content": {
+                  "type": "string",
+                  "description": "What the task is — a short imperative line."
+                },
+                "status": {
+                  "type": "string",
+                  "description": "pending (not started) | in_progress (now) | completed (done).",
+                  "enum": [
+                    "pending",
+                    "in_progress",
+                    "completed"
+                  ]
+                }
+              },
+              "required": [
+                "content",
+                "status"
+              ]
+            }
+          }
+        },
+        "required": [
+          "todos"
+        ]
+      }
+    },
+    {
+      "name": "workflow",
+      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "script": {
+            "type": "string",
+            "description": "The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return <json-value>`)."
+          },
+          "meta": {
+            "type": "object",
+            "description": "The workflow identity block (plain JSON — never code).",
+            "properties": {
+              "name": {
+                "type": "string",
+                "description": "Short kebab-case workflow name."
+              },
+              "description": {
+                "type": "string",
+                "description": "One-line description of what the workflow does."
+              },
+              "whenToUse": {
+                "type": "string",
+                "description": "Optional guidance on when this workflow applies."
+              },
+              "phases": {
+                "type": "array",
+                "description": "Optional phase declarations matched by phase() calls.",
+                "items": {
+                  "type": "object",
+                  "properties": {
+                    "title": {
+                      "type": "string",
+                      "description": "The phase title phase() calls match by exact string."
+                    },
+                    "detail": {
+                      "type": "string",
+                      "description": "Optional one-line description of the phase."
+                    },
+                    "provider": {
+                      "type": "string",
+                      "description": "Optional provider override this phase is expected to use."
+                    },
+                    "model": {
+                      "type": "string",
+                      "description": "Optional model override this phase is expected to use."
+                    }
+                  },
+                  "required": [
+                    "title"
+                  ]
+                }
+              }
+            },
+            "required": [
+              "name",
+              "description"
+            ]
+          },
+          "args": {
+            "type": "object",
+            "description": "Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]})."
+          }
+        },
+        "required": [
+          "script",
+          "meta"
+        ]
+      }
+    }
+  ],
+  "changes": [
+    [
+      {
+        "name": "bash",
+        "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "command": {
+              "type": "string",
+              "description": "The bash command to execute."
+            },
+            "description": {
+              "type": "string",
+              "description": "Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."
+            },
+            "timeoutMs": {
+              "type": "number",
+              "description": "Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."
+            },
+            "workdir": {
+              "type": "string",
+              "description": "Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."
+            },
+            "run_in_background": {
+              "type": "boolean",
+              "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
+            },
+            "sandbox_permissions": {
+              "type": "string",
+              "description": "The wider sandbox mode this command needs. Only valid as a one-shot retry of a command the sandbox just denied; requires justification and user approval.",
+              "enum": [
+                "workspace-write",
+                "danger-full-access"
+              ]
+            },
+            "justification": {
+              "type": "string",
+              "description": "Required with sandbox_permissions: one sentence for the user explaining why this exact command needs the wider access."
+            }
+          },
+          "required": [
+            "command",
+            "description"
+          ]
+        }
+      },
+      {
+        "name": "skill",
+        "description": "Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "name": {
+              "type": "string",
+              "description": "The exact skill name from the available skills list."
+            }
+          },
+          "required": [
+            "name"
+          ]
+        }
+      },
+      {
+        "name": "subagent",
+        "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "description": {
+              "type": "string",
+              "description": "A short (3-5 word) description of the delegated task, for display."
+            },
+            "prompt": {
+              "type": "string",
+              "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+            },
+            "run_in_background": {
+              "type": "boolean",
+              "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
+            }
+          },
+          "required": [
+            "description",
+            "prompt"
+          ]
+        }
+      },
+      {
+        "name": "subagent_fork",
+        "description": "Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "description": {
+              "type": "string",
+              "description": "A short (3-5 word) description of the delegated task, for display."
+            },
+            "prompt": {
+              "type": "string",
+              "description": "The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."
+            },
+            "run_in_background": {
+              "type": "boolean",
+              "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
+            }
+          },
+          "required": [
+            "description",
+            "prompt"
+          ]
+        }
+      },
+      {
+        "name": "task_kill",
+        "description": "Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "task_id": {
+              "type": "string",
+              "description": "Task id returned by the tool that started the background work."
+            },
+            "reason": {
+              "type": "string",
+              "description": "Optional short reason, recorded in the log and forwarded to the task."
+            }
+          },
+          "required": [
+            "task_id"
+          ]
+        }
+      },
+      {
+        "name": "task_list",
+        "description": "List your background tasks (running and finished) with their ids, kinds, and statuses.",
+        "parameters": {
+          "type": "object",
+          "properties": {}
+        }
+      },
+      {
+        "name": "task_output",
+        "description": "Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "task_id": {
+              "type": "string",
+              "description": "Task id returned by the tool that started the background work."
+            },
+            "wait": {
+              "type": "boolean",
+              "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+            },
+            "timeout_ms": {
+              "type": "number",
+              "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+            }
+          },
+          "required": [
+            "task_id"
+          ]
+        }
+      },
+      {
+        "name": "todo_write",
+        "description": "Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "todos": {
+              "type": "array",
+              "description": "The COMPLETE task list, replacing any previous list.",
+              "items": {
+                "type": "object",
+                "properties": {
+                  "content": {
+                    "type": "string",
+                    "description": "What the task is — a short imperative line."
+                  },
+                  "status": {
+                    "type": "string",
+                    "description": "pending (not started) | in_progress (now) | completed (done).",
+                    "enum": [
+                      "pending",
+                      "in_progress",
+                      "completed"
+                    ]
+                  }
+                },
+                "required": [
+                  "content",
+                  "status"
+                ]
+              }
+            }
+          },
+          "required": [
+            "todos"
+          ]
+        }
+      },
+      {
+        "name": "workflow",
+        "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "script": {
+              "type": "string",
+              "description": "The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return <json-value>`)."
+            },
+            "meta": {
+              "type": "object",
+              "description": "The workflow identity block (plain JSON — never code).",
+              "properties": {
+                "name": {
+                  "type": "string",
+                  "description": "Short kebab-case workflow name."
+                },
+                "description": {
+                  "type": "string",
+                  "description": "One-line description of what the workflow does."
+                },
+                "whenToUse": {
+                  "type": "string",
+                  "description": "Optional guidance on when this workflow applies."
+                },
+                "phases": {
+                  "type": "array",
+                  "description": "Optional phase declarations matched by phase() calls.",
+                  "items": {
+                    "type": "object",
+                    "properties": {
+                      "title": {
+                        "type": "string",
+                        "description": "The phase title phase() calls match by exact string."
+                      },
+                      "detail": {
+                        "type": "string",
+                        "description": "Optional one-line description of the phase."
+                      },
+                      "provider": {
+                        "type": "string",
+                        "description": "Optional provider override this phase is expected to use."
+                      },
+                      "model": {
+                        "type": "string",
+                        "description": "Optional model override this phase is expected to use."
+                      }
+                    },
+                    "required": [
+                      "title"
+                    ]
+                  }
+                }
+              },
+              "required": [
+                "name",
+                "description"
+              ]
+            },
+            "args": {
+              "type": "object",
+              "description": "Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]})."
+            }
+          },
+          "required": [
+            "script",
+            "meta"
+          ]
+        }
+      }
+    ]
+  ]
+}
diff --git a/examples/acp-agent/tests/snapshots/multi-turn/session.jsonl b/examples/acp-agent/tests/snapshots/multi-turn/session.jsonl
index 3b8c470ded..cd32072eaa 100644
--- a/examples/acp-agent/tests/snapshots/multi-turn/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/multi-turn/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352113765,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352113765,"data":{"content":[{"type":"text","text":"Reply with exactly the word: ONE. No tools."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352113767,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352113768,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352113768,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352114428,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352114428,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352114542,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -28,7 +28,7 @@
 {"type":"assistant/chunk","seq":26,"time":1783352114687,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"ONE"}}}}
 {"type":"assistant/chunk","seq":27,"time":1783352114687,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2864,"outputTokens":20,"cacheReadTokens":0,"reasoningTokens":18}}}}
 {"type":"assistant/chunk","seq":28,"time":1783352114687,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":29,"time":1783352114690,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word \"ONE\" and use no tools."},{"type":"text","text":"ONE"}],"usage":{"inputTokens":2864,"outputTokens":20,"cacheReadTokens":0,"reasoningTokens":18}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28],"surfaceOp":"append"}
+{"type":"assistant/message","seq":29,"time":1783352114690,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word \"ONE\" and use no tools."},{"type":"text","text":"ONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2864,"outputTokens":20,"cacheReadTokens":0,"reasoningTokens":18}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28],"surfaceOp":"append"}
 {"type":"step/end","seq":30,"time":1783352114690,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":31,"time":1783352114690,"data":{"turn":1,"reason":{"kind":"completed"}}}
 {"type":"turn/start","seq":32,"time":1783352114699,"data":{"turn":2,"trigger":{"kind":"message","source":{"kind":"user"}}}}
@@ -60,6 +60,6 @@
 {"type":"assistant/chunk","seq":58,"time":1783352115610,"data":{"turn":2,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"TWO"}}}}
 {"type":"assistant/chunk","seq":59,"time":1783352115610,"data":{"turn":2,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":64,"outputTokens":21,"cacheReadTokens":2816,"reasoningTokens":18}}}}
 {"type":"assistant/chunk","seq":60,"time":1783352115610,"data":{"turn":2,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":61,"time":1783352115611,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word \"TWO\" and no tools."},{"type":"text","text":"TWO"}],"usage":{"inputTokens":64,"outputTokens":21,"cacheReadTokens":2816,"reasoningTokens":18}},"sourceEventSeqs":[35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60],"surfaceOp":"append"}
+{"type":"assistant/message","seq":61,"time":1783352115611,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word \"TWO\" and no tools."},{"type":"text","text":"TWO"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":64,"outputTokens":21,"cacheReadTokens":2816,"reasoningTokens":18}},"sourceEventSeqs":[35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60],"surfaceOp":"append"}
 {"type":"step/end","seq":62,"time":1783352115611,"data":{"turn":2,"step":1}}
 {"type":"turn/end","seq":63,"time":1783352115611,"data":{"turn":2,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/multi-turn/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/multi-turn/stdout.golden.jsonl
index 1d9d45954a..5c38cacf55 100644
--- a/examples/acp-agent/tests/snapshots/multi-turn/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/multi-turn/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/parallel-tool-calls/input.json b/examples/acp-agent/tests/snapshots/parallel-tool-calls/input.json
new file mode 100644
index 0000000000..e5356e4af5
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/parallel-tool-calls/input.json
@@ -0,0 +1,7 @@
+{
+  "steps": [
+    { "op": "initialize" },
+    { "op": "newSession" },
+    { "op": "prompt", "text": "Use the read tool twice in the same assistant message: read a.txt and b.txt. Then reply DONE." }
+  ]
+}
diff --git a/examples/acp-agent/tests/snapshots/parallel-tool-calls/session.jsonl b/examples/acp-agent/tests/snapshots/parallel-tool-calls/session.jsonl
new file mode 100644
index 0000000000..81503dc4cf
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/parallel-tool-calls/session.jsonl
@@ -0,0 +1,28 @@
+{"type":"session","version":0,"id":"{{sessionId}}","createdAt":0,"cwd":"{{cwd}}"}
+{"type":"turn/start","seq":0,"time":0,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
+{"type":"user/message","seq":1,"time":0,"data":{"content":[{"type":"text","text":"Use the read tool twice in the same assistant message: read a.txt and b.txt. Then reply DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
+{"type":"step/start","seq":2,"time":0,"data":{"turn":1,"step":1}}
+{"type":"request/header","seq":3,"time":0,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"assistant/chunk","seq":4,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
+{"type":"assistant/chunk","seq":5,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"call_read_a","name":"read","argumentsDelta":"{\"file_path\":\"a.txt\"}"}}}
+{"type":"assistant/chunk","seq":6,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_read_a","name":"read","arguments":"{\"file_path\":\"a.txt\"}"}}}}
+{"type":"assistant/chunk","seq":7,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":1,"blockType":"tool-call"}}}
+{"type":"assistant/chunk","seq":8,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":1,"id":"call_read_b","name":"read","argumentsDelta":"{\"file_path\":\"b.txt\"}"}}}
+{"type":"assistant/chunk","seq":9,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_read_b","name":"read","arguments":"{\"file_path\":\"b.txt\"}"}}}}
+{"type":"assistant/chunk","seq":10,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
+{"type":"assistant/chunk","seq":11,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
+{"type":"assistant/message","seq":12,"time":0,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"call_read_a","name":"read","arguments":"{\"file_path\":\"a.txt\"}"},{"type":"tool-call","id":"call_read_b","name":"read","arguments":"{\"file_path\":\"b.txt\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[4,5,6,7,8,9,10,11],"surfaceOp":"append"}
+{"type":"tool/call","seq":13,"time":0,"data":{"turn":1,"step":1,"callId":"call_read_a","name":"read","arguments":"{\"file_path\":\"a.txt\"}"}}
+{"type":"tool/call","seq":14,"time":0,"data":{"turn":1,"step":1,"callId":"call_read_b","name":"read","arguments":"{\"file_path\":\"b.txt\"}"}}
+{"type":"tool/result","seq":15,"time":0,"data":{"turn":1,"step":1,"callId":"call_read_a","content":[{"type":"text","text":"<path>{{cwd}}/a.txt</path>\n<type>file</type>\n<content>\n1: alpha\n\n(End of file - total 1 lines)\n</content>"}],"isError":false},"sourceEventSeqs":[13],"surfaceOp":"append"}
+{"type":"tool/result","seq":16,"time":0,"data":{"turn":1,"step":1,"callId":"call_read_b","content":[{"type":"text","text":"<path>{{cwd}}/b.txt</path>\n<type>file</type>\n<content>\n1: beta\n\n(End of file - total 1 lines)\n</content>"}],"isError":false},"sourceEventSeqs":[14],"surfaceOp":"append"}
+{"type":"step/end","seq":17,"time":0,"data":{"turn":1,"step":1}}
+{"type":"step/start","seq":18,"time":0,"data":{"turn":1,"step":2}}
+{"type":"assistant/chunk","seq":19,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}
+{"type":"assistant/chunk","seq":20,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":0,"text":"DONE"}}}
+{"type":"assistant/chunk","seq":21,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"DONE"}}}}
+{"type":"assistant/chunk","seq":22,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":1}}}}
+{"type":"assistant/chunk","seq":23,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
+{"type":"assistant/message","seq":24,"time":0,"data":{"turn":1,"step":2,"content":[{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":1}},"sourceEventSeqs":[19,20,21,22,23],"surfaceOp":"append"}
+{"type":"step/end","seq":25,"time":0,"data":{"turn":1,"step":2}}
+{"type":"turn/end","seq":26,"time":0,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/parallel-tool-calls/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/parallel-tool-calls/stdout.golden.jsonl
new file mode 100644
index 0000000000..6d7bc199e8
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/parallel-tool-calls/stdout.golden.jsonl
@@ -0,0 +1,8 @@
+{"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call","toolCallId":"call_read_a","title":"Read a.txt","kind":"read","status":"in_progress","locations":[{"path":"a.txt","line":1}]}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call","toolCallId":"call_read_b","title":"Read b.txt","kind":"read","status":"in_progress","locations":[{"path":"b.txt","line":1}]}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call_update","toolCallId":"call_read_a","status":"completed","content":[{"type":"content","content":{"type":"text","text":"<path>{{cwd}}/a.txt</path>\n<type>file</type>\n<content>\n1: alpha\n\n(End of file - total 1 lines)\n</content>"}}]}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call_update","toolCallId":"call_read_b","status":"completed","content":[{"type":"content","content":{"type":"text","text":"<path>{{cwd}}/b.txt</path>\n<type>file</type>\n<content>\n1: beta\n\n(End of file - total 1 lines)\n</content>"}}]}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"DONE"}}}}
+{"jsonrpc":"2.0","id":3,"result":{"stopReason":"end_turn"}}
diff --git a/examples/acp-agent/tests/snapshots/parallel-tool-calls/workspace/a.txt b/examples/acp-agent/tests/snapshots/parallel-tool-calls/workspace/a.txt
new file mode 100644
index 0000000000..4a58007052
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/parallel-tool-calls/workspace/a.txt
@@ -0,0 +1 @@
+alpha
diff --git a/examples/acp-agent/tests/snapshots/parallel-tool-calls/workspace/b.txt b/examples/acp-agent/tests/snapshots/parallel-tool-calls/workspace/b.txt
new file mode 100644
index 0000000000..65b2df87f7
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/parallel-tool-calls/workspace/b.txt
@@ -0,0 +1 @@
+beta
diff --git a/examples/acp-agent/tests/snapshots/permission-switching/session.jsonl b/examples/acp-agent/tests/snapshots/permission-switching/session.jsonl
index eefadc2be4..4a62de6642 100644
--- a/examples/acp-agent/tests/snapshots/permission-switching/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/permission-switching/session.jsonl
@@ -5,7 +5,7 @@
 {"type":"approval/policy","seq":3,"time":1783962244578,"data":{"policy":"ask"}}
 {"type":"user/message","seq":4,"time":1783962244578,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly this one command in a single call: printf 'before\\n' > out.txt && cat out.txt. Then reply with the single word DONE and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":5,"time":1783962244579,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":6,"time":1783962244580,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":6,"time":1783962244580,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":7,"time":1783860667444,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":8,"time":1783860667445,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":9,"time":1783860667445,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -63,7 +63,7 @@
 {"type":"assistant/chunk","seq":61,"time":1783962244582,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_E1vtulcKU1LKUgLahxdR3767","name":"bash","arguments":"{\"command\": \"printf 'before\\\\n' > out.txt && cat out.txt\", \"description\": \"Write and then read out.txt\"}"}}}}
 {"type":"assistant/chunk","seq":62,"time":1783962244582,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":1411,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":63,"time":1783962244582,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":64,"time":1783962244582,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a specific command and then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_E1vtulcKU1LKUgLahxdR3767","name":"bash","arguments":"{\"command\": \"printf 'before\\\\n' > out.txt && cat out.txt\", \"description\": \"Write and then read out.txt\"}"}],"usage":{"inputTokens":1411,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63],"surfaceOp":"append"}
+{"type":"assistant/message","seq":64,"time":1783962244582,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a specific command and then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_E1vtulcKU1LKUgLahxdR3767","name":"bash","arguments":"{\"command\": \"printf 'before\\\\n' > out.txt && cat out.txt\", \"description\": \"Write and then read out.txt\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1411,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63],"surfaceOp":"append"}
 {"type":"tool/call","seq":65,"time":1783962244582,"data":{"turn":1,"step":1,"callId":"call_00_E1vtulcKU1LKUgLahxdR3767","name":"bash","arguments":"{\"command\": \"printf 'before\\\\n' > out.txt && cat out.txt\", \"description\": \"Write and then read out.txt\"}"}}
 {"type":"tool/result","seq":66,"time":1783962244599,"data":{"turn":1,"step":1,"callId":"call_00_E1vtulcKU1LKUgLahxdR3767","content":[{"type":"text","text":"before\n"}],"isError":false},"sourceEventSeqs":[65],"surfaceOp":"append"}
 {"type":"step/end","seq":67,"time":1783962244599,"data":{"turn":1,"step":1}}
@@ -97,7 +97,7 @@
 {"type":"assistant/chunk","seq":95,"time":1783962244601,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":96,"time":1783962244601,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":109,"outputTokens":24,"cacheReadTokens":1408,"reasoningTokens":21}}}}
 {"type":"assistant/chunk","seq":97,"time":1783962244601,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":98,"time":1783962244601,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The command ran successfully, output \"before\". Now I need to reply with just the word DONE."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":109,"outputTokens":24,"cacheReadTokens":1408,"reasoningTokens":21}},"sourceEventSeqs":[69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97],"surfaceOp":"append"}
+{"type":"assistant/message","seq":98,"time":1783962244601,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The command ran successfully, output \"before\". Now I need to reply with just the word DONE."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":109,"outputTokens":24,"cacheReadTokens":1408,"reasoningTokens":21}},"sourceEventSeqs":[69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97],"surfaceOp":"append"}
 {"type":"step/end","seq":99,"time":1783962244601,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":100,"time":1783962244601,"data":{"turn":1,"reason":{"kind":"completed"}}}
 {"type":"turn/start","seq":101,"time":1783962244623,"data":{"turn":2,"trigger":{"kind":"message","source":{"kind":"user"}}}}
@@ -107,7 +107,7 @@
 {"type":"user/message","seq":105,"time":1783962244624,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly: cat out.txt. Then reply with the single word DONE and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"context/message","seq":106,"time":1783962244624,"data":{"content":[{"type":"text","text":"The approval policy changed from \"ask\" to \"never\" (changed by the user)."}],"source":{"kind":"plugin","plugin":"user-approval"}},"surfaceOp":"append"}
 {"type":"step/start","seq":107,"time":1783962244624,"data":{"turn":2,"step":1}}
-{"type":"request/header-delta","seq":108,"time":1783962244624,"data":{"system":{"keepStart":9,"keepEnd":2,"insert":["{{system}}","{{system}}"]}}}
+{"type":"request/header","seq":108,"time":1784000791271,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"change"}}
 {"type":"assistant/chunk","seq":109,"time":1783860671025,"data":{"turn":2,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":110,"time":1783860671026,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":111,"time":1783860671026,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -155,7 +155,7 @@
 {"type":"assistant/chunk","seq":153,"time":1783962244626,"data":{"turn":2,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_7Jb7FWHNjIBVML49dEJl1990","name":"bash","arguments":"{\"description\": \"Read out.txt\", \"command\": \"cat out.txt\"}"}}}}
 {"type":"assistant/chunk","seq":154,"time":1783962244626,"data":{"turn":2,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":1608,"outputTokens":82,"cacheReadTokens":0,"reasoningTokens":19}}}}
 {"type":"assistant/chunk","seq":155,"time":1783962244626,"data":{"turn":2,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":156,"time":1783962244626,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run `cat out.txt` and then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_7Jb7FWHNjIBVML49dEJl1990","name":"bash","arguments":"{\"description\": \"Read out.txt\", \"command\": \"cat out.txt\"}"}],"usage":{"inputTokens":1608,"outputTokens":82,"cacheReadTokens":0,"reasoningTokens":19}},"sourceEventSeqs":[109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155],"surfaceOp":"append"}
+{"type":"assistant/message","seq":156,"time":1783962244626,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run `cat out.txt` and then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_7Jb7FWHNjIBVML49dEJl1990","name":"bash","arguments":"{\"description\": \"Read out.txt\", \"command\": \"cat out.txt\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1608,"outputTokens":82,"cacheReadTokens":0,"reasoningTokens":19}},"sourceEventSeqs":[109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155],"surfaceOp":"append"}
 {"type":"tool/call","seq":157,"time":1783962244626,"data":{"turn":2,"step":1,"callId":"call_00_7Jb7FWHNjIBVML49dEJl1990","name":"bash","arguments":"{\"description\": \"Read out.txt\", \"command\": \"cat out.txt\"}"}}
 {"type":"tool/result","seq":158,"time":1783962244631,"data":{"turn":2,"step":1,"callId":"call_00_7Jb7FWHNjIBVML49dEJl1990","content":[{"type":"text","text":"before\n"}],"isError":false},"sourceEventSeqs":[157],"surfaceOp":"append"}
 {"type":"step/end","seq":159,"time":1783962244631,"data":{"turn":2,"step":1}}
@@ -166,7 +166,7 @@
 {"type":"assistant/chunk","seq":164,"time":1783962244632,"data":{"turn":2,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":165,"time":1783962244632,"data":{"turn":2,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":40,"outputTokens":3,"cacheReadTokens":1664,"reasoningTokens":0}}}}
 {"type":"assistant/chunk","seq":166,"time":1783962244632,"data":{"turn":2,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":167,"time":1783962244632,"data":{"turn":2,"step":2,"content":[{"type":"text","text":"DONE"}],"usage":{"inputTokens":40,"outputTokens":3,"cacheReadTokens":1664,"reasoningTokens":0}},"sourceEventSeqs":[161,162,163,164,165,166],"surfaceOp":"append"}
+{"type":"assistant/message","seq":167,"time":1783962244632,"data":{"turn":2,"step":2,"content":[{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":40,"outputTokens":3,"cacheReadTokens":1664,"reasoningTokens":0}},"sourceEventSeqs":[161,162,163,164,165,166],"surfaceOp":"append"}
 {"type":"step/end","seq":168,"time":1783962244632,"data":{"turn":2,"step":2}}
 {"type":"turn/end","seq":169,"time":1783962244632,"data":{"turn":2,"reason":{"kind":"completed"}}}
 {"type":"turn/start","seq":170,"time":1783962244637,"data":{"turn":3,"trigger":{"kind":"message","source":{"kind":"user"}}}}
@@ -233,6 +233,6 @@
 {"type":"assistant/chunk","seq":231,"time":1783962244639,"data":{"turn":3,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"My approval policy is set to \"never\" — no actions requiring approval are permitted, and they are rejected automatically."}}}}
 {"type":"assistant/chunk","seq":232,"time":1783962244639,"data":{"turn":3,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":64,"outputTokens":56,"cacheReadTokens":1664,"reasoningTokens":32}}}}
 {"type":"assistant/chunk","seq":233,"time":1783962244639,"data":{"turn":3,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":234,"time":1783962244640,"data":{"turn":3,"step":1,"content":[{"type":"reasoning","text":"The user wants me to state the current approval policy without using any tools. From the context provided, the approval policy changed from \"ask\" to \"never\"."},{"type":"text","text":"My approval policy is set to \"never\" — no actions requiring approval are permitted, and they are rejected automatically."}],"usage":{"inputTokens":64,"outputTokens":56,"cacheReadTokens":1664,"reasoningTokens":32}},"sourceEventSeqs":[173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233],"surfaceOp":"append"}
+{"type":"assistant/message","seq":234,"time":1783962244640,"data":{"turn":3,"step":1,"content":[{"type":"reasoning","text":"The user wants me to state the current approval policy without using any tools. From the context provided, the approval policy changed from \"ask\" to \"never\"."},{"type":"text","text":"My approval policy is set to \"never\" — no actions requiring approval are permitted, and they are rejected automatically."}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":64,"outputTokens":56,"cacheReadTokens":1664,"reasoningTokens":32}},"sourceEventSeqs":[173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233],"surfaceOp":"append"}
 {"type":"step/end","seq":235,"time":1783962244640,"data":{"turn":3,"step":1}}
 {"type":"turn/end","seq":236,"time":1783962244640,"data":{"turn":3,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/permission-switching/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/permission-switching/stdout.golden.jsonl
index 651897e850..be9ca10cf5 100644
--- a/examples/acp-agent/tests/snapshots/permission-switching/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/permission-switching/stdout.golden.jsonl
@@ -1,6 +1,6 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
-{"jsonrpc":"2.0","id":3,"result":{"configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"workspace-write","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":3,"result":{"configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"workspace-write","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
@@ -44,7 +44,7 @@
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"D"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"ONE"}}}}
 {"jsonrpc":"2.0","id":4,"result":{"stopReason":"end_turn"}}
-{"jsonrpc":"2.0","id":5,"result":{"configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":5,"result":{"configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/permission-switching/system-prompt.golden.md b/examples/acp-agent/tests/snapshots/permission-switching/system-prompt.golden.md
index 3da863d3b8..622bc4e23a 100644
--- a/examples/acp-agent/tests/snapshots/permission-switching/system-prompt.golden.md
+++ b/examples/acp-agent/tests/snapshots/permission-switching/system-prompt.golden.md
@@ -7,11 +7,26 @@ Verify your work by running the code or tests. Keep answers brief and factual.
 
 Check the [exit code: N] marker on every bash result; investigate failures before moving on.
 
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
 <!-- dsh-user-approval-policy:ask -->
 
 Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.
 
-<!-- request/header-delta 1: keepStart=9, keepEnd=2 -->
+<!-- request/header change 1 -->
+
+You are an AI agent powered by the DeepSeek Harness SDK.
+
+You are a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}. Your bash tool runs under a file sandbox — a `[sandbox: file access denied …]` result is policy, not a command bug.
+
+Verify your work by running the code or tests. Keep answers brief and factual.
+
+
+Check the [exit code: N] marker on every bash result; investigate failures before moving on.
+
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
 
 Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
 <!-- dsh-user-approval-policy:never -->
+
+Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.
diff --git a/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.golden.json b/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.golden.json
index 68d4b037d7..7c814257fb 100644
--- a/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.golden.json
+++ b/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.golden.json
@@ -2,7 +2,7 @@
   "initial": [
     {
       "name": "bash",
-      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; poll it with `bash_output` and stop it with `bash_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
+      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -24,7 +24,7 @@
           },
           "run_in_background": {
             "type": "boolean",
-            "description": "Run in the background and return a task id immediately. No timeout applies."
+            "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
           },
           "sandbox_permissions": {
             "type": "string",
@@ -45,38 +45,6 @@
         ]
       }
     },
-    {
-      "name": "bash_kill",
-      "description": "Ask the executor to kill a running background bash task by task id.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
-    {
-      "name": "bash_output",
-      "description": "Read new output from a background bash task started with `bash` + `run_in_background`. Returns only output produced since the previous bash_output call, plus the task status. Tasks keep running while you do other work; poll again later for more output.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
     {
       "name": "skill",
       "description": "Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.",
@@ -95,7 +63,7 @@
     },
     {
       "name": "subagent",
-      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation.",
+      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -106,6 +74,10 @@
           "prompt": {
             "type": "string",
             "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -116,7 +88,7 @@
     },
     {
       "name": "subagent_fork",
-      "description": "Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps.",
+      "description": "Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -127,6 +99,10 @@
           "prompt": {
             "type": "string",
             "description": "The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -135,6 +111,58 @@
         ]
       }
     },
+    {
+      "name": "task_kill",
+      "description": "Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "reason": {
+            "type": "string",
+            "description": "Optional short reason, recorded in the log and forwarded to the task."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
+    {
+      "name": "task_list",
+      "description": "List your background tasks (running and finished) with their ids, kinds, and statuses.",
+      "parameters": {
+        "type": "object",
+        "properties": {}
+      }
+    },
+    {
+      "name": "task_output",
+      "description": "Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "wait": {
+            "type": "boolean",
+            "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+          },
+          "timeout_ms": {
+            "type": "number",
+            "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
     {
       "name": "todo_write",
       "description": "Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).",
@@ -175,7 +203,7 @@
     },
     {
       "name": "workflow",
-      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), `model` (override). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
+      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -213,6 +241,10 @@
                       "type": "string",
                       "description": "Optional one-line description of the phase."
                     },
+                    "provider": {
+                      "type": "string",
+                      "description": "Optional provider override this phase is expected to use."
+                    },
                     "model": {
                       "type": "string",
                       "description": "Optional model override this phase is expected to use."
@@ -241,5 +273,280 @@
       }
     }
   ],
-  "deltas": []
+  "changes": [
+    [
+      {
+        "name": "bash",
+        "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "command": {
+              "type": "string",
+              "description": "The bash command to execute."
+            },
+            "description": {
+              "type": "string",
+              "description": "Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."
+            },
+            "timeoutMs": {
+              "type": "number",
+              "description": "Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."
+            },
+            "workdir": {
+              "type": "string",
+              "description": "Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."
+            },
+            "run_in_background": {
+              "type": "boolean",
+              "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
+            },
+            "sandbox_permissions": {
+              "type": "string",
+              "description": "The wider sandbox mode this command needs. Only valid as a one-shot retry of a command the sandbox just denied; requires justification and user approval.",
+              "enum": [
+                "workspace-write",
+                "danger-full-access"
+              ]
+            },
+            "justification": {
+              "type": "string",
+              "description": "Required with sandbox_permissions: one sentence for the user explaining why this exact command needs the wider access."
+            }
+          },
+          "required": [
+            "command",
+            "description"
+          ]
+        }
+      },
+      {
+        "name": "skill",
+        "description": "Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "name": {
+              "type": "string",
+              "description": "The exact skill name from the available skills list."
+            }
+          },
+          "required": [
+            "name"
+          ]
+        }
+      },
+      {
+        "name": "subagent",
+        "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "description": {
+              "type": "string",
+              "description": "A short (3-5 word) description of the delegated task, for display."
+            },
+            "prompt": {
+              "type": "string",
+              "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+            },
+            "run_in_background": {
+              "type": "boolean",
+              "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
+            }
+          },
+          "required": [
+            "description",
+            "prompt"
+          ]
+        }
+      },
+      {
+        "name": "subagent_fork",
+        "description": "Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "description": {
+              "type": "string",
+              "description": "A short (3-5 word) description of the delegated task, for display."
+            },
+            "prompt": {
+              "type": "string",
+              "description": "The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."
+            },
+            "run_in_background": {
+              "type": "boolean",
+              "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
+            }
+          },
+          "required": [
+            "description",
+            "prompt"
+          ]
+        }
+      },
+      {
+        "name": "task_kill",
+        "description": "Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "task_id": {
+              "type": "string",
+              "description": "Task id returned by the tool that started the background work."
+            },
+            "reason": {
+              "type": "string",
+              "description": "Optional short reason, recorded in the log and forwarded to the task."
+            }
+          },
+          "required": [
+            "task_id"
+          ]
+        }
+      },
+      {
+        "name": "task_list",
+        "description": "List your background tasks (running and finished) with their ids, kinds, and statuses.",
+        "parameters": {
+          "type": "object",
+          "properties": {}
+        }
+      },
+      {
+        "name": "task_output",
+        "description": "Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "task_id": {
+              "type": "string",
+              "description": "Task id returned by the tool that started the background work."
+            },
+            "wait": {
+              "type": "boolean",
+              "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+            },
+            "timeout_ms": {
+              "type": "number",
+              "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+            }
+          },
+          "required": [
+            "task_id"
+          ]
+        }
+      },
+      {
+        "name": "todo_write",
+        "description": "Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "todos": {
+              "type": "array",
+              "description": "The COMPLETE task list, replacing any previous list.",
+              "items": {
+                "type": "object",
+                "properties": {
+                  "content": {
+                    "type": "string",
+                    "description": "What the task is — a short imperative line."
+                  },
+                  "status": {
+                    "type": "string",
+                    "description": "pending (not started) | in_progress (now) | completed (done).",
+                    "enum": [
+                      "pending",
+                      "in_progress",
+                      "completed"
+                    ]
+                  }
+                },
+                "required": [
+                  "content",
+                  "status"
+                ]
+              }
+            }
+          },
+          "required": [
+            "todos"
+          ]
+        }
+      },
+      {
+        "name": "workflow",
+        "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
+        "parameters": {
+          "type": "object",
+          "properties": {
+            "script": {
+              "type": "string",
+              "description": "The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return <json-value>`)."
+            },
+            "meta": {
+              "type": "object",
+              "description": "The workflow identity block (plain JSON — never code).",
+              "properties": {
+                "name": {
+                  "type": "string",
+                  "description": "Short kebab-case workflow name."
+                },
+                "description": {
+                  "type": "string",
+                  "description": "One-line description of what the workflow does."
+                },
+                "whenToUse": {
+                  "type": "string",
+                  "description": "Optional guidance on when this workflow applies."
+                },
+                "phases": {
+                  "type": "array",
+                  "description": "Optional phase declarations matched by phase() calls.",
+                  "items": {
+                    "type": "object",
+                    "properties": {
+                      "title": {
+                        "type": "string",
+                        "description": "The phase title phase() calls match by exact string."
+                      },
+                      "detail": {
+                        "type": "string",
+                        "description": "Optional one-line description of the phase."
+                      },
+                      "provider": {
+                        "type": "string",
+                        "description": "Optional provider override this phase is expected to use."
+                      },
+                      "model": {
+                        "type": "string",
+                        "description": "Optional model override this phase is expected to use."
+                      }
+                    },
+                    "required": [
+                      "title"
+                    ]
+                  }
+                }
+              },
+              "required": [
+                "name",
+                "description"
+              ]
+            },
+            "args": {
+              "type": "object",
+              "description": "Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]})."
+            }
+          },
+          "required": [
+            "script",
+            "meta"
+          ]
+        }
+      }
+    ]
+  ]
 }
diff --git a/examples/acp-agent/tests/snapshots/repeat-tool-guard/session.jsonl b/examples/acp-agent/tests/snapshots/repeat-tool-guard/session.jsonl
index 7e50e71b3b..20d18973d9 100644
--- a/examples/acp-agent/tests/snapshots/repeat-tool-guard/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/repeat-tool-guard/session.jsonl
@@ -2,13 +2,13 @@
 {"type":"turn/start","seq":0,"time":0,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":0,"data":{"content":[{"type":"text","text":"Write the todo list 'watch the kettle boil' five times in a row without changing it, then reply DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":0,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":0,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":0,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
 {"type":"assistant/chunk","seq":5,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"call_1","name":"todo_write","argumentsDelta":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}}
 {"type":"assistant/chunk","seq":6,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_1","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}}}
 {"type":"assistant/chunk","seq":7,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
 {"type":"assistant/chunk","seq":8,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":9,"time":0,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"call_1","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}],"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[4,5,6,7,8],"surfaceOp":"append"}
+{"type":"assistant/message","seq":9,"time":0,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"call_1","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[4,5,6,7,8],"surfaceOp":"append"}
 {"type":"tool/call","seq":10,"time":0,"data":{"turn":1,"step":1,"callId":"call_1","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}
 {"type":"todo/write","seq":11,"time":0,"data":{"todos":[{"content":"watch the kettle boil","status":"in_progress"}]}}
 {"type":"tool/result","seq":12,"time":0,"data":{"turn":1,"step":1,"callId":"call_1","content":[{"type":"text","text":"Updated todo list: 0 pending, 1 in progress, 0 completed."}],"isError":false},"sourceEventSeqs":[10],"surfaceOp":"append"}
@@ -19,7 +19,7 @@
 {"type":"assistant/chunk","seq":17,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_2","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}}}
 {"type":"assistant/chunk","seq":18,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
 {"type":"assistant/chunk","seq":19,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":20,"time":0,"data":{"turn":1,"step":2,"content":[{"type":"tool-call","id":"call_2","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}],"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[15,16,17,18,19],"surfaceOp":"append"}
+{"type":"assistant/message","seq":20,"time":0,"data":{"turn":1,"step":2,"content":[{"type":"tool-call","id":"call_2","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[15,16,17,18,19],"surfaceOp":"append"}
 {"type":"tool/call","seq":21,"time":0,"data":{"turn":1,"step":2,"callId":"call_2","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}
 {"type":"todo/write","seq":22,"time":0,"data":{"todos":[{"content":"watch the kettle boil","status":"in_progress"}]}}
 {"type":"tool/result","seq":23,"time":0,"data":{"turn":1,"step":2,"callId":"call_2","content":[{"type":"text","text":"Updated todo list: 0 pending, 1 in progress, 0 completed."}],"isError":false},"sourceEventSeqs":[21],"surfaceOp":"append"}
@@ -30,7 +30,7 @@
 {"type":"assistant/chunk","seq":28,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_3","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}}}
 {"type":"assistant/chunk","seq":29,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
 {"type":"assistant/chunk","seq":30,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":31,"time":0,"data":{"turn":1,"step":3,"content":[{"type":"tool-call","id":"call_3","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}],"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[26,27,28,29,30],"surfaceOp":"append"}
+{"type":"assistant/message","seq":31,"time":0,"data":{"turn":1,"step":3,"content":[{"type":"tool-call","id":"call_3","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[26,27,28,29,30],"surfaceOp":"append"}
 {"type":"tool/call","seq":32,"time":0,"data":{"turn":1,"step":3,"callId":"call_3","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}
 {"type":"todo/write","seq":33,"time":0,"data":{"todos":[{"content":"watch the kettle boil","status":"in_progress"}]}}
 {"type":"tool/result","seq":34,"time":0,"data":{"turn":1,"step":3,"callId":"call_3","content":[{"type":"text","text":"Updated todo list: 0 pending, 1 in progress, 0 completed."}],"isError":false},"sourceEventSeqs":[32],"surfaceOp":"append"}
@@ -42,7 +42,7 @@
 {"type":"assistant/chunk","seq":40,"time":0,"data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_4","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}}}
 {"type":"assistant/chunk","seq":41,"time":0,"data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
 {"type":"assistant/chunk","seq":42,"time":0,"data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":43,"time":0,"data":{"turn":1,"step":4,"content":[{"type":"tool-call","id":"call_4","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}],"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[38,39,40,41,42],"surfaceOp":"append"}
+{"type":"assistant/message","seq":43,"time":0,"data":{"turn":1,"step":4,"content":[{"type":"tool-call","id":"call_4","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[38,39,40,41,42],"surfaceOp":"append"}
 {"type":"tool/call","seq":44,"time":0,"data":{"turn":1,"step":4,"callId":"call_4","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}
 {"type":"todo/write","seq":45,"time":0,"data":{"todos":[{"content":"watch the kettle boil","status":"in_progress"}]}}
 {"type":"tool/result","seq":46,"time":0,"data":{"turn":1,"step":4,"callId":"call_4","content":[{"type":"text","text":"Updated todo list: 0 pending, 1 in progress, 0 completed."}],"isError":false},"sourceEventSeqs":[44],"surfaceOp":"append"}
@@ -53,7 +53,7 @@
 {"type":"assistant/chunk","seq":51,"time":0,"data":{"turn":1,"step":5,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_5","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}}}
 {"type":"assistant/chunk","seq":52,"time":0,"data":{"turn":1,"step":5,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
 {"type":"assistant/chunk","seq":53,"time":0,"data":{"turn":1,"step":5,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":54,"time":0,"data":{"turn":1,"step":5,"content":[{"type":"tool-call","id":"call_5","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}],"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[49,50,51,52,53],"surfaceOp":"append"}
+{"type":"assistant/message","seq":54,"time":0,"data":{"turn":1,"step":5,"content":[{"type":"tool-call","id":"call_5","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[49,50,51,52,53],"surfaceOp":"append"}
 {"type":"tool/call","seq":55,"time":0,"data":{"turn":1,"step":5,"callId":"call_5","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"watch the kettle boil\", \"status\": \"in_progress\"}]}"}}
 {"type":"todo/write","seq":56,"time":0,"data":{"todos":[{"content":"watch the kettle boil","status":"in_progress"}]}}
 {"type":"tool/result","seq":57,"time":0,"data":{"turn":1,"step":5,"callId":"call_5","content":[{"type":"text","text":"Updated todo list: 0 pending, 1 in progress, 0 completed."}],"isError":false},"sourceEventSeqs":[55],"surfaceOp":"append"}
@@ -65,6 +65,6 @@
 {"type":"assistant/chunk","seq":63,"time":0,"data":{"turn":1,"step":6,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"DONE."}}}}
 {"type":"assistant/chunk","seq":64,"time":0,"data":{"turn":1,"step":6,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":3}}}}
 {"type":"assistant/chunk","seq":65,"time":0,"data":{"turn":1,"step":6,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":66,"time":0,"data":{"turn":1,"step":6,"content":[{"type":"text","text":"DONE."}],"usage":{"inputTokens":10,"outputTokens":3}},"sourceEventSeqs":[61,62,63,64,65],"surfaceOp":"append"}
+{"type":"assistant/message","seq":66,"time":0,"data":{"turn":1,"step":6,"content":[{"type":"text","text":"DONE."}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":3}},"sourceEventSeqs":[61,62,63,64,65],"surfaceOp":"append"}
 {"type":"step/end","seq":67,"time":0,"data":{"turn":1,"step":6}}
 {"type":"turn/end","seq":68,"time":0,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/repeat-tool-guard/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/repeat-tool-guard/stdout.golden.jsonl
index a0901b6297..3a900cfe34 100644
--- a/examples/acp-agent/tests/snapshots/repeat-tool-guard/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/repeat-tool-guard/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call","toolCallId":"call_1","title":"Update todo list","kind":"other","status":"in_progress","rawInput":[{"content":"watch the kettle boil","status":"in_progress"}]}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"plan","entries":[{"content":"watch the kettle boil","priority":"medium","status":"in_progress"}]}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call_update","toolCallId":"call_1","status":"completed","content":[{"type":"content","content":{"type":"text","text":"Updated todo list: 0 pending, 1 in progress, 0 completed."}}]}}}
diff --git a/examples/acp-agent/tests/snapshots/skill-load/session.jsonl b/examples/acp-agent/tests/snapshots/skill-load/session.jsonl
index 62555f1e79..0fe1a5152c 100644
--- a/examples/acp-agent/tests/snapshots/skill-load/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/skill-load/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783654655602,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783654655603,"data":{"content":[{"type":"text","text":"Load the snapshot-skill skill with the skill tool, then reply DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783654655608,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783654655608,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}","messagePrefix":[{"role":"user","content":[{"type":"text","text":"<system-reminder>\nA skill is a reusable set of task-specific instructions. The following skills are available in this session:\n\n<available_skills>\n- `snapshot-skill`: Exercise project skill discovery and loading in snapshot tests.\n</available_skills>\n\nIf the user names a skill, or the task clearly matches a skill's description, call the `skill` tool with the exact skill name before taking task actions. Load all applicable skills, then follow their full instructions. This catalog contains summaries only; do not infer or follow a skill's instructions until it has been loaded.\n</system-reminder>"}]}]},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783654655608,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}","messagePrefix":[{"role":"user","content":[{"type":"text","text":"<system-reminder>\nA skill is a reusable set of task-specific instructions. The following skills are available in this session:\n\n<available_skills>\n- `snapshot-skill`: Exercise project skill discovery and loading in snapshot tests.\n</available_skills>\n\nIf the user names a skill, or the task clearly matches a skill's description, call the `skill` tool with the exact skill name before taking task actions. Load all applicable skills, then follow their full instructions. This catalog contains summaries only; do not infer or follow a skill's instructions until it has been loaded.\n</system-reminder>"}]}]},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783654655609,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783654655609,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"Load the requested skill."}}}
 {"type":"assistant/chunk","seq":6,"time":1783654655609,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":1,"blockType":"tool-call"}}}
@@ -11,7 +11,7 @@
 {"type":"assistant/chunk","seq":9,"time":1783654655609,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_skill_load","name":"skill","arguments":"{\"name\":\"snapshot-skill\"}"}}}}
 {"type":"assistant/chunk","seq":10,"time":1783654655609,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":100,"outputTokens":20,"cacheReadTokens":0,"reasoningTokens":5}}}}
 {"type":"assistant/chunk","seq":11,"time":1783654655609,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":12,"time":1783654655609,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"Load the requested skill."},{"type":"tool-call","id":"call_skill_load","name":"skill","arguments":"{\"name\":\"snapshot-skill\"}"}],"usage":{"inputTokens":100,"outputTokens":20,"cacheReadTokens":0,"reasoningTokens":5}},"sourceEventSeqs":[4,5,6,7,8,9,10,11],"surfaceOp":"append"}
+{"type":"assistant/message","seq":12,"time":1783654655609,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"Load the requested skill."},{"type":"tool-call","id":"call_skill_load","name":"skill","arguments":"{\"name\":\"snapshot-skill\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":100,"outputTokens":20,"cacheReadTokens":0,"reasoningTokens":5}},"sourceEventSeqs":[4,5,6,7,8,9,10,11],"surfaceOp":"append"}
 {"type":"tool/call","seq":13,"time":1783654655609,"data":{"turn":1,"step":1,"callId":"call_skill_load","name":"skill","arguments":"{\"name\":\"snapshot-skill\"}"}}
 {"type":"tool/result","seq":14,"time":1783654655610,"data":{"turn":1,"step":1,"callId":"call_skill_load","content":[{"type":"text","text":"<skill_content name=\"snapshot-skill\">\n<skill_resources>\nBase directory for this skill: /var/folders/_g/59jgff8x2gqd39f5vy1wnbfc0000gn/T/acp-snap-cwd-DhYwNW/.dsh/skills/snapshot-skill\nResolve relative paths mentioned by this skill against the base directory before using them. Load referenced resources only as needed.\n</skill_resources>\n\n<skill_instructions>\nFollow these snapshot-only instructions.\nResolve referenced resources relative to this skill directory.\n</skill_instructions>\n</skill_content>"}],"isError":false},"sourceEventSeqs":[13],"surfaceOp":"append"}
 {"type":"step/end","seq":15,"time":1783654655610,"data":{"turn":1,"step":1}}
@@ -24,6 +24,6 @@
 {"type":"assistant/chunk","seq":22,"time":1783654655611,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":23,"time":1783654655611,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":180,"outputTokens":10,"cacheReadTokens":0,"reasoningTokens":4}}}}
 {"type":"assistant/chunk","seq":24,"time":1783654655611,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":25,"time":1783654655611,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The skill is loaded."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":180,"outputTokens":10,"cacheReadTokens":0,"reasoningTokens":4}},"sourceEventSeqs":[17,18,19,20,21,22,23,24],"surfaceOp":"append"}
+{"type":"assistant/message","seq":25,"time":1783654655611,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The skill is loaded."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":180,"outputTokens":10,"cacheReadTokens":0,"reasoningTokens":4}},"sourceEventSeqs":[17,18,19,20,21,22,23,24],"surfaceOp":"append"}
 {"type":"step/end","seq":26,"time":1783654655611,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":27,"time":1783654655611,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/skill-load/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/skill-load/stdout.golden.jsonl
index 10198918d6..6a3dfcc4c4 100644
--- a/examples/acp-agent/tests/snapshots/skill-load/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/skill-load/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"Load the requested skill."}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call","toolCallId":"call_skill_load","title":"Load skill snapshot-skill","kind":"read","status":"in_progress","rawInput":"snapshot-skill"}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call_update","toolCallId":"call_skill_load","status":"completed","content":[{"type":"content","content":{"type":"text","text":"<skill_content name=\"snapshot-skill\">\n<skill_resources>\nBase directory for this skill: {{cwd}}/.dsh/skills/snapshot-skill\nResolve relative paths mentioned by this skill against the base directory before using them. Load referenced resources only as needed.\n</skill_resources>\n\n<skill_instructions>\nFollow these snapshot-only instructions.\nResolve referenced resources relative to this skill directory.\n</skill_instructions>\n</skill_content>"}}]}}}
diff --git a/examples/acp-agent/tests/snapshots/skill-load/system-prompt.golden.md b/examples/acp-agent/tests/snapshots/skill-load/system-prompt.golden.md
index d0ce1272d0..43ecb9746f 100644
--- a/examples/acp-agent/tests/snapshots/skill-load/system-prompt.golden.md
+++ b/examples/acp-agent/tests/snapshots/skill-load/system-prompt.golden.md
@@ -7,6 +7,8 @@ Verify your work by running the code or tests. Keep answers brief and factual.
 
 Check the [exit code: N] marker on every bash result; investigate failures before moving on.
 
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
 Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
 <!-- dsh-user-approval-policy:never -->
 
diff --git a/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.golden.json b/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.golden.json
index 68d4b037d7..e422a063da 100644
--- a/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.golden.json
+++ b/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.golden.json
@@ -2,7 +2,7 @@
   "initial": [
     {
       "name": "bash",
-      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; poll it with `bash_output` and stop it with `bash_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
+      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -24,7 +24,7 @@
           },
           "run_in_background": {
             "type": "boolean",
-            "description": "Run in the background and return a task id immediately. No timeout applies."
+            "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
           },
           "sandbox_permissions": {
             "type": "string",
@@ -45,38 +45,6 @@
         ]
       }
     },
-    {
-      "name": "bash_kill",
-      "description": "Ask the executor to kill a running background bash task by task id.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
-    {
-      "name": "bash_output",
-      "description": "Read new output from a background bash task started with `bash` + `run_in_background`. Returns only output produced since the previous bash_output call, plus the task status. Tasks keep running while you do other work; poll again later for more output.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
     {
       "name": "skill",
       "description": "Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.",
@@ -95,7 +63,7 @@
     },
     {
       "name": "subagent",
-      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation.",
+      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -106,6 +74,10 @@
           "prompt": {
             "type": "string",
             "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -116,7 +88,7 @@
     },
     {
       "name": "subagent_fork",
-      "description": "Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps.",
+      "description": "Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -127,6 +99,10 @@
           "prompt": {
             "type": "string",
             "description": "The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -135,6 +111,58 @@
         ]
       }
     },
+    {
+      "name": "task_kill",
+      "description": "Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "reason": {
+            "type": "string",
+            "description": "Optional short reason, recorded in the log and forwarded to the task."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
+    {
+      "name": "task_list",
+      "description": "List your background tasks (running and finished) with their ids, kinds, and statuses.",
+      "parameters": {
+        "type": "object",
+        "properties": {}
+      }
+    },
+    {
+      "name": "task_output",
+      "description": "Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "wait": {
+            "type": "boolean",
+            "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+          },
+          "timeout_ms": {
+            "type": "number",
+            "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
     {
       "name": "todo_write",
       "description": "Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).",
@@ -175,7 +203,7 @@
     },
     {
       "name": "workflow",
-      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), `model` (override). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
+      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -213,6 +241,10 @@
                       "type": "string",
                       "description": "Optional one-line description of the phase."
                     },
+                    "provider": {
+                      "type": "string",
+                      "description": "Optional provider override this phase is expected to use."
+                    },
                     "model": {
                       "type": "string",
                       "description": "Optional model override this phase is expected to use."
@@ -241,5 +273,5 @@
       }
     }
   ],
-  "deltas": []
+  "changes": []
 }
diff --git a/examples/acp-agent/tests/snapshots/subagent-fork/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-fork/session.1.jsonl
index 409fa1428d..4ecb61b60f 100644
--- a/examples/acp-agent/tests/snapshots/subagent-fork/session.1.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-fork/session.1.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352134837,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352134838,"data":{"content":[{"type":"text","text":"Remember this fact for later: the project codeword is MARMALADE. Reply with the single word OK and stop. Do not use any tools."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352134840,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352134840,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352134840,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352135465,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352135465,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352135621,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -33,13 +33,13 @@
 {"type":"assistant/chunk","seq":31,"time":1783352135771,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"OK"}}}}
 {"type":"assistant/chunk","seq":32,"time":1783352135771,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2885,"outputTokens":25,"cacheReadTokens":0,"reasoningTokens":23}}}}
 {"type":"assistant/chunk","seq":33,"time":1783352135771,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":34,"time":1783352135773,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to remember the codeword \"MARMALADE\" and reply with just \"OK\"."},{"type":"text","text":"OK"}],"usage":{"inputTokens":2885,"outputTokens":25,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33],"surfaceOp":"append"}
+{"type":"assistant/message","seq":34,"time":1783352135773,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to remember the codeword \"MARMALADE\" and reply with just \"OK\"."},{"type":"text","text":"OK"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2885,"outputTokens":25,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33],"surfaceOp":"append"}
 {"type":"step/end","seq":35,"time":1783352135773,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":36,"time":1783352135773,"data":{"turn":1,"reason":{"kind":"completed"}}}
 {"type":"turn/start","seq":37,"time":1783352137162,"data":{"turn":2,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":38,"time":1783352137163,"data":{"content":[{"type":"text","text":"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":39,"time":1783352137163,"data":{"turn":2,"step":1}}
-{"type":"request/header","seq":40,"time":1783352137163,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"resume"}}
+{"type":"request/header","seq":40,"time":1783352137163,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"resume"}}
 {"type":"assistant/chunk","seq":41,"time":1783352137783,"data":{"turn":2,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":42,"time":1783352137783,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":43,"time":1783352137961,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -84,6 +84,6 @@
 {"type":"assistant/chunk","seq":82,"time":1783352138307,"data":{"turn":2,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"MARMALADE"}}}}
 {"type":"assistant/chunk","seq":83,"time":1783352138307,"data":{"turn":2,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":97,"outputTokens":39,"cacheReadTokens":2816,"reasoningTokens":34}}}}
 {"type":"assistant/chunk","seq":84,"time":1783352138307,"data":{"turn":2,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":85,"time":1783352138308,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user asked me to remember the project codeword \"MARMALADE\" and now they're asking what it is. I should just reply with that word."},{"type":"text","text":"MARMALADE"}],"usage":{"inputTokens":97,"outputTokens":39,"cacheReadTokens":2816,"reasoningTokens":34}},"sourceEventSeqs":[41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84],"surfaceOp":"append"}
+{"type":"assistant/message","seq":85,"time":1783352138308,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user asked me to remember the project codeword \"MARMALADE\" and now they're asking what it is. I should just reply with that word."},{"type":"text","text":"MARMALADE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":97,"outputTokens":39,"cacheReadTokens":2816,"reasoningTokens":34}},"sourceEventSeqs":[41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84],"surfaceOp":"append"}
 {"type":"step/end","seq":86,"time":1783352138308,"data":{"turn":2,"step":1}}
 {"type":"turn/end","seq":87,"time":1783352138308,"data":{"turn":2,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-fork/session.jsonl b/examples/acp-agent/tests/snapshots/subagent-fork/session.jsonl
index d61503ffbb..64da6e60e5 100644
--- a/examples/acp-agent/tests/snapshots/subagent-fork/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-fork/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352134837,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352134838,"data":{"content":[{"type":"text","text":"Remember this fact for later: the project codeword is MARMALADE. Reply with the single word OK and stop. Do not use any tools."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352134840,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352134840,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352134840,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352135465,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352135465,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352135621,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -33,7 +33,7 @@
 {"type":"assistant/chunk","seq":31,"time":1783352135771,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"OK"}}}}
 {"type":"assistant/chunk","seq":32,"time":1783352135771,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2885,"outputTokens":25,"cacheReadTokens":0,"reasoningTokens":23}}}}
 {"type":"assistant/chunk","seq":33,"time":1783352135771,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":34,"time":1783352135773,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to remember the codeword \"MARMALADE\" and reply with just \"OK\"."},{"type":"text","text":"OK"}],"usage":{"inputTokens":2885,"outputTokens":25,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33],"surfaceOp":"append"}
+{"type":"assistant/message","seq":34,"time":1783352135773,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to remember the codeword \"MARMALADE\" and reply with just \"OK\"."},{"type":"text","text":"OK"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2885,"outputTokens":25,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33],"surfaceOp":"append"}
 {"type":"step/end","seq":35,"time":1783352135773,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":36,"time":1783352135773,"data":{"turn":1,"reason":{"kind":"completed"}}}
 {"type":"turn/start","seq":37,"time":1783352135780,"data":{"turn":2,"trigger":{"kind":"message","source":{"kind":"user"}}}}
@@ -149,7 +149,7 @@
 {"type":"assistant/chunk","seq":147,"time":1783352137158,"data":{"turn":2,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_sAtKUseRzHRBvL4CF7XF1334","name":"subagent_fork","arguments":"{\"description\": \"Recall project codeword\", \"prompt\": \"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else.\"}"}}}}
 {"type":"assistant/chunk","seq":148,"time":1783352137158,"data":{"turn":2,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":158,"outputTokens":147,"cacheReadTokens":2816,"reasoningTokens":59}}}}
 {"type":"assistant/chunk","seq":149,"time":1783352137159,"data":{"turn":2,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":150,"time":1783352137159,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use subagent_fork to delegate a question to a child agent. The child agent inherits this conversation and should be able to answer: the project codeword is MARMALADE. After the subagent returns, I should reply with PARENT_DONE."},{"type":"tool-call","id":"call_00_sAtKUseRzHRBvL4CF7XF1334","name":"subagent_fork","arguments":"{\"description\": \"Recall project codeword\", \"prompt\": \"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else.\"}"}],"usage":{"inputTokens":158,"outputTokens":147,"cacheReadTokens":2816,"reasoningTokens":59}},"sourceEventSeqs":[40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149],"surfaceOp":"append"}
+{"type":"assistant/message","seq":150,"time":1783352137159,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use subagent_fork to delegate a question to a child agent. The child agent inherits this conversation and should be able to answer: the project codeword is MARMALADE. After the subagent returns, I should reply with PARENT_DONE."},{"type":"tool-call","id":"call_00_sAtKUseRzHRBvL4CF7XF1334","name":"subagent_fork","arguments":"{\"description\": \"Recall project codeword\", \"prompt\": \"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else.\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":158,"outputTokens":147,"cacheReadTokens":2816,"reasoningTokens":59}},"sourceEventSeqs":[40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149],"surfaceOp":"append"}
 {"type":"tool/call","seq":151,"time":1783352137159,"data":{"turn":2,"step":1,"callId":"call_00_sAtKUseRzHRBvL4CF7XF1334","name":"subagent_fork","arguments":"{\"description\": \"Recall project codeword\", \"prompt\": \"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else.\"}"}}
 {"type":"tool/result","seq":152,"time":1783352138315,"data":{"turn":2,"step":1,"callId":"call_00_sAtKUseRzHRBvL4CF7XF1334","content":[{"type":"text","text":"MARMALADE"}],"isError":false},"sourceEventSeqs":[151],"surfaceOp":"append"}
 {"type":"step/end","seq":153,"time":1783352138316,"data":{"turn":2,"step":1}}
@@ -189,6 +189,6 @@
 {"type":"assistant/chunk","seq":187,"time":1783352139274,"data":{"turn":2,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"PARENT_DONE"}}}}
 {"type":"assistant/chunk","seq":188,"time":1783352139274,"data":{"turn":2,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":65,"outputTokens":30,"cacheReadTokens":3072,"reasoningTokens":25}}}}
 {"type":"assistant/chunk","seq":189,"time":1783352139274,"data":{"turn":2,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":190,"time":1783352139274,"data":{"turn":2,"step":2,"content":[{"type":"reasoning","text":"The forked child agent correctly returned \"MARMALADE\". Now I need to reply with \"PARENT_DONE\"."},{"type":"text","text":"PARENT_DONE"}],"usage":{"inputTokens":65,"outputTokens":30,"cacheReadTokens":3072,"reasoningTokens":25}},"sourceEventSeqs":[155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189],"surfaceOp":"append"}
+{"type":"assistant/message","seq":190,"time":1783352139274,"data":{"turn":2,"step":2,"content":[{"type":"reasoning","text":"The forked child agent correctly returned \"MARMALADE\". Now I need to reply with \"PARENT_DONE\"."},{"type":"text","text":"PARENT_DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":65,"outputTokens":30,"cacheReadTokens":3072,"reasoningTokens":25}},"sourceEventSeqs":[155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189],"surfaceOp":"append"}
 {"type":"step/end","seq":191,"time":1783352139274,"data":{"turn":2,"step":2}}
 {"type":"turn/end","seq":192,"time":1783352139274,"data":{"turn":2,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-fork/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/subagent-fork/stdout.golden.jsonl
index 44fc4402fc..0574f2e6c4 100644
--- a/examples/acp-agent/tests/snapshots/subagent-fork/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-fork/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl
index 85a6405535..c99a5681e2 100644
--- a/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-mixed/session.1.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352145224,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352145224,"data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352145224,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352145224,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352145224,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352145820,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352145821,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352145985,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -31,6 +31,6 @@
 {"type":"assistant/chunk","seq":29,"time":1783352146129,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"ALPHA"}}}}
 {"type":"assistant/chunk","seq":30,"time":1783352146129,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":48,"outputTokens":23,"cacheReadTokens":2816,"reasoningTokens":19}}}}
 {"type":"assistant/chunk","seq":31,"time":1783352146129,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":32,"time":1783352146130,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user asked me to reply with exactly the word \"ALPHA\" and nothing else."},{"type":"text","text":"ALPHA"}],"usage":{"inputTokens":48,"outputTokens":23,"cacheReadTokens":2816,"reasoningTokens":19}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],"surfaceOp":"append"}
+{"type":"assistant/message","seq":32,"time":1783352146130,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user asked me to reply with exactly the word \"ALPHA\" and nothing else."},{"type":"text","text":"ALPHA"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":48,"outputTokens":23,"cacheReadTokens":2816,"reasoningTokens":19}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],"surfaceOp":"append"}
 {"type":"step/end","seq":33,"time":1783352146130,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":34,"time":1783352146130,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl b/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl
index b9b6b48a9a..adfcb6f60e 100644
--- a/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-mixed/session.2.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352142834,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352142834,"data":{"content":[{"type":"text","text":"Remember this fact for later: the project codeword is SAFFRON. Reply with the single word OK and stop. Do not use any tools."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352142835,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352142836,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352142836,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352143493,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352143494,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352143621,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -27,13 +27,13 @@
 {"type":"assistant/chunk","seq":25,"time":1783352143768,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"OK"}}}}
 {"type":"assistant/chunk","seq":26,"time":1783352143768,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2883,"outputTokens":19,"cacheReadTokens":0,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":27,"time":1783352143768,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":28,"time":1783352143771,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to remember a codeword and just reply with \"OK\"."},{"type":"text","text":"OK"}],"usage":{"inputTokens":2883,"outputTokens":19,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27],"surfaceOp":"append"}
+{"type":"assistant/message","seq":28,"time":1783352143771,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to remember a codeword and just reply with \"OK\"."},{"type":"text","text":"OK"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2883,"outputTokens":19,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27],"surfaceOp":"append"}
 {"type":"step/end","seq":29,"time":1783352143771,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":30,"time":1783352143771,"data":{"turn":1,"reason":{"kind":"completed"}}}
 {"type":"turn/start","seq":31,"time":1783352147508,"data":{"turn":2,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":32,"time":1783352147509,"data":{"content":[{"type":"text","text":"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":33,"time":1783352147509,"data":{"turn":2,"step":1}}
-{"type":"request/header","seq":34,"time":1783352147509,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"resume"}}
+{"type":"request/header","seq":34,"time":1783352147509,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"resume"}}
 {"type":"assistant/chunk","seq":35,"time":1783352147925,"data":{"turn":2,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":36,"time":1783352147925,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":37,"time":1783352148019,"data":{"turn":2,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -74,6 +74,6 @@
 {"type":"assistant/chunk","seq":72,"time":1783352148345,"data":{"turn":2,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"SAFFRON"}}}}
 {"type":"assistant/chunk","seq":73,"time":1783352148345,"data":{"turn":2,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":95,"outputTokens":35,"cacheReadTokens":2816,"reasoningTokens":31}}}}
 {"type":"assistant/chunk","seq":74,"time":1783352148345,"data":{"turn":2,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":75,"time":1783352148345,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user is asking me to recall the project codeword that was mentioned earlier in the conversation. I was told to remember it: SAFFRON."},{"type":"text","text":"SAFFRON"}],"usage":{"inputTokens":95,"outputTokens":35,"cacheReadTokens":2816,"reasoningTokens":31}},"sourceEventSeqs":[35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74],"surfaceOp":"append"}
+{"type":"assistant/message","seq":75,"time":1783352148345,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"The user is asking me to recall the project codeword that was mentioned earlier in the conversation. I was told to remember it: SAFFRON."},{"type":"text","text":"SAFFRON"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":95,"outputTokens":35,"cacheReadTokens":2816,"reasoningTokens":31}},"sourceEventSeqs":[35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74],"surfaceOp":"append"}
 {"type":"step/end","seq":76,"time":1783352148345,"data":{"turn":2,"step":1}}
 {"type":"turn/end","seq":77,"time":1783352148345,"data":{"turn":2,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-mixed/session.jsonl b/examples/acp-agent/tests/snapshots/subagent-mixed/session.jsonl
index ff5cfda975..1ea4f541e1 100644
--- a/examples/acp-agent/tests/snapshots/subagent-mixed/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-mixed/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352142834,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352142834,"data":{"content":[{"type":"text","text":"Remember this fact for later: the project codeword is SAFFRON. Reply with the single word OK and stop. Do not use any tools."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352142835,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352142836,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352142836,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352143493,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352143494,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352143621,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -27,7 +27,7 @@
 {"type":"assistant/chunk","seq":25,"time":1783352143768,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"OK"}}}}
 {"type":"assistant/chunk","seq":26,"time":1783352143768,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2883,"outputTokens":19,"cacheReadTokens":0,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":27,"time":1783352143768,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":28,"time":1783352143771,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to remember a codeword and just reply with \"OK\"."},{"type":"text","text":"OK"}],"usage":{"inputTokens":2883,"outputTokens":19,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27],"surfaceOp":"append"}
+{"type":"assistant/message","seq":28,"time":1783352143771,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to remember a codeword and just reply with \"OK\"."},{"type":"text","text":"OK"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2883,"outputTokens":19,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27],"surfaceOp":"append"}
 {"type":"step/end","seq":29,"time":1783352143771,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":30,"time":1783352143771,"data":{"turn":1,"reason":{"kind":"completed"}}}
 {"type":"turn/start","seq":31,"time":1783352143779,"data":{"turn":2,"trigger":{"kind":"message","source":{"kind":"user"}}}}
@@ -108,7 +108,7 @@
 {"type":"assistant/chunk","seq":106,"time":1783352145221,"data":{"turn":2,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_YvHr2bGomk5HhpgDTvE81896","name":"subagent","arguments":"{\"description\": \"Reply ALPHA only\", \"prompt\": \"Reply with exactly the word ALPHA and nothing else.\"}"}}}}
 {"type":"assistant/chunk","seq":107,"time":1783352145221,"data":{"turn":2,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":185,"outputTokens":110,"cacheReadTokens":2816,"reasoningTokens":35}}}}
 {"type":"assistant/chunk","seq":108,"time":1783352145221,"data":{"turn":2,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":109,"time":1783352145221,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"Let me do these two delegations one at a time as requested.\n\nFirst, I'll use the subagent tool (fresh child) to reply with \"ALPHA\"."},{"type":"tool-call","id":"call_00_YvHr2bGomk5HhpgDTvE81896","name":"subagent","arguments":"{\"description\": \"Reply ALPHA only\", \"prompt\": \"Reply with exactly the word ALPHA and nothing else.\"}"}],"usage":{"inputTokens":185,"outputTokens":110,"cacheReadTokens":2816,"reasoningTokens":35}},"sourceEventSeqs":[34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108],"surfaceOp":"append"}
+{"type":"assistant/message","seq":109,"time":1783352145221,"data":{"turn":2,"step":1,"content":[{"type":"reasoning","text":"Let me do these two delegations one at a time as requested.\n\nFirst, I'll use the subagent tool (fresh child) to reply with \"ALPHA\"."},{"type":"tool-call","id":"call_00_YvHr2bGomk5HhpgDTvE81896","name":"subagent","arguments":"{\"description\": \"Reply ALPHA only\", \"prompt\": \"Reply with exactly the word ALPHA and nothing else.\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":185,"outputTokens":110,"cacheReadTokens":2816,"reasoningTokens":35}},"sourceEventSeqs":[34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108],"surfaceOp":"append"}
 {"type":"tool/call","seq":110,"time":1783352145222,"data":{"turn":2,"step":1,"callId":"call_00_YvHr2bGomk5HhpgDTvE81896","name":"subagent","arguments":"{\"description\": \"Reply ALPHA only\", \"prompt\": \"Reply with exactly the word ALPHA and nothing else.\"}"}}
 {"type":"tool/result","seq":111,"time":1783352146133,"data":{"turn":2,"step":1,"callId":"call_00_YvHr2bGomk5HhpgDTvE81896","content":[{"type":"text","text":"ALPHA"}],"isError":false},"sourceEventSeqs":[110],"surfaceOp":"append"}
 {"type":"step/end","seq":112,"time":1783352146134,"data":{"turn":2,"step":1}}
@@ -204,7 +204,7 @@
 {"type":"assistant/chunk","seq":202,"time":1783352147502,"data":{"turn":2,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_JSr5rhREq23wSmwSkCP77184","name":"subagent_fork","arguments":"{\"description\": \"Recall project codeword\", \"prompt\": \"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else.\"}"}}}}
 {"type":"assistant/chunk","seq":203,"time":1783352147502,"data":{"turn":2,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":54,"outputTokens":128,"cacheReadTokens":3072,"reasoningTokens":40}}}}
 {"type":"assistant/chunk","seq":204,"time":1783352147502,"data":{"turn":2,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":205,"time":1783352147503,"data":{"turn":2,"step":2,"content":[{"type":"reasoning","text":"The first subagent returned \"ALPHA\". Now I need to use the subagent_fork tool (forked child that inherits this conversation) to ask about the project codeword."},{"type":"tool-call","id":"call_00_JSr5rhREq23wSmwSkCP77184","name":"subagent_fork","arguments":"{\"description\": \"Recall project codeword\", \"prompt\": \"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else.\"}"}],"usage":{"inputTokens":54,"outputTokens":128,"cacheReadTokens":3072,"reasoningTokens":40}},"sourceEventSeqs":[114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204],"surfaceOp":"append"}
+{"type":"assistant/message","seq":205,"time":1783352147503,"data":{"turn":2,"step":2,"content":[{"type":"reasoning","text":"The first subagent returned \"ALPHA\". Now I need to use the subagent_fork tool (forked child that inherits this conversation) to ask about the project codeword."},{"type":"tool-call","id":"call_00_JSr5rhREq23wSmwSkCP77184","name":"subagent_fork","arguments":"{\"description\": \"Recall project codeword\", \"prompt\": \"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else.\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":54,"outputTokens":128,"cacheReadTokens":3072,"reasoningTokens":40}},"sourceEventSeqs":[114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204],"surfaceOp":"append"}
 {"type":"tool/call","seq":206,"time":1783352147503,"data":{"turn":2,"step":2,"callId":"call_00_JSr5rhREq23wSmwSkCP77184","name":"subagent_fork","arguments":"{\"description\": \"Recall project codeword\", \"prompt\": \"What is the project codeword mentioned earlier in this conversation? Reply with exactly that one word and nothing else.\"}"}}
 {"type":"tool/result","seq":207,"time":1783352148348,"data":{"turn":2,"step":2,"callId":"call_00_JSr5rhREq23wSmwSkCP77184","content":[{"type":"text","text":"SAFFRON"}],"isError":false},"sourceEventSeqs":[206],"surfaceOp":"append"}
 {"type":"step/end","seq":208,"time":1783352148348,"data":{"turn":2,"step":2}}
@@ -283,6 +283,6 @@
 {"type":"assistant/chunk","seq":281,"time":1783352149821,"data":{"turn":2,"step":3,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"PARENT_DONE"}}}}
 {"type":"assistant/chunk","seq":282,"time":1783352149821,"data":{"turn":2,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":69,"outputTokens":69,"cacheReadTokens":3200,"reasoningTokens":64}}}}
 {"type":"assistant/chunk","seq":283,"time":1783352149821,"data":{"turn":2,"step":3,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":284,"time":1783352149822,"data":{"turn":2,"step":3,"content":[{"type":"reasoning","text":"Both subagents returned:\n1. First (fresh child): \"ALPHA\"\n2. Second (forked child): \"SAFFRON\" - correctly inherited the conversation context where I was asked to remember the codeword \"SAFFRON\".\n\nNow I reply with \"PARENT_DONE\" as instructed."},{"type":"text","text":"PARENT_DONE"}],"usage":{"inputTokens":69,"outputTokens":69,"cacheReadTokens":3200,"reasoningTokens":64}},"sourceEventSeqs":[210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251,252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283],"surfaceOp":"append"}
+{"type":"assistant/message","seq":284,"time":1783352149822,"data":{"turn":2,"step":3,"content":[{"type":"reasoning","text":"Both subagents returned:\n1. First (fresh child): \"ALPHA\"\n2. Second (forked child): \"SAFFRON\" - correctly inherited the conversation context where I was asked to remember the codeword \"SAFFRON\".\n\nNow I reply with \"PARENT_DONE\" as instructed."},{"type":"text","text":"PARENT_DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":69,"outputTokens":69,"cacheReadTokens":3200,"reasoningTokens":64}},"sourceEventSeqs":[210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251,252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283],"surfaceOp":"append"}
 {"type":"step/end","seq":285,"time":1783352149822,"data":{"turn":2,"step":3}}
 {"type":"turn/end","seq":286,"time":1783352149822,"data":{"turn":2,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-mixed/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/subagent-mixed/stdout.golden.jsonl
index 08ad14dc9c..bfc195aa36 100644
--- a/examples/acp-agent/tests/snapshots/subagent-mixed/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-mixed/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl
index e046c226d0..86c481c5ff 100644
--- a/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-multi/session.1.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352127670,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352127670,"data":{"content":[{"type":"text","text":"Reply with exactly the word ALPHA and nothing else."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352127671,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352127671,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352127671,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352128125,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352128125,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352128240,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -31,6 +31,6 @@
 {"type":"assistant/chunk","seq":29,"time":1783352128365,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"ALPHA"}}}}
 {"type":"assistant/chunk","seq":30,"time":1783352128365,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":49,"outputTokens":23,"cacheReadTokens":2816,"reasoningTokens":19}}}}
 {"type":"assistant/chunk","seq":31,"time":1783352128365,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":32,"time":1783352128365,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word \"ALPHA\" and nothing else."},{"type":"text","text":"ALPHA"}],"usage":{"inputTokens":49,"outputTokens":23,"cacheReadTokens":2816,"reasoningTokens":19}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],"surfaceOp":"append"}
+{"type":"assistant/message","seq":32,"time":1783352128365,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word \"ALPHA\" and nothing else."},{"type":"text","text":"ALPHA"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":49,"outputTokens":23,"cacheReadTokens":2816,"reasoningTokens":19}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],"surfaceOp":"append"}
 {"type":"step/end","seq":33,"time":1783352128365,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":34,"time":1783352128366,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl b/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl
index e113159d39..483e687a14 100644
--- a/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-multi/session.2.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352129662,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352129662,"data":{"content":[{"type":"text","text":"Reply with exactly the word BETA and nothing else."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352129663,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352129663,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352129663,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352130236,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352130236,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352130375,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -29,6 +29,6 @@
 {"type":"assistant/chunk","seq":27,"time":1783352130527,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"BETA"}}}}
 {"type":"assistant/chunk","seq":28,"time":1783352130527,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":48,"outputTokens":21,"cacheReadTokens":2816,"reasoningTokens":18}}}}
 {"type":"assistant/chunk","seq":29,"time":1783352130527,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":30,"time":1783352130528,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word \"BETA\" and nothing else."},{"type":"text","text":"BETA"}],"usage":{"inputTokens":48,"outputTokens":21,"cacheReadTokens":2816,"reasoningTokens":18}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29],"surfaceOp":"append"}
+{"type":"assistant/message","seq":30,"time":1783352130528,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word \"BETA\" and nothing else."},{"type":"text","text":"BETA"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":48,"outputTokens":21,"cacheReadTokens":2816,"reasoningTokens":18}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29],"surfaceOp":"append"}
 {"type":"step/end","seq":31,"time":1783352130528,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":32,"time":1783352130528,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-multi/session.jsonl b/examples/acp-agent/tests/snapshots/subagent-multi/session.jsonl
index c7a079391f..a8093fba04 100644
--- a/examples/acp-agent/tests/snapshots/subagent-multi/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-multi/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352126251,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352126251,"data":{"content":[{"type":"text","text":"Use the subagent tool TWICE, once at a time, to delegate two subtasks to child agents. First subtask: 'Reply with exactly the word ALPHA and nothing else.' Second subtask (after the first returns): 'Reply with exactly the word BETA and nothing else.' After both subagents return, reply with the single word PARENT_DONE and stop. Do not use the bash tool."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352126252,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352126253,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352126253,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352126729,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352126729,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352126848,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -92,7 +92,7 @@
 {"type":"assistant/chunk","seq":90,"time":1783352127665,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_7zDCCjdsQgrk5LR2bAEQ1010","name":"subagent","arguments":"{\"description\": \"Return ALPHA only\", \"prompt\": \"Reply with exactly the word ALPHA and nothing else.\"}"}}}}
 {"type":"assistant/chunk","seq":91,"time":1783352127665,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2938,"outputTokens":124,"cacheReadTokens":0,"reasoningTokens":49}}}}
 {"type":"assistant/chunk","seq":92,"time":1783352127665,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":93,"time":1783352127668,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the subagent tool twice, sequentially (one at a time). First subagent should reply with \"ALPHA\", second with \"BETA\". After both return, I reply with \"PARENT_DONE\"."},{"type":"tool-call","id":"call_00_7zDCCjdsQgrk5LR2bAEQ1010","name":"subagent","arguments":"{\"description\": \"Return ALPHA only\", \"prompt\": \"Reply with exactly the word ALPHA and nothing else.\"}"}],"usage":{"inputTokens":2938,"outputTokens":124,"cacheReadTokens":0,"reasoningTokens":49}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92],"surfaceOp":"append"}
+{"type":"assistant/message","seq":93,"time":1783352127668,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the subagent tool twice, sequentially (one at a time). First subagent should reply with \"ALPHA\", second with \"BETA\". After both return, I reply with \"PARENT_DONE\"."},{"type":"tool-call","id":"call_00_7zDCCjdsQgrk5LR2bAEQ1010","name":"subagent","arguments":"{\"description\": \"Return ALPHA only\", \"prompt\": \"Reply with exactly the word ALPHA and nothing else.\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2938,"outputTokens":124,"cacheReadTokens":0,"reasoningTokens":49}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92],"surfaceOp":"append"}
 {"type":"tool/call","seq":94,"time":1783352127668,"data":{"turn":1,"step":1,"callId":"call_00_7zDCCjdsQgrk5LR2bAEQ1010","name":"subagent","arguments":"{\"description\": \"Return ALPHA only\", \"prompt\": \"Reply with exactly the word ALPHA and nothing else.\"}"}}
 {"type":"tool/result","seq":95,"time":1783352128371,"data":{"turn":1,"step":1,"callId":"call_00_7zDCCjdsQgrk5LR2bAEQ1010","content":[{"type":"text","text":"ALPHA"}],"isError":false},"sourceEventSeqs":[94],"surfaceOp":"append"}
 {"type":"step/end","seq":96,"time":1783352128371,"data":{"turn":1,"step":1}}
@@ -158,7 +158,7 @@
 {"type":"assistant/chunk","seq":156,"time":1783352129661,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_FudNKuJ0fchSptGy3Scw1411","name":"subagent","arguments":"{\"description\": \"Return BETA only\", \"prompt\": \"Reply with exactly the word BETA and nothing else.\"}"}}}}
 {"type":"assistant/chunk","seq":157,"time":1783352129661,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":133,"outputTokens":96,"cacheReadTokens":2944,"reasoningTokens":23}}}}
 {"type":"assistant/chunk","seq":158,"time":1783352129661,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":159,"time":1783352129661,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"First subagent returned \"ALPHA\". Now I'll call the second subagent to return \"BETA\"."},{"type":"tool-call","id":"call_00_FudNKuJ0fchSptGy3Scw1411","name":"subagent","arguments":"{\"description\": \"Return BETA only\", \"prompt\": \"Reply with exactly the word BETA and nothing else.\"}"}],"usage":{"inputTokens":133,"outputTokens":96,"cacheReadTokens":2944,"reasoningTokens":23}},"sourceEventSeqs":[98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158],"surfaceOp":"append"}
+{"type":"assistant/message","seq":159,"time":1783352129661,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"First subagent returned \"ALPHA\". Now I'll call the second subagent to return \"BETA\"."},{"type":"tool-call","id":"call_00_FudNKuJ0fchSptGy3Scw1411","name":"subagent","arguments":"{\"description\": \"Return BETA only\", \"prompt\": \"Reply with exactly the word BETA and nothing else.\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":133,"outputTokens":96,"cacheReadTokens":2944,"reasoningTokens":23}},"sourceEventSeqs":[98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158],"surfaceOp":"append"}
 {"type":"tool/call","seq":160,"time":1783352129661,"data":{"turn":1,"step":2,"callId":"call_00_FudNKuJ0fchSptGy3Scw1411","name":"subagent","arguments":"{\"description\": \"Return BETA only\", \"prompt\": \"Reply with exactly the word BETA and nothing else.\"}"}}
 {"type":"tool/result","seq":161,"time":1783352130531,"data":{"turn":1,"step":2,"callId":"call_00_FudNKuJ0fchSptGy3Scw1411","content":[{"type":"text","text":"BETA"}],"isError":false},"sourceEventSeqs":[160],"surfaceOp":"append"}
 {"type":"step/end","seq":162,"time":1783352130531,"data":{"turn":1,"step":2}}
@@ -203,6 +203,6 @@
 {"type":"assistant/chunk","seq":201,"time":1783352131242,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"PARENT_DONE"}}}}
 {"type":"assistant/chunk","seq":202,"time":1783352131242,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":115,"outputTokens":35,"cacheReadTokens":3072,"reasoningTokens":30}}}}
 {"type":"assistant/chunk","seq":203,"time":1783352131242,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":204,"time":1783352131243,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"Both subagents have returned: first with \"ALPHA\", second with \"BETA\". Now I should reply with \"PARENT_DONE\"."},{"type":"text","text":"PARENT_DONE"}],"usage":{"inputTokens":115,"outputTokens":35,"cacheReadTokens":3072,"reasoningTokens":30}},"sourceEventSeqs":[164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203],"surfaceOp":"append"}
+{"type":"assistant/message","seq":204,"time":1783352131243,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"Both subagents have returned: first with \"ALPHA\", second with \"BETA\". Now I should reply with \"PARENT_DONE\"."},{"type":"text","text":"PARENT_DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":115,"outputTokens":35,"cacheReadTokens":3072,"reasoningTokens":30}},"sourceEventSeqs":[164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203],"surfaceOp":"append"}
 {"type":"step/end","seq":205,"time":1783352131243,"data":{"turn":1,"step":3}}
 {"type":"turn/end","seq":206,"time":1783352131243,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-multi/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/subagent-multi/stdout.golden.jsonl
index 93b38e33cb..7f9d2c51fa 100644
--- a/examples/acp-agent/tests/snapshots/subagent-multi/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-multi/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-spawn/session.1.jsonl b/examples/acp-agent/tests/snapshots/subagent-spawn/session.1.jsonl
index fedd7cbfb4..8dd26c4e70 100644
--- a/examples/acp-agent/tests/snapshots/subagent-spawn/session.1.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-spawn/session.1.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352120856,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352120856,"data":{"content":[{"type":"text","text":"Reply with exactly the word CHILD_OK and nothing else."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352120856,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352120856,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352120856,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352121437,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352121438,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352121635,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -29,6 +29,6 @@
 {"type":"assistant/chunk","seq":27,"time":1783352121777,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"CHILD_OK"}}}}
 {"type":"assistant/chunk","seq":28,"time":1783352121777,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":48,"outputTokens":21,"cacheReadTokens":2816,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":29,"time":1783352121777,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":30,"time":1783352121777,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word CHILD_OK and nothing else."},{"type":"text","text":"CHILD_OK"}],"usage":{"inputTokens":48,"outputTokens":21,"cacheReadTokens":2816,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29],"surfaceOp":"append"}
+{"type":"assistant/message","seq":30,"time":1783352121777,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word CHILD_OK and nothing else."},{"type":"text","text":"CHILD_OK"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":48,"outputTokens":21,"cacheReadTokens":2816,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29],"surfaceOp":"append"}
 {"type":"step/end","seq":31,"time":1783352121778,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":32,"time":1783352121778,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-spawn/session.jsonl b/examples/acp-agent/tests/snapshots/subagent-spawn/session.jsonl
index b550a8cb19..6a87cfabb5 100644
--- a/examples/acp-agent/tests/snapshots/subagent-spawn/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-spawn/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352119273,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352119274,"data":{"content":[{"type":"text","text":"Use the subagent tool exactly once to delegate this subtask to a child agent: 'Reply with exactly the word CHILD_OK and nothing else.' After the subagent returns, reply with the single word PARENT_DONE and stop. Do not use the bash tool."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352119275,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352119281,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352119281,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352119925,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352119925,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352120053,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -110,7 +110,7 @@
 {"type":"assistant/chunk","seq":108,"time":1783352120851,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_gVbLWC12Qu8JheZpVRRz8749","name":"subagent","arguments":"{\"description\": \"Reply with CHILD_OK\", \"prompt\": \"Reply with exactly the word CHILD_OK and nothing else.\"}"}}}}
 {"type":"assistant/chunk","seq":109,"time":1783352120852,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2907,"outputTokens":142,"cacheReadTokens":0,"reasoningTokens":67}}}}
 {"type":"assistant/chunk","seq":110,"time":1783352120852,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":111,"time":1783352120854,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to:\n1. Use the subagent tool exactly once to delegate the task: \"Reply with exactly the word CHILD_OK and nothing else.\"\n2. After the subagent returns, reply with the single word PARENT_DONE and stop.\n3. Do not use the bash tool.\n\nLet me do this."},{"type":"tool-call","id":"call_00_gVbLWC12Qu8JheZpVRRz8749","name":"subagent","arguments":"{\"description\": \"Reply with CHILD_OK\", \"prompt\": \"Reply with exactly the word CHILD_OK and nothing else.\"}"}],"usage":{"inputTokens":2907,"outputTokens":142,"cacheReadTokens":0,"reasoningTokens":67}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110],"surfaceOp":"append"}
+{"type":"assistant/message","seq":111,"time":1783352120854,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to:\n1. Use the subagent tool exactly once to delegate the task: \"Reply with exactly the word CHILD_OK and nothing else.\"\n2. After the subagent returns, reply with the single word PARENT_DONE and stop.\n3. Do not use the bash tool.\n\nLet me do this."},{"type":"tool-call","id":"call_00_gVbLWC12Qu8JheZpVRRz8749","name":"subagent","arguments":"{\"description\": \"Reply with CHILD_OK\", \"prompt\": \"Reply with exactly the word CHILD_OK and nothing else.\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2907,"outputTokens":142,"cacheReadTokens":0,"reasoningTokens":67}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110],"surfaceOp":"append"}
 {"type":"tool/call","seq":112,"time":1783352120854,"data":{"turn":1,"step":1,"callId":"call_00_gVbLWC12Qu8JheZpVRRz8749","name":"subagent","arguments":"{\"description\": \"Reply with CHILD_OK\", \"prompt\": \"Reply with exactly the word CHILD_OK and nothing else.\"}"}}
 {"type":"tool/result","seq":113,"time":1783352121784,"data":{"turn":1,"step":1,"callId":"call_00_gVbLWC12Qu8JheZpVRRz8749","content":[{"type":"text","text":"CHILD_OK"}],"isError":false},"sourceEventSeqs":[112],"surfaceOp":"append"}
 {"type":"step/end","seq":114,"time":1783352121784,"data":{"turn":1,"step":1}}
@@ -155,6 +155,6 @@
 {"type":"assistant/chunk","seq":153,"time":1783352122731,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"PARENT_DONE"}}}}
 {"type":"assistant/chunk","seq":154,"time":1783352122732,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":120,"outputTokens":35,"cacheReadTokens":2944,"reasoningTokens":30}}}}
 {"type":"assistant/chunk","seq":155,"time":1783352122732,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":156,"time":1783352122732,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The subagent returned \"CHILD_OK\" as expected. Now I need to reply with the single word \"PARENT_DONE\" and stop."},{"type":"text","text":"PARENT_DONE"}],"usage":{"inputTokens":120,"outputTokens":35,"cacheReadTokens":2944,"reasoningTokens":30}},"sourceEventSeqs":[116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155],"surfaceOp":"append"}
+{"type":"assistant/message","seq":156,"time":1783352122732,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The subagent returned \"CHILD_OK\" as expected. Now I need to reply with the single word \"PARENT_DONE\" and stop."},{"type":"text","text":"PARENT_DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":120,"outputTokens":35,"cacheReadTokens":2944,"reasoningTokens":30}},"sourceEventSeqs":[116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155],"surfaceOp":"append"}
 {"type":"step/end","seq":157,"time":1783352122732,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":158,"time":1783352122732,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/subagent-spawn/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/subagent-spawn/stdout.golden.jsonl
index 26d45699cc..4a07c313e0 100644
--- a/examples/acp-agent/tests/snapshots/subagent-spawn/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/subagent-spawn/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/text-turn/session.jsonl b/examples/acp-agent/tests/snapshots/text-turn/session.jsonl
index 59e803f394..04ac2f9794 100644
--- a/examples/acp-agent/tests/snapshots/text-turn/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/text-turn/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783600629541,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783600629541,"data":{"content":[{"type":"text","text":"Reply with exactly the word: PONG. Do not use any tools."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783600629542,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783600629542,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783600629542,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783600630819,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783600630820,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783600630822,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -31,6 +31,6 @@
 {"type":"assistant/chunk","seq":29,"time":1783600631009,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"PONG"}}}}
 {"type":"assistant/chunk","seq":30,"time":1783600631009,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3091,"outputTokens":23,"cacheReadTokens":0,"reasoningTokens":20}}}}
 {"type":"assistant/chunk","seq":31,"time":1783600631009,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":32,"time":1783600631011,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word \"PONG\" and not use any tools."},{"type":"text","text":"PONG"}],"usage":{"inputTokens":3091,"outputTokens":23,"cacheReadTokens":0,"reasoningTokens":20}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],"surfaceOp":"append"}
+{"type":"assistant/message","seq":32,"time":1783600631011,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly the word \"PONG\" and not use any tools."},{"type":"text","text":"PONG"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3091,"outputTokens":23,"cacheReadTokens":0,"reasoningTokens":20}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],"surfaceOp":"append"}
 {"type":"step/end","seq":33,"time":1783600631011,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":34,"time":1783600631011,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/text-turn/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/text-turn/stdout.golden.jsonl
index 1059a9cc6c..29528d9984 100644
--- a/examples/acp-agent/tests/snapshots/text-turn/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/text-turn/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/text-turn/system-prompt.golden.md b/examples/acp-agent/tests/snapshots/text-turn/system-prompt.golden.md
index d0ce1272d0..43ecb9746f 100644
--- a/examples/acp-agent/tests/snapshots/text-turn/system-prompt.golden.md
+++ b/examples/acp-agent/tests/snapshots/text-turn/system-prompt.golden.md
@@ -7,6 +7,8 @@ Verify your work by running the code or tests. Keep answers brief and factual.
 
 Check the [exit code: N] marker on every bash result; investigate failures before moving on.
 
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
 Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
 <!-- dsh-user-approval-policy:never -->
 
diff --git a/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.golden.json b/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.golden.json
index 68d4b037d7..e422a063da 100644
--- a/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.golden.json
+++ b/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.golden.json
@@ -2,7 +2,7 @@
   "initial": [
     {
       "name": "bash",
-      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; poll it with `bash_output` and stop it with `bash_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
+      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -24,7 +24,7 @@
           },
           "run_in_background": {
             "type": "boolean",
-            "description": "Run in the background and return a task id immediately. No timeout applies."
+            "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
           },
           "sandbox_permissions": {
             "type": "string",
@@ -45,38 +45,6 @@
         ]
       }
     },
-    {
-      "name": "bash_kill",
-      "description": "Ask the executor to kill a running background bash task by task id.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
-    {
-      "name": "bash_output",
-      "description": "Read new output from a background bash task started with `bash` + `run_in_background`. Returns only output produced since the previous bash_output call, plus the task status. Tasks keep running while you do other work; poll again later for more output.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
     {
       "name": "skill",
       "description": "Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.",
@@ -95,7 +63,7 @@
     },
     {
       "name": "subagent",
-      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation.",
+      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -106,6 +74,10 @@
           "prompt": {
             "type": "string",
             "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -116,7 +88,7 @@
     },
     {
       "name": "subagent_fork",
-      "description": "Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps.",
+      "description": "Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -127,6 +99,10 @@
           "prompt": {
             "type": "string",
             "description": "The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -135,6 +111,58 @@
         ]
       }
     },
+    {
+      "name": "task_kill",
+      "description": "Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "reason": {
+            "type": "string",
+            "description": "Optional short reason, recorded in the log and forwarded to the task."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
+    {
+      "name": "task_list",
+      "description": "List your background tasks (running and finished) with their ids, kinds, and statuses.",
+      "parameters": {
+        "type": "object",
+        "properties": {}
+      }
+    },
+    {
+      "name": "task_output",
+      "description": "Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "wait": {
+            "type": "boolean",
+            "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+          },
+          "timeout_ms": {
+            "type": "number",
+            "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
     {
       "name": "todo_write",
       "description": "Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).",
@@ -175,7 +203,7 @@
     },
     {
       "name": "workflow",
-      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), `model` (override). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
+      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -213,6 +241,10 @@
                       "type": "string",
                       "description": "Optional one-line description of the phase."
                     },
+                    "provider": {
+                      "type": "string",
+                      "description": "Optional provider override this phase is expected to use."
+                    },
                     "model": {
                       "type": "string",
                       "description": "Optional model override this phase is expected to use."
@@ -241,5 +273,5 @@
       }
     }
   ],
-  "deltas": []
+  "changes": []
 }
diff --git a/examples/acp-agent/tests/snapshots/todo-plan/session.jsonl b/examples/acp-agent/tests/snapshots/todo-plan/session.jsonl
index 8e93132d30..909afc44cd 100644
--- a/examples/acp-agent/tests/snapshots/todo-plan/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/todo-plan/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352057655,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352057655,"data":{"content":[{"type":"text","text":"Use the todo_write tool to record a plan with exactly three todos: \"read the code\" (in_progress), \"write the fix\" (pending), \"run the tests\" (pending). Send all three in one todo_write call. Then reply with the single word DONE and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352057657,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352057657,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352057657,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352058320,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352058320,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352058426,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -94,7 +94,7 @@
 {"type":"assistant/chunk","seq":92,"time":1783352059096,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_fjAnBThbDjxepBtp3hDt3264","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"read the code\", \"status\": \"in_progress\"}, {\"content\": \"write the fix\", \"status\": \"pending\"}, {\"content\": \"run the tests\", \"status\": \"pending\"}]}"}}}}
 {"type":"assistant/chunk","seq":93,"time":1783352059096,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2913,"outputTokens":121,"cacheReadTokens":0,"reasoningTokens":31}}}}
 {"type":"assistant/chunk","seq":94,"time":1783352059096,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":95,"time":1783352059099,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the todo_write tool to record a plan with exactly three todos in the specified statuses, then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_fjAnBThbDjxepBtp3hDt3264","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"read the code\", \"status\": \"in_progress\"}, {\"content\": \"write the fix\", \"status\": \"pending\"}, {\"content\": \"run the tests\", \"status\": \"pending\"}]}"}],"usage":{"inputTokens":2913,"outputTokens":121,"cacheReadTokens":0,"reasoningTokens":31}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94],"surfaceOp":"append"}
+{"type":"assistant/message","seq":95,"time":1783352059099,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the todo_write tool to record a plan with exactly three todos in the specified statuses, then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_fjAnBThbDjxepBtp3hDt3264","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"read the code\", \"status\": \"in_progress\"}, {\"content\": \"write the fix\", \"status\": \"pending\"}, {\"content\": \"run the tests\", \"status\": \"pending\"}]}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2913,"outputTokens":121,"cacheReadTokens":0,"reasoningTokens":31}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94],"surfaceOp":"append"}
 {"type":"tool/call","seq":96,"time":1783352059099,"data":{"turn":1,"step":1,"callId":"call_00_fjAnBThbDjxepBtp3hDt3264","name":"todo_write","arguments":"{\"todos\": [{\"content\": \"read the code\", \"status\": \"in_progress\"}, {\"content\": \"write the fix\", \"status\": \"pending\"}, {\"content\": \"run the tests\", \"status\": \"pending\"}]}"}}
 {"type":"todo/write","seq":97,"time":1783352059100,"data":{"todos":[{"content":"read the code","status":"in_progress"},{"content":"write the fix","status":"pending"},{"content":"run the tests","status":"pending"}]}}
 {"type":"tool/result","seq":98,"time":1783352059101,"data":{"turn":1,"step":1,"callId":"call_00_fjAnBThbDjxepBtp3hDt3264","content":[{"type":"text","text":"Updated todo list: 2 pending, 1 in progress, 0 completed."}],"isError":false},"sourceEventSeqs":[96],"surfaceOp":"append"}
@@ -129,6 +129,6 @@
 {"type":"assistant/chunk","seq":127,"time":1783352059980,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":128,"time":1783352059980,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":237,"outputTokens":24,"cacheReadTokens":2816,"reasoningTokens":21}}}}
 {"type":"assistant/chunk","seq":129,"time":1783352059980,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":130,"time":1783352059981,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The todos have been written successfully. Now I just need to reply with the single word \"DONE\"."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":237,"outputTokens":24,"cacheReadTokens":2816,"reasoningTokens":21}},"sourceEventSeqs":[101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129],"surfaceOp":"append"}
+{"type":"assistant/message","seq":130,"time":1783352059981,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The todos have been written successfully. Now I just need to reply with the single word \"DONE\"."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":237,"outputTokens":24,"cacheReadTokens":2816,"reasoningTokens":21}},"sourceEventSeqs":[101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129],"surfaceOp":"append"}
 {"type":"step/end","seq":131,"time":1783352059981,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":132,"time":1783352059981,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/todo-plan/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/todo-plan/stdout.golden.jsonl
index f060b6b92f..fb2cd46879 100644
--- a/examples/acp-agent/tests/snapshots/todo-plan/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/todo-plan/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/tool-call-turn/session.jsonl b/examples/acp-agent/tests/snapshots/tool-call-turn/session.jsonl
index 20a11443a7..f631c9bf54 100644
--- a/examples/acp-agent/tests/snapshots/tool-call-turn/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/tool-call-turn/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352044771,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352044771,"data":{"content":[{"type":"text","text":"Use the bash tool to run exactly: echo SNAPSHOT_OK. Then reply with the single word DONE and stop."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352044773,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352044773,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352044773,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352045294,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352045294,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352045396,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -57,7 +57,7 @@
 {"type":"assistant/chunk","seq":55,"time":1783352045866,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_Rn2Mz1y8uZN62ukEXiNO2077","name":"bash","arguments":"{\"command\": \"echo SNAPSHOT_OK\", \"description\": \"Run echo SNAPSHOT_OK\"}"}}}}
 {"type":"assistant/chunk","seq":56,"time":1783352045866,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2879,"outputTokens":89,"cacheReadTokens":0,"reasoningTokens":17}}}}
 {"type":"assistant/chunk","seq":57,"time":1783352045866,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":58,"time":1783352045867,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a specific bash command and then reply with DONE."},{"type":"tool-call","id":"call_00_Rn2Mz1y8uZN62ukEXiNO2077","name":"bash","arguments":"{\"command\": \"echo SNAPSHOT_OK\", \"description\": \"Run echo SNAPSHOT_OK\"}"}],"usage":{"inputTokens":2879,"outputTokens":89,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57],"surfaceOp":"append"}
+{"type":"assistant/message","seq":58,"time":1783352045867,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a specific bash command and then reply with DONE."},{"type":"tool-call","id":"call_00_Rn2Mz1y8uZN62ukEXiNO2077","name":"bash","arguments":"{\"command\": \"echo SNAPSHOT_OK\", \"description\": \"Run echo SNAPSHOT_OK\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2879,"outputTokens":89,"cacheReadTokens":0,"reasoningTokens":17}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57],"surfaceOp":"append"}
 {"type":"tool/call","seq":59,"time":1783352045867,"data":{"turn":1,"step":1,"callId":"call_00_Rn2Mz1y8uZN62ukEXiNO2077","name":"bash","arguments":"{\"command\": \"echo SNAPSHOT_OK\", \"description\": \"Run echo SNAPSHOT_OK\"}"}}
 {"type":"tool/result","seq":60,"time":1783352045879,"data":{"turn":1,"step":1,"callId":"call_00_Rn2Mz1y8uZN62ukEXiNO2077","content":[{"type":"text","text":"SNAPSHOT_OK\n"}],"isError":false},"sourceEventSeqs":[59],"surfaceOp":"append"}
 {"type":"step/end","seq":61,"time":1783352045880,"data":{"turn":1,"step":1}}
@@ -95,6 +95,6 @@
 {"type":"assistant/chunk","seq":93,"time":1783352047156,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":94,"time":1783352047156,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":170,"outputTokens":28,"cacheReadTokens":2816,"reasoningTokens":25}}}}
 {"type":"assistant/chunk","seq":95,"time":1783352047156,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":96,"time":1783352047158,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The command executed successfully and printed SNAPSHOT_OK. Now I need to reply with the single word DONE."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":170,"outputTokens":28,"cacheReadTokens":2816,"reasoningTokens":25}},"sourceEventSeqs":[63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95],"surfaceOp":"append"}
+{"type":"assistant/message","seq":96,"time":1783352047158,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The command executed successfully and printed SNAPSHOT_OK. Now I need to reply with the single word DONE."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":170,"outputTokens":28,"cacheReadTokens":2816,"reasoningTokens":25}},"sourceEventSeqs":[63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95],"surfaceOp":"append"}
 {"type":"step/end","seq":97,"time":1783352047158,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":98,"time":1783352047158,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/tool-call-turn/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/tool-call-turn/stdout.golden.jsonl
index 041ea02703..6abf7023b5 100644
--- a/examples/acp-agent/tests/snapshots/tool-call-turn/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/tool-call-turn/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/workflow-run/session.1.jsonl b/examples/acp-agent/tests/snapshots/workflow-run/session.1.jsonl
index a8c1d6018b..3d89428bbd 100644
--- a/examples/acp-agent/tests/snapshots/workflow-run/session.1.jsonl
+++ b/examples/acp-agent/tests/snapshots/workflow-run/session.1.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783600636316,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783600636316,"data":{"content":[{"type":"text","text":"Reply with exactly the word WF_CHILD_OK and nothing else."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783600636316,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783600636317,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783600636317,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783600638073,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783600638073,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783600638173,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -31,6 +31,6 @@
 {"type":"assistant/chunk","seq":29,"time":1783600638280,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"WF_CHILD_OK"}}}}
 {"type":"assistant/chunk","seq":30,"time":1783600638280,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":17,"outputTokens":23,"cacheReadTokens":3072,"reasoningTokens":18}}}}
 {"type":"assistant/chunk","seq":31,"time":1783600638280,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":32,"time":1783600638281,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly \"WF_CHILD_OK\" and nothing else."},{"type":"text","text":"WF_CHILD_OK"}],"usage":{"inputTokens":17,"outputTokens":23,"cacheReadTokens":3072,"reasoningTokens":18}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],"surfaceOp":"append"}
+{"type":"assistant/message","seq":32,"time":1783600638281,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to reply with exactly \"WF_CHILD_OK\" and nothing else."},{"type":"text","text":"WF_CHILD_OK"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":17,"outputTokens":23,"cacheReadTokens":3072,"reasoningTokens":18}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],"surfaceOp":"append"}
 {"type":"step/end","seq":33,"time":1783600638281,"data":{"turn":1,"step":1}}
 {"type":"turn/end","seq":34,"time":1783600638281,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/workflow-run/session.jsonl b/examples/acp-agent/tests/snapshots/workflow-run/session.jsonl
index ab53f20550..3e0ae3da73 100644
--- a/examples/acp-agent/tests/snapshots/workflow-run/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/workflow-run/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783600631838,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783600631838,"data":{"content":[{"type":"text","text":"Use the workflow tool exactly once, with args omitted, meta set to { \"name\": \"snapshot-flow\", \"description\": \"one child for the snapshot\" }, and this EXACT script body (copy it verbatim):\nphase('Run')\nconst reply = await agent('Reply with exactly the word WF_CHILD_OK and nothing else.')\nreturn { reply }\nAfter the workflow returns, reply with the single word WORKFLOW_DONE and stop. Do not use any other tool."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783600631839,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783600631839,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783600631839,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783600634643,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783600634643,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783600635634,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -158,7 +158,7 @@
 {"type":"assistant/chunk","seq":156,"time":1783600636246,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_dD2BLuNeJCTh2iiYC1QR3449","name":"workflow","arguments":"{\"meta\": {\"name\": \"snapshot-flow\", \"description\": \"one child for the snapshot\"}, \"script\": \"\\nphase('Run')\\nconst reply = await agent('Reply with exactly the word WF_CHILD_OK and nothing else.')\\nreturn { reply }\\n\"}"}}}}
 {"type":"assistant/chunk","seq":157,"time":1783600636246,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":3174,"outputTokens":191,"cacheReadTokens":0,"reasoningTokens":88}}}}
 {"type":"assistant/chunk","seq":158,"time":1783600636246,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":159,"time":1783600636247,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the workflow tool exactly once with specific parameters. Let me carefully follow the instructions:\n\n1. args omitted (so I don't include it)\n2. meta = { \"name\": \"snapshot-flow\", \"description\": \"one child for the snapshot\" }\n3. script = as given verbatim\n4. After it returns, reply with \"WORKFLOW_DONE\"\n\nLet me do exactly that."},{"type":"tool-call","id":"call_00_dD2BLuNeJCTh2iiYC1QR3449","name":"workflow","arguments":"{\"meta\": {\"name\": \"snapshot-flow\", \"description\": \"one child for the snapshot\"}, \"script\": \"\\nphase('Run')\\nconst reply = await agent('Reply with exactly the word WF_CHILD_OK and nothing else.')\\nreturn { reply }\\n\"}"}],"usage":{"inputTokens":3174,"outputTokens":191,"cacheReadTokens":0,"reasoningTokens":88}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158],"surfaceOp":"append"}
+{"type":"assistant/message","seq":159,"time":1783600636247,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the workflow tool exactly once with specific parameters. Let me carefully follow the instructions:\n\n1. args omitted (so I don't include it)\n2. meta = { \"name\": \"snapshot-flow\", \"description\": \"one child for the snapshot\" }\n3. script = as given verbatim\n4. After it returns, reply with \"WORKFLOW_DONE\"\n\nLet me do exactly that."},{"type":"tool-call","id":"call_00_dD2BLuNeJCTh2iiYC1QR3449","name":"workflow","arguments":"{\"meta\": {\"name\": \"snapshot-flow\", \"description\": \"one child for the snapshot\"}, \"script\": \"\\nphase('Run')\\nconst reply = await agent('Reply with exactly the word WF_CHILD_OK and nothing else.')\\nreturn { reply }\\n\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3174,"outputTokens":191,"cacheReadTokens":0,"reasoningTokens":88}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158],"surfaceOp":"append"}
 {"type":"tool/call","seq":160,"time":1783600636247,"data":{"turn":1,"step":1,"callId":"call_00_dD2BLuNeJCTh2iiYC1QR3449","name":"workflow","arguments":"{\"meta\": {\"name\": \"snapshot-flow\", \"description\": \"one child for the snapshot\"}, \"script\": \"\\nphase('Run')\\nconst reply = await agent('Reply with exactly the word WF_CHILD_OK and nothing else.')\\nreturn { reply }\\n\"}"}}
 {"type":"tool/result","seq":161,"time":1783600638304,"data":{"turn":1,"step":1,"callId":"call_00_dD2BLuNeJCTh2iiYC1QR3449","content":[{"type":"text","text":"workflow \"snapshot-flow\" completed (1 agent).\nReturn value:\n{\n  \"reply\": \"WF_CHILD_OK\"\n}"}],"isError":false},"sourceEventSeqs":[160],"surfaceOp":"append"}
 {"type":"step/end","seq":162,"time":1783600638304,"data":{"turn":1,"step":1}}
@@ -204,6 +204,6 @@
 {"type":"assistant/chunk","seq":202,"time":1783600640865,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"WORKFLOW_DONE"}}}}
 {"type":"assistant/chunk","seq":203,"time":1783600640865,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":328,"outputTokens":36,"cacheReadTokens":3072,"reasoningTokens":30}}}}
 {"type":"assistant/chunk","seq":204,"time":1783600640865,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":205,"time":1783600640865,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The workflow returned successfully with the reply \"WF_CHILD_OK\". Now I need to reply with exactly \"WORKFLOW_DONE\" and stop."},{"type":"text","text":"WORKFLOW_DONE"}],"usage":{"inputTokens":328,"outputTokens":36,"cacheReadTokens":3072,"reasoningTokens":30}},"sourceEventSeqs":[164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204],"surfaceOp":"append"}
+{"type":"assistant/message","seq":205,"time":1783600640865,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The workflow returned successfully with the reply \"WF_CHILD_OK\". Now I need to reply with exactly \"WORKFLOW_DONE\" and stop."},{"type":"text","text":"WORKFLOW_DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":328,"outputTokens":36,"cacheReadTokens":3072,"reasoningTokens":30}},"sourceEventSeqs":[164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204],"surfaceOp":"append"}
 {"type":"step/end","seq":206,"time":1783600640865,"data":{"turn":1,"step":2}}
 {"type":"turn/end","seq":207,"time":1783600640865,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/workflow-run/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/workflow-run/stdout.golden.jsonl
index 9c0bbd37be..4566071090 100644
--- a/examples/acp-agent/tests/snapshots/workflow-run/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/workflow-run/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/workspace-context/input.json b/examples/acp-agent/tests/snapshots/workspace-context/input.json
new file mode 100644
index 0000000000..94fd9dae92
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/workspace-context/input.json
@@ -0,0 +1,7 @@
+{
+  "steps": [
+    { "op": "initialize" },
+    { "op": "newSession" },
+    { "op": "prompt", "text": "Read nested/task.txt with the read tool, then reply DONE." }
+  ]
+}
diff --git a/examples/acp-agent/tests/snapshots/workspace-context/replay.override.json b/examples/acp-agent/tests/snapshots/workspace-context/replay.override.json
new file mode 100644
index 0000000000..ef70491338
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/workspace-context/replay.override.json
@@ -0,0 +1,22 @@
+[
+  {
+    "kind": "chunks",
+    "chunks": [
+      { "type": "block-start", "index": 0, "blockType": "tool-call" },
+      { "type": "tool-call-delta", "index": 0, "id": "call_workspace_read", "name": "read", "argumentsDelta": "{\"file_path\":\"nested/task.txt\"}" },
+      { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_workspace_read", "name": "read", "arguments": "{\"file_path\":\"nested/task.txt\"}" } },
+      { "type": "usage", "usage": { "inputTokens": 10, "outputTokens": 5 } },
+      { "type": "finish", "reason": { "kind": "tool-calls" } }
+    ]
+  },
+  {
+    "kind": "chunks",
+    "chunks": [
+      { "type": "block-start", "index": 0, "blockType": "text" },
+      { "type": "text-delta", "index": 0, "text": "DONE" },
+      { "type": "block-end", "index": 0, "block": { "type": "text", "text": "DONE" } },
+      { "type": "usage", "usage": { "inputTokens": 10, "outputTokens": 2 } },
+      { "type": "finish", "reason": { "kind": "stop" } }
+    ]
+  }
+]
diff --git a/examples/acp-agent/tests/snapshots/workspace-context/session.jsonl b/examples/acp-agent/tests/snapshots/workspace-context/session.jsonl
new file mode 100644
index 0000000000..b133708b9a
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/workspace-context/session.jsonl
@@ -0,0 +1,24 @@
+{"type":"session","version":0,"id":"{{sessionId}}","createdAt":0,"cwd":"{{cwd}}"}
+{"type":"turn/start","seq":0,"time":1783778297065,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
+{"type":"user/message","seq":1,"time":1783778297066,"data":{"content":[{"type":"text","text":"Read nested/task.txt with the read tool, then reply DONE."}],"source":{"kind":"user"}},"surfaceOp":"append"}
+{"type":"step/start","seq":2,"time":1783778297069,"data":{"turn":1,"step":1}}
+{"type":"request/header","seq":3,"time":1783778297070,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}","messagePrefix":[{"role":"user","content":[{"type":"text","text":"<system-reminder>\nThe following workspace instructions may be relevant to your work. Use them as guidance when applicable. More specific instructions take precedence over broader ones. They do not override system, developer, or direct user instructions.\n\nInstructions from: AGENTS.md\n\nRoot snapshot instruction.\n\n</system-reminder>"}]}]},"reason":"initial"}}
+{"type":"assistant/chunk","seq":4,"time":1783778297070,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
+{"type":"assistant/chunk","seq":5,"time":1783778297070,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"call_workspace_read","name":"read","argumentsDelta":"{\"file_path\":\"nested/task.txt\"}"}}}
+{"type":"assistant/chunk","seq":6,"time":1783778297070,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_workspace_read","name":"read","arguments":"{\"file_path\":\"nested/task.txt\"}"}}}}
+{"type":"assistant/chunk","seq":7,"time":1783778297070,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":5}}}}
+{"type":"assistant/chunk","seq":8,"time":1783778297070,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
+{"type":"assistant/message","seq":9,"time":1783778297070,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"call_workspace_read","name":"read","arguments":"{\"file_path\":\"nested/task.txt\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[4,5,6,7,8],"surfaceOp":"append"}
+{"type":"tool/call","seq":10,"time":1783778297070,"data":{"turn":1,"step":1,"callId":"call_workspace_read","name":"read","arguments":"{\"file_path\":\"nested/task.txt\"}"}}
+{"type":"tool/result","seq":11,"time":1783778297072,"data":{"turn":1,"step":1,"callId":"call_workspace_read","content":[{"type":"text","text":"<path>{{cwd}}/nested/task.txt</path>\n<type>file</type>\n<content>\n1: snapshot task\n\n(End of file - total 1 lines)\n</content>"}],"isError":false},"sourceEventSeqs":[10],"surfaceOp":"append"}
+{"type":"context/message","seq":12,"time":1783778297072,"data":{"content":[{"type":"text","text":"<system-reminder>\nAdditional instructions from: nested/AGENTS.md\n\nThese instructions apply to work under `nested`. Use them as guidance when relevant; more specific instructions take precedence. They do not override system, developer, or direct user instructions.\n\nNested snapshot instruction.\n\n</system-reminder>"}],"source":{"kind":"plugin","plugin":"workspace-context"},"envelope":"raw","meta":{"kind":"workspace-instructions","version":1,"changes":[{"action":"set","scope":"nested","path":"nested/AGENTS.md","digest":"c446df9a85c7e73a3055f394a4822a19ac9ead5a"}]}},"surfaceOp":"append"}
+{"type":"step/end","seq":13,"time":1783778297072,"data":{"turn":1,"step":1}}
+{"type":"step/start","seq":14,"time":1783778297072,"data":{"turn":1,"step":2}}
+{"type":"assistant/chunk","seq":15,"time":1783778297073,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}
+{"type":"assistant/chunk","seq":16,"time":1783778297073,"data":{"turn":1,"step":2,"chunk":{"type":"text-delta","index":0,"text":"DONE"}}}
+{"type":"assistant/chunk","seq":17,"time":1783778297073,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"DONE"}}}}
+{"type":"assistant/chunk","seq":18,"time":1783778297073,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":2}}}}
+{"type":"assistant/chunk","seq":19,"time":1783778297073,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
+{"type":"assistant/message","seq":20,"time":1783778297073,"data":{"turn":1,"step":2,"content":[{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":2}},"sourceEventSeqs":[15,16,17,18,19],"surfaceOp":"append"}
+{"type":"step/end","seq":21,"time":1783778297073,"data":{"turn":1,"step":2}}
+{"type":"turn/end","seq":22,"time":1783778297073,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/workspace-context/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/workspace-context/stdout.golden.jsonl
new file mode 100644
index 0000000000..f99eb73eef
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/workspace-context/stdout.golden.jsonl
@@ -0,0 +1,6 @@
+{"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call","toolCallId":"call_workspace_read","title":"Read nested/task.txt","kind":"read","status":"in_progress","locations":[{"path":"nested/task.txt","line":1}]}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"tool_call_update","toolCallId":"call_workspace_read","status":"completed","content":[{"type":"content","content":{"type":"text","text":"<path>{{cwd}}/nested/task.txt</path>\n<type>file</type>\n<content>\n1: snapshot task\n\n(End of file - total 1 lines)\n</content>"}}]}}}
+{"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_message_chunk","content":{"type":"text","text":"DONE"}}}}
+{"jsonrpc":"2.0","id":3,"result":{"stopReason":"end_turn"}}
diff --git a/examples/acp-agent/tests/snapshots/workspace-context/system-prompt.golden.md b/examples/acp-agent/tests/snapshots/workspace-context/system-prompt.golden.md
new file mode 100644
index 0000000000..b1fc71924b
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/workspace-context/system-prompt.golden.md
@@ -0,0 +1,21 @@
+You are an AI agent powered by the DeepSeek Harness SDK.
+
+You are a coding assistant powered by the deepseek-v4-flash model. Your working directory is {{cwd}}.
+
+Verify your work by running the code or tests. Keep answers brief and factual.
+
+
+Use the read tool — not shell commands like cat — to inspect text files. Results include line numbers. Use offset and limit to continue reading large files.
+
+Use the write tool to create files or completely replace file contents. Existing files are overwritten, so read an existing file first (the default fs-policy requires it) and prefer edit for targeted changes.
+
+Use the edit tool for targeted changes to existing UTF-8 text files. It replaces literal old_string with new_string; by default old_string must appear exactly once. If old_string appears multiple times, provide a more specific old_string or set replace_all to true. Read the file first (the default fs-policy requires it), unless you just created or edited it in this session.
+
+Check the [exit code: N] marker on every bash result; investigate failures before moving on.
+
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
+Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
+<!-- dsh-user-approval-policy:never -->
+
+Use the workflow tool ONLY when the user explicitly asks for a workflow or for large multi-agent orchestration: you write a JavaScript script (the tool description documents the exact format) that fans work out across many subagents with phases and structured results. For one or two delegations, prefer plain subagent calls.
diff --git a/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.golden.json b/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.golden.json
new file mode 100644
index 0000000000..4b08a1e365
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.golden.json
@@ -0,0 +1,352 @@
+{
+  "initial": [
+    {
+      "name": "bash",
+      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "command": {
+            "type": "string",
+            "description": "The bash command to execute."
+          },
+          "description": {
+            "type": "string",
+            "description": "Clear, concise description of what this command does in active voice, 5-10 words (shown in the UI). Examples: \"ls\" → \"List files in current directory\"; \"git status\" → \"Show working tree status\"; \"npm install\" → \"Install package dependencies\"."
+          },
+          "timeoutMs": {
+            "type": "number",
+            "description": "Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry."
+          },
+          "workdir": {
+            "type": "string",
+            "description": "Working directory for this command. Defaults to the session workspace; a relative path is resolved against it."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
+          },
+          "sandbox_permissions": {
+            "type": "string",
+            "description": "The wider sandbox mode this command needs. Only valid as a one-shot retry of a command the sandbox just denied; requires justification and user approval.",
+            "enum": [
+              "workspace-write",
+              "danger-full-access"
+            ]
+          },
+          "justification": {
+            "type": "string",
+            "description": "Required with sandbox_permissions: one sentence for the user explaining why this exact command needs the wider access."
+          }
+        },
+        "required": [
+          "command",
+          "description"
+        ]
+      }
+    },
+    {
+      "name": "edit",
+      "description": "Edit an existing UTF-8 text file by replacing literal text.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "file_path": {
+            "type": "string",
+            "description": "Path to edit, resolved by the filesystem backend."
+          },
+          "old_string": {
+            "type": "string",
+            "description": "Literal text to replace. Must match exactly."
+          },
+          "new_string": {
+            "type": "string",
+            "description": "Literal replacement text. Use an empty string to delete the match."
+          },
+          "replace_all": {
+            "type": "boolean",
+            "description": "Replace all matches. Defaults to false; when false, old_string must appear exactly once."
+          }
+        },
+        "required": [
+          "file_path",
+          "old_string",
+          "new_string"
+        ]
+      }
+    },
+    {
+      "name": "read",
+      "description": "Read a UTF-8 text file and return line-numbered content.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "file_path": {
+            "type": "string",
+            "description": "Path to read, resolved by the filesystem backend."
+          },
+          "offset": {
+            "type": "number",
+            "description": "1-based first line to return. Defaults to 1."
+          },
+          "limit": {
+            "type": "number",
+            "description": "Maximum number of lines to return. Defaults to 2000."
+          }
+        },
+        "required": [
+          "file_path"
+        ]
+      }
+    },
+    {
+      "name": "skill",
+      "description": "Load the full instructions for an available skill. Call this with the exact skill name from the session skill catalog before acting on a task that names or clearly matches that skill.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "name": {
+            "type": "string",
+            "description": "The exact skill name from the available skills list."
+          }
+        },
+        "required": [
+          "name"
+        ]
+      }
+    },
+    {
+      "name": "subagent",
+      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "description": {
+            "type": "string",
+            "description": "A short (3-5 word) description of the delegated task, for display."
+          },
+          "prompt": {
+            "type": "string",
+            "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
+          }
+        },
+        "required": [
+          "description",
+          "prompt"
+        ]
+      }
+    },
+    {
+      "name": "subagent_fork",
+      "description": "Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "description": {
+            "type": "string",
+            "description": "A short (3-5 word) description of the delegated task, for display."
+          },
+          "prompt": {
+            "type": "string",
+            "description": "The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
+          }
+        },
+        "required": [
+          "description",
+          "prompt"
+        ]
+      }
+    },
+    {
+      "name": "task_kill",
+      "description": "Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "reason": {
+            "type": "string",
+            "description": "Optional short reason, recorded in the log and forwarded to the task."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
+    {
+      "name": "task_list",
+      "description": "List your background tasks (running and finished) with their ids, kinds, and statuses.",
+      "parameters": {
+        "type": "object",
+        "properties": {}
+      }
+    },
+    {
+      "name": "task_output",
+      "description": "Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "wait": {
+            "type": "boolean",
+            "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+          },
+          "timeout_ms": {
+            "type": "number",
+            "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
+    {
+      "name": "todo_write",
+      "description": "Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "todos": {
+            "type": "array",
+            "description": "The COMPLETE task list, replacing any previous list.",
+            "items": {
+              "type": "object",
+              "properties": {
+                "content": {
+                  "type": "string",
+                  "description": "What the task is — a short imperative line."
+                },
+                "status": {
+                  "type": "string",
+                  "description": "pending (not started) | in_progress (now) | completed (done).",
+                  "enum": [
+                    "pending",
+                    "in_progress",
+                    "completed"
+                  ]
+                }
+              },
+              "required": [
+                "content",
+                "status"
+              ]
+            }
+          }
+        },
+        "required": [
+          "todos"
+        ]
+      }
+    },
+    {
+      "name": "workflow",
+      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "script": {
+            "type": "string",
+            "description": "The plain-JS workflow script body (top-level await allowed; NO `export const meta` statement; end with `return <json-value>`)."
+          },
+          "meta": {
+            "type": "object",
+            "description": "The workflow identity block (plain JSON — never code).",
+            "properties": {
+              "name": {
+                "type": "string",
+                "description": "Short kebab-case workflow name."
+              },
+              "description": {
+                "type": "string",
+                "description": "One-line description of what the workflow does."
+              },
+              "whenToUse": {
+                "type": "string",
+                "description": "Optional guidance on when this workflow applies."
+              },
+              "phases": {
+                "type": "array",
+                "description": "Optional phase declarations matched by phase() calls.",
+                "items": {
+                  "type": "object",
+                  "properties": {
+                    "title": {
+                      "type": "string",
+                      "description": "The phase title phase() calls match by exact string."
+                    },
+                    "detail": {
+                      "type": "string",
+                      "description": "Optional one-line description of the phase."
+                    },
+                    "provider": {
+                      "type": "string",
+                      "description": "Optional provider override this phase is expected to use."
+                    },
+                    "model": {
+                      "type": "string",
+                      "description": "Optional model override this phase is expected to use."
+                    }
+                  },
+                  "required": [
+                    "title"
+                  ]
+                }
+              }
+            },
+            "required": [
+              "name",
+              "description"
+            ]
+          },
+          "args": {
+            "type": "object",
+            "description": "Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]})."
+          }
+        },
+        "required": [
+          "script",
+          "meta"
+        ]
+      }
+    },
+    {
+      "name": "write",
+      "description": "Create or fully replace a UTF-8 text file.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "file_path": {
+            "type": "string",
+            "description": "Path to write, resolved by the filesystem backend."
+          },
+          "content": {
+            "type": "string",
+            "description": "Full UTF-8 text content to write."
+          }
+        },
+        "required": [
+          "file_path",
+          "content"
+        ]
+      }
+    }
+  ],
+  "changes": []
+}
diff --git a/examples/acp-agent/tests/snapshots/workspace-context/workspace/.dsh-project b/examples/acp-agent/tests/snapshots/workspace-context/workspace/.dsh-project
new file mode 100644
index 0000000000..8ce6fed8d7
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/workspace-context/workspace/.dsh-project
@@ -0,0 +1 @@
+snapshot root marker
diff --git a/examples/acp-agent/tests/snapshots/workspace-context/workspace/AGENTS.md b/examples/acp-agent/tests/snapshots/workspace-context/workspace/AGENTS.md
new file mode 100644
index 0000000000..a66cf16a13
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/workspace-context/workspace/AGENTS.md
@@ -0,0 +1 @@
+Root snapshot instruction.
diff --git a/examples/acp-agent/tests/snapshots/workspace-context/workspace/nested/AGENTS.md b/examples/acp-agent/tests/snapshots/workspace-context/workspace/nested/AGENTS.md
new file mode 100644
index 0000000000..862c12a235
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/workspace-context/workspace/nested/AGENTS.md
@@ -0,0 +1 @@
+Nested snapshot instruction.
diff --git a/examples/acp-agent/tests/snapshots/workspace-context/workspace/nested/task.txt b/examples/acp-agent/tests/snapshots/workspace-context/workspace/nested/task.txt
new file mode 100644
index 0000000000..39e2106a6f
--- /dev/null
+++ b/examples/acp-agent/tests/snapshots/workspace-context/workspace/nested/task.txt
@@ -0,0 +1 @@
+snapshot task
diff --git a/examples/acp-agent/tests/snapshots/workspace-edit/session.jsonl b/examples/acp-agent/tests/snapshots/workspace-edit/session.jsonl
index 9a908f24a3..6b9b03a95e 100644
--- a/examples/acp-agent/tests/snapshots/workspace-edit/session.jsonl
+++ b/examples/acp-agent/tests/snapshots/workspace-edit/session.jsonl
@@ -2,7 +2,7 @@
 {"type":"turn/start","seq":0,"time":1783352264080,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}
 {"type":"user/message","seq":1,"time":1783352264081,"data":{"content":[{"type":"text","text":"A file named greeting.txt in the current directory contains one word. Use the bash tool to append a second line containing the word WORLD to it (so it has two lines), then read the file back with `cat greeting.txt` to confirm, and reply with the single word DONE. Use a single bash call per action."}],"source":{"kind":"user"}},"surfaceOp":"append"}
 {"type":"step/start","seq":2,"time":1783352264082,"data":{"turn":1,"step":1}}
-{"type":"request/header","seq":3,"time":1783352264083,"data":{"header":{"config":{"model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
+{"type":"request/header","seq":3,"time":1783352264083,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
 {"type":"assistant/chunk","seq":4,"time":1783352264544,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}}
 {"type":"assistant/chunk","seq":5,"time":1783352264544,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":"The"}}}
 {"type":"assistant/chunk","seq":6,"time":1783352264642,"data":{"turn":1,"step":1,"chunk":{"type":"reasoning-delta","index":0,"text":" user"}}}
@@ -77,7 +77,7 @@
 {"type":"assistant/chunk","seq":75,"time":1783352265489,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_OjRFB4zvxu6UALDjytZD0978","name":"read","arguments":"{\"file_path\": \"greeting.txt\"}"}}}}
 {"type":"assistant/chunk","seq":76,"time":1783352265489,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":2918,"outputTokens":101,"cacheReadTokens":0,"reasoningTokens":55}}}}
 {"type":"assistant/chunk","seq":77,"time":1783352265489,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":78,"time":1783352265491,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to:\n1. Read the file greeting.txt\n2. Append the word WORLD as a second line\n3. Read the file back with cat to confirm\n4. Reply with DONE\n\nLet me start by reading the file to see its contents."},{"type":"tool-call","id":"call_00_OjRFB4zvxu6UALDjytZD0978","name":"read","arguments":"{\"file_path\": \"greeting.txt\"}"}],"usage":{"inputTokens":2918,"outputTokens":101,"cacheReadTokens":0,"reasoningTokens":55}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77],"surfaceOp":"append"}
+{"type":"assistant/message","seq":78,"time":1783352265491,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to:\n1. Read the file greeting.txt\n2. Append the word WORLD as a second line\n3. Read the file back with cat to confirm\n4. Reply with DONE\n\nLet me start by reading the file to see its contents."},{"type":"tool-call","id":"call_00_OjRFB4zvxu6UALDjytZD0978","name":"read","arguments":"{\"file_path\": \"greeting.txt\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2918,"outputTokens":101,"cacheReadTokens":0,"reasoningTokens":55}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77],"surfaceOp":"append"}
 {"type":"tool/call","seq":79,"time":1783352265491,"data":{"turn":1,"step":1,"callId":"call_00_OjRFB4zvxu6UALDjytZD0978","name":"read","arguments":"{\"file_path\": \"greeting.txt\"}"}}
 {"type":"tool/result","seq":80,"time":1783352265504,"data":{"turn":1,"step":1,"callId":"call_00_OjRFB4zvxu6UALDjytZD0978","content":[{"type":"text","text":"<path>/tmp/acp-snap-cwd-rxbEpP/greeting.txt</path>\n<type>file</type>\n<content>\n1: hello\n\n(End of file - total 1 lines)\n</content>"}],"isError":false},"sourceEventSeqs":[79],"surfaceOp":"append"}
 {"type":"step/end","seq":81,"time":1783352265504,"data":{"turn":1,"step":1}}
@@ -154,7 +154,7 @@
 {"type":"assistant/chunk","seq":152,"time":1783352267302,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_IUUvbNiPcnwhVL8ErEFS4806","name":"bash","arguments":"{\"command\": \"printf '\\\\nWORLD' >> greeting.txt\", \"description\": \"Append newline and WORLD to greeting.txt\"}"}}}}
 {"type":"assistant/chunk","seq":153,"time":1783352267302,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":261,"outputTokens":107,"cacheReadTokens":2816,"reasoningTokens":32}}}}
 {"type":"assistant/chunk","seq":154,"time":1783352267302,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":155,"time":1783352267302,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The file contains \"hello\" on one line. Now I need to append a second line with \"WORLD\" to it. Then cat it to confirm."},{"type":"tool-call","id":"call_00_IUUvbNiPcnwhVL8ErEFS4806","name":"bash","arguments":"{\"command\": \"printf '\\\\nWORLD' >> greeting.txt\", \"description\": \"Append newline and WORLD to greeting.txt\"}"}],"usage":{"inputTokens":261,"outputTokens":107,"cacheReadTokens":2816,"reasoningTokens":32}},"sourceEventSeqs":[83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154],"surfaceOp":"append"}
+{"type":"assistant/message","seq":155,"time":1783352267302,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The file contains \"hello\" on one line. Now I need to append a second line with \"WORLD\" to it. Then cat it to confirm."},{"type":"tool-call","id":"call_00_IUUvbNiPcnwhVL8ErEFS4806","name":"bash","arguments":"{\"command\": \"printf '\\\\nWORLD' >> greeting.txt\", \"description\": \"Append newline and WORLD to greeting.txt\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":261,"outputTokens":107,"cacheReadTokens":2816,"reasoningTokens":32}},"sourceEventSeqs":[83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154],"surfaceOp":"append"}
 {"type":"tool/call","seq":156,"time":1783352267302,"data":{"turn":1,"step":2,"callId":"call_00_IUUvbNiPcnwhVL8ErEFS4806","name":"bash","arguments":"{\"command\": \"printf '\\\\nWORLD' >> greeting.txt\", \"description\": \"Append newline and WORLD to greeting.txt\"}"}}
 {"type":"tool/result","seq":157,"time":1783352267330,"data":{"turn":1,"step":2,"callId":"call_00_IUUvbNiPcnwhVL8ErEFS4806","content":[{"type":"text","text":"(no output)"}],"isError":false},"sourceEventSeqs":[156],"surfaceOp":"append"}
 {"type":"step/end","seq":158,"time":1783352267330,"data":{"turn":1,"step":2}}
@@ -201,7 +201,7 @@
 {"type":"assistant/chunk","seq":199,"time":1783352268414,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":1,"block":{"type":"tool-call","id":"call_00_Wo4H7tFNheZJWKVDrAHK5851","name":"bash","arguments":"{\"command\": \"cat greeting.txt\", \"description\": \"Read greeting.txt to confirm\"}"}}}}
 {"type":"assistant/chunk","seq":200,"time":1783352268414,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":126,"outputTokens":77,"cacheReadTokens":3072,"reasoningTokens":12}}}}
 {"type":"assistant/chunk","seq":201,"time":1783352268414,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}
-{"type":"assistant/message","seq":202,"time":1783352268415,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"Good, now let me read the file back with cat."},{"type":"tool-call","id":"call_00_Wo4H7tFNheZJWKVDrAHK5851","name":"bash","arguments":"{\"command\": \"cat greeting.txt\", \"description\": \"Read greeting.txt to confirm\"}"}],"usage":{"inputTokens":126,"outputTokens":77,"cacheReadTokens":3072,"reasoningTokens":12}},"sourceEventSeqs":[160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201],"surfaceOp":"append"}
+{"type":"assistant/message","seq":202,"time":1783352268415,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"Good, now let me read the file back with cat."},{"type":"tool-call","id":"call_00_Wo4H7tFNheZJWKVDrAHK5851","name":"bash","arguments":"{\"command\": \"cat greeting.txt\", \"description\": \"Read greeting.txt to confirm\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":126,"outputTokens":77,"cacheReadTokens":3072,"reasoningTokens":12}},"sourceEventSeqs":[160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201],"surfaceOp":"append"}
 {"type":"tool/call","seq":203,"time":1783352268415,"data":{"turn":1,"step":3,"callId":"call_00_Wo4H7tFNheZJWKVDrAHK5851","name":"bash","arguments":"{\"command\": \"cat greeting.txt\", \"description\": \"Read greeting.txt to confirm\"}"}}
 {"type":"tool/result","seq":204,"time":1783352268429,"data":{"turn":1,"step":3,"callId":"call_00_Wo4H7tFNheZJWKVDrAHK5851","content":[{"type":"text","text":"hello\n\nWORLD"}],"isError":false},"sourceEventSeqs":[203],"surfaceOp":"append"}
 {"type":"step/end","seq":205,"time":1783352268429,"data":{"turn":1,"step":3}}
@@ -236,6 +236,6 @@
 {"type":"assistant/chunk","seq":234,"time":1783352269538,"data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":1,"block":{"type":"text","text":"DONE"}}}}
 {"type":"assistant/chunk","seq":235,"time":1783352269538,"data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":219,"outputTokens":25,"cacheReadTokens":3072,"reasoningTokens":22}}}}
 {"type":"assistant/chunk","seq":236,"time":1783352269538,"data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}
-{"type":"assistant/message","seq":237,"time":1783352269538,"data":{"turn":1,"step":4,"content":[{"type":"reasoning","text":"The file now has two lines:\n1. hello\n2. WORLD\n\nI can reply with DONE."},{"type":"text","text":"DONE"}],"usage":{"inputTokens":219,"outputTokens":25,"cacheReadTokens":3072,"reasoningTokens":22}},"sourceEventSeqs":[207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236],"surfaceOp":"append"}
+{"type":"assistant/message","seq":237,"time":1783352269538,"data":{"turn":1,"step":4,"content":[{"type":"reasoning","text":"The file now has two lines:\n1. hello\n2. WORLD\n\nI can reply with DONE."},{"type":"text","text":"DONE"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":219,"outputTokens":25,"cacheReadTokens":3072,"reasoningTokens":22}},"sourceEventSeqs":[207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236],"surfaceOp":"append"}
 {"type":"step/end","seq":238,"time":1783352269538,"data":{"turn":1,"step":4}}
 {"type":"turn/end","seq":239,"time":1783352269539,"data":{"turn":1,"reason":{"kind":"completed"}}}
diff --git a/examples/acp-agent/tests/snapshots/workspace-edit/stdout.golden.jsonl b/examples/acp-agent/tests/snapshots/workspace-edit/stdout.golden.jsonl
index 5f6c7f4c52..1c76f1ac61 100644
--- a/examples/acp-agent/tests/snapshots/workspace-edit/stdout.golden.jsonl
+++ b/examples/acp-agent/tests/snapshots/workspace-edit/stdout.golden.jsonl
@@ -1,5 +1,5 @@
 {"jsonrpc":"2.0","id":1,"result":{"protocolVersion":1,"agentInfo":{"name":"deepseek-harness-acp","version":"0.0.1"},"agentCapabilities":{"loadSession":true,"promptCapabilities":{"image":false,"audio":false,"embeddedContext":false}},"authMethods":[]}}
-{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
+{"jsonrpc":"2.0","id":2,"result":{"sessionId":"{{sessionId}}","configOptions":[{"id":"model","name":"Model","description":"Sets this session's provider and model.","category":"model","type":"select","currentValue":"[\"deepseek\",\"deepseek-v4-flash\"]","options":[{"value":"[\"deepseek\",\"deepseek-v4-flash\"]","name":"deepseek-v4-flash"},{"value":"[\"deepseek\",\"deepseek-v4-pro\"]","name":"deepseek-v4-pro"}]},{"id":"permission","name":"Permissions","description":"Sets this session's sandbox and approval behavior.","category":"mode","type":"select","currentValue":"danger-full-access","options":[{"value":"workspace-write","name":"workspace-write","description":"Write inside the workspace and permitted temporary directories; wider retries require approval."},{"value":"danger-full-access","name":"danger-full-access","description":"Full file access without approval prompts."}]}]}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":"The"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" user"}}}}
 {"jsonrpc":"2.0","method":"session/update","params":{"sessionId":"{{sessionId}}","update":{"sessionUpdate":"agent_thought_chunk","content":{"type":"text","text":" wants"}}}}
diff --git a/examples/acp-agent/tests/snapshots/workspace-edit/system-prompt.golden.md b/examples/acp-agent/tests/snapshots/workspace-edit/system-prompt.golden.md
index 6bb634b339..ddf502a773 100644
--- a/examples/acp-agent/tests/snapshots/workspace-edit/system-prompt.golden.md
+++ b/examples/acp-agent/tests/snapshots/workspace-edit/system-prompt.golden.md
@@ -13,6 +13,8 @@ Use the edit tool for targeted changes to existing UTF-8 text files. It replaces
 
 Check the [exit code: N] marker on every bash result; investigate failures before moving on.
 
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+
 Approval prompts are disabled in this session: actions that require approval are rejected automatically — do not request sandbox escalation (do not set `sandbox_permissions`).
 <!-- dsh-user-approval-policy:never -->
 
diff --git a/examples/acp-agent/tests/snapshots/workspace-edit/tool-schemas.golden.json b/examples/acp-agent/tests/snapshots/workspace-edit/tool-schemas.golden.json
index 18e285888b..4b08a1e365 100644
--- a/examples/acp-agent/tests/snapshots/workspace-edit/tool-schemas.golden.json
+++ b/examples/acp-agent/tests/snapshots/workspace-edit/tool-schemas.golden.json
@@ -2,7 +2,7 @@
   "initial": [
     {
       "name": "bash",
-      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; poll it with `bash_output` and stop it with `bash_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
+      "description": "Execute a bash command (`bash -c`) and return its stdout/stderr. Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. Current harness environment facts are exposed through managed `$DSH_*` variables; inspect them when needed. Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`. Attempting a command the sandbox may deny is safe and expected: run it and read the marker rather than assuming the denial. When a command is denied and a wider mode would let it succeed, escalate immediately in the same turn — the one sanctioned exception to a denial: retry the exact same command once with `sandbox_permissions` (the narrowest wider mode that suffices) plus a one-sentence `justification`. Do not detour through chat to ask permission first — the approval prompt raised by that retry is how the user consents. If the session states approval prompts are disabled, there is no exception: a denial is final — do not set `sandbox_permissions`. Never escalate speculatively: ground the request in a real denial — normally the one this command just hit; escalating up front is fine only when this session already denied the same access. A rejected escalation is final for that command — stop and explain, never work around it — but it does not forbid attempting or escalating other commands later.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -24,7 +24,7 @@
           },
           "run_in_background": {
             "type": "boolean",
-            "description": "Run in the background and return a task id immediately. No timeout applies."
+            "description": "Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies."
           },
           "sandbox_permissions": {
             "type": "string",
@@ -45,38 +45,6 @@
         ]
       }
     },
-    {
-      "name": "bash_kill",
-      "description": "Ask the executor to kill a running background bash task by task id.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
-    {
-      "name": "bash_output",
-      "description": "Read new output from a background bash task started with `bash` + `run_in_background`. Returns only output produced since the previous bash_output call, plus the task status. Tasks keep running while you do other work; poll again later for more output.",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "task_id": {
-            "type": "string",
-            "description": "Task id returned by the bash tool."
-          }
-        },
-        "required": [
-          "task_id"
-        ]
-      }
-    },
     {
       "name": "edit",
       "description": "Edit an existing UTF-8 text file by replacing literal text.",
@@ -149,7 +117,7 @@
     },
     {
       "name": "subagent",
-      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation.",
+      "description": "Delegate a self-contained task to a subagent (a separate agent that works in its own context) and return its final result. Use this to offload focused, independent work — research, a scoped implementation, an analysis — so it does not consume this conversation's context. The subagent runs to completion and you receive only its final answer, not its intermediate steps. Give it a complete, standalone prompt: it does not see this conversation. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -160,6 +128,10 @@
           "prompt": {
             "type": "string",
             "description": "The complete, self-contained task for the subagent. It does not share this conversation's context, so include everything it needs."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -170,7 +142,7 @@
     },
     {
       "name": "subagent_fork",
-      "description": "Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps.",
+      "description": "Delegate a task to a subagent that inherits this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps. Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -181,6 +153,10 @@
           "prompt": {
             "type": "string",
             "description": "The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new."
+          },
+          "run_in_background": {
+            "type": "boolean",
+            "description": "Run as a background task and return its id; collect with task_output or stop with task_kill."
           }
         },
         "required": [
@@ -189,6 +165,58 @@
         ]
       }
     },
+    {
+      "name": "task_kill",
+      "description": "Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "reason": {
+            "type": "string",
+            "description": "Optional short reason, recorded in the log and forwarded to the task."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
+    {
+      "name": "task_list",
+      "description": "List your background tasks (running and finished) with their ids, kinds, and statuses.",
+      "parameters": {
+        "type": "object",
+        "properties": {}
+      }
+    },
+    {
+      "name": "task_output",
+      "description": "Read a background task. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.",
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "task_id": {
+            "type": "string",
+            "description": "Task id returned by the tool that started the background work."
+          },
+          "wait": {
+            "type": "boolean",
+            "description": "Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive."
+          },
+          "timeout_ms": {
+            "type": "number",
+            "description": "Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum."
+          }
+        },
+        "required": [
+          "task_id"
+        ]
+      }
+    },
     {
       "name": "todo_write",
       "description": "Record and update a structured task list for the current work. Send the ENTIRE list every call — it REPLACES the previous list (there are no partial updates, no per-item edits). Use it to plan multi-step work and show progress: add one todo per concrete step before you start. Keep AT MOST ONE todo `in_progress` at a time; while work remains, exactly one active task should be `in_progress`. Mark a todo `completed` the moment it is done (do not batch completions), and allow no `in_progress` item only once all work is complete. Skip the list for trivial single-step tasks. Statuses: `pending` (not started), `in_progress` (being worked on now), `completed` (finished).",
@@ -229,7 +257,7 @@
     },
     {
       "name": "workflow",
-      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), `model` (override). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
+      "description": "Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.\n\nThe workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return <value>` — the value must be JSON-serializable and is this tool's result.\n\nScript-body hooks:\n- `agent(prompt, opts?): Promise<any>` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly.\n- `pipeline(items, ...stages): Promise<any[]>` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages.\n- `parallel(thunks): Promise<any[]>` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`.\n- `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim.\n\nMisused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`.\n\nConstraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes.",
       "parameters": {
         "type": "object",
         "properties": {
@@ -267,6 +295,10 @@
                       "type": "string",
                       "description": "Optional one-line description of the phase."
                     },
+                    "provider": {
+                      "type": "string",
+                      "description": "Optional provider override this phase is expected to use."
+                    },
                     "model": {
                       "type": "string",
                       "description": "Optional model override this phase is expected to use."
@@ -316,5 +348,5 @@
       }
     }
   ],
-  "deltas": []
+  "changes": []
 }
diff --git a/examples/acp-agent/workspace-context.cordis.snapshot.yml b/examples/acp-agent/workspace-context.cordis.snapshot.yml
new file mode 100644
index 0000000000..e0b02cbb21
--- /dev/null
+++ b/examples/acp-agent/workspace-context.cordis.snapshot.yml
@@ -0,0 +1,37 @@
+# Keyless replay counterpart of workspace-context.cordis.yml. Patches do not
+# compose across includes, so this applies the scenario config and model swap
+# directly to the live tree.
+- id: base
+  name: '@cordisjs/plugin-include'
+  config:
+    path: ./cordis.yml
+    patches:
+      - id: llm-deepseek
+        name: '@deepseek-ai/dsh-llm-deepseek'
+        disabled: true
+      - id: acp-agent
+        name: '@deepseek-ai/dsh-acp-demo'
+        config:
+          provider: deepseek
+          model: deepseek-v4-flash
+          persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+          workspaceContext:
+            maxBytes: 65536
+            dshHome: !!js process.cwd() + '/.dsh'
+            projectRootMarkers:
+              - .dsh-project
+          persona: |
+            You are a coding assistant powered by the {{model}} model. Your working directory is {{cwd}}.
+
+            Verify your work by running the code or tests. Keep answers brief and factual.
+      - insert:
+          - id: fs-local
+            name: '@deepseek-ai/dsh-fs-local'
+            config:
+              cwd: !!js process.cwd()
+          - id: fs-policy
+            name: '@deepseek-ai/dsh-fs-policy'
+          - id: tool-fs
+            name: '@deepseek-ai/dsh-tool-fs'
+          - id: llm-replay
+            name: '@deepseek-ai/dsh-llm-replay'
diff --git a/examples/acp-agent/workspace-context.cordis.yml b/examples/acp-agent/workspace-context.cordis.yml
new file mode 100644
index 0000000000..1b8c0279f0
--- /dev/null
+++ b/examples/acp-agent/workspace-context.cordis.yml
@@ -0,0 +1,32 @@
+# Workspace-context snapshot overlay: keep project-root and user-global
+# discovery inside the scenario's temporary cwd. The app config patch replaces
+# the whole base config, so the base fields are restated verbatim.
+- id: base
+  name: '@cordisjs/plugin-include'
+  config:
+    path: ./cordis.yml
+    patches:
+      - id: acp-agent
+        name: '@deepseek-ai/dsh-acp-demo'
+        config:
+          provider: deepseek
+          model: deepseek-v4-flash
+          persistenceRoot: !!js process.env.DSH_SNAPSHOT_SESSIONS_ROOT ?? './.sessions'
+          workspaceContext:
+            maxBytes: 65536
+            dshHome: !!js process.cwd() + '/.dsh'
+            projectRootMarkers:
+              - .dsh-project
+          persona: |
+            You are a coding assistant powered by the {{model}} model. Your working directory is {{cwd}}.
+
+            Verify your work by running the code or tests. Keep answers brief and factual.
+      - insert:
+          - id: fs-local
+            name: '@deepseek-ai/dsh-fs-local'
+            config:
+              cwd: !!js process.cwd()
+          - id: fs-policy
+            name: '@deepseek-ai/dsh-fs-policy'
+          - id: tool-fs
+            name: '@deepseek-ai/dsh-tool-fs'
diff --git a/examples/coding-agent/README.md b/examples/coding-agent/README.md
index 380ab8134f..9e627324d4 100644
--- a/examples/coding-agent/README.md
+++ b/examples/coding-agent/README.md
@@ -11,7 +11,7 @@ The REPL agent demo wiring: DeepSeek V4 + the `read`/`write`/`edit` filesystem t
 pnpm run demo:repl
 ```
 
-Type a coding task. The agent works through the `read`/`write`/`edit` filesystem tools for ordinary file operations and `bash` (+ `bash_output` / `bash_kill` for background tasks) for shell commands, searches, and test runs, each in a fresh `bash -c` (the system prompt tells the model to pass `workdir` instead of `cd`). Both the fs tools and bash resolve relative paths against the session workspace. It can also delegate with `subagent`/`subagent_fork` and track multi-step work with `todo_write` (a whole-list task tracker rendered as a checklist). Reasoning streams dimmed; tool calls/results render inline.
+Type a coding task. The agent works through the `read`/`write`/`edit` filesystem tools for ordinary file operations and `bash` (+ the generic `task_output` / `task_list` / `task_kill` for background tasks) for shell commands, searches, and test runs, each in a fresh `bash -c` (the system prompt tells the model to pass `workdir` instead of `cd`). Both the fs tools and bash resolve relative paths against the session workspace. It can also delegate with `subagent`/`subagent_fork` and track multi-step work with `todo_write` (a whole-list task tracker rendered as a checklist). Reasoning streams dimmed; tool calls/results render inline.
 
 ```
 > fix the failing test in /path/to/project
@@ -54,8 +54,8 @@ This example is a thin leaf `cordis.yml`: it picks the swappable backends, loads
 |---|---|
 | `hmr` (`@cordisjs/plugin-hmr`) | the dev/demo edit-reload loop — a **leaf** entry (not baked into the app) because it is Loader-only and needs `node --expose-internals`, which `demo:repl` passes |
 | `llm-deepseek` | real `LlmAdapter` via config (`!!js process.env.…` secrets); swap one line to `@deepseek-ai/dsh-llm-pi-ai` for the library-backed twin |
-| `bash` (`dsh-bash-local`) | the executor implementation — the swappable half of the bash seam. The model-facing `bash`/`bash_output`/`bash_kill` tool schemas (`tool-bash`) come from `agent-core`, so only the executor is a leaf choice |
-| `stdio-agent` (`@deepseek-ai/dsh-stdio-demo`) | the app bundle: the agent-core spine + console logger + JSONL persistence + readline UI + a pre-created `main` agent. Its config carries the model, system prompt, `persistenceRoot` (`./.sessions`), and `resumeSessionId` — so persistence and the agent are configured here, not wired as separate leaf plugins |
+| `bash` (`dsh-bash-local`) | the executor implementation — the swappable half of the bash seam. The model-facing `bash` schema (`tool-bash`) and generic `task_*` controls (`tool-tasks`) come from `dsh-agent-spine-demo`, so only the executor is a leaf choice |
+| `stdio-agent` (`@deepseek-ai/dsh-stdio-demo`) | the app bundle: the agent-spine demo + console logger + JSONL persistence + readline UI + a pre-created `main` agent. Its config carries the model, system prompt, `persistenceRoot` (`./.sessions`), and `resumeSessionId` — so persistence and the agent are configured here, not wired as separate leaf plugins |
 | `subagent`, `subagent-spawn`, `subagent-fork` | the subagent provider registry plus the two in-process backends: a fresh child and a child seeded with the parent's completed-turn prefix |
 | `tool-subagent`, `tool-subagent-fork` | two model-facing `dsh-tool-subagent` loads, each bound to a different provider and exposed under a distinct tool name (`subagent`, `subagent_fork`) |
 | `tool-todo` | the model-facing `todo_write` tool; writes the whole task list to the session log and renders as a checklist in stdio |
diff --git a/examples/coding-agent/code-mode.cordis.yml b/examples/coding-agent/code-mode.cordis.yml
index 5d7198124c..3fec40e0f3 100644
--- a/examples/coding-agent/code-mode.cordis.yml
+++ b/examples/coding-agent/code-mode.cordis.yml
@@ -11,9 +11,12 @@
       - id: stdio-agent
         name: '@deepseek-ai/dsh-stdio-demo'
         config:
+          provider: deepseek
           model: deepseek-v4-flash
           resumeSessionId: !!js process.env.RESUME_SESSION_ID
           persistenceRoot: './.sessions'
+          workspaceContext:
+            maxBytes: 65536
           tools:
             mode: code
           welcome: 'code-mode agent ready. Give it a multi-tool task.'
diff --git a/examples/coding-agent/composition.md b/examples/coding-agent/composition.md
index 90ff8222b9..cc71975961 100644
--- a/examples/coding-agent/composition.md
+++ b/examples/coding-agent/composition.md
@@ -23,6 +23,8 @@ flowchart LR
   bundle_agent_core --> spine_sessions["ctx.sessions"]
   bundle_agent_core --> spine_tools["ctx.tools + tool-bash"]
   bundle_agent_core --> spine_loop["ctx.agents + ctx.agentLoop"]
+  plugin_coding_token_meter["token-meter<br/>@deepseek-ai/dsh-token-meter"]
+  cfg --> plugin_coding_token_meter
   plugin_coding_compact_basic["compact-basic<br/>@deepseek-ai/dsh-compact-basic"]
   cfg --> plugin_coding_compact_basic
   plugin_coding_subagent["subagent<br/>@deepseek-ai/dsh-subagent"]
@@ -47,6 +49,14 @@ flowchart LR
   cfg --> plugin_coding_fs_policy
   plugin_coding_tool_fs["tool-fs<br/>@deepseek-ai/dsh-tool-fs"]
   cfg --> plugin_coding_tool_fs
+  plugin_coding_tool_fs_search["tool-fs-search<br/>@deepseek-ai/dsh-tool-fs-search"]
+  cfg --> plugin_coding_tool_fs_search
+  plugin_coding_timeout_policy["timeout-policy<br/>@deepseek-ai/dsh-timeout-policy"]
+  cfg --> plugin_coding_timeout_policy
+  plugin_coding_spill_local["spill-local<br/>@deepseek-ai/dsh-spill-local"]
+  cfg --> plugin_coding_spill_local
+  plugin_coding_spill_policy["spill-policy<br/>@deepseek-ai/dsh-spill-policy"]
+  cfg --> plugin_coding_spill_policy
 ```
 
 | Plugin id | Package / module |
@@ -55,6 +65,7 @@ flowchart LR
 | `llm-deepseek` | `@deepseek-ai/dsh-llm-deepseek` |
 | `bash` | `@deepseek-ai/dsh-bash-local` |
 | `stdio-agent` | `@deepseek-ai/dsh-stdio-demo` |
+| `token-meter` | `@deepseek-ai/dsh-token-meter` |
 | `compact-basic` | `@deepseek-ai/dsh-compact-basic` |
 | `subagent` | `@deepseek-ai/dsh-subagent` |
 | `subagent-spawn` | `@deepseek-ai/dsh-subagent-spawn` |
@@ -67,6 +78,10 @@ flowchart LR
 | `fs-local` | `@deepseek-ai/dsh-fs-local` |
 | `fs-policy` | `@deepseek-ai/dsh-fs-policy` |
 | `tool-fs` | `@deepseek-ai/dsh-tool-fs` |
+| `tool-fs-search` | `@deepseek-ai/dsh-tool-fs-search` |
+| `timeout-policy` | `@deepseek-ai/dsh-timeout-policy` |
+| `spill-local` | `@deepseek-ai/dsh-spill-local` |
+| `spill-policy` | `@deepseek-ai/dsh-spill-policy` |
 
 Source config: [`examples/coding-agent/cordis.yml`](cordis.yml).
 
diff --git a/examples/coding-agent/cordis.yml b/examples/coding-agent/cordis.yml
index 600ffda160..bc3b542592 100644
--- a/examples/coding-agent/cordis.yml
+++ b/examples/coding-agent/cordis.yml
@@ -1,5 +1,6 @@
 # REPL agent with swappable DeepSeek and local-bash backends. `dsh-stdio-demo`
-# supplies the agent-core spine, logging, JSONL persistence, readline UI, and `main` agent.
+# supplies the agent spine, workspace instructions, generic task controls,
+# logging, JSONL persistence, readline UI, and `main` agent.
 # HMR remains a leaf because it requires Loader internals; `demo:repl` passes
 # `--expose-internals`. The app bin loads the gitignored root `.env`; this file
 # reads `DEEPSEEK_API_KEY` and optional `DEEPSEEK_BASE_URL` through `!!js`.
@@ -10,16 +11,12 @@
   config:
     root: ['.']
 
-# The DeepSeek adapter. Swap to '@deepseek-ai/dsh-llm-pi-ai' for the pi-ai-backed
-# twin (same config shape; `reasoning: high` replaces thinking/reasoningEffort).
+# The native DeepSeek adapter.
 - id: llm-deepseek
   name: '@deepseek-ai/dsh-llm-deepseek'
   config:
     apiKey: !!js process.env.DEEPSEEK_API_KEY
     baseURL: !!js process.env.DEEPSEEK_BASE_URL
-    models:
-      - deepseek-v4-pro
-      - deepseek-v4-flash
 
 # Local executor for the app bundle's bash tool.
 - id: bash
@@ -31,11 +28,14 @@
 - id: stdio-agent
   name: '@deepseek-ai/dsh-stdio-demo'
   config:
+    provider: deepseek
     model: deepseek-v4-flash
     # Set RESUME_SESSION_ID to continue a prior persisted session (the ids live
     # under ./.sessions); unset starts a fresh session each run.
     resumeSessionId: !!js process.env.RESUME_SESSION_ID
     persistenceRoot: './.sessions'
+    workspaceContext:
+      maxBytes: 65536
     welcome: 'agent REPL ready. Give it a coding task.'
     # Keep the persona to identity and behavior; tool plugins own tool guidance.
     # The loop resolves {{model}} from this agent's configuration.
@@ -45,17 +45,14 @@
       Verify your work by running the code or tests. Keep answers brief and
       factual.
 
-# Summarize an older range when derived history approaches the context window.
-# This leaf consumes `ctx.llm` and the app's `agent/pre-step` seam.
+# Replay-aware request pressure with one service-wide context window.
+- id: token-meter
+  name: '@deepseek-ai/dsh-token-meter'
+
+# Summarize an older range when measured history approaches the context window.
+# Service-wide policy provides the ordinary threshold and retained-tail defaults.
 - id: compact-basic
   name: '@deepseek-ai/dsh-compact-basic'
-  config:
-    contextWindow: 128000
-    thresholdRatio: 0.8
-    retainTokens: 20480
-    summarizationModel: ''
-    maxTokens: 8192
-    compactionRetries: 1
 
 # Expose fresh-child `spawn` and completed-prefix `fork` through independent
 # in-process backends. Each tool instance needs a distinct `toolName`; the registry
@@ -111,3 +108,29 @@
 
 - id: tool-fs
   name: '@deepseek-ai/dsh-tool-fs'
+
+# Bash-backed discovery tools (glob/grep): fixed ripgrep commands through the
+# local bash executor above — not ctx.fs. Capped results save the complete
+# formatted list through the spill backend below (ctx.spillStore, optional).
+- id: tool-fs-search
+  name: '@deepseek-ai/dsh-tool-fs-search'
+
+# The tool-call timeout enforcer: arms each declared ToolDefinition.timeoutMs
+# (the search tools above declare 30s) as a deadline on exec.signal. Without
+# it a declared budget is advisory and only the bash executor's own timeout
+# backstop applies.
+- id: timeout-policy
+  name: '@deepseek-ai/dsh-timeout-policy'
+
+# Tool-output spill stack: a local backend that saves oversized tool text under
+# a private session-scoped dir, and the tools/post-execute policy that replaces
+# an over-budget plain-text result with a preview + the spill locator/retrieval
+# hint. A leaf pair after the app (needs ctx.tools). The policy is a no-op until
+# a tool returns more than maxInlineBytes of plain text.
+- id: spill-local
+  name: '@deepseek-ai/dsh-spill-local'
+
+- id: spill-policy
+  name: '@deepseek-ai/dsh-spill-policy'
+  config:
+    maxInlineBytes: 50000
diff --git a/examples/coding-agent/tests/code-mode.e2e.ts b/examples/coding-agent/tests/code-mode.e2e.ts
index 3f6e109e3f..b6b55048e4 100644
--- a/examples/coding-agent/tests/code-mode.e2e.ts
+++ b/examples/coding-agent/tests/code-mode.e2e.ts
@@ -1,10 +1,10 @@
-import { mkdtemp, readFile, rm } from 'node:fs/promises'
+import { mkdtemp, mkdir, readFile, rm, writeFile } from 'node:fs/promises'
 import { tmpdir } from 'node:os'
 import { join } from 'node:path'
 import { afterEach, describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
 import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
+import SessionStore, { SessionId } from '@deepseek-ai/dsh-session'
 import type { SessionEvent } from '@deepseek-ai/dsh-session'
 import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
 import ToolRegistry, { RUN_CODE_NAME } from '@deepseek-ai/dsh-tools'
@@ -14,6 +14,9 @@ import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
 import * as ToolBash from '@deepseek-ai/dsh-tool-bash'
 import * as LlmDeepSeek from '@deepseek-ai/dsh-llm-deepseek'
 import { WorkerCodeRuntime } from '@deepseek-ai/dsh-code-runtime-worker'
+import LocalFileSystem from '@deepseek-ai/dsh-fs-local'
+import * as ToolFs from '@deepseek-ai/dsh-tool-fs'
+import * as WorkspaceContext from '@deepseek-ai/dsh-workspace-context'
 
 /**
  * With-key Code Mode proof: a real model receives only `run_code`, composes two
@@ -23,6 +26,7 @@ import { WorkerCodeRuntime } from '@deepseek-ai/dsh-code-runtime-worker'
 
 const PERSONA = 'You are coding-agent. You work by writing TypeScript programs for run_code: '
   + 'batch related tool work into one program and print or return ONLY the findings that matter.'
+const WORKSPACE_PROBE = 'dragonfruit-8675309'
 
 let ctx: Context | undefined
 let workdir: string | undefined
@@ -45,13 +49,29 @@ async function codeModeHarness(cwd: string): Promise<Context> {
   await harness.plugin(ToolRegistry, { mode: 'code' })
   await harness.plugin(AgentRegistry)
   await harness.plugin(AgentLoop, { agents: [] })
-  await harness.plugin(LlmDeepSeek, { models: ['deepseek-v4-flash'] })
+  await harness.plugin(LlmDeepSeek)
   await harness.plugin(LocalBashExecutor, { cwd, timeoutMs: 30_000 })
   await harness.plugin(ToolBash)
   await harness.plugin(WorkerCodeRuntime, {})
   return harness
 }
 
+async function workspaceCodeModeHarness(): Promise<Context> {
+  const harness = new Context()
+  await harness.plugin(LlmService)
+  await harness.plugin(SessionStore)
+  await harness.plugin(SystemPrompt, { persona: PERSONA })
+  await harness.plugin(ToolRegistry, { mode: 'code' })
+  await harness.plugin(AgentRegistry)
+  await harness.plugin(LocalFileSystem, { cwd: '/' })
+  await harness.plugin(ToolFs)
+  await harness.plugin(WorkspaceContext, { maxBytes: 65536 })
+  await harness.plugin(AgentLoop, { agents: [] })
+  await harness.plugin(LlmDeepSeek, { models: [{ id: 'deepseek-v4-flash' }] })
+  await harness.plugin(WorkerCodeRuntime, {})
+  return harness
+}
+
 function waitForIdle(harness: Context, agent: ReactLoopAgent): Promise<void> {
   return new Promise((resolve) => {
     const dispose = harness.on('agent/status', (subject, status) => {
@@ -67,7 +87,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('Code Mode: real model writes a p
   it('collapses the wire tool list to [run_code], bridges sub-calls, and returns curated output', async () => {
     workdir = await mkdtemp(join(tmpdir(), 'dsh-code-mode-e2e-'))
     ctx = await codeModeHarness(workdir)
-    const agent = ctx.agentLoop.create(AgentId('e2e-code-mode'), { model: 'deepseek-v4-flash' })
+    const agent = ctx.agentLoop.create(AgentId('e2e-code-mode'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     agent.send([{
       type: 'text',
@@ -107,4 +127,43 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('Code Mode: real model writes a p
     expect(finalText).toContain('alpha-7')
     expect(finalText).toContain('beta-9')
   }, 180_000)
+
+  it('delivers nested workspace instructions discovered by an fs sub-call after the outer result', async () => {
+    workdir = await mkdtemp(join(tmpdir(), 'dsh-code-mode-workspace-e2e-'))
+    await mkdir(join(workdir, '.git'), { recursive: true })
+    await mkdir(join(workdir, 'pkg/deep'), { recursive: true })
+    await writeFile(join(workdir, 'pkg/AGENTS.md'), `If asked for the Code Mode workspace handshake, reply with exactly ${WORKSPACE_PROBE} and nothing else.\n`)
+    await writeFile(join(workdir, 'pkg/deep/task.txt'), 'Touch this file to discover the nested instructions.\n')
+    ctx = await workspaceCodeModeHarness()
+    const handle = await ctx.agents.create({
+      agentId: AgentId('e2e-code-mode-workspace'),
+      sessionId: SessionId('e2e-code-mode-workspace-session'),
+      meta: { cwd: workdir },
+      agentOptions: { provider: 'deepseek', model: 'deepseek-v4-flash' },
+    })
+
+    handle.agent.send([{
+      type: 'text',
+      text: 'Use one run_code program to call tools.read on pkg/deep/task.txt. After it finishes, answer: Code Mode workspace handshake?',
+    }])
+    await waitForIdle(ctx, handle.agent as ReactLoopAgent)
+
+    const events: SessionEvent[] = [...handle.agent.session.events]
+    const dispatch = events.find(event => event.type === 'tool/code-dispatch' && event.data.name === 'read')
+    const outerResult = events.find(event => event.type === 'tool/result')
+    const workspaceContext = events.find(event => event.type === 'context/message'
+      && typeof event.data.meta === 'object'
+      && event.data.meta !== null
+      && !Array.isArray(event.data.meta)
+      && event.data.meta.kind === 'workspace-instructions')
+    expect(dispatch).toBeDefined()
+    expect(outerResult).toBeDefined()
+    expect(workspaceContext).toBeDefined()
+    expect(workspaceContext!.seq).toBeGreaterThan(outerResult!.seq)
+    const finalMessage = events.findLast(event => event.type === 'assistant/message')
+    const answer = finalMessage?.type === 'assistant/message'
+      ? finalMessage.data.content.filter(block => block.type === 'text').map(block => block.text).join('')
+      : ''
+    expect(answer).toContain(WORKSPACE_PROBE)
+  }, 180_000)
 })
diff --git a/examples/coding-agent/tests/coding-task.e2e.ts b/examples/coding-agent/tests/coding-task.e2e.ts
index ce716bdb2c..d53688835e 100644
--- a/examples/coding-agent/tests/coding-task.e2e.ts
+++ b/examples/coding-agent/tests/coding-task.e2e.ts
@@ -54,7 +54,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('coding task: fix a failing test
     expect(before.status).not.toBe(0)
 
     ctx = await codingHarness(workdir, { persona: SYSTEM_PROMPT })
-    const agent = ctx.agentLoop.create(AgentId('e2e-task'), { model: 'deepseek-v4-flash' })
+    const agent = ctx.agentLoop.create(AgentId('e2e-task'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     agent.send([{
       type: 'text',
diff --git a/examples/coding-agent/tests/compaction.e2e.ts b/examples/coding-agent/tests/compaction.e2e.ts
index 932209115f..6ba2908ad3 100644
--- a/examples/coding-agent/tests/compaction.e2e.ts
+++ b/examples/coding-agent/tests/compaction.e2e.ts
@@ -33,17 +33,20 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('compaction: a long session compa
     // Reasoning tokens require a larger generation cap than the retained checkpoint.
     ctx = await codingHarness(workdir, {
       persona: SYSTEM_PROMPT,
-      compact: {
+      tokenMeter: {
         contextWindow: 2000,
+      },
+      compact: {
         thresholdRatio: 0.5,
         retainTokens: 400,
+        summarizationProvider: '',
         summarizationModel: '',
         maxTokens: 1024,
         compactionRetries: 1,
       },
       persistenceRoot: join(workdir, '.sessions'),
     })
-    const agent = ctx.agentLoop.create(AgentId('e2e-compaction'), { model: 'deepseek-v4-flash' })
+    const agent = ctx.agentLoop.create(AgentId('e2e-compaction'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     agent.send([{
       type: 'text',
diff --git a/examples/coding-agent/tests/full-loop.e2e.ts b/examples/coding-agent/tests/full-loop.e2e.ts
index 8718139ced..7314b7ab21 100644
--- a/examples/coding-agent/tests/full-loop.e2e.ts
+++ b/examples/coding-agent/tests/full-loop.e2e.ts
@@ -28,7 +28,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('full loop: real model + real bas
   it('runs a bash command on request and reports its output', async () => {
     workdir = await mkdtemp(join(tmpdir(), 'dsh-full-loop-e2e-'))
     ctx = await codingHarness(workdir, { persona: SYSTEM_PROMPT })
-    const agent = ctx.agentLoop.create(AgentId('e2e-loop'), { model: 'deepseek-v4-flash' })
+    const agent = ctx.agentLoop.create(AgentId('e2e-loop'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     agent.send([{ type: 'text', text: 'Run `echo e2e-ok` with the bash tool and tell me its exact output.' }])
     await waitForIdle(ctx, agent)
diff --git a/examples/coding-agent/tests/harness.ts b/examples/coding-agent/tests/harness.ts
index dd0bc42a1b..944d01a0b9 100644
--- a/examples/coding-agent/tests/harness.ts
+++ b/examples/coding-agent/tests/harness.ts
@@ -1,15 +1,13 @@
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
 import type { SessionEvent } from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry from '@deepseek-ai/dsh-agent'
 import AgentLoop, { ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
 import * as ToolBash from '@deepseek-ai/dsh-tool-bash'
 import * as ToolTodo from '@deepseek-ai/dsh-tool-todo'
 import * as LlmDeepSeek from '@deepseek-ai/dsh-llm-deepseek'
+import TokenMeterService from '@deepseek-ai/dsh-token-meter'
+import type { TokenMeterConfig } from '@deepseek-ai/dsh-token-meter'
 import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
 import { BasicCompactService } from '@deepseek-ai/dsh-compact-basic'
 import type { BasicCompactConfig } from '@deepseek-ai/dsh-compact-basic'
@@ -46,23 +44,26 @@ export interface CodingHarnessOptions {
    * compaction plugin (the default suites run without it).
    */
   compact?: BasicCompactConfig
+  /** Optional token-meter capacity loaded before compact-basic. */
+  tokenMeter?: TokenMeterConfig
 }
 
 export async function codingHarness(workdir: string, options: CodingHarnessOptions = {}): Promise<Context> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt, { persona: options.persona ?? '' })
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx, {
+    systemPrompt: { persona: options.persona ?? '' },
+  })
   await ctx.plugin(AgentLoop, { agents: [] })
-  await ctx.plugin(LlmDeepSeek, { models: ['deepseek-v4-flash'] })
+  await ctx.plugin(LlmDeepSeek)
   await ctx.plugin(LocalBashExecutor, { cwd: workdir, timeoutMs: 30_000 })
   await ctx.plugin(ToolBash)
   await ctx.plugin(ToolTodo)
-  // Compaction is opt-in: only the compaction e2e loads it, with a lowered
-  // contextWindow/retainTokens so a short real session crosses the threshold.
-  if (options.compact !== undefined) await ctx.plugin(BasicCompactService, options.compact)
+  // Compaction is opt-in: only the compaction e2e loads the reusable meter and
+  // backend, with a lower context window so a short real session crosses the threshold.
+  if (options.compact !== undefined) {
+    await ctx.plugin(TokenMeterService, options.tokenMeter)
+    await ctx.plugin(BasicCompactService, options.compact)
+  }
   // Durable JSONL persistence is opt-in: only the resume e2e needs it, and the
   // other suites stay file-free. Loaded last so a resume's deferred
   // `ctx.inject(['sessionPersistence'])` resolves once this is present.
diff --git a/examples/coding-agent/tests/resume.e2e.ts b/examples/coding-agent/tests/resume.e2e.ts
index 70382b4beb..c9081c659f 100644
--- a/examples/coding-agent/tests/resume.e2e.ts
+++ b/examples/coding-agent/tests/resume.e2e.ts
@@ -42,7 +42,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('resume: continue a persisted ses
     const first = (await ctx.agents.create({
       agentId: AgentId('resume-1'),
       sessionId: SESSION_ID,
-      agentOptions: { model: 'deepseek-v4-flash' },
+      agentOptions: { provider: 'deepseek', model: 'deepseek-v4-flash' },
     })).agent as ReactLoopAgent
     first.send([{ type: 'text', text: `Remember this code for later: ${SECRET}. Just acknowledge it.` }])
     await waitForIdle(ctx, first)
@@ -56,7 +56,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('resume: continue a persisted ses
     const resumed = (await ctx.agents.resume({
       agentId: AgentId('resume-2'),
       resumeSessionId: SESSION_ID,
-      agentOptions: { model: 'deepseek-v4-flash' },
+      agentOptions: { provider: 'deepseek', model: 'deepseek-v4-flash' },
     })).agent as ReactLoopAgent
     expect(resumed.session.id).toBe(SESSION_ID)
     // The prior user turn is in the rehydrated log before the model is asked.
diff --git a/examples/coding-agent/tests/todo-write.e2e.ts b/examples/coding-agent/tests/todo-write.e2e.ts
index 698fbd9e9e..a16f15528d 100644
--- a/examples/coding-agent/tests/todo-write.e2e.ts
+++ b/examples/coding-agent/tests/todo-write.e2e.ts
@@ -26,7 +26,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('todo_write: real model records a
   it('appends a todo/write event with the model-produced task list', async () => {
     workdir = await mkdtemp(join(tmpdir(), 'dsh-todo-write-e2e-'))
     ctx = await codingHarness(workdir, { persona: TODO_SYSTEM_PROMPT })
-    const agent = ctx.agentLoop.create(AgentId('e2e-todo'), { model: 'deepseek-v4-flash' })
+    const agent = ctx.agentLoop.create(AgentId('e2e-todo'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     agent.send([{ type: 'text', text:
       'Use the todo_write tool to record a plan of exactly two steps: first '
diff --git a/examples/cordis-agent/cordis.yml b/examples/cordis-agent/cordis.yml
index 8cd989e9e4..a567705cbd 100644
--- a/examples/cordis-agent/cordis.yml
+++ b/examples/cordis-agent/cordis.yml
@@ -19,11 +19,8 @@
   config:
     apiKey: !!js process.env.DEEPSEEK_API_KEY
     baseURL: !!js process.env.DEEPSEEK_BASE_URL
-    models:
-      - deepseek-v4-pro
-      - deepseek-v4-flash
 
-# Local bash executor for agent-core's tool-bash schema — gives the agent an
+# Local bash executor for agent-spine-demo's tool-bash schema — gives the agent an
 # ordinary tool whose calls make the mounted listeners observably fire.
 - id: bash
   name: '@deepseek-ai/dsh-bash-local'
@@ -51,9 +48,12 @@
 - id: stdio-agent
   name: '@deepseek-ai/dsh-stdio-demo'
   config:
+    provider: deepseek
     model: deepseek-v4-flash
     resumeSessionId: !!js process.env.RESUME_SESSION_ID
     persistenceRoot: './.sessions'
+    workspaceContext:
+      maxBytes: 65536
     welcome: 'cordis-agent ready. Ask it to inspect its runtime, mount a listener, or invent a tool for itself.'
     persona: |
       You are cordis-agent, a self-referential harness demo powered by the
diff --git a/examples/cordis-agent/tests/cordis-tools.e2e.ts b/examples/cordis-agent/tests/cordis-tools.e2e.ts
index 32eb90c1ac..68fe88038b 100644
--- a/examples/cordis-agent/tests/cordis-tools.e2e.ts
+++ b/examples/cordis-agent/tests/cordis-tools.e2e.ts
@@ -38,7 +38,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('cordis tools: a real model modif
   it('mounts a status listener whose tagged output actually fires, then unmounts it', async () => {
     ctx = await cordisHarness()
     const log = vi.spyOn(console, 'log').mockImplementation(() => {})
-    const agent = ctx.agentLoop.create(AgentId('cordis-e2e-listener'), { model: 'deepseek-v4-flash' })
+    const agent = ctx.agentLoop.create(AgentId('cordis-e2e-listener'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     agent.send([{
       type: 'text',
@@ -66,7 +66,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('cordis tools: a real model modif
 
   it('builds itself a reverse_text tool and actually calls it', async () => {
     ctx = await cordisHarness()
-    const agent = ctx.agentLoop.create(AgentId('cordis-e2e-selftool'), { model: 'deepseek-v4-flash' })
+    const agent = ctx.agentLoop.create(AgentId('cordis-e2e-selftool'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     agent.send([{
       type: 'text',
@@ -113,7 +113,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('cordis tools: a real model modif
 
   it('composes two mounts through provide/inject, and unmounting the provider parks the consumer', async () => {
     ctx = await cordisHarness()
-    const agent = ctx.agentLoop.create(AgentId('cordis-e2e-compose'), { model: 'deepseek-v4-flash' })
+    const agent = ctx.agentLoop.create(AgentId('cordis-e2e-compose'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     agent.send([{
       type: 'text',
diff --git a/examples/cordis-agent/tests/harness.ts b/examples/cordis-agent/tests/harness.ts
index 78e5b0bb93..19868af14c 100644
--- a/examples/cordis-agent/tests/harness.ts
+++ b/examples/cordis-agent/tests/harness.ts
@@ -1,10 +1,6 @@
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry from '@deepseek-ai/dsh-agent'
 import AgentLoop, { ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as LlmDeepSeek from '@deepseek-ai/dsh-llm-deepseek'
 import * as ToolCordis from '@deepseek-ai/dsh-tool-cordis'
 
@@ -23,13 +19,11 @@ const PERSONA = 'You are cordis-agent, a self-referential harness demo. '
 
 export async function cordisHarness(): Promise<Context> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt, { persona: PERSONA })
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx, {
+    systemPrompt: { persona: PERSONA },
+  })
   await ctx.plugin(AgentLoop, { agents: [] })
-  await ctx.plugin(LlmDeepSeek, { models: ['deepseek-v4-flash'] })
+  await ctx.plugin(LlmDeepSeek)
   await ctx.plugin(ToolCordis)
   return ctx
 }
diff --git a/examples/echo-agent/composition.md b/examples/echo-agent/composition.md
index 5160bd20b7..9c9f04cb50 100644
--- a/examples/echo-agent/composition.md
+++ b/examples/echo-agent/composition.md
@@ -16,6 +16,8 @@ flowchart LR
   cfg --> plugin_echo_echo_tool
   plugin_echo_bash["bash<br/>@deepseek-ai/dsh-bash-local"]
   cfg --> plugin_echo_bash
+  plugin_echo_fs_local["fs-local<br/>@deepseek-ai/dsh-fs-local"]
+  cfg --> plugin_echo_fs_local
   plugin_echo_stdio_agent["stdio-agent<br/>@deepseek-ai/dsh-stdio-demo"]
   cfg --> plugin_echo_stdio_agent
   plugin_echo_stdio_agent --> bundle_agent_core["@deepseek-ai/dsh-agent-spine-demo"]
@@ -33,6 +35,7 @@ flowchart LR
 | `mock-llm` | `./src/mock-llm.ts` |
 | `echo-tool` | `./src/echo-tool.ts` |
 | `bash` | `@deepseek-ai/dsh-bash-local` |
+| `fs-local` | `@deepseek-ai/dsh-fs-local` |
 | `stdio-agent` | `@deepseek-ai/dsh-stdio-demo` |
 
 Source config: [`examples/echo-agent/cordis.yml`](cordis.yml).
diff --git a/examples/echo-agent/cordis.yml b/examples/echo-agent/cordis.yml
index fbf998e770..429ca118d6 100644
--- a/examples/echo-agent/cordis.yml
+++ b/examples/echo-agent/cordis.yml
@@ -16,17 +16,27 @@
 - id: echo-tool
   name: './src/echo-tool.ts'
 
-# Local bash executor: agent-core ships the `tool-bash` consumer schema, so the
+# Local bash executor: agent-spine-demo ships the `tool-bash` consumer schema, so the
 # leaf provides the executor it runs on (the echo demo doesn't drive bash, but
 # the tool is part of the shared spine).
 - id: bash
   name: '@deepseek-ai/dsh-bash-local'
 
+# Local filesystem provider for agent-spine-demo's workspace-context loader. This
+# does not expose model-facing read/write/edit tools in the echo demo.
+- id: fs-local
+  name: '@deepseek-ai/dsh-fs-local'
+  config:
+    cwd: !!js process.cwd()
+
 # The app pre-creates `main` on the mock model and supplies logging, persistence, and readline UI.
 - id: stdio-agent
   name: '@deepseek-ai/dsh-stdio-demo'
   config:
+    provider: mock
     model: mock-echo
     persona: 'You are echo-agent, a demo agent.'
     welcome: 'echo-agent ready. Type a message ("echo <text>" triggers the tool).'
     persistenceRoot: './.sessions'
+    workspaceContext:
+      maxBytes: 65536
diff --git a/examples/echo-agent/src/mock-llm.ts b/examples/echo-agent/src/mock-llm.ts
index 93132711de..1f61dc4ee3 100644
--- a/examples/echo-agent/src/mock-llm.ts
+++ b/examples/echo-agent/src/mock-llm.ts
@@ -55,5 +55,5 @@ export const name = 'mock-llm'
 export const inject = ['llm']
 
 export function apply(ctx: Context) {
-  ctx.llm.registerAdapter(['mock-echo'], new MockEchoAdapter())
+  ctx.llm.registerAdapter(['mock'], new MockEchoAdapter())
 }
diff --git a/packages/context/time-context/tests/fixtures/cordis.yml b/examples/echo-agent/tests/fixtures/context/time-context/cordis.yml
similarity index 83%
rename from packages/context/time-context/tests/fixtures/cordis.yml
rename to examples/echo-agent/tests/fixtures/context/time-context/cordis.yml
index da18fc11df..9b59e2ded9 100644
--- a/packages/context/time-context/tests/fixtures/cordis.yml
+++ b/examples/echo-agent/tests/fixtures/context/time-context/cordis.yml
@@ -1,6 +1,6 @@
 # Test-only composition: keep time-context opt-in while exercising its real Loader/app path.
 - id: mock-llm
-  name: '../../../../../examples/echo-agent/src/mock-llm.ts'
+  name: '../../../../src/mock-llm.ts'
 
 - id: bash
   name: '@deepseek-ai/dsh-bash-local'
@@ -11,7 +11,9 @@
 - id: stdio-agent
   name: '@deepseek-ai/dsh-stdio-demo'
   config:
+    provider: mock
     model: mock-echo
     persona: 'Test the time-context plugin.'
     welcome: 'time-context e2e ready.'
     persistenceRoot: './.sessions'
+    workspaceContext: false
diff --git a/examples/package.json b/examples/package.json
new file mode 100644
index 0000000000..8d77731a05
--- /dev/null
+++ b/examples/package.json
@@ -0,0 +1,46 @@
+{
+  "name": "dsh-examples",
+  "private": true,
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Workspace umbrella for runnable demos and example-owned test compositions: declares their cordis.yml packages so plain Node resolves real exports→lib. Not a build target.",
+  "dependencies": {
+    "@cordisjs/plugin-hmr": "workspace:*",
+    "@cordisjs/plugin-include": "workspace:*",
+    "@deepseek-ai/dsh-acp-demo": "workspace:*",
+    "@deepseek-ai/dsh-bash-local": "workspace:*",
+    "@deepseek-ai/dsh-bash-sandbox": "workspace:*",
+    "@deepseek-ai/dsh-code-runtime-worker": "workspace:*",
+    "@deepseek-ai/dsh-compact-basic": "workspace:*",
+    "@deepseek-ai/dsh-fs-local": "workspace:*",
+    "@deepseek-ai/dsh-fs-policy": "workspace:*",
+    "@deepseek-ai/dsh-hooks-claude": "workspace:*",
+    "@deepseek-ai/dsh-hooks-codex": "workspace:*",
+    "@deepseek-ai/dsh-llm": "workspace:*",
+    "@deepseek-ai/dsh-llm-deepseek": "workspace:*",
+    "@deepseek-ai/dsh-llm-replay": "workspace:*",
+    "@deepseek-ai/dsh-permission": "workspace:*",
+    "@deepseek-ai/dsh-repeat-tool-guard": "workspace:*",
+    "@deepseek-ai/dsh-sandbox-local": "workspace:*",
+    "@deepseek-ai/dsh-spill-local": "workspace:*",
+    "@deepseek-ai/dsh-spill-policy": "workspace:*",
+    "@deepseek-ai/dsh-stdio-demo": "workspace:*",
+    "@deepseek-ai/dsh-subagent": "workspace:*",
+    "@deepseek-ai/dsh-subagent-fork": "workspace:*",
+    "@deepseek-ai/dsh-subagent-spawn": "workspace:*",
+    "@deepseek-ai/dsh-time-context": "workspace:*",
+    "@deepseek-ai/dsh-timeout-policy": "workspace:*",
+    "@deepseek-ai/dsh-token-meter": "workspace:*",
+    "@deepseek-ai/dsh-tool-cordis": "workspace:*",
+    "@deepseek-ai/dsh-tool-fs": "workspace:*",
+    "@deepseek-ai/dsh-tool-fs-search": "workspace:*",
+    "@deepseek-ai/dsh-tool-subagent": "workspace:*",
+    "@deepseek-ai/dsh-tool-todo": "workspace:*",
+    "@deepseek-ai/dsh-tool-workflow": "workspace:*",
+    "@deepseek-ai/dsh-tools": "workspace:*",
+    "@deepseek-ai/dsh-user-approval": "workspace:*",
+    "@deepseek-ai/dsh-web": "workspace:*",
+    "@deepseek-ai/dsh-web-fetch-local": "workspace:*",
+    "@deepseek-ai/dsh-workflow-workerthread": "workspace:*"
+  }
+}
diff --git a/knip.json b/knip.json
index 46bafd8751..eee0e347b1 100644
--- a/knip.json
+++ b/knip.json
@@ -5,15 +5,16 @@
   "ignoreWorkspaces": ["vendor/*", "python/sdk-runtime"],
   "workspaces": {
     ".": {
+      "project": ["scripts/**/*.ts"]
+    },
+    "examples": {
       "entry": [
-        "examples/echo-agent/src/*.ts",
-        "examples/echo-agent/tests/**/*.e2e.ts",
-        "examples/coding-agent/tests/**/*.e2e.ts",
-        "examples/cordis-agent/tests/**/*.e2e.ts",
-        "examples/acp-agent/tests/**/*.e2e.ts",
-        "examples/*/tests/**/*.snapshot.ts"
+        "echo-agent/src/*.ts",
+        "*/tests/**/*.e2e.ts",
+        "*/tests/**/*.snapshot.ts"
       ],
-      "project": ["scripts/**/*.ts", "examples/**/*.ts"]
+      "project": ["**/*.ts"],
+      "ignoreDependencies": ["@deepseek-ai/.+", "@cordisjs/.+"]
     },
     "website": {
       "project": ["**/*.ts"],
@@ -45,15 +46,24 @@
       "project": ["src/**/*.ts"],
       "ignoreDependencies": ["cordis"]
     },
+    "packages/util/home": {
+      "entry": ["tests/**/*.spec.ts"],
+      "project": ["src/**/*.ts", "tests/**/*.ts"],
+      "ignoreDependencies": ["cordis"]
+    },
     "packages/util/timeout": {
       "entry": ["tests/**/*.spec.ts"],
       "project": ["src/**/*.ts", "tests/**/*.ts"],
       "ignoreDependencies": ["cordis"]
     },
+    "packages/util/retention": {
+      "entry": ["tests/**/*.spec.ts"],
+      "project": ["src/**/*.ts", "tests/**/*.ts"],
+      "ignoreDependencies": ["cordis"]
+    },
     "packages/support/acp-snapshot": {
       "entry": ["tests/**/*.spec.ts", "tests/fixtures/fake-acp-agent.ts"],
-      "project": ["src/**/*.ts", "tests/**/*.ts"],
-      "ignoreDependencies": ["cordis"]
+      "project": ["src/**/*.ts", "tests/**/*.ts"]
     },
     "packages/support/loader-smoke": {
       "entry": ["tests/**/*.spec.ts", "tests/fixtures/*.ts"],
@@ -76,6 +86,15 @@
       "entry": ["tests/**/*.spec.ts", "tests/**/*.e2e.ts"],
       "project": ["src/**/*.ts", "tests/**/*.ts"]
     },
+    "packages/context/workspace-context": {
+      "entry": ["tests/**/*.spec.ts", "tests/**/*.e2e.ts"],
+      "project": ["src/**/*.ts", "tests/**/*.ts"]
+    },
+    "packages/util/paths": {
+      "entry": ["tests/**/*.spec.ts"],
+      "project": ["src/**/*.ts", "tests/**/*.ts"],
+      "ignoreDependencies": ["cordis"]
+    },
     "packages/web/web-search-exa": {
       "entry": ["tests/**/*.spec.ts", "tests/**/*.e2e.ts"],
       "project": ["src/**/*.ts", "tests/**/*.ts"]
@@ -107,6 +126,15 @@
     "packages/examples/jsonrpc-demo": {
       "project": ["src/**/*.ts"]
     },
+    "packages/sdk/create-sdk": {
+      "entry": ["src/bin.ts", "tests/**/*.spec.ts", "tests/**/*.e2e.ts", "tests/**/*.snapshot.ts"],
+      "project": ["src/**/*.ts", "tests/**/*.ts"]
+    },
+    "packages/sdk/scripts": {
+      "entry": ["src/bin.ts", "tests/**/*.spec.ts", "tests/**/*.snapshot.ts"],
+      "project": ["src/**/*.ts", "tests/**/*.ts"],
+      "ignoreDependencies": ["node-addon-require-builtin"]
+    },
     "packages/subagent/subagent-spawn": {
       "entry": ["tests/**/*.spec.ts", "tests/**/*.e2e.ts"],
       "project": ["src/**/*.ts", "tests/**/*.ts"]
@@ -124,6 +152,11 @@
       "entry": ["tests/**/*.spec.ts", "tests/**/*.e2e.ts"],
       "project": ["src/**/*.ts", "tests/**/*.ts"]
     },
+    "packages/fs/tool-fs-search": {
+      "entry": ["tests/**/*.spec.ts"],
+      "project": ["src/**/*.ts", "tests/**/*.ts"],
+      "ignoreBinaries": ["rg"]
+    },
     "packages/mcp/mcp-client": {
       "entry": ["tests/**/*.spec.ts", "tests/**/*.e2e.ts", "tests/fixture-server.ts"],
       "project": ["src/**/*.ts", "tests/**/*.ts"],
diff --git a/package.json b/package.json
index 8c18188c09..d7e0e24446 100644
--- a/package.json
+++ b/package.json
@@ -45,12 +45,15 @@
     "verify-rfc-classification": "tsx scripts/verify-rfc-classification.ts",
     "verify-rfc-format": "tsx scripts/verify-rfc-format.ts",
     "verify-type-equiv": "tsx scripts/verify-type-equiv.ts",
+    "verify-translation-prompt": "tsx scripts/verify-translation-prompt.ts",
     "verify-translation-pairing": "tsx scripts/verify-translation-pairing.ts",
     "verify-doc-budgets": "tsx scripts/verify-doc-budgets.ts",
     "docs:dev": "pnpm --filter @deepseek-ai/website run dev",
     "docs:build": "pnpm --filter @deepseek-ai/website run build",
     "docs:preview": "pnpm --filter @deepseek-ai/website run preview",
     "docs:check": "pnpm exec vitest run scripts/project-doc-site.spec.ts && pnpm run docs:build",
+    "website:dev": "pnpm run docs:dev",
+    "website:build": "pnpm run docs:build",
     "verify-package-readme-limitations": "tsx scripts/verify-package-readme-limitations.ts",
     "verify-node-next-types": "tsx scripts/verify-node-next-types.ts",
     "verify-runtime-closure": "tsx scripts/verify-runtime-closure.ts",
@@ -74,7 +77,7 @@
     "verify-scoped-events": "tsx scripts/gen-scoped-events.ts --check",
     "verify-module-graph": "tsx scripts/gen-module-graph.ts --check",
     "constraints": "tsx scripts/check-workspace-constraints.ts",
-    "doc-sync": "pnpm run doc-typecheck && pnpm run verify-cordis-catalog && pnpm run verify-cordis-api && pnpm run verify-export-jsdoc && pnpm run verify-tool-catalog && pnpm run verify-config-catalog && pnpm run verify-persistence-catalog && pnpm run verify-doc-graphs && pnpm run verify-scoped-events && pnpm run verify-md-wrap && pnpm run verify-md-links && pnpm run verify-doc-refs && pnpm run verify-package-paths && pnpm run verify-package-readme-model-experience && pnpm run verify-mermaid && pnpm run verify-rfc-classification && pnpm run verify-rfc-format && pnpm run verify-type-equiv && pnpm run verify-translation-pairing && pnpm run verify-doc-budgets && pnpm run verify-package-readme-limitations && pnpm run docs:check",
+    "doc-sync": "pnpm run doc-typecheck && pnpm run verify-cordis-catalog && pnpm run verify-cordis-api && pnpm run verify-export-jsdoc && pnpm run verify-tool-catalog && pnpm run verify-config-catalog && pnpm run verify-persistence-catalog && pnpm run verify-doc-graphs && pnpm run verify-scoped-events && pnpm run verify-md-wrap && pnpm run verify-md-links && pnpm run verify-doc-refs && pnpm run verify-package-paths && pnpm run verify-package-readme-model-experience && pnpm run verify-mermaid && pnpm run verify-rfc-classification && pnpm run verify-rfc-format && pnpm run verify-type-equiv && pnpm run verify-translation-prompt && pnpm run verify-translation-pairing && pnpm run verify-doc-budgets && pnpm run verify-package-readme-limitations && pnpm run docs:check",
     "hygiene": "pnpm run knip && pnpm run publint && pnpm run constraints && pnpm run verify-cordis-config && pnpm run verify-node-next-types && pnpm run verify-runtime-closure",
     "demo:echo": "node --expose-internals --import tsx packages/examples/stdio-demo/src/bin.ts examples/echo-agent/cordis.yml",
     "demo:repl": "node --expose-internals --import tsx packages/examples/stdio-demo/src/bin.ts examples/coding-agent/cordis.yml",
@@ -104,6 +107,7 @@
     "mermaid": "11.16.0",
     "micromark-extension-gfm": "^3.0.0",
     "publint": "^0.3.21",
+    "saxes": "^6.0.0",
     "tsdown": "^0.22.2",
     "tsx": "^4.22.4",
     "typescript": "^6.0.3",
diff --git a/packages/README.md b/packages/README.md
index db364421b1..6c83abfc91 100644
--- a/packages/README.md
+++ b/packages/README.md
@@ -4,7 +4,7 @@ Packages use the `@deepseek-ai/dsh-*` scope. Each is a Cordis `Service` subclass
 
 ## Hierarchy
 
-Packages are grouped by modular role at `packages/<group>/<pkg>/`. The group directory is a pure container (no `package.json`); the package name stays `@deepseek-ai/dsh-<pkg>` regardless of group. **Each group README is the canonical per-package map** — package roles, ctx keys, and the product-vs-support split live there, next to the code.
+Packages live at `packages/<group>/<pkg>/`; groups are containers, while names remain `@deepseek-ai/dsh-<pkg>`. **Each group README is the canonical package/ctx-key map.**
 
 | Group | Role | Release expectation |
 |---|---|---|
@@ -13,30 +13,33 @@ Packages are grouped by modular role at `packages/<group>/<pkg>/`. The group dir
 | [`bash/`](bash/README.md) | Bash capability family: the executor seam, a local impl, and the model-facing tool | Product — stable surface |
 | [`code-runtime/`](code-runtime/README.md) | Code-execution capability family: the abstract runtime seam for model-written programs + a worker-thread backend | Product — stable surface |
 | [`sandbox/`](sandbox/README.md) | Process-confinement seam; bwrap/Landlock/Seatbelt backends | Product — stable surface |
-| [`fs/`](fs/README.md) | Filesystem capability family: the abstract seam, a local impl, and the model-facing file tools | Product — stable surface |
+| [`fs/`](fs/README.md) | Filesystem capability family: the abstract seam, a local impl, the model-facing file tools, and the bash-backed discovery tools | Product — stable surface |
 | [`skill/`](skill/README.md) | Skill capability family: the provider registry, local provider, and model-facing catalog/loader | Product — stable surface |
 | [`compact/`](compact/README.md) | Compaction capability family: the abstract seam + a basic backend (tool deferred) | Product — stable surface |
-| [`context/`](context/README.md) | Opt-in request-context enrichment | Product — stable surface |
+| [`context/`](context/README.md) | Model-visible request context, including workspace instructions and time context | Product — stable surface |
 | [`subagent/`](subagent/README.md) | Subagent capability family: the provider-registry seam and the model-facing delegation tool | Product — stable surface |
+| [`tasks/`](tasks/README.md) | Generic background-task runtime and model-facing `task_*` control tools | Product — stable surface |
 | [`workflow/`](workflow/README.md) | Workflow capability family: the script-engine seam, the worker-thread engine, and the model-facing `workflow` tool | Product — stable surface |
 | [`web/`](web/README.md) | Web capability family: the abstract seam, search/fetch provider impls, and the model-facing web tools | Product — stable surface |
-| [`timeout/`](timeout/README.md) | Tool-call timeout policy: the `tools/execute` deadline enforcer | Product — stable surface |
+| [`spill/`](spill/README.md) | Spill capability family: the storage seam, a local impl, and the tool-result spill policy | Product — stable surface |
 | [`todo/`](todo/README.md) | Todo/planning family: the model-facing `todo_write` tool | Product — stable surface |
+| [`timeout/`](timeout/README.md) | Tool-call timeout policy: the `tools/execute` deadline enforcer | Product — stable surface |
 | [`guard/`](guard/README.md) | Loop-hygiene guards: advisory repeat-call reminders | Product — stable surface |
 | [`cordis/`](cordis/README.md) | Self-referential runtime toolset: inspect the live runtime's plugins and services, mount/unmount model-written plugins ([design](../docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md)) | Product — stable surface |
 | [`hooks/`](hooks/README.md) | Hook bridges + the shared Claude Code / Codex wire-protocol library | Product — stable surface |
 | [`session-persistence/`](session-persistence/README.md) | Persistence capability family: the seam + JSONL/SQLite backends | Product — stable surface |
-| [`session-query/`](session-query/README.md) | Session retrieval family: logical corpus, surface records, and bounded exact reads | Product — stable surface |
+| [`session-query/`](session-query/README.md) | Session retrieval: logical corpus, bounded reads, lineage, and event relationships | Product — stable surface |
+| [`sdk/`](sdk/README.md) | Project SDK tooling | Product — stable surface |
 | [`ui/`](ui/README.md) | Editor/client integration surfaces: ACP bridge, JSON-RPC SDK server, user-approval/user-interaction seams, ask-user tool | Product — stable surface |
 | [`examples/`](examples/README.md) | Demo bundles (agent-spine + stdio/ACP/JSON-RPC bins) the leaves load | Support — example infra |
-| [`support/`](support/README.md) | Support infrastructure (invariants, replay, Loader smokes) | Support — lower compatibility expectations |
-| [`util/`](util/README.md) | Low-level zero-dependency utilities shared across groups (the `Branded<B>` primitive) | Support — small, stable, harness-dep-free |
+| [`support/`](support/README.md) | Support infrastructure (testkits, invariants, replay, Loader smokes, subagent mock) | Support — lower compatibility expectations |
+| [`util/`](util/README.md) | Low-level zero-dependency utilities shared across groups (`Branded<B>`, Harness home/path helpers, timeout, retention) | Support — small, stable, harness-dep-free |
 
-The split is the point: a package's group says whether it is part of the product API or support/test/example infrastructure, so release and removal decisions do not treat every package as an equal public contract. New packages join an existing group; adding a new top-level group is a deliberate act (extend the group READMEs and this table).
+Groups distinguish product API from support infrastructure. New packages join an existing group; a new group updates its README and this table.
 
 ## Dependencies
 
-The inter-package dependency graph is generated: [docs/module-graph.md](../docs/module-graph.md) (`pnpm run gen-module-graph`, freshness-gated in CI).
+The dependency graph is generated: [docs/module-graph.md](../docs/module-graph.md) (`pnpm run gen-module-graph`, freshness-gated in CI).
 
 The rule it must obey: **extension plugins depend on interfaces, never on the concrete loop.** `dsh-agent-loop` is swappable — UI/hook/tool plugins keep working against the `dsh-agent` vocabulary if the loop is replaced. The sanctioned exception is a **composition/bundle** package like `dsh-agent-spine-demo`, whose whole job is to assemble the concrete spine: it depends on `dsh-agent-loop` (and the other concrete spine plugins) on purpose. The rule constrains plugins that EXTEND the system, not the bundle that COMPOSES it. A swappable capability splits into interface / implementation / consumer packages (the bash trio is the template — see [capability seams](../docs/rfc/implemented/architecture/2026-06-13-capability-seams.md)).
 
diff --git a/packages/bash/README.md b/packages/bash/README.md
index c850ce2c15..20c11c5dff 100644
--- a/packages/bash/README.md
+++ b/packages/bash/README.md
@@ -7,6 +7,6 @@ The canonical three-package capability seam (see [capability seams](../../docs/r
 | `bash/` | Abstract bash executor seam (interface + vocabulary; sandbox result facts carry the [`sandbox/`](../sandbox/README.md) seam's mode/enforcement vocabulary) | `ctx.bash` |
 | `bash-local/` | Local-subprocess `BashExecutor` implementation | (registers `ctx.bash`) |
 | `bash-sandbox/` | Sandbox-consuming `BashExecutor` (wraps every command argv via `ctx.sandbox`, stamps denial/enforcement facts; extends `bash-local`'s mechanics) | (registers `ctx.bash`) |
-| `tool-bash/` | Model-facing `bash`/`bash_output`/`bash_kill` tool schemas | (registers on `ctx.tools`) |
+| `tool-bash/` | Model-facing `bash` schema; background processes register with the generic [`tasks/`](../tasks/README.md) runtime | (registers on `ctx.tools`) |
 
 The interface lives at `bash/bash/`. `bash-sandbox` replacing `bash-local` without touching the interface or the tool is the split doing exactly what it exists for — a leaf `cordis.yml` picks one executor entry, plus a `ctx.sandbox` provider entry for the confined one (see [the acp-agent example's default composition](../../examples/acp-agent/)).
diff --git a/packages/bash/bash-local/README.md b/packages/bash/bash-local/README.md
index cf919ab836..afa5000767 100644
--- a/packages/bash/bash-local/README.md
+++ b/packages/bash/bash-local/README.md
@@ -23,13 +23,13 @@ Design surveyed against the bash tools of Claude Code, OpenCode, Codex, and pi;
 
 - **Spawn per call, no shell state** — every call is a fresh non-login `bash -c` (deterministic; no rc files). All four surveyed tools spawn per call. `XXX(stateful-shell)` in `src/run.ts` records the two proven stateful designs (Claude Code's cwd-only persistence; Codex's PTY exec sessions) for when real workflows demand them.
 - **Process-group kills with escalation** — children are spawned `detached` (own process group); kills send SIGTERM to the group, then SIGKILL after the `graceMs` grace (default 3s — OpenCode's escalation; pipelines and subshells die with the parent). ESRCH is tolerated; daemons that re-parent away from the group can still survive — same caveat as the surveyed tools.
-- **Tail-keep truncation + spill files** — output beyond `maxOutputBytes` keeps the in-memory TAIL (errors/results cluster at the end — pi/OpenCode rationale) while the FULL stream is appended to a temp file whose path is reported when available. If the final spill close reports a delayed writeback failure, the executor still returns the tail but withholds the path rather than advertising a possibly incomplete file.
-- **Model-friendly environment** — ambient credential-shaped variables are removed before noninteractive terminal defaults and explicit caller entries are applied. Supplied stdin is written and closed; otherwise fd 0 is `/dev/null`. Trusted plugins use `env` and `stdin`, but the model-facing tool does not expose them. See the [bash stdin/env RFC](../../../docs/rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md).
-- **Background tasks** — `start()` returns immediately, no timeout applies (Claude Code detaches timeouts when backgrounding), `readOutput()` is incremental with whole-stream byte offsets, and disposal kills everything. The spec's opaque `owner` token is stored on the tracked task and returned by `ownerOf(id)` — the executor never interprets it (the consumer's access policy does), and because it lives with the task here it survives a `tool-bash` HMR reload.
+- **Tail-keep truncation + spill files** — output beyond `maxOutputBytes` keeps the in-memory TAIL (errors/results cluster at the end — pi/OpenCode rationale) while the FULL stream is appended to a temp file whose path is reported when available. A foreground `BashExecRequest.stdoutMaxBytes` can raise stdout's capture budget for one trusted caller; stderr and background tasks still use `maxOutputBytes`. If the final spill close reports a delayed writeback failure, the executor still returns the tail but withholds the path rather than advertising a possibly incomplete file.
+- **Model-friendly env + credential scrub** — `process.env` minus credential-shaped vars (`*KEY*`/`*SECRET*`/`*TOKEN*`) and all ambient `DSH_*` names, then `NO_COLOR=1 TERM=dumb PAGER=cat GIT_PAGER=cat` (Codex's hardcoded set) so pagers and ANSI color don't garble results. A spec's ordinary `env` is merged after the scrub but rejects `DSH_*`; managed `dshEnv` rejects ordinary names and merges last, preventing stale nested-harness identity. Supplied stdin is written and closed; otherwise fd 0 is `/dev/null`. See the [stdin/env RFC](../../../docs/rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md) and [managed environment RFC](../../../docs/rfc/implemented/feature/2026-07-10-agent-session-identity-and-log-location.md).
+- **Background processes** — `start()` returns a live `BashProcess` handle immediately, no timeout applies (Claude Code detaches timeouts when backgrounding), the handle's `readOutput()` is incremental with whole-stream byte offsets, and disposal kills every running process and awaits its exit. Everything task-shaped (ids, ownership, polling, notices) lives in the generic [`ctx.tasks` runtime](../../tasks/tasks/README.md), which the tool layer registers the handle with — this executor never sees a session or a registry.
 
 ## Model Experience
 
-Indirectly, through `dsh-tool-bash`, which renders this executor's bounded stdout/stderr tails, background-task deltas and state, spill-file path, exact `Error: unknown bash task "<taskId>"` and `Error: aborted before spawn: <reason>` failures, and retains each resulting tool message until compaction.
+Indirectly, through `dsh-tool-bash`, which renders this executor's bounded stdout/stderr tails, background-process deltas, spill-file paths, and infrastructure failures.
 
 ## Known Limitations and Deferred Work
 
@@ -38,6 +38,5 @@ Indirectly, through `dsh-tool-bash`, which renders this executor's bounded stdou
 - **POSIX-only** — the `bash` binary, detached process groups, group kills, and SIGTERM→SIGKILL escalation are hardcoded; Windows is unsupported.
 - **The credential scrub is a name heuristic** — `*KEY*`/`*SECRET*`/`*TOKEN*` only; differently-named secrets (e.g. `*PASSWORD*`) pass through, and a whitelist for over-scrubbed vars is noted future work.
 - **Spill files are never deleted** — full-output recovery files (and the private per-process spill dir) accumulate under the OS tmpdir until something external cleans them.
-- **Finished background tasks are never evicted** — they stay in the task map, retaining their in-memory output tails, until executor disposal.
 
 The raw process handling lives in `src/run.ts`; `src/index.ts` is the service wiring.
diff --git a/packages/bash/bash-local/src/index.ts b/packages/bash/bash-local/src/index.ts
index f6dfcea636..587c33b9a2 100644
--- a/packages/bash/bash-local/src/index.ts
+++ b/packages/bash/bash-local/src/index.ts
@@ -1,15 +1,14 @@
 /**
- * Local-subprocess implementation of the bash seam. Each call runs in its own
- * process group, background tasks are tracked, and disposal kills and awaits
- * them. Execution policy belongs in `tools/pre-execute` or a sandboxing
- * executor, not this local process layer.
+ * Local-subprocess implementation of the bash executor seam. Each command runs
+ * as `bash -c` in its own process group; disposal kills and joins live groups.
+ * Execution policy belongs in `tools/pre-execute` or a sandboxing executor.
  * @module @deepseek-ai/dsh-bash-local
  */
 
 import { Context } from 'cordis'
 import z from 'schemastery'
-import { BashExecutor, BashTaskId } from '@deepseek-ai/dsh-bash'
-import type { BashExecRequest, BashExecSpec, BashRunResult, BashTask, BashTaskRead, OwnerToken } from '@deepseek-ai/dsh-bash'
+import { BashExecutor } from '@deepseek-ai/dsh-bash'
+import type { BashExecRequest, BashExecSpec, BashProcess, BashProcessRead, BashRunResult } from '@deepseek-ai/dsh-bash'
 import { clampTimeout, deadline, timeoutOf } from '@deepseek-ai/dsh-timeout'
 import { DEFAULT_GRACE_MS, runBash } from './run.ts'
 import type { RunInternals, RunningBash } from './run.ts'
@@ -37,20 +36,9 @@ function assertPositiveFinite(name: string, value: number): void {
   }
 }
 
-interface TrackedTask extends BashTask {
-  running: RunningBash
-  /** Whole-stream byte offsets already delivered via {@link LocalBashExecutor.readOutput}. */
-  stdoutOffset: number
-  stderrOffset: number
-  /** Opaque owner token from the {@link BashExecSpec} (the consumer's isolation key). */
-  owner: OwnerToken | undefined
-}
-
 /**
- * Local-subprocess bash executor. Defaults follow the agent-tool survey
- * consensus: 120s default / 600s max timeout (Claude Code, OpenCode), 64KB
- * in-memory output with full-stream spill files (pi, OpenCode),
- * process-group SIGTERM→SIGKILL kills with a 3s grace (OpenCode).
+ * Local bash executor with bounded output, spill files, and process-group
+ * `SIGTERM` to `SIGKILL` escalation.
  */
 export class LocalBashExecutor extends BashExecutor {
   static Config: z<Config> = z.object({
@@ -61,8 +49,8 @@ export class LocalBashExecutor extends BashExecutor {
     graceMs: z.number().default(DEFAULT_GRACE_MS),
   })
 
-  private tasks = new Map<BashTaskId, TrackedTask>()
-  private nextTaskId = 1
+  /** Live processes retained only so disposal can kill and join them. */
+  private live = new Map<BashProcess, RunningBash>()
   /** Test seam: spill knobs forwarded to runBash. */
   internals: RunInternals = {}
 
@@ -71,26 +59,21 @@ export class LocalBashExecutor extends BashExecutor {
 
   constructor(ctx: Context, config: Config) {
     super(ctx)
-    // schemastery (static Config) has already filled the defaulted fields;
-    // the cast records that runtime fact for exactOptionalPropertyTypes.
+    // Schemastery fills these fields before construction; the type does not encode that step.
     this.config = config as ResolvedConfig
     assertPositiveFinite('timeoutMs', this.config.timeoutMs)
     assertPositiveFinite('maxTimeoutMs', this.config.maxTimeoutMs)
     assertPositiveFinite('maxOutputBytes', this.config.maxOutputBytes)
     assertPositiveFinite('graceMs', this.config.graceMs)
     ctx.effect(() => async () => {
-      // Kill every live process group and WAIT for the processes to close so nothing outlives
-      // the fiber (HMR safety) — a TERM-trapping child is held until the SIGKILL escalation
-      // lands.
+      // Await closure so even a TERM-trapping child cannot outlive the fiber.
       const pending: Promise<void>[] = []
-      for (const task of this.tasks.values()) {
-        if (task.status === 'running') {
-          task.status = 'killed'
-          task.running.kill()
-          pending.push(task.done)
-        }
+      for (const [proc, running] of this.live) {
+        proc.status = 'killed'
+        running.kill()
+        pending.push(proc.done)
       }
-      this.tasks.clear()
+      this.live.clear()
       await Promise.all(pending)
     }, 'local bash teardown')
   }
@@ -109,18 +92,19 @@ export class LocalBashExecutor extends BashExecutor {
       this.config.maxTimeoutMs,
       'bash-local: request.timeoutMs',
     )
+    const stdoutMaxBytes = request.stdoutMaxBytes ?? this.config.maxOutputBytes
+    assertPositiveFinite('request.stdoutMaxBytes', stdoutMaxBytes)
     return {
       command: request.command,
       workdir: request.workdir ?? this.config.cwd ?? process.cwd(),
       timeoutMs,
+      stdoutMaxBytes,
       ...request.signal ? { signal: request.signal } : {},
-      // Carry stdin/env through verbatim — optional, no config default (absent
-      // means none). env merges AFTER the scrub in run.ts.
+      // Carry stdin/ordinary env/trusted dshEnv through verbatim — optional,
+      // no config default. run.ts owns the scrub and merge order.
       ...request.stdin !== undefined ? { stdin: request.stdin } : {},
       ...request.env !== undefined ? { env: request.env } : {},
-      // Carry the owner through verbatim (required-but-nullable on the spec):
-      // the executor never interprets it — the consumer's access policy does.
-      owner: request.owner,
+      ...request.dshEnv !== undefined ? { dshEnv: request.dshEnv } : {},
       // Carry a sandbox-mode override through verbatim: this executor never
       // confines, so the field is inert here (the seam contract) — a
       // sandboxing subclass overrides resolve() to stamp its default instead.
@@ -129,128 +113,99 @@ export class LocalBashExecutor extends BashExecutor {
   }
 
   async run(spec: BashExecSpec): Promise<BashRunResult> {
-    // One fused deadline drives both the timeout and upstream cancellation;
-    // runBash listens on d.signal and runs the SIGTERM→grace→SIGKILL kill.
-    // `using` clears the timer across the awaited process lifetime.
+    // One deadline combines timeout and upstream cancellation; disposal clears its timer.
     using d = deadline(spec.signal, spec.timeoutMs, 'BASH_TIMEOUT')
     const outcome = await runBash({
       command: spec.command,
       cwd: spec.workdir,
-      maxOutputBytes: this.config.maxOutputBytes,
+      stdoutMaxBytes: spec.stdoutMaxBytes,
+      stderrMaxBytes: this.config.maxOutputBytes,
       graceMs: this.config.graceMs,
       signal: d.signal,
       stdin: spec.stdin,
       env: spec.env,
+      dshEnv: spec.dshEnv,
     }, this.internals).done
-    // Classify the FIRST abort reason: a BASH_TIMEOUT TimeoutReason means our timeout cut the
-    // command short; any other abort — an upstream cancel, or a foreign (outer) deadline's
-    // timeout under nesting — is aborted.
+    // Only this executor's timeout reason counts as timedOut; outer deadlines count as aborts.
     const timedOut = timeoutOf(d.signal, 'BASH_TIMEOUT') !== undefined
     const aborted = d.signal.aborted && !timedOut
     return { ...outcome, timedOut, aborted, timeoutMs: spec.timeoutMs }
   }
 
-  start(spec: BashExecSpec): BashTask {
-    // No timeout for background tasks (matches Claude Code, which detaches the timeout when
-    // backgrounding); callers stop tasks via kill() — or via spec.signal, which the seam
-    // contract honors for background runs too (runBash wires it to the group kill).
+  start(spec: BashExecSpec): BashProcess {
+    // Background runs ignore timeoutMs; callers stop them through kill() or spec.signal.
     const running = runBash({
       command: spec.command,
       cwd: spec.workdir,
-      maxOutputBytes: this.config.maxOutputBytes,
+      stdoutMaxBytes: this.config.maxOutputBytes,
+      stderrMaxBytes: this.config.maxOutputBytes,
       graceMs: this.config.graceMs,
       signal: spec.signal,
       stdin: spec.stdin,
       env: spec.env,
+      dshEnv: spec.dshEnv,
     }, this.internals)
 
-    const id = BashTaskId(`bash-${this.nextTaskId++}`)
-    const task: TrackedTask = {
-      id,
+    let stdoutOffset = 0
+    let stderrOffset = 0
+    const proc: BashProcess = {
       status: 'running',
       exitCode: null,
       signal: null,
-      owner: spec.owner,
-      running,
-      stdoutOffset: 0,
-      stderrOffset: 0,
       done: running.done.then((outcome) => {
-        // Abort-killed tasks report as killed, not completed. Background runs
-        // forward only the upstream signal (no timeout), so its aborted state
-        // is the authoritative "was this cancelled" signal.
-        if (task.status === 'running') task.status = spec.signal?.aborted === true ? 'killed' : 'completed'
-        task.exitCode = outcome.exitCode
-        task.signal = outcome.signal
-        this.notifyTaskDone(task)
+        // Any signal termination is killed, including a command signaling itself.
+        if (proc.status === 'running') {
+          proc.status = spec.signal?.aborted === true || outcome.signal !== null ? 'killed' : 'completed'
+        }
+        proc.exitCode = outcome.exitCode
+        proc.signal = outcome.signal
+        this.onProcessDone(proc, running.stderr.readFrom(0).text)
+        this.live.delete(proc)
       }, (error: unknown) => {
-        // Spawn-level failure (bad workdir, …): the task never ran. String()
-        // suffices — runBash only rejects with Error instances.
-        task.status = 'killed'
-        task.running.stderr.push(Buffer.from(`spawn failed: ${String(error)}`))
-        this.notifyTaskDone(task)
+        // Background spawn failures settle as killed and surface through the read path.
+        proc.status = 'killed'
+        running.stderr.push(Buffer.from(`spawn failed: ${String(error)}`))
+        this.onProcessDone(proc, running.stderr.readFrom(0).text)
+        this.live.delete(proc)
       }),
-    }
-    this.tasks.set(id, task)
-    return task
-  }
+      readOutput: (): BashProcessRead => {
+        const out = running.stdout.readFrom(stdoutOffset)
+        const err = running.stderr.readFrom(stderrOffset)
+        stdoutOffset = out.nextOffset
+        stderrOffset = err.nextOffset
 
-  get(id: BashTaskId): BashTask | undefined {
-    return this.tasks.get(id)
+        // Single newline between sections: stdout chunks usually end with one
+        // already; add it only when missing.
+        const separator = out.text.length > 0 && !out.text.endsWith('\n') ? '\n' : ''
+        const delta = out.text
+          + (err.text.length > 0 ? `${separator}[stderr]\n${err.text}` : '')
+        return {
+          delta,
+          lossy: out.lossy || err.lossy,
+          ...out.spillPath !== undefined ? { stdoutSpillPath: out.spillPath } : {},
+          ...err.spillPath !== undefined ? { stderrSpillPath: err.spillPath } : {},
+        }
+      },
+      kill: (): boolean => {
+        if (proc.status !== 'running') return false
+        proc.status = 'killed'
+        running.kill()
+        return true
+      },
+    }
+    this.live.set(proc, running)
+    return proc
   }
 
   /**
-   * Full collected stderr of a tracked task from stream start (bounded by the
-   * in-memory cap; bytes only in the spill file are not re-read). A protected
-   * seam for subclasses that classify a settled task's outcome — reading here
-   * does NOT advance the consumer's {@link readOutput} cursor. An unknown id
-   * (a task already dropped by disposal) reads as empty.
+   * Settlement hook for subclasses that attach execution facts to a process.
+   * Called after exit facts or spawn-failure output are stamped and before
+   * {@link BashProcess.done} resolves. The base implementation is intentionally
+   * empty.
+   * @param _proc - the settled process handle.
+   * @param _stderr - the process's retained stderr tail used by subclasses for settlement classification.
    */
-  protected collectedStderr(id: BashTaskId): string {
-    const task = this.tasks.get(id)
-    return task === undefined ? '' : task.running.stderr.readFrom(0).text
-  }
-
-  ownerOf(id: BashTaskId): OwnerToken | undefined {
-    // Unknown id and known-but-ownerless both read as undefined — the consumer
-    // treats undefined as "open" and a truly unknown id fails at readOutput/kill.
-    return this.tasks.get(id)?.owner
-  }
-
-  list(): BashTask[] {
-    return [...this.tasks.values()]
-  }
-
-  readOutput(id: BashTaskId): BashTaskRead {
-    const task = this.tasks.get(id)
-    if (!task) throw new Error(`unknown bash task "${id}"`)
-
-    const out = task.running.stdout.readFrom(task.stdoutOffset)
-    const err = task.running.stderr.readFrom(task.stderrOffset)
-    task.stdoutOffset = out.nextOffset
-    task.stderrOffset = err.nextOffset
-
-    // Single newline between sections: stdout chunks usually end with one
-    // already; add it only when missing.
-    const separator = out.text.length > 0 && !out.text.endsWith('\n') ? '\n' : ''
-    const delta = out.text
-      + (err.text.length > 0 ? `${separator}[stderr]\n${err.text}` : '')
-    return {
-      task,
-      delta,
-      lossy: out.lossy || err.lossy,
-      ...out.spillPath !== undefined ? { stdoutSpillPath: out.spillPath } : {},
-      ...err.spillPath !== undefined ? { stderrSpillPath: err.spillPath } : {},
-    }
-  }
-
-  kill(id: BashTaskId): boolean {
-    const task = this.tasks.get(id)
-    if (!task) throw new Error(`unknown bash task "${id}"`)
-    if (task.status !== 'running') return false
-    task.status = 'killed'
-    task.running.kill()
-    return true
-  }
+  protected onProcessDone(_proc: BashProcess, _stderr: string): void {}
 }
 
 export default LocalBashExecutor
diff --git a/packages/bash/bash-local/src/run.ts b/packages/bash/bash-local/src/run.ts
index 02e7a963be..fa4dae73d6 100644
--- a/packages/bash/bash-local/src/run.ts
+++ b/packages/bash/bash-local/src/run.ts
@@ -11,7 +11,8 @@ import { randomBytes } from 'node:crypto'
 import { closeSync, mkdtempSync, openSync, writeSync } from 'node:fs'
 import { tmpdir } from 'node:os'
 import { join } from 'node:path'
-import type { CollectedOutput } from '@deepseek-ai/dsh-bash'
+import { DSH_ENV_PREFIX } from '@deepseek-ai/dsh-bash'
+import type { CollectedOutput, DshEnvironment } from '@deepseek-ai/dsh-bash'
 
 /**
  * Model-friendly environment overrides: disable colors, pagers, and
@@ -34,26 +35,43 @@ export const ENV_OVERRIDES = {
 export const SENSITIVE_ENV_PATTERN = /KEY|SECRET|TOKEN/i
 
 /**
- * Build a child environment by scrubbing credential-shaped ambient variables,
- * applying model-friendly overrides, then merging trusted caller entries last.
- *
- * @param extra - caller-supplied entries merged last; an explicit entry wins even against the scrub and the overrides.
+ * Build a child environment from scrubbed ambient values, terminal overrides,
+ * ordinary caller entries, and a managed `DSH_*` snapshot. Ambient managed
+ * names are removed; ordinary and managed entries reject the other channel's
+ * namespace before `dshEnv` merges last.
+ * @param extra - caller entries; `DSH_*` names are rejected.
+ * @param dshEnv - managed entries; non-`DSH_*` names are rejected.
  * @returns the environment to hand to `spawn` for the child process.
  */
-export function childEnv(extra?: Record<string, string>): NodeJS.ProcessEnv {
+export function childEnv(
+  extra?: Readonly<Record<string, string>>,
+  dshEnv?: DshEnvironment,
+): NodeJS.ProcessEnv {
   const env: NodeJS.ProcessEnv = {}
   for (const [key, value] of Object.entries(process.env)) {
-    if (!SENSITIVE_ENV_PATTERN.test(key)) env[key] = value
+    if (!SENSITIVE_ENV_PATTERN.test(key) && !key.startsWith(DSH_ENV_PREFIX)) env[key] = value
   }
-  return { ...env, ...ENV_OVERRIDES, ...extra }
+  for (const key of Object.keys(extra ?? {})) {
+    if (key.startsWith(DSH_ENV_PREFIX)) {
+      throw new Error(`ordinary bash env cannot set reserved variable "${key}"; use dshEnv`)
+    }
+  }
+  for (const key of Object.keys(dshEnv ?? {})) {
+    if (!key.startsWith(DSH_ENV_PREFIX)) {
+      throw new Error(`managed bash env cannot set ordinary variable "${key}"; use env`)
+    }
+  }
+  return { ...env, ...ENV_OVERRIDES, ...extra, ...dshEnv }
 }
 
 /** What to run and under which limits (resolved — no defaults in here). */
 export interface SpawnSpec {
   command: string
   cwd: string
-  /** Per-stream in-memory cap; overflow spills to disk (tail kept in memory). */
-  maxOutputBytes: number
+  /** Stdout in-memory cap; overflow spills to disk (tail kept in memory). */
+  stdoutMaxBytes: number
+  /** Stderr in-memory cap; overflow spills to disk (tail kept in memory). */
+  stderrMaxBytes: number
   /** Grace period between the SIGTERM and the SIGKILL escalation on a kill. */
   graceMs: number
   /**
@@ -71,12 +89,12 @@ export interface SpawnSpec {
    */
   stdin?: string | undefined
   /**
-   * Extra environment entries, merged onto the scrubbed env AFTER the
-   * credential scrub and the model-friendly overrides (so an explicit entry
-   * wins). Set by in-process plugins; the model-facing tool does not forward
-   * model input here.
+   * Ordinary environment entries merged after the credential scrub and
+   * terminal overrides. `DSH_*` names are rejected and belong in `dshEnv`.
    */
   env?: Record<string, string> | undefined
+  /** Harness-owned entries; non-`DSH_*` names are rejected before spawn. */
+  dshEnv?: DshEnvironment | undefined
 }
 
 /**
@@ -278,13 +296,13 @@ export function runBash(spec: SpawnSpec, internals: RunInternals = {}): RunningB
   }
 
   // Keep absent stdin as /dev/null; literal tuples preserve non-null output types.
-  const env = childEnv(spec.env)
+  const env = childEnv(spec.env, spec.dshEnv)
   const child: ChildProcessByStdio<Writable | null, Readable, Readable> = spec.stdin !== undefined
     ? spawn('bash', ['-c', spec.command], { cwd: spec.cwd, env, stdio: ['pipe', 'pipe', 'pipe'], detached: true })
     : spawn('bash', ['-c', spec.command], { cwd: spec.cwd, env, stdio: ['ignore', 'pipe', 'pipe'], detached: true })
 
-  const stdout = new OutputCollector(spec.maxOutputBytes, 'stdout', spillDir)
-  const stderr = new OutputCollector(spec.maxOutputBytes, 'stderr', spillDir)
+  const stdout = new OutputCollector(spec.stdoutMaxBytes, 'stdout', spillDir)
+  const stderr = new OutputCollector(spec.stderrMaxBytes, 'stderr', spillDir)
   child.stdout.on('data', (chunk: Buffer) => { stdout.push(chunk) })
   child.stderr.on('data', (chunk: Buffer) => { stderr.push(chunk) })
 
diff --git a/packages/bash/bash-local/tests/executor.spec.ts b/packages/bash/bash-local/tests/executor.spec.ts
index 8f24b4b017..9db0c2eebd 100644
--- a/packages/bash/bash-local/tests/executor.spec.ts
+++ b/packages/bash/bash-local/tests/executor.spec.ts
@@ -1,11 +1,10 @@
 import { mkdtempSync } from 'node:fs'
 import { tmpdir } from 'node:os'
 import { join } from 'node:path'
-import { describe, expect, it, vi } from 'vitest'
+import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
 import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
-import { BashTaskId } from '@deepseek-ai/dsh-bash'
-import type { BashTaskRead } from '@deepseek-ai/dsh-bash'
+import type { BashProcess } from '@deepseek-ai/dsh-bash'
 
 const spillDir = mkdtempSync(join(tmpdir(), 'dsh-bash-exec-spec-'))
 
@@ -18,36 +17,20 @@ async function setup(config: ConstructorParameters<typeof LocalBashExecutor>[1]
   return { ctx, bash }
 }
 
-/** Poll until a pid no longer exists. */
-async function waitGone(pid: number, timeoutMs = 5_000): Promise<void> {
+/**
+ * Poll a handle's consuming readOutput until the ACCUMULATED delta contains
+ * `expected`; returns the accumulation (reads never re-deliver, so the caller
+ * gets everything produced up to the match).
+ */
+async function readUntil(proc: BashProcess, expected: string, timeoutMs = 5_000): Promise<string> {
   const deadline = Date.now() + timeoutMs
+  let all = ''
   while (Date.now() < deadline) {
-    try {
-      process.kill(pid, 0)
-    } catch {
-      return
-    }
+    all += proc.readOutput().delta
+    if (all.includes(expected)) return all
     await new Promise(resolve => setTimeout(resolve, 20))
   }
-  throw new Error(`pid ${pid} still alive after ${timeoutMs}ms`)
-}
-
-async function readUntil(
-  bash: LocalBashExecutor,
-  id: BashTaskId,
-  expected: string,
-  timeoutMs = 5_000,
-): Promise<BashTaskRead> {
-  const deadline = Date.now() + timeoutMs
-  let last: BashTaskRead | undefined
-  let delta = ''
-  while (Date.now() < deadline) {
-    last = bash.readOutput(id)
-    delta += last.delta
-    if (delta.includes(expected)) return { ...last, delta }
-    await new Promise(resolve => setTimeout(resolve, 20))
-  }
-  throw new Error(`task ${id} output did not include ${JSON.stringify(expected)}; output was ${JSON.stringify(delta)}, last delta was ${JSON.stringify(last?.delta ?? '')}`)
+  throw new Error(`process output did not include ${JSON.stringify(expected)}; accumulated ${JSON.stringify(all)}`)
 }
 
 describe('LocalBashExecutor.run', () => {
@@ -88,15 +71,23 @@ describe('LocalBashExecutor.run', () => {
     const { bash } = await setup()
     expect(() => bash.resolve({ command: 'true', timeoutMs: Number.NaN })).toThrow(/request\.timeoutMs/)
     expect(() => bash.resolve({ command: 'true', timeoutMs: -1 })).toThrow(/request\.timeoutMs/)
+    expect(() => bash.resolve({ command: 'true', stdoutMaxBytes: Number.NaN })).toThrow(/request\.stdoutMaxBytes/)
+    expect(() => bash.resolve({ command: 'true', stdoutMaxBytes: -1 })).toThrow(/request\.stdoutMaxBytes/)
   })
 
-  it('kill escalation uses the configured graceMs (a TERM-trapping task dies by SIGKILL)', async () => {
-    const { bash } = await setup() // setup pins graceMs: 200 via config
-    const task = bash.start(bash.resolve({ command: 'trap \'\' TERM; echo ready; while :; do sleep 60 & wait $!; done' }))
-    await readUntil(bash, task.id, 'ready\n')
-    bash.kill(task.id)
-    await task.done
-    expect(task.signal).toBe('SIGKILL')
+  it('defaults stdoutMaxBytes to maxOutputBytes and lets foreground callers raise stdout only', async () => {
+    const { bash } = await setup({ maxOutputBytes: 100 })
+    expect(bash.resolve({ command: 'true' }).stdoutMaxBytes).toBe(100)
+
+    const result = await bash.run(bash.resolve({
+      command: 'printf "%.0sx" $(seq 1 500); printf "%.0se" $(seq 1 500) >&2',
+      stdoutMaxBytes: 500,
+    }))
+
+    expect(result.stdout.truncated).toBe(false)
+    expect(result.stdout.text).toBe('x'.repeat(500))
+    expect(result.stderr.truncated).toBe(true)
+    expect(result.stderr.text.length).toBeLessThanOrEqual(100)
   })
 
   it('per-call timeout takes precedence under the cap and kills on expiry', async () => {
@@ -136,247 +127,209 @@ describe('LocalBashExecutor.run', () => {
     await expect(bash.run(bash.resolve({ command: 'true', workdir: '/nonexistent-dsh' }))).rejects.toThrow(/ENOENT/)
   })
 
-  it('resolve() carries stdin/env onto the spec, and run() threads them to the command', async () => {
+  it('resolve() carries stdin/env/dshEnv onto the spec, and run() threads them to the command', async () => {
     const { bash } = await setup()
-    const spec = bash.resolve({ command: 'cat; echo "[$DSH_SEAM_VAR]"', stdin: 'piped\n', env: { DSH_SEAM_VAR: 'env-ok' } })
-    // resolve() keeps the stdin/env fields verbatim (optional, no default).
+    const spec = bash.resolve({
+      command: 'cat; echo "[$SEAM_VAR][$DSH_SEAM_VAR]"',
+      stdin: 'piped\n',
+      env: { SEAM_VAR: 'env-ok' },
+      dshEnv: { DSH_SEAM_VAR: 'dsh-ok' },
+    })
+    // resolve() keeps the optional input/environment fields verbatim.
     expect(spec.stdin).toBe('piped\n')
-    expect(spec.env).toEqual({ DSH_SEAM_VAR: 'env-ok' })
+    expect(spec.env).toEqual({ SEAM_VAR: 'env-ok' })
+    expect(spec.dshEnv).toEqual({ DSH_SEAM_VAR: 'dsh-ok' })
     const result = await bash.run(spec)
-    expect(result.stdout.text).toBe('piped\n[env-ok]\n')
+    expect(result.stdout.text).toBe('piped\n[env-ok][dsh-ok]\n')
   })
 
-  it('resolve() omits stdin/env when the request supplies neither', async () => {
+  it('resolve() omits stdin/env/dshEnv when the request supplies none', async () => {
     const { bash } = await setup()
     const spec = bash.resolve({ command: 'true' })
     expect('stdin' in spec).toBe(false)
     expect('env' in spec).toBe(false)
+    expect('dshEnv' in spec).toBe(false)
   })
 })
 
-describe('LocalBashExecutor background tasks', () => {
-  it('start returns immediately with a registered running task', async () => {
+describe('LocalBashExecutor.start (background process handles)', () => {
+  it('start returns immediately with a running handle that settles as completed', async () => {
     const { bash } = await setup()
     const before = Date.now()
-    const task = bash.start(bash.resolve({ command: 'sleep 0.2; echo done' }))
+    const proc = bash.start(bash.resolve({ command: 'sleep 0.2; echo done' }))
     expect(Date.now() - before).toBeLessThan(150)
-    expect(task.status).toBe('running')
-    expect(bash.get(task.id)).toBe(task)
-    expect(bash.list()).toContain(task)
-    await task.done
-    expect(task.status).toBe('completed')
-    expect(task.exitCode).toBe(0)
+    expect(proc.status).toBe('running')
+    await proc.done
+    expect(proc.status).toBe('completed')
+    expect(proc.exitCode).toBe(0)
   })
 
-  it('assigns sequential ids', async () => {
+  it('threads stdin and extra env into a background process', async () => {
     const { bash } = await setup()
-    const first = bash.start(bash.resolve({ command: 'true' }))
-    const second = bash.start(bash.resolve({ command: 'true' }))
-    expect(first.id).toBe('bash-1')
-    expect(second.id).toBe('bash-2')
-    await Promise.all([first.done, second.done])
-  })
-
-  it('threads stdin and extra env into a background task', async () => {
-    const { bash } = await setup()
-    const task = bash.start(bash.resolve({
-      command: 'cat; echo "[$DSH_BG_VAR]"',
+    const proc = bash.start(bash.resolve({
+      command: 'cat; echo "[$BG_VAR][$DSH_BG_VAR]"',
       stdin: 'bg-stdin\n',
-      env: { DSH_BG_VAR: 'bg-env' },
+      env: { BG_VAR: 'bg-env' },
+      dshEnv: { DSH_BG_VAR: 'bg-dsh-env' },
     }))
-    const read = await readUntil(bash, task.id, '[bg-env]')
-    expect(read.delta).toContain('bg-stdin')
-    await task.done
-    expect(task.exitCode).toBe(0)
+    const output = await readUntil(proc, '[bg-env][bg-dsh-env]')
+    expect(output).toContain('bg-stdin')
+    await proc.done
+    expect(proc.exitCode).toBe(0)
   })
 
-  it('readOutput returns increments without re-delivery', async () => {
+  it('readOutput is consuming: increments are never re-delivered, and reads stay valid after exit', async () => {
     const { bash } = await setup()
-    const task = bash.start(bash.resolve({ command: 'echo first; sleep 1; echo second' }))
-    const first = await readUntil(bash, task.id, 'first\n')
-    expect(first.delta).toBe('first\n')
-    expect(first.lossy).toBe(false)
-    await task.done
-    const second = bash.readOutput(task.id)
+    const proc = bash.start(bash.resolve({ command: 'echo first; sleep 1; echo second' }))
+    const first = await readUntil(proc, 'first\n')
+    expect(first).toBe('first\n')
+    await proc.done
+    // Read-after-exit returns the remaining buffered output — once.
+    const second = proc.readOutput()
     expect(second.delta).toBe('second\n')
-    const third = bash.readOutput(task.id)
-    expect(third.delta).toBe('')
+    expect(second.lossy).toBe(false)
+    expect(proc.readOutput().delta).toBe('')
   })
 
   it('readOutput marks stderr sections', async () => {
     const { bash } = await setup()
-    const task = bash.start(bash.resolve({ command: 'echo out; echo err >&2' }))
-    await task.done
-    const read = bash.readOutput(task.id)
-    expect(read.delta).toBe('out\n[stderr]\nerr\n')
+    const proc = bash.start(bash.resolve({ command: 'echo out; echo err >&2' }))
+    await proc.done
+    expect(proc.readOutput().delta).toBe('out\n[stderr]\nerr\n')
   })
 
   it('readOutput reports stderr-only deltas without a leading newline', async () => {
     const { bash } = await setup()
-    const task = bash.start(bash.resolve({ command: 'echo err >&2' }))
-    await task.done
-    expect(bash.readOutput(task.id).delta).toBe('[stderr]\nerr\n')
+    const proc = bash.start(bash.resolve({ command: 'echo err >&2' }))
+    await proc.done
+    expect(proc.readOutput().delta).toBe('[stderr]\nerr\n')
   })
 
-  it('readOutput flags lossy reads and reports spill paths', async () => {
+  it('readOutput adds a separator only when stdout lacks a trailing newline', async () => {
+    const { bash } = await setup()
+    const proc = bash.start(bash.resolve({ command: 'printf out; echo err >&2' }))
+    await proc.done
+    expect(proc.readOutput().delta).toBe('out\n[stderr]\nerr\n')
+  })
+
+  it('readOutput flags lossy reads and reports stdout spill paths', async () => {
     const { bash } = await setup({ maxOutputBytes: 100 })
-    const task = bash.start(bash.resolve({ command: 'for i in $(seq 1 100); do printf "line-%04d\\n" $i; done' }))
-    await task.done
-    const read = bash.readOutput(task.id)
+    const proc = bash.start(bash.resolve({ command: 'for i in $(seq 1 100); do printf "line-%04d\\n" $i; done' }))
+    await proc.done
+    const read = proc.readOutput()
     // Window slid past offset 0 → lossy, spill path points at the full stream.
     expect(read.lossy).toBe(true)
     expect(read.stdoutSpillPath).toBeDefined()
   })
 
-  it('readOutput throws for unknown ids', async () => {
-    const { bash } = await setup()
-    expect(() => bash.readOutput(BashTaskId('nope'))).toThrow(/unknown bash task "nope"/)
-  })
-
-  it('kill terminates the process group and reports status killed', async () => {
-    const { bash } = await setup()
-    const task = bash.start(bash.resolve({ command: 'sleep 60' }))
-    expect(bash.kill(task.id)).toBe(true)
-    await task.done
-    expect(task.status).toBe('killed')
-    expect(task.signal).toBe('SIGTERM')
-  })
-
-  it('kill returns false for finished tasks and throws for unknown ids', async () => {
-    const { bash } = await setup()
-    const task = bash.start(bash.resolve({ command: 'true' }))
-    await task.done
-    expect(bash.kill(task.id)).toBe(false)
-    expect(() => bash.kill(BashTaskId('nope'))).toThrow(/unknown bash task "nope"/)
-  })
-
-  it('notifies onTaskDone listeners on completion', async () => {
-    const { bash } = await setup()
-    const seen: [string, string][] = []
-    bash.onTaskDone(task => void seen.push([task.id, task.status]))
-    const task = bash.start(bash.resolve({ command: 'true' }))
-    await task.done
-    expect(seen).toEqual([[task.id, 'completed']])
-  })
-
-  it('notifies onTaskDone for killed tasks too', async () => {
-    const { bash } = await setup()
-    const listener = vi.fn()
-    bash.onTaskDone(listener)
-    const task = bash.start(bash.resolve({ command: 'sleep 60' }))
-    bash.kill(task.id)
-    await task.done
-    expect(listener).toHaveBeenCalledWith(task)
-    expect(task.status).toBe('killed')
-  })
-
-  it('marks tasks killed when the background spawn itself fails', async () => {
-    const { bash } = await setup()
-    const listener = vi.fn()
-    bash.onTaskDone(listener)
-    const task = bash.start(bash.resolve({ command: 'true', workdir: '/nonexistent-dsh' }))
-    await task.done
-    expect(task.status).toBe('killed')
-    expect(listener).toHaveBeenCalledWith(task)
-    expect(bash.readOutput(task.id).delta).toContain('spawn failed')
-  })
-
-  it('readOutput adds a separator only when stdout lacks a trailing newline', async () => {
-    const { bash } = await setup()
-    const task = bash.start(bash.resolve({ command: 'printf out; echo err >&2' }))
-    await task.done
-    expect(bash.readOutput(task.id).delta).toBe('out\n[stderr]\nerr\n')
-  })
-
   it('readOutput reports stderr spill paths', async () => {
     const { bash } = await setup({ maxOutputBytes: 100 })
-    const task = bash.start(bash.resolve({ command: 'for i in $(seq 1 100); do printf "line-%04d\\n" $i >&2; done' }))
-    await task.done
-    const read = bash.readOutput(task.id)
+    const proc = bash.start(bash.resolve({ command: 'for i in $(seq 1 100); do printf "line-%04d\\n" $i >&2; done' }))
+    await proc.done
+    const read = proc.readOutput()
     expect(read.lossy).toBe(true)
     expect(read.stderrSpillPath).toBeDefined()
     expect(read.delta).toContain('[stderr]')
   })
 
-  it('disposing with already-finished tasks only kills the running ones', async () => {
+  it('kill() terminates the process group: true once, false after settlement', async () => {
+    const { bash } = await setup()
+    const proc = bash.start(bash.resolve({ command: 'sleep 60' }))
+    expect(proc.kill()).toBe(true)
+    await proc.done
+    expect(proc.status).toBe('killed')
+    expect(proc.signal).toBe('SIGTERM')
+    expect(proc.kill()).toBe(false)
+  })
+
+  it('kill() returns false for a naturally completed process', async () => {
+    const { bash } = await setup()
+    const proc = bash.start(bash.resolve({ command: 'true' }))
+    await proc.done
+    expect(proc.status).toBe('completed')
+    expect(proc.kill()).toBe(false)
+  })
+
+  it('kill escalation uses the configured graceMs (a TERM-trapping process dies by SIGKILL)', async () => {
+    const { bash } = await setup() // setup pins graceMs: 200 via config
+    // The child echoes AFTER arming the trap, so waiting for the marker
+    // guarantees SIGTERM is already ignored when the kill lands (a fixed sleep
+    // is load-flaky: a slow spawn would take the SIGTERM before the trap).
+    const proc = bash.start(bash.resolve({ command: 'trap \'\' TERM; echo armed; sleep 60' }))
+    await readUntil(proc, 'armed')
+    proc.kill()
+    await proc.done
+    expect(proc.status).toBe('killed')
+    expect(proc.signal).toBe('SIGKILL')
+  })
+
+  it('a spec.signal abort settles the handle as killed, not completed', async () => {
+    const { bash } = await setup()
+    const controller = new AbortController()
+    const proc = bash.start(bash.resolve({ command: 'sleep 60', signal: controller.signal }))
+    controller.abort()
+    await proc.done
+    expect(proc.status).toBe('killed')
+    expect(proc.signal).toBe('SIGTERM')
+  })
+
+  it('a self-signal exit settles the handle as killed, not completed', async () => {
+    const { bash } = await setup()
+    const proc = bash.start(bash.resolve({ command: 'kill -TERM $$' }))
+    await proc.done
+    expect(proc.status).toBe('killed')
+    expect(proc.exitCode).toBeNull()
+    expect(proc.signal).toBe('SIGTERM')
+  })
+
+  it('a background spawn failure settles as killed with the error readable on stderr', async () => {
+    const { bash } = await setup()
+    const proc = bash.start(bash.resolve({ command: 'true', workdir: '/nonexistent-dsh' }))
+    // done resolves (never rejects) even though the process never ran.
+    await expect(proc.done).resolves.toBeUndefined()
+    expect(proc.status).toBe('killed')
+    expect(proc.readOutput().delta).toContain('spawn failed:')
+  })
+})
+
+describe('LocalBashExecutor disposal', () => {
+  it('disposing the fiber kills running processes and AWAITS their exit (no orphans, SIGKILL escalation included)', async () => {
     const ctx = new Context()
     const fiber = await ctx.plugin(LocalBashExecutor, { graceMs: 200 })
     const bash = ctx.bash as LocalBashExecutor
     bash.internals = { spillDir }
 
-    const finished = bash.start(bash.resolve({ command: 'true' }))
+    // The child prints its own pid ($$ = the detached bash group leader) so
+    // the test can probe liveness through the public read surface alone.
+    const proc = bash.start(bash.resolve({ command: 'trap \'\' TERM; echo $$; sleep 60' }))
+    const pid = Number((await readUntil(proc, '\n')).trim())
+    expect(Number.isInteger(pid) && pid > 0).toBe(true)
+
+    await fiber.dispose()
+    // Disposal itself waited: the pid must already be gone, no grace left —
+    // even for a TERM-trapping child held until the SIGKILL escalation landed.
+    expect(() => process.kill(pid, 0)).toThrow()
+    expect(proc.status).toBe('killed')
+    await proc.done
+  })
+
+  it('settled processes already left the live map: dispose does not touch them', async () => {
+    const ctx = new Context()
+    const fiber = await ctx.plugin(LocalBashExecutor, { graceMs: 200 })
+    const bash = ctx.bash as LocalBashExecutor
+    bash.internals = { spillDir }
+
+    const finished = bash.start(bash.resolve({ command: 'echo done' }))
     await finished.done
+    expect(finished.status).toBe('completed')
     const running = bash.start(bash.resolve({ command: 'sleep 60' }))
 
     await fiber.dispose()
-    await running.done
+    // The teardown marks every LIVE entry killed; a settled process had
+    // already left the map, so its status stays completed.
     expect(finished.status).toBe('completed')
+    expect(running.status).toBe('killed')
+    await running.done
     expect(running.signal).toBe('SIGTERM')
-    expect(bash.list()).toEqual([])
-  })
-
-  it('disposing the executor fiber kills running tasks (no orphans)', async () => {
-    const ctx = new Context()
-    const fiber = await ctx.plugin(LocalBashExecutor, { graceMs: 200 })
-    const bash = ctx.bash as LocalBashExecutor
-    bash.internals = { spillDir }
-    const listener = vi.fn()
-    bash.onTaskDone(listener)
-
-    const task = bash.start(bash.resolve({ command: 'sleep 60' }))
-    const running = bash.get(task.id)!
-    await new Promise(resolve => setTimeout(resolve, 50))
-
-    // Grab the pid before dispose clears the registry.
-    const pid = (running as unknown as { running: { pid: number } }).running.pid
-    await fiber.dispose()
-    await waitGone(pid)
-    expect(bash.list()).toEqual([])
-    // Listener silenced by base-class teardown — no late notifications.
-    expect(listener).not.toHaveBeenCalled()
-  })
-})
-
-describe('executor cancellation, callback, and disposal contracts', () => {
-  it('start honors a pre-aborted or later-aborted AbortSignal', async () => {
-    const { bash } = await setup()
-    const controller = new AbortController()
-    const task = bash.start(bash.resolve({ command: 'sleep 60', signal: controller.signal }))
-    controller.abort()
-    await task.done
-    expect(task.status).toBe('killed')
-    expect(task.signal).toBe('SIGTERM')
-  })
-
-  it('a throwing onTaskDone listener does not reject task.done or starve later listeners', async () => {
-    const { bash } = await setup()
-    const errorSpy = vi.spyOn(console, 'error').mockImplementation(() => undefined)
-    const second = vi.fn()
-    try {
-      bash.onTaskDone(() => { throw new Error('listener bug') })
-      bash.onTaskDone(second)
-      const task = bash.start(bash.resolve({ command: 'true' }))
-      await expect(task.done).resolves.toBeUndefined()
-      expect(second).toHaveBeenCalledWith(task)
-      expect(errorSpy).toHaveBeenCalled()
-    } finally {
-      errorSpy.mockRestore()
-    }
-  })
-
-  it('dispose AWAITS a TERM-trapping process (SIGKILL escalation included)', async () => {
-    const ctx = new Context()
-    const fiber = await ctx.plugin(LocalBashExecutor, { graceMs: 200 })
-    const bash = ctx.bash as LocalBashExecutor
-    bash.internals = { spillDir }
-
-    const task = bash.start(bash.resolve({ command: 'trap \'\' TERM; sleep 60' }))
-    await new Promise(resolve => setTimeout(resolve, 100))
-    const pid = (task as unknown as { running: { pid: number } }).running.pid
-
-    await fiber.dispose()
-    // Disposal itself waited: the pid must already be gone, no grace left.
-    expect(() => process.kill(pid, 0)).toThrow()
-    expect(task.status).toBe('killed')
   })
 })
diff --git a/packages/bash/bash-local/tests/run.spec.ts b/packages/bash/bash-local/tests/run.spec.ts
index 923b3adf7b..e65500a4b5 100644
--- a/packages/bash/bash-local/tests/run.spec.ts
+++ b/packages/bash/bash-local/tests/run.spec.ts
@@ -2,6 +2,7 @@ import { mkdtempSync, readFileSync, statSync } from 'node:fs'
 import { tmpdir } from 'node:os'
 import { dirname, join } from 'node:path'
 import { describe, expect, it, vi } from 'vitest'
+import type { DshEnvironment } from '@deepseek-ai/dsh-bash'
 import { killGroup, OutputCollector, runBash } from '../src/run.ts'
 import type { RunningBash } from '../src/run.ts'
 
@@ -26,7 +27,8 @@ function spec(command: string, overrides: Partial<Parameters<typeof runBash>[0]>
   return {
     command,
     cwd: process.cwd(),
-    maxOutputBytes: 64_000,
+    stdoutMaxBytes: 64_000,
+    stderrMaxBytes: 64_000,
     graceMs: 3_000,
     ...overrides,
   }
@@ -197,19 +199,19 @@ describe('stdin and extra env (set by in-process plugins)', () => {
     expect(piped.stdout.text).toBe('socket\n')
   })
 
-  it('merges extra env entries onto the scrubbed environment', async () => {
-    const result = await runBash(spec('echo "$DSH_EXTRA_ONE/$DSH_EXTRA_TWO"', {
-      env: { DSH_EXTRA_ONE: 'alpha', DSH_EXTRA_TWO: 'beta' },
+  it('merges ordinary extra env entries onto the scrubbed environment', async () => {
+    const result = await runBash(spec('echo "$EXTRA_ONE/$EXTRA_TWO"', {
+      env: { EXTRA_ONE: 'alpha', EXTRA_TWO: 'beta' },
     })).done
     expect(result.stdout.text).toBe('alpha/beta\n')
   })
 
   it('an explicit extra env entry overrides the model-friendly override and the scrub', async () => {
     // TERM is a model-friendly OVERRIDE (dumb); an explicit extra entry wins.
-    // DSH_OVERRIDE_KEY matches the credential scrub pattern, yet an explicit
+    // EXPLICIT_OVERRIDE_KEY matches the credential scrub pattern, yet an explicit
     // entry is still honored — the scrub only drops AMBIENT process.env creds.
-    const result = await runBash(spec('echo "$TERM/$DSH_OVERRIDE_KEY"', {
-      env: { TERM: 'xterm-256color', DSH_OVERRIDE_KEY: 'explicit-wins' },
+    const result = await runBash(spec('echo "$TERM/$EXPLICIT_OVERRIDE_KEY"', {
+      env: { TERM: 'xterm-256color', EXPLICIT_OVERRIDE_KEY: 'explicit-wins' },
     })).done
     expect(result.stdout.text).toBe('xterm-256color/explicit-wins\n')
   })
@@ -224,10 +226,24 @@ describe('stdin and extra env (set by in-process plugins)', () => {
 })
 
 describe('output truncation and spill', () => {
+  it('applies stdout and stderr caps independently', async () => {
+    const result = await runBash(
+      spec('printf "%.0sx" $(seq 1 500); printf "%.0se" $(seq 1 500) >&2', {
+        stdoutMaxBytes: 500,
+        stderrMaxBytes: 100,
+      }),
+      { spillDir },
+    ).done
+    expect(result.stdout.truncated).toBe(false)
+    expect(result.stdout.text).toBe('x'.repeat(500))
+    expect(result.stderr.truncated).toBe(true)
+    expect(result.stderr.text.length).toBeLessThanOrEqual(100)
+  })
+
   it('keeps the tail and spills the full stream to disk', async () => {
     // 200 numbered lines of ~10 bytes; cap at 500 bytes keeps a late tail.
     const result = await runBash(
-      spec('for i in $(seq 1 200); do printf "line-%04d\\n" $i; done', { maxOutputBytes: 500 }),
+      spec('for i in $(seq 1 200); do printf "line-%04d\\n" $i; done', { stdoutMaxBytes: 500, stderrMaxBytes: 500 }),
       { spillDir },
     ).done
     expect(result.stdout.truncated).toBe(true)
@@ -242,7 +258,7 @@ describe('output truncation and spill', () => {
 
   it('does not truncate output exactly at the cap', async () => {
     const result = await runBash(
-      spec('printf "%.0sx" $(seq 1 500)', { maxOutputBytes: 500 }),
+      spec('printf "%.0sx" $(seq 1 500)', { stdoutMaxBytes: 500, stderrMaxBytes: 500 }),
       { spillDir },
     ).done
     expect(result.stdout.truncated).toBe(false)
@@ -253,7 +269,7 @@ describe('output truncation and spill', () => {
   it('settles with the tail and no spill path when final spill close fails', async () => {
     failNextClose.value = true
     const result = await runBash(
-      spec('for i in $(seq 1 200); do printf "line-%04d\\n" $i; done', { maxOutputBytes: 500 }),
+      spec('for i in $(seq 1 200); do printf "line-%04d\\n" $i; done', { stdoutMaxBytes: 500, stderrMaxBytes: 500 }),
       { spillDir },
     ).done
     expect(failNextClose.value).toBe(false)
@@ -348,13 +364,13 @@ describe('abort edge cases', () => {
 })
 
 describe('environment and spill-file hardening', () => {
-  it('scrubs credential-shaped env vars from child processes', async () => {
+  it('scrubs credential-shaped and ambient DSH env vars from child processes', async () => {
     process.env.DSH_TEST_API_KEY = 'super-secret'
     process.env.DSH_TEST_TOKEN = 'also-secret'
     process.env.DSH_TEST_PLAIN = 'visible'
     try {
       const result = await runBash(spec('echo "[${DSH_TEST_API_KEY:-absent}|${DSH_TEST_TOKEN:-absent}|${DSH_TEST_PLAIN:-absent}]"')).done
-      expect(result.stdout.text.trim()).toBe('[absent|absent|visible]')
+      expect(result.stdout.text.trim()).toBe('[absent|absent|absent]')
     } finally {
       delete process.env.DSH_TEST_API_KEY
       delete process.env.DSH_TEST_TOKEN
@@ -362,9 +378,32 @@ describe('environment and spill-file hardening', () => {
     }
   })
 
+  it('injects only the current trusted DSH environment after scrubbing ambient values', async () => {
+    process.env.DSH_STALE = 'old-value'
+    try {
+      const result = await runBash(spec('echo "[${DSH_STALE:-absent}|$DSH_SHELL|$DSH_SESSION_ID]"', {
+        dshEnv: { DSH_SHELL: '1', DSH_SESSION_ID: 'current-session' },
+      })).done
+      expect(result.stdout.text.trim()).toBe('[absent|1|current-session]')
+    } finally {
+      delete process.env.DSH_STALE
+    }
+  })
+
+  it('rejects DSH variables on the ordinary env channel', () => {
+    expect(() => runBash(spec('true', { env: { DSH_WRONG_CHANNEL: 'bad' } })))
+      .toThrow(/DSH_WRONG_CHANNEL.*dshEnv/)
+  })
+
+  it('rejects ordinary variables on the managed env channel', () => {
+    const invalid = { PATH: '/wrong-channel' } as unknown as DshEnvironment
+    expect(() => runBash(spec('true', { dshEnv: invalid })))
+      .toThrow(/managed bash env.*PATH.*use env/)
+  })
+
   it('creates spill files with owner-only permissions and random names', async () => {
     const result = await runBash(
-      spec('for i in $(seq 1 200); do printf "line-%04d\\n" $i; done', { maxOutputBytes: 500 }),
+      spec('for i in $(seq 1 200); do printf "line-%04d\\n" $i; done', { stdoutMaxBytes: 500, stderrMaxBytes: 500 }),
       { spillDir },
     ).done
     const path = result.stdout.spillPath!
@@ -375,7 +414,7 @@ describe('environment and spill-file hardening', () => {
 
   it('defaults spills into a private per-process directory', async () => {
     const result = await runBash(
-      spec('for i in $(seq 1 200); do printf "line-%04d\\n" $i; done', { maxOutputBytes: 500 }),
+      spec('for i in $(seq 1 200); do printf "line-%04d\\n" $i; done', { stdoutMaxBytes: 500, stderrMaxBytes: 500 }),
     ).done
     const dir = dirname(result.stdout.spillPath!)
     expect(dir).toMatch(/dsh-bash-/)
diff --git a/packages/bash/bash-sandbox/README.md b/packages/bash/bash-sandbox/README.md
index 0d0ebca439..347aefb679 100644
--- a/packages/bash/bash-sandbox/README.md
+++ b/packages/bash/bash-sandbox/README.md
@@ -2,21 +2,23 @@
 
 Sandbox-consuming implementation of the [`@deepseek-ai/dsh-bash`](../bash/) executor seam. Load it **instead of** `@deepseek-ai/dsh-bash-local`, together with a [`ctx.sandbox`](../../sandbox/sandbox/) provider (e.g. [`@deepseek-ai/dsh-sandbox-local`](../../sandbox/sandbox-local/)) — no alternate tool plugin is needed; `dsh-tool-bash` detects the executor's `sandboxMode` capability and adds the escalation fields.
 
+The package root exports the default and named `SandboxBashExecutor` plugin plus its `Config`; quoting and result-classification helpers stay internal.
+
 Every command is confined by handing the provider the exact `['bash', '-c', command]` argv this executor is about to spawn and spawning the returned (wrapped) argv instead. WHICH platform runner confines it — and whether one is usable at all (fail closed with a structured `SANDBOX_UNAVAILABLE` error, never a silent unconfined run) — is the provider's concern; this package owns the bash side only.
 
 | Mode | File effects |
 |---|---|
 | `read-only` (default) | No writes anywhere (of `/dev`, only the `/dev/null` node is writable, so `>/dev/null` keeps working) |
 | `workspace-write` | Writes only under `workspaceRoot` + `/tmp` (ephemeral under bwrap, the host `/tmp` under Landlock, `/private/tmp` plus the per-user temp dir under Seatbelt) |
-| `danger-full-access` | No confinement; the provider is never consulted. Execution is `dsh-bash-local`'s verbatim — foreground results still carry `sandbox: { mode, denied: false }` (no `enforcement`: nothing was confined), background tasks carry no sandbox facts |
+| `danger-full-access` | No confinement; the provider is never consulted. Foreground results carry `sandbox: { mode, denied: false }`; background process handles carry no sandbox facts. |
 
 Semantics:
 
 - **Denials are result facts.** A failed run whose stderr carries the selected backend's own denial dialect — the signatures the provider stamps on every wrap (EROFS text under bwrap, EACCES under Landlock, EPERM under Seatbelt) — is reported as `BashRunResult.sandbox.denied: true` (conservative classification, read from the collected stderr tail); every CONFINED run also carries the mode it executed under (`result.sandbox.mode`) and the provider's enforcement completeness (`result.sandbox.enforcement`: `full`, or `partial` on an older Landlock ABI).
-- **Runner failures are sandbox failures, never task failures.** A failed run matching the wrap's `runnerFailureSignatures` (the runner's own error prefix — also what the shell prints for a missing runner) means the sandbox itself broke and the command NEVER RAN; the check outranks denial classification because a runner's error text can contain denial words. The foreground path re-throws it as the structured fail-closed `SANDBOX_UNAVAILABLE` error, with the runner's first stderr line as the cause; a settled background task stamps `task.sandbox.runnerFailed` instead (no error channel remains after settle), which `bash_output` renders as its own marker.
-- **Config default, per-call override.** `resolve()` stamps the configured sandbox mode onto each spec unless an approved request supplies a wider mode. That override affects only its call or background task. `ctx.bash.sandboxMode` reports the default so the tool advertises escalation only when supported; results report the effective mode. The model learns standing mode only from tool/result facts, not a system-prompt announcement.
+- **Runner failures are sandbox failures, never command failures.** Foreground execution throws `SANDBOX_UNAVAILABLE`; a settled background process stamps `process.sandbox.runnerFailed`, which the bash producer renders through generic `task_output`. Spawn failures also pass through settlement, so confined background handles retain their mode/enforcement facts and release per-process accounting.
+- **Config-time default, per-call policy.** The DEFAULT mode is fixed by this entry's config for the executor's lifetime; `resolve()` stamps it onto every spec, and an explicit request-level `sandboxMode` override — set by the tool layer only for a call whose wider mode a human granted through `ctx.approval` ([the sandbox RFC § Escalation](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md)) — makes THAT call run, classify, and report under its own mode while every neighbor keeps the default (background facts are stamped per task at settle). The capability fact `ctx.bash.sandboxMode` reports the configured default so the tool layer advertises escalation only when this executor is mounted. The model learns of the sandbox only through result facts — the static bash tool description explains the denial marker; there is no current-mode statement in the system prompt.
 - **File effects only.** Network and process visibility are deliberately not restricted — the mode vocabulary does not pretend to cover what the backend does not enforce.
-- Process mechanics (spawn, process-group kills, output collection/spill, background tasks, credential scrub) are inherited verbatim from [`dsh-bash-local`](../bash-local/); the runner ladder, probes, and the per-platform Landlock launcher packages live with [`dsh-sandbox-local`](../../sandbox/sandbox-local/).
+- Process mechanics (spawn, process-group kills, output collection/spill, background handles, credential scrub) are inherited from [`dsh-bash-local`](../bash-local/); runner selection lives in [`dsh-sandbox-local`](../../sandbox/sandbox-local/).
 
 Deny-only at the seam: a denial is a reported fact, and this executor never negotiates permissions itself — the approval question lives in the tool layer (`dsh-tool-bash`), which drives the override this package honors.
 
@@ -56,5 +58,5 @@ The keyless consumer-integration proofs are `tests/bwrap.e2e.ts`, `tests/landloc
 
 - **Confinement covers file effects only** — network access and process visibility are unchanged, so the modes are not a general-purpose security sandbox.
 - **Denials are inferred from failed-command stderr** — backend signatures make the inference portable, but a matching application error can be classified as a denial and a denial omitted from the retained tail can be missed.
-- **A background runner failure has no immediate error channel** — it is recorded on the settled task and surfaces when the caller polls with `bash_output`.
+- **A background runner failure has no immediate error channel** — it is recorded on the settled process and surfaces when the caller reads the generic task with `task_output`.
 - **`danger-full-access` deliberately bypasses `ctx.sandbox`** — it is an explicit unconfined mode, not a wider sandbox profile.
diff --git a/packages/bash/bash-sandbox/src/helpers.ts b/packages/bash/bash-sandbox/src/helpers.ts
new file mode 100644
index 0000000000..a98f47216e
--- /dev/null
+++ b/packages/bash/bash-sandbox/src/helpers.ts
@@ -0,0 +1,49 @@
+/**
+ * Internal shell-quoting and sandbox-result classification helpers.
+ *
+ * @module @deepseek-ai/dsh-bash-sandbox/helpers
+ */
+
+import type { BashRunResult } from '@deepseek-ai/dsh-bash'
+
+/**
+ * Quote one string as a single-quoted POSIX shell word.
+ * @param text - raw argv element to preserve through the outer shell parse.
+ * @returns the quoted shell word.
+ */
+export function shellQuote(text: string): string {
+  return `'${text.replaceAll("'", String.raw`'\''`)}'`
+}
+
+/**
+ * Classify a failed run against the selected backend's denial dialect.
+ * @param result - settled foreground run.
+ * @param signatures - case-insensitive denial substrings from the active wrap.
+ * @returns whether the failed run matches that denial dialect.
+ */
+export function classifyDenial(result: BashRunResult, signatures: readonly string[]): boolean {
+  return matchesSignature(result.exitCode, result.stderr.text, signatures)
+}
+
+/**
+ * Classify a failed run against the selected backend's runner-failure dialect.
+ * @param result - settled foreground run.
+ * @param signatures - case-insensitive runner-failure substrings from the active wrap.
+ * @returns whether the failed run matches that runner-failure dialect.
+ */
+export function classifyRunnerFailure(result: BashRunResult, signatures: readonly string[]): boolean {
+  return matchesSignature(result.exitCode, result.stderr.text, signatures)
+}
+
+/**
+ * Match a non-zero exit against case-insensitive stderr signatures.
+ * @param exitCode - process exit code; null means signal termination.
+ * @param stderr - collected stderr text.
+ * @param signatures - substrings identifying the selected backend's dialect.
+ * @returns whether this is a non-zero exit whose stderr matches a signature.
+ */
+export function matchesSignature(exitCode: number | null, stderr: string, signatures: readonly string[]): boolean {
+  if (exitCode === null || exitCode === 0) return false
+  const lowered = stderr.toLowerCase()
+  return signatures.some(signature => lowered.includes(signature.toLowerCase()))
+}
diff --git a/packages/bash/bash-sandbox/src/index.ts b/packages/bash/bash-sandbox/src/index.ts
index 3b8837a7b6..124f6b6a9f 100644
--- a/packages/bash/bash-sandbox/src/index.ts
+++ b/packages/bash/bash-sandbox/src/index.ts
@@ -3,24 +3,25 @@
  * `ctx.sandbox`, inherits local process mechanics, and reports the selected
  * mode, enforcement, and denial facts. Runner failure means the command never
  * ran: foreground calls throw `SANDBOX_UNAVAILABLE`, while settled background
- * tasks carry `runnerFailed`. The tool owns approval and passes per-call modes.
+ * processes carry `runnerFailed`. The tool owns approval and passes per-call modes.
  * @module @deepseek-ai/dsh-bash-sandbox
  */
 
 import { resolve } from 'node:path'
 import { Context } from 'cordis'
 import z from 'schemastery'
-import type { BashExecRequest, BashExecSpec, BashRunResult, BashTask, BashTaskId } from '@deepseek-ai/dsh-bash'
+import type { BashExecRequest, BashExecSpec, BashProcess, BashRunResult } from '@deepseek-ai/dsh-bash'
 import { SandboxUnavailableError } from '@deepseek-ai/dsh-sandbox'
 import type { ConfinedSandboxMode, SandboxEnforcement, SandboxMode } from '@deepseek-ai/dsh-sandbox'
 import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
 import type { Config as LocalConfig } from '@deepseek-ai/dsh-bash-local'
+import { classifyDenial, classifyRunnerFailure, matchesSignature, shellQuote } from './helpers.ts'
 
 /**
  * Plugin config: the local executor's knobs plus the sandbox policy. All
  * optional — `static Config` supplies the defaults (`mode: 'read-only'` is the
  * fail-safe default; an example that wants a workspace-writable agent opts in
- * explicitly). The runner choice is NOT configured here: which platform
+ * explicitly). The runner choice is not configured here: which platform
  * backend confines the command is the `ctx.sandbox` provider's config.
  */
 export interface Config extends LocalConfig {
@@ -33,54 +34,6 @@ export interface Config extends LocalConfig {
   workspaceRoot?: string
 }
 
-/**
- * Quote one string as a single-quoted POSIX shell word (embedded single
- * quotes become `'\''`), so a wrapped argv element survives the outer
- * `bash -c` re-parse byte-for-byte.
- * @param text - the raw argv element to quote.
- * @returns the single-quoted shell word.
- */
-export function shellQuote(text: string): string {
-  return `'${text.replaceAll("'", String.raw`'\''`)}'`
-}
-
-/**
- * Conservatively classify a nonzero, non-signal run using only the selected
- * backend's denial signatures. Text inference may miss a denial or match
- * unrelated stderr in that dialect; it never uses another backend's terms.
- * @param result - the settled foreground run to classify.
- * @param signatures - the active wrap's denial dialect, case-insensitive stderr substrings.
- * @returns whether the run's failure reads as a sandbox denial.
- */
-export function classifyDenial(result: BashRunResult, signatures: readonly string[]): boolean {
-  return matchesSignature(result.exitCode, result.stderr.text, signatures)
-}
-
-/**
- * Classify a nonzero run using the selected backend's runner-failure
- * signatures. Callers check this before denial because runner diagnostics may
- * contain denial words; the command did not run.
- * @param result - the settled foreground run to classify.
- * @param signatures - the active wrap's runner-failure signatures,
- *   case-insensitive stderr substrings.
- * @returns whether the run's failure reads as the runner itself failing.
- */
-export function classifyRunnerFailure(result: BashRunResult, signatures: readonly string[]): boolean {
-  return matchesSignature(result.exitCode, result.stderr.text, signatures)
-}
-
-/**
- * The classifier core shared by foreground results and settled background
- * tasks: failed AND signature present. Lowercases BOTH sides — the seam
- * declares its signatures case-insensitive, and producers compose them from
- * runtime data of any case (an `argv0` path, `No such file or directory`).
- */
-function matchesSignature(exitCode: number | null, stderr: string, signatures: readonly string[]): boolean {
-  if (exitCode === null || exitCode === 0) return false
-  const lowered = stderr.toLowerCase()
-  return signatures.some(signature => lowered.includes(signature.toLowerCase()))
-}
-
 /**
  * Registers as `ctx.bash` in place of the local executor and requires a
  * `ctx.sandbox` provider; the tool layer is unchanged. The configured mode is
@@ -104,11 +57,12 @@ export class SandboxBashExecutor extends LocalBashExecutor {
   private readonly mode: SandboxMode
   private readonly workspaceRoot: string
   /**
-   * Per-task mode and wrap facts retained until settlement. Overlapping tasks
-   * may use different modes or provider facts, so one latest-wrap field would
-   * misclassify earlier completions.
+   * Per-process confinement facts retained until settlement. Providers may
+   * vary enforcement and diagnostic dialect between overlapping calls, so a
+   * shared latest-wrap value would classify a process against the wrong facts.
+   * Unconfined processes have no entry.
    */
-  private readonly taskFacts = new Map<BashTaskId, {
+  private readonly processFacts = new Map<BashProcess, {
     mode: ConfinedSandboxMode
     enforcement: SandboxEnforcement
     denialSignatures: readonly string[]
@@ -117,10 +71,7 @@ export class SandboxBashExecutor extends LocalBashExecutor {
 
   constructor(ctx: Context, config: Config) {
     super(ctx, config)
-    // schemastery (static Config) already filled the defaulted fields — the
-    // cast records that runtime fact (mirrors LocalBashExecutor's config
-    // cast). `workspaceRoot` and `cwd` have NO schema default, so their
-    // fallback chain is real branching.
+    // Schemastery fills mode before construction; workspaceRoot and cwd retain runtime fallbacks.
     this.mode = config.mode as SandboxMode
     this.workspaceRoot = resolve(config.workspaceRoot ?? config.cwd ?? process.cwd())
   }
@@ -158,39 +109,36 @@ export class SandboxBashExecutor extends LocalBashExecutor {
     return { ...result, sandbox: { mode, denied: classifyDenial(result, confined.denialSignatures), enforcement: confined.enforcement } }
   }
 
-  override start(spec: BashExecSpec): BashTask {
+  override start(spec: BashExecSpec): BashProcess {
     // Same stamped-by-resolve invariant as run().
     const mode = spec.sandboxMode as SandboxMode
     if (mode === 'danger-full-access') return super.start(spec)
-    // Classification needs settled stderr. Store facts synchronously after
-    // spawn, before the earliest process completion can be observed.
+    // Install facts synchronously; promise settlement cannot run before start() returns.
     const confined = this.confine(spec.command, mode)
-    const task = super.start({ ...spec, command: confined.command })
+    const proc = super.start({ ...spec, command: confined.command })
     const { enforcement, denialSignatures, runnerFailureSignatures } = confined
-    this.taskFacts.set(task.id, { mode, enforcement, denialSignatures, runnerFailureSignatures })
-    return task
+    this.processFacts.set(proc, { mode, enforcement, denialSignatures, runnerFailureSignatures })
+    return proc
   }
 
   /**
-   * Stamp per-task sandbox facts before completion listeners and `done` settle.
-   * Full-access tasks have no facts; signal deaths are not denials.
+   * Stamp per-process sandbox facts before `done` settles. Full-access processes
+   * have no facts; signal deaths are not denials.
    */
-  protected override notifyTaskDone(task: BashTask): void {
-    const facts = this.taskFacts.get(task.id)
+  protected override onProcessDone(proc: BashProcess, stderr: string): void {
+    const facts = this.processFacts.get(proc)
     if (facts !== undefined) {
-      this.taskFacts.delete(task.id)
-      const stderr = this.collectedStderr(task.id)
-      // Runner failure outranks denial. Background settlement has no throw
-      // channel, so this fact is its counterpart to the foreground exception.
-      const runnerFailed = matchesSignature(task.exitCode, stderr, facts.runnerFailureSignatures)
-      task.sandbox = {
+      this.processFacts.delete(proc)
+      // Runner failure outranks denial because its diagnostics may contain denial terms.
+      const runnerFailed = matchesSignature(proc.exitCode, stderr, facts.runnerFailureSignatures)
+      proc.sandbox = {
         mode: facts.mode,
-        denied: !runnerFailed && matchesSignature(task.exitCode, stderr, facts.denialSignatures),
+        denied: !runnerFailed && matchesSignature(proc.exitCode, stderr, facts.denialSignatures),
         enforcement: facts.enforcement,
         ...(runnerFailed ? { runnerFailed } : {}),
       }
     }
-    super.notifyTaskDone(task)
+    super.onProcessDone(proc, stderr)
   }
 
   /**
diff --git a/packages/bash/bash-sandbox/tests/bwrap.e2e.ts b/packages/bash/bash-sandbox/tests/bwrap.e2e.ts
index 9a5da8fdb4..237619ae06 100644
--- a/packages/bash/bash-sandbox/tests/bwrap.e2e.ts
+++ b/packages/bash/bash-sandbox/tests/bwrap.e2e.ts
@@ -5,7 +5,8 @@ import { homedir } from 'node:os'
 import { join } from 'node:path'
 import { afterEach, describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import { bwrapProfileArgs, LocalSandboxProvider } from '@deepseek-ai/dsh-sandbox-local'
+import { LocalSandboxProvider } from '@deepseek-ai/dsh-sandbox-local'
+import { bwrapProfileArgs } from '@deepseek-ai/dsh-sandbox-local/src/profiles.ts'
 import { SandboxBashExecutor } from '@deepseek-ai/dsh-bash-sandbox'
 
 /**
diff --git a/packages/bash/bash-sandbox/tests/sandbox.spec.ts b/packages/bash/bash-sandbox/tests/sandbox.spec.ts
index 63fea52bd6..619bc05151 100644
--- a/packages/bash/bash-sandbox/tests/sandbox.spec.ts
+++ b/packages/bash/bash-sandbox/tests/sandbox.spec.ts
@@ -13,7 +13,8 @@ import { Context } from 'cordis'
 import type { BashRunResult, CollectedOutput } from '@deepseek-ai/dsh-bash'
 import { SANDBOX_UNAVAILABLE, SandboxProvider, SandboxUnavailableError } from '@deepseek-ai/dsh-sandbox'
 import type { ConfinedArgv, SandboxPolicy } from '@deepseek-ai/dsh-sandbox'
-import { classifyDenial, classifyRunnerFailure, SandboxBashExecutor, shellQuote } from '@deepseek-ai/dsh-bash-sandbox'
+import { SandboxBashExecutor } from '@deepseek-ai/dsh-bash-sandbox'
+import { classifyDenial, classifyRunnerFailure, shellQuote } from '../src/helpers.ts'
 import type { Config } from '@deepseek-ai/dsh-bash-sandbox'
 
 const spillDir = mkdtempSync(join(tmpdir(), 'dsh-bash-sandbox-spec-'))
@@ -132,7 +133,7 @@ describe('danger-full-access', () => {
     const task = bash.start(bash.resolve({ command: 'echo free-bg' }))
     await task.done
     expect(task.sandbox).toBeUndefined()
-    expect(bash.readOutput(task.id).delta).toContain('free-bg')
+    expect(task.readOutput().delta).toContain('free-bg')
     expect(calls).toHaveLength(0)
   })
 })
@@ -184,7 +185,7 @@ describe('per-call sandboxMode override (the escalation mechanism)', () => {
     const task = bash.start(bash.resolve({ command: 'echo bg-free', sandboxMode: 'danger-full-access' }))
     await task.done
     expect(task.sandbox).toBeUndefined()
-    expect(bash.readOutput(task.id).delta).toContain('bg-free')
+    expect(task.readOutput().delta).toContain('bg-free')
     expect(calls).toHaveLength(0)
   })
 })
@@ -243,6 +244,20 @@ describe('result facts', () => {
 })
 
 describe('background sandbox facts', () => {
+  it('stamps facts and releases accounting when background spawn fails', async () => {
+    const { bash } = await setup()
+    const missingWorkdir = join(mkdtempSync(join(tmpdir(), 'dsh-sandbox-missing-cwd-')), 'missing')
+    const task = bash.start(bash.resolve({ command: 'true', workdir: missingWorkdir }))
+
+    await task.done
+
+    expect(task.status).toBe('killed')
+    expect(task.readOutput().delta).toContain('spawn failed:')
+    expect(task.sandbox).toEqual({ mode: 'read-only', denied: false, enforcement: 'full' })
+    const accounting = (bash as unknown as { processFacts: Map<unknown, unknown> }).processFacts
+    expect(accounting.size).toBe(0)
+  })
+
   it('stamps a settled denial: nonzero exit + permission stderr under a confined mode', async () => {
     const { bash } = await setup()
     const task = bash.start(bash.resolve({ command: 'echo "x: Permission denied" >&2; exit 1' }))
@@ -273,15 +288,6 @@ describe('background sandbox facts', () => {
     expect(task.sandbox).toEqual({ mode: 'read-only', denied: false, enforcement: 'full', runnerFailed: true })
   })
 
-  it('completion listeners already see the stamped facts (stamp precedes notify)', async () => {
-    const { ctx, bash } = await setup({}, argv => ({ argv: [...argv], enforcement: 'partial', denialSignatures: UNIX_SIGNATURES, runnerFailureSignatures: RUNNER_FAILURE }))
-    const seen: unknown[] = []
-    ctx.bash.onTaskDone((task) => { seen.push(task.sandbox) })
-    const task = bash.start(bash.resolve({ command: 'echo "x: Permission denied" >&2; exit 1' }))
-    await task.done
-    expect(seen).toEqual([{ mode: 'read-only', denied: true, enforcement: 'partial' }])
-  })
-
   it('overlapping background tasks keep their OWN wrap facts (per-task, not latest-wrap)', async () => {
     // Facts belong to each wrap and may vary between calls. The slow task settles after the
     // quick task starts; a shared latest-wrap field would classify and stamp it with the wrong
@@ -308,8 +314,8 @@ describe('background sandbox facts', () => {
     const task = bash.start(bash.resolve({ command: 'echo "Permission denied" >&2; sleep 30' }))
     // Let the stderr land before the kill so the classifier sees the
     // signature and must still refuse it on the null exit code alone.
-    await vi.waitFor(() => { expect(bash.readOutput(task.id).delta).toContain('Permission denied') })
-    bash.kill(task.id)
+    await vi.waitFor(() => { expect(task.readOutput().delta).toContain('Permission denied') })
+    task.kill()
     await task.done
     expect(task.sandbox).toEqual({ mode: 'read-only', denied: false, enforcement: 'full' })
   })
diff --git a/packages/bash/bash-sandbox/tests/seatbelt.e2e.ts b/packages/bash/bash-sandbox/tests/seatbelt.e2e.ts
index 7cc58202fc..cc3703d742 100644
--- a/packages/bash/bash-sandbox/tests/seatbelt.e2e.ts
+++ b/packages/bash/bash-sandbox/tests/seatbelt.e2e.ts
@@ -5,7 +5,8 @@ import { homedir } from 'node:os'
 import { join } from 'node:path'
 import { afterEach, describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import { LocalSandboxProvider, seatbeltProfileArgs } from '@deepseek-ai/dsh-sandbox-local'
+import { LocalSandboxProvider } from '@deepseek-ai/dsh-sandbox-local'
+import { seatbeltProfileArgs } from '@deepseek-ai/dsh-sandbox-local/src/profiles.ts'
 import { SandboxBashExecutor } from '@deepseek-ai/dsh-bash-sandbox'
 
 /**
diff --git a/packages/bash/bash/README.md b/packages/bash/bash/README.md
index 540eef700c..e4b5bf1952 100644
--- a/packages/bash/bash/README.md
+++ b/packages/bash/bash/README.md
@@ -1,6 +1,6 @@
 # @deepseek-ai/dsh-bash
 
-The **bash executor seam**: an abstract `BashExecutor` service (`ctx.bash`) defining WHAT a bash backend does — run commands, manage background tasks — without saying HOW.
+The **bash executor seam**: an abstract `BashExecutor` service (`ctx.bash`) defining WHAT a bash backend does — run foreground commands and start background processes — without saying HOW. Task ids, ownership, collection, cancellation, and notices belong to the generic `ctx.tasks` runtime.
 
 This package is the interface quarter of the bash capability, split so each concern can evolve (and be swapped) independently:
 
@@ -18,23 +18,20 @@ The split mirrors the LLM seam (`LlmService`/`LlmAdapter`) and the agent-tool su
 | Member | Semantics |
 |---|---|
 | `run(spec)` | Foreground execution. Resolves when the command finishes. **Rejects only for infrastructure failures** (unusable workdir, missing shell, pre-aborted signal); nonzero exits, timeout kills, and abort kills resolve with a descriptive `BashRunResult`. |
-| `start(spec)` | Background execution. Returns a `BashTask` handle immediately; **no timeout applies** (stop tasks via `kill`). |
-| `get(id)` / `list()` | Task lookup. |
+| `start(spec)` | Background execution. Returns a task-free `BashProcess` handle immediately; **no timeout applies**. The caller may adapt it into `ctx.tasks`. |
 | `sandboxMode` | The capability fact for the tool layer: the default mode a SANDBOXING executor confines under (`undefined` in the base class — "this executor does not sandbox"). `dsh-tool-bash` reads it at registration to advertise the escalation fields only when the composition honors them. |
-| `ownerOf(id)` | The opaque OWNER token recorded for a background task at `start` (from the spec's `owner`), or `undefined` for an unknown id OR a known-but-ownerless task. The executor stores/returns it verbatim and NEVER interprets it — the access POLICY lives in the consumer (`dsh-tool-bash`), which compares `ownerOf(id)` to the caller's token. Storing ownership here (disposed with the executor's fiber) is what makes it survive a consumer HMR reload. |
-| `readOutput(id)` | **Incremental** output read — consecutive reads never re-deliver. Reads that lost data to buffer bounds flag `lossy` and point at full-stream spill files. Throws for unknown ids. |
-| `kill(id)` | Kill a running task. Returns `false` when it already finished; throws for unknown ids. |
-| `onTaskDone(listener)` | Completion listener (effect-based, disposer returned). Fires exactly once per task; never after the service is disposed. |
+| `BashProcess.readOutput()` | **Incremental** output read — consecutive reads never re-deliver. Reads that lost data to buffer bounds flag `lossy` and point at full-stream spill files. |
+| `BashProcess.kill()` | Kill the process group. Returns `false` when it already finished. |
 
-Implementations subclass `BashExecutor`, implement the abstract methods, and call `notifyTaskDone(task)` on background completion. Disposal must kill every running task (no orphan processes) — see the HMR-safety tests.
+Implementations subclass `BashExecutor` and implement the abstract methods. Disposal must kill every running process and await its exit — see the HMR-safety tests.
 
 ## Vocabulary
 
-`BashExecRequest` (command, workdir?, timeoutMs?, signal?, stdin?, env?, owner?, sandboxMode?) resolves to `BashExecSpec` (command, workdir, timeoutMs, signal?, stdin?, env?, owner, sandboxMode) before execution; `owner` and `sandboxMode` are optional on the request and **required-but-nullable** on the resolved spec, so a forgotten one is a visible `undefined` rather than a silently-absent property. `sandboxMode` is the explicit per-call sandbox-policy input: an escalation grant a human just issued ([the sandbox RFC § Escalation](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md), which outranks) or the session's standing override ([the sandbox RFC § Per-session mode switching](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md)); a sandboxing executor's `resolve()` stamps its configured default when the request carries none, and a non-sandboxing executor carries the field verbatim and confines nothing.
+`BashExecRequest` (command, workdir?, timeoutMs?, stdoutMaxBytes?, signal?, stdin?, env?, dshEnv?, sandboxMode?) resolves to `BashExecSpec` (command, workdir, timeoutMs, stdoutMaxBytes, signal?, stdin?, env?, dshEnv?, sandboxMode) before execution. `stdoutMaxBytes` is a trusted foreground-run capture budget for consumers that must parse complete bounded stdout; the model-facing bash tool does not expose it. `sandboxMode` is optional on the request and required-but-nullable on the resolved spec: it carries an approved one-shot escalation or the session's standing override; a sandboxing executor stamps its configured default when absent, while a non-sandboxing executor carries the field and confines nothing.
 
-The seam owns per-session sandbox overrides through the log-only `bash/sandbox-mode` event, `effectiveSandboxMode`, and `setSandboxMode`; writers preserve turn enclosure, and replay restores the last override. `BashTaskId` and `OwnerToken` are distinct brands. Foreground `run` returns exit, timeout, cancellation, output, and optional sandbox facts; background `start` and `readOutput` use task records. A sandboxing executor reports the executed mode, conservative denial classification, and enforcement completeness. See [core-data-structures/bash.md](../../../docs/core-data-structures/bash.md) for full shapes.
+The seam also owns the per-session mode override vocabulary: the log-only `'bash/sandbox-mode'` session event, the pure `effectiveSandboxMode(events)` fold, and the `setSandboxMode(session, mode)` write path. `run()` returns `BashRunResult`; `start()` returns `BashProcess`, whose incremental read and kill methods are adapted by `dsh-tool-bash` into a generic task registration. A sandboxing executor stamps `BashSandboxInfo` on foreground results and settled process handles. See `src/types.ts` and [core-data-structures/bash.md](../../../docs/core-data-structures/bash.md).
 
-`stdin` and `env` are set by in-process plugins (the hooks bridges, native plugins) to feed a hook command its JSON payload on stdin and its `CLAUDE_PROJECT_DIR`/`CLAUDE_PLUGIN_ROOT` env. The model-facing `dsh-tool-bash` tool does not expose them as parameters — a model already has equivalent power through shell syntax (`FOO=bar cmd`, a heredoc), so they would be redundant tool params. This is not a security boundary: the implementation's credential scrub (not these fields) is what keeps the harness's ambient secrets out of a spawned command. They are plain optionals on the resolved spec (unlike `owner`'s required-but-nullable): a missing one means "none", the safe default. See [the bash-stdin-env RFC](../../../docs/rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md).
+`stdin` and ordinary `env` are set by in-process plugins (the hooks bridges, native plugins) to feed a hook command its JSON payload and `CLAUDE_PROJECT_DIR`/`CLAUDE_PLUGIN_ROOT` values. `dshEnv` is a separate trusted overlay restricted by type to managed keys; the exported `DSH_ENV_PREFIX` is the single source for that namespace, its `DshEnvironmentKey` template type, executor scrubbing, registry validation, derived built-in names, and model guidance. Model bash uses the current snapshot collected by `ctx.bashEnv`. Implementations remove inherited managed keys, reject those names in ordinary `env`, then merge `dshEnv`, so an omitted current fact cannot fall back to stale ambient state. The model-facing tool exposes none of these as parameters. All three remain optional on the resolved spec; absent means no input/overlay. See [the bash-stdin-env RFC](../../../docs/rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md) and [the session environment RFC](../../../docs/rfc/implemented/feature/2026-07-10-agent-session-identity-and-log-location.md).
 
 ## Model Experience
 
diff --git a/packages/bash/bash/package.json b/packages/bash/bash/package.json
index 2ae0566142..0783995161 100644
--- a/packages/bash/bash/package.json
+++ b/packages/bash/bash/package.json
@@ -22,13 +22,11 @@
   ],
   "license": "BSD-3-Clause",
   "peerDependencies": {
-    "@deepseek-ai/dsh-brand": "^0.0.1",
     "@deepseek-ai/dsh-sandbox": "^0.0.1",
     "@deepseek-ai/dsh-session": "^0.0.1",
     "cordis": "^4.0.0-rc.7"
   },
   "devDependencies": {
-    "@deepseek-ai/dsh-brand": "workspace:^",
     "@deepseek-ai/dsh-sandbox": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
     "cordis": "^4.0.0-rc.7"
diff --git a/packages/bash/bash/src/index.ts b/packages/bash/bash/src/index.ts
index 67e36385f0..75a5f7f230 100644
--- a/packages/bash/bash/src/index.ts
+++ b/packages/bash/bash/src/index.ts
@@ -1,25 +1,27 @@
 /**
- * The bash executor seam (`ctx.bash`): an abstract service defining what a bash backend does —
- * run commands, manage background tasks — without saying how.
+ * The `ctx.bash` executor seam for foreground commands and background process
+ * handles. Task ids, ownership, polling, and notices belong to
+ * `@deepseek-ai/dsh-tasks`, keeping executors independent of sessions.
  * @module @deepseek-ai/dsh-bash
  */
 
 import { Context, Service } from 'cordis'
 import type { SandboxMode } from '@deepseek-ai/dsh-sandbox'
-import type { BashExecRequest, BashExecSpec, BashRunResult, BashTask, BashTaskId, BashTaskListener, BashTaskRead, OwnerToken } from './types.ts'
+import type { BashExecRequest, BashExecSpec, BashProcess, BashRunResult } from './types.ts'
 
-export { BashTaskId, OwnerToken } from './types.ts'
+export { DSH_ENV_PREFIX } from './types.ts'
 export { SANDBOX_MODES, effectiveSandboxMode, setSandboxMode } from './session-mode.ts'
 export type {
   BashExecRequest,
   BashExecSpec,
+  BashProcess,
+  BashProcessRead,
+  BashProcessStatus,
   BashRunResult,
   BashSandboxInfo,
-  BashTask,
-  BashTaskListener,
-  BashTaskRead,
-  BashTaskStatus,
   CollectedOutput,
+  DshEnvironment,
+  DshEnvironmentKey,
 } from './types.ts'
 
 declare module 'cordis' {
@@ -29,34 +31,30 @@ declare module 'cordis' {
 }
 
 /**
- * Registers one `ctx.bash` implementation. Runtime command failures resolve as
- * {@link BashRunResult}; only infrastructure failures reject. Background starts
- * return immediately without a timeout, report completion exactly once while
- * live, and remain cancellable by signal or {@link kill}. Output reads are
- * incremental and flag lost buffered data; disposal kills and awaits all tasks.
+ * Abstract bash execution service. Subclass, implement the abstract methods,
+ * and load the subclass as a plugin — it registers as `ctx.bash` (one
+ * implementation per context; loading a second throws, which is cordis'
+ * standard duplicate-service behavior).
+ *
+ * Implementations must honor these semantics:
+ * - {@link run} rejects only for infrastructure failures. Nonzero exits,
+ *   timeout kills, and abort kills resolve with a {@link BashRunResult}.
+ * - {@link start} returns immediately; no timeout applies to background
+ *   processes. `done` settles at process close and never rejects; spawn
+ *   failures settle as `killed` with the error on stderr.
+ * - {@link BashProcess.readOutput} is incremental: consecutive reads never
+ *   repeat output. Lossy reads report truncation and available spill files.
+ * - Disposal kills all running background processes and awaits their exit.
  */
 export abstract class BashExecutor extends Service {
-  private listeners = new Set<BashTaskListener>()
-  private listenersClosed = false
-
   constructor(ctx: Context) {
     super(ctx, 'bash')
-    ctx.effect(() => () => {
-      // Close the listener registry before subclass teardown so late task
-      // completions (e.g. from kills issued during dispose) stay silent.
-      this.listenersClosed = true
-      this.listeners.clear()
-    }, 'bash listener teardown')
   }
 
   /**
-   * The sandbox mode this executor confines commands under BY DEFAULT, or `undefined` when it
-   * does not sandbox at all — the capability fact the tool and ACP layers read to advertise
-   * sandbox controls honestly.
-   * A session or call may override this default, so widening is evaluated per
-   * execution rather than encoded in this getter.
-   * @returns the configured default mode of a sandboxing executor;
-   *   `undefined` for an executor that never confines.
+   * The sandbox mode this executor applies by default, or `undefined` when it
+   * does not sandbox commands.
+   * @returns the configured default sandbox mode, when supported.
    */
   get sandboxMode(): SandboxMode | undefined {
     return undefined
@@ -79,79 +77,11 @@ export abstract class BashExecutor extends Service {
   abstract run(spec: BashExecSpec): Promise<BashRunResult>
 
   /**
-   * Start a background task and return its handle immediately.
+   * Start a background process and return its handle immediately.
    * @param spec - a resolved spec from {@link resolve}, never a raw request.
-   * @returns the live task handle; completion fires {@link onTaskDone}.
+   * @returns the live process handle (reads, kill, quiescence promise).
    */
-  abstract start(spec: BashExecSpec): BashTask
-
-  /**
-   * Look up a background task by id.
-   * @param id - the task id to look up.
-   * @returns the tracked task, or undefined for an id this executor never issued.
-   */
-  abstract get(id: BashTaskId): BashTask | undefined
-
-  /**
-   * The opaque OWNER token recorded for a background task at {@link start} (from the {@link
-   * BashExecSpec}'s `owner`), or `undefined` for an unknown id OR a known-but-ownerless task.
-   * The executor stores the token without interpreting policy; keeping it here
-   * lets ownership survive a consumer-plugin reload.
-   * @param id - the background task id to look up ownership for.
-   * @returns the token recorded at start, verbatim; undefined for an unknown
-   *   id or a known-but-ownerless task.
-   */
-  abstract ownerOf(id: BashTaskId): OwnerToken | undefined
-
-  /**
-   * All tracked background tasks (insertion order).
-   * @returns every task this executor started, running or finished.
-   */
-  abstract list(): BashTask[]
-
-  /**
-   * Read output produced since the previous read. Throws for unknown ids.
-   * @param id - the task to read from.
-   * @returns the incremental read; consecutive reads never re-deliver output.
-   */
-  abstract readOutput(id: BashTaskId): BashTaskRead
-
-  /**
-   * Kill a running background task. Returns false when it had already
-   * finished (no-op). Throws for unknown ids.
-   * @param id - the task to kill.
-   * @returns true when this call killed it, false when it had already finished.
-   */
-  abstract kill(id: BashTaskId): boolean
-
-  /**
-   * Register a background-task completion listener (disposed with the
-   * calling fiber). Listeners never fire after this service is disposed.
-   * @param listener - called exactly once per task completion.
-   * @returns the disposer that unregisters the listener.
-   */
-  onTaskDone(listener: BashTaskListener): () => void {
-    const dispose = this.ctx.effect(() => {
-      this.listeners.add(listener)
-      return () => this.listeners.delete(listener)
-    }, 'bash.onTaskDone()')
-    return () => void dispose()
-  }
-
-  /** For implementations: notify listeners that `task` completed. Listener
-   * exceptions are contained (logged) — one bad listener must not reject
-   * `BashTask.done` or starve the listeners after it. */
-  protected notifyTaskDone(task: BashTask): void {
-    if (this.listenersClosed) return
-    for (const listener of this.listeners) {
-      try {
-        listener(task)
-      } catch (error: unknown) {
-        // Listener bugs are reported, never propagated into task.done.
-        console.error('bash onTaskDone listener threw:', error)
-      }
-    }
-  }
+  abstract start(spec: BashExecSpec): BashProcess
 }
 
 export default BashExecutor
diff --git a/packages/bash/bash/src/types.ts b/packages/bash/bash/src/types.ts
index 9fe9c2efed..6e0ddca91e 100644
--- a/packages/bash/bash/src/types.ts
+++ b/packages/bash/bash/src/types.ts
@@ -1,79 +1,33 @@
 /**
- * Execution vocabulary for the bash executor seam. Types only — the abstract
- * service lives in `./index.ts`, implementations in sibling packages
- * (`@deepseek-ai/dsh-bash-local` first).
- *
+ * Execution types for the bash executor seam. Background task semantics belong
+ * to `@deepseek-ai/dsh-tasks`; this seam exposes only process handles.
  * @module dsh-bash/types
  */
 
-import type { Branded } from '@deepseek-ai/dsh-brand'
 import type { SandboxEnforcement, SandboxMode } from '@deepseek-ai/dsh-sandbox'
 
-/** Identifies one background task within an executor (generated `bash-N`). */
-export type BashTaskId = Branded<'BashTaskId'>
+/** Namespace prefix reserved for DeepSeek Harness-managed child environment facts. */
+export const DSH_ENV_PREFIX = 'DSH_' as const
+
+/** One environment key inside the managed {@link DSH_ENV_PREFIX} namespace. */
+export type DshEnvironmentKey = `${typeof DSH_ENV_PREFIX}${string}`
+
+/** Trusted DeepSeek Harness variables for one bash execution. */
+export type DshEnvironment = Readonly<Record<DshEnvironmentKey, string>>
 
 /**
- * Brand a string as a {@link BashTaskId}.
- * @param id - the raw task-id string (the executor generates `bash-N`).
- * @returns the same string, branded; no validation is performed.
- */
-export function BashTaskId(id: string): BashTaskId {
-  return id as BashTaskId
-}
-
-/**
- * A background task's opaque isolation key — the CONSUMER's owner identity, not
- * the bash seam's. The executor stores and returns it verbatim and never
- * interprets it; the access policy lives in the consumer (`dsh-tool-bash`),
- * which is the single boundary that casts its own id vocabulary into one. A
- * DISTINCT brand (not a `SessionId` alias) keeps the seam decoupled — a
- * sandboxed/remote executor inherits no session dependency.
- */
-export type OwnerToken = Branded<'OwnerToken'>
-
-/**
- * Brand a string as an {@link OwnerToken}. Only the consuming boundary
- * (`dsh-tool-bash`) should cast its own id vocabulary in — see the type's doc.
- * @param id - the consumer's raw owner identity (the tool layer passes the owning agent's session id).
- * @returns the same string, branded; no validation is performed.
- */
-export function OwnerToken(id: string): OwnerToken {
-  return id as OwnerToken
-}
-
-/**
- * Sandbox facts for one foreground run — present on {@link BashRunResult} iff
- * a sandboxing executor ran the command (an unsandboxed executor reports no
- * `sandbox` field at all). Reported independently of `exitCode`/`signal`
- * (orthogonal outcomes), so a caller can tell "the command failed on its own"
- * from "the sandbox blocked a file operation". The mode/enforcement
- * vocabulary lives on the `@deepseek-ai/dsh-sandbox` seam; this shape is the
- * bash seam's result-fact carrier for it.
+ * Sandbox facts for one run, present iff a sandboxing executor handled it.
+ * Facts are reported independently of process exit status so callers can
+ * distinguish command failures from policy denials and runner failures.
  */
 export interface BashSandboxInfo {
   /** The mode the command actually ran under. */
   mode: SandboxMode
-  /**
-   * True when the executor classifies this run's failure as the sandbox
-   * denying a file operation. The classification is CONSERVATIVE (a failed
-   * exit whose stderr carries a filesystem-permission signature) and reads
-   * the COLLECTED stderr — the bounded in-memory tail per
-   * {@link CollectedOutput} semantics, so a signature that survives only in a
-   * spill file is missed toward `denied: false`. A plain command failure
-   * keeps `denied: false` even under a sandboxed mode.
-   */
+  /** Whether the sandbox denied a file operation. */
   denied: boolean
-  /**
-   * How completely the runner enforced `mode`'s file effects — see
-   * {@link SandboxEnforcement}. Absent exactly when `mode` is
-   * `danger-full-access`: nothing is confined, so there is no enforcement to
-   * report.
-   */
+  /** How completely the selected runner enforced the requested mode. */
   enforcement?: SandboxEnforcement
-  /**
-   * The sandbox runner failed before executing the command. Set only on settled
-   * background tasks; foreground runs throw `SANDBOX_UNAVAILABLE` instead.
-   */
+  /** Whether the sandbox runner failed before the command could run. */
   runnerFailed?: boolean
 }
 
@@ -89,6 +43,13 @@ export interface BashExecRequest {
   workdir?: string | undefined
   /** Timeout override in milliseconds (implementations cap it). */
   timeoutMs?: number | undefined
+  /**
+   * Foreground stdout capture budget in bytes. Absent uses the executor's
+   * default output cap. Trusted in-process consumers use this when they must
+   * parse complete stdout up to their own bounded limit; the model-facing bash
+   * tool does not expose it as a parameter.
+   */
+  stdoutMaxBytes?: number | undefined
   /** Abort signal — implementations kill the command when it fires. */
   signal?: AbortSignal | undefined
   /**
@@ -100,80 +61,52 @@ export interface BashExecRequest {
    */
   stdin?: string | undefined
   /**
-   * Extra environment entries for the command, merged AFTER the
-   * implementation's credential scrub (so an explicit entry here is honored even
-   * when its name matches the scrub pattern — the caller named a value it holds,
-   * not the harness's ambient secret). Set by in-process plugins (the hooks
-   * bridges set `CLAUDE_PROJECT_DIR`, `CLAUDE_PLUGIN_ROOT`, …); the model-facing
-   * bash tool does not expose it as a parameter (a model that needs an env var
-   * uses shell syntax like `FOO=bar cmd`).
+   * Ordinary environment entries for the command, merged after the credential
+   * scrub. `DSH_*` is reserved for {@link dshEnv} and implementations reject it
+   * here. Set by in-process plugins (the hooks bridges set
+   * `CLAUDE_PROJECT_DIR`, `CLAUDE_PLUGIN_ROOT`, …); the model-facing bash tool
+   * does not expose it as a parameter.
    */
   env?: Record<string, string> | undefined
   /**
-   * Opaque OWNER token for a background task — the consumer's isolation key
-   * (the tool layer passes the owning agent's `session.header.id`). The
-   * executor stores it on the task and exposes it via {@link BashExecutor.ownerOf};
-   * the executor itself NEVER interprets it (no access policy lives in the
-   * seam — that is the consumer's job). Absent for foreground runs and for an
-   * ownerless background start (a non-agent caller).
-   */
-  owner?: OwnerToken | undefined
-  /**
-   * Explicit per-call sandbox policy. The tool stamps a session override or a
-   * one-shot approved escalation, with the grant taking precedence. Sandboxing
-   * executors honor it for this call; non-sandboxing executors do not confine.
+   * Harness-owned `DSH_*` variables for this execution. Executors discard
+   * ambient `DSH_*` entries before merging this snapshot, so an unavailable
+   * current fact cannot inherit a stale value from the harness process, and
+   * reject non-`DSH_*` names supplied through this managed channel.
    */
+  dshEnv?: DshEnvironment | undefined
+  /** Explicit per-call sandbox mode override. */
   sandboxMode?: SandboxMode | undefined
 }
 
 /**
- * A fully-resolved execution SPEC — exactly what {@link BashExecutor.run} /
- * {@link BashExecutor.start} act on. `workdir` and `timeoutMs` are REQUIRED:
- * defaulting and capping already happened in {@link BashExecutor.resolve}, so
- * the executor never hides a `?? config` fallback (explicit > implicit). For
- * background tasks, `start()` ignores `timeoutMs` (background runs have no
- * timeout) — the field is still required because the type is shared.
+ * A resolved execution spec. {@link BashExecutor.resolve} fills and caps the
+ * required fields; {@link BashExecutor.start} ignores `timeoutMs` because
+ * background processes have no executor timeout.
  */
 export interface BashExecSpec {
   command: string
   workdir: string
   timeoutMs: number
+  /**
+   * Resolved foreground stdout capture budget in bytes. `run()` uses it for
+   * stdout; background tasks and stderr keep the executor's own output cap.
+   */
+  stdoutMaxBytes: number
   /** Abort signal — implementations kill the command when it fires. */
   signal?: AbortSignal | undefined
-  /**
-   * Bytes to write to the command's stdin (then close it), carried through
-   * verbatim from {@link BashExecRequest.stdin}. OPTIONAL on the resolved spec
-   * (unlike `owner`): it has no config default, so a missing one means "no
-   * stdin" — the safe, ordinary case — not a silent footgun, so it stays a
-   * plain optional rather than required-but-nullable (see the request field).
-   */
+  /** Bytes to write to stdin before closing it; absent means no stdin. */
   stdin?: string | undefined
   /**
-   * Extra environment entries, carried through verbatim from
-   * {@link BashExecRequest.env} and merged by the implementation AFTER its
-   * credential scrub (an explicit entry wins even when its name matches the
-   * scrub pattern). OPTIONAL on the spec for the same reason as `stdin` — no
-   * config default, absent means "no extra env".
+   * Ordinary environment entries carried through from
+   * {@link BashExecRequest.env}. `DSH_*` remains reserved for {@link dshEnv}.
+   * OPTIONAL on the spec for the same reason as `stdin`: absent means no
+   * ordinary extra environment.
    */
   env?: Record<string, string> | undefined
-  /**
-   * Opaque owner token, REQUIRED-but-nullable (mirrors `workdir`/`timeoutMs`
-   * being required on the resolved spec): {@link BashExecutor.resolve} carries
-   * the request's `owner` through, defaulting a missing one to `undefined`. A
-   * required field makes a forgotten owner a VISIBLE `undefined` rather than a
-   * silently-absent property that yields an unowned (cross-session-readable)
-   * task. `start()` stores it; `run()` (foreground) ignores it.
-   */
-  owner: OwnerToken | undefined
-  /**
-   * The sandbox mode this call executes under, REQUIRED-but-nullable for the
-   * same visibility reason as `owner`. A sandboxing executor's `resolve()`
-   * stamps the effective mode (the request's explicit override, else its
-   * configured default) so `run()`/`start()` read the spec, never the config;
-   * a non-sandboxing executor carries the request value through verbatim and
-   * ignores it (`undefined` under such an executor means what its README says:
-   * unconfined execution).
-   */
+  /** Managed `DSH_*` snapshot; implementations reject ordinary names. */
+  dshEnv?: DshEnvironment | undefined
+  /** Resolved sandbox mode; ignored by executors that do not confine. */
   sandboxMode: SandboxMode | undefined
 }
 
@@ -193,50 +126,33 @@ export interface BashRunResult {
   exitCode: number | null
   /** Terminating signal (e.g. 'SIGTERM'); null on normal exit. */
   signal: NodeJS.Signals | null
-  /** True when the executor's own timeout killed the command. */
+  /**
+   * True when the executor's own timeout was the FIRST cause to cut the command
+   * short. Mutually exclusive with {@link aborted}: one fused deadline drives
+   * both the timeout and the caller's cancellation, so a timeout and an abort
+   * racing before process close report the single first-abort cause, not both
+   * (see the [timeout-library RFC](../../../../docs/rfc/implemented/architecture/2026-07-06-timeout-deadline-library.md)).
+   */
   timedOut: boolean
-  /** True when the caller's AbortSignal killed the command. */
+  /**
+   * True when the caller's `AbortSignal` was the FIRST cause to kill the command
+   * (and it was not the executor's own timeout). Mutually exclusive with
+   * {@link timedOut} — see there for the first-cause classification.
+   */
   aborted: boolean
   /** The effective timeout applied to this run (after defaulting/capping). */
   timeoutMs: number
   stdout: CollectedOutput
   stderr: CollectedOutput
-  /**
-   * Sandbox facts, present iff a sandboxing executor ran the command — an
-   * unsandboxed executor (e.g. `dsh-bash-local`) never sets it. See
-   * {@link BashSandboxInfo} for the `denied` classification semantics.
-   */
+  /** Sandbox execution facts, absent for an unsandboxed executor. */
   sandbox?: BashSandboxInfo
 }
 
-/** Lifecycle of a background task. */
-export type BashTaskStatus = 'running' | 'completed' | 'killed'
+/** Lifecycle of a background process. */
+export type BashProcessStatus = 'running' | 'completed' | 'killed'
 
-/** A tracked background task handle. */
-export interface BashTask {
-  readonly id: BashTaskId
-  status: BashTaskStatus
-  /** Exit code once finished (null = killed by signal / still running). */
-  exitCode: number | null
-  /** Terminating signal name, when signal-killed. */
-  signal: NodeJS.Signals | null
-  /** Resolves when the underlying process closes (never rejects). */
-  readonly done: Promise<void>
-  /**
-   * Sandbox facts for this task's execution, stamped by a sandboxing executor
-   * once the task settles and BEFORE completion listeners are notified — an
-   * `onTaskDone` consumer and a `done` awaiter both see it. Denial
-   * classification runs against the settled task's collected stderr, so the
-   * field cannot exist earlier: absent while the task is running and under an
-   * executor that does not sandbox. See {@link BashSandboxInfo} for the
-   * `denied` semantics.
-   */
-  sandbox?: BashSandboxInfo
-}
-
-/** One incremental {@link BashExecutor.readOutput} read. */
-export interface BashTaskRead {
-  task: BashTask
+/** One incremental {@link BashProcess.readOutput} read. */
+export interface BashProcessRead {
   /** Output produced since the previous read (stderr in a marked section). */
   delta: string
   /** True when truncation dropped unread bytes the delta cannot include. */
@@ -247,5 +163,31 @@ export interface BashTaskRead {
   stderrSpillPath?: string
 }
 
-/** Completion callback for background tasks. */
-export type BashTaskListener = (task: BashTask) => void
+/**
+ * A background process handle returned by {@link BashExecutor.start}. It is the
+ * only access path; buffered output remains readable after exit. Executor
+ * disposal kills running processes and awaits {@link done}.
+ */
+export interface BashProcess {
+  /** Process lifecycle state (settled exactly once). */
+  status: BashProcessStatus
+  /** Exit code once finished (null = killed by signal / still running). */
+  exitCode: number | null
+  /** Terminating signal name, when signal-killed. */
+  signal: NodeJS.Signals | null
+  /** Resolves when the underlying process closes (never rejects — a spawn failure settles as `killed` with the error on stderr). */
+  readonly done: Promise<void>
+  /** Sandbox facts, stamped once a confined process settles. */
+  sandbox?: BashSandboxInfo
+  /**
+   * Read output produced since the previous read (consuming — consecutive
+   * reads never re-deliver). Reads that lost data flag `lossy` and point at
+   * full-stream spill files when available.
+   */
+  readOutput(): BashProcessRead
+  /**
+   * Kill the process group. Returns false when it had already finished
+   * (no-op); idempotent.
+   */
+  kill(): boolean
+}
diff --git a/packages/bash/bash/tests/service.spec.ts b/packages/bash/bash/tests/service.spec.ts
index bdee15aedd..63d9533410 100644
--- a/packages/bash/bash/tests/service.spec.ts
+++ b/packages/bash/bash/tests/service.spec.ts
@@ -1,150 +1,84 @@
-import { describe, expect, it, vi } from 'vitest'
+import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import { BashExecutor, BashTaskId, OwnerToken } from '@deepseek-ai/dsh-bash'
-import type { BashExecRequest, BashExecSpec, BashRunResult, BashTask, BashTaskRead } from '@deepseek-ai/dsh-bash'
+import { BashExecutor } from '@deepseek-ai/dsh-bash'
+import type { BashExecRequest, BashExecSpec, BashProcess, BashProcessRead, BashRunResult } from '@deepseek-ai/dsh-bash'
 
-/** Minimal concrete executor: records calls, lets tests drive completions. */
+/**
+ * Minimal concrete executor: canned foreground results, a hand-built process
+ * handle. The seam is TASK-FREE (start returns a {@link BashProcess} handle;
+ * task semantics live in `ctx.tasks`), so this stub is all an implementation
+ * owes the abstract class.
+ */
 class StubExecutor extends BashExecutor {
-  tasks = new Map<BashTaskId, BashTask>()
-  private owners = new Map<BashTaskId, OwnerToken | undefined>()
-
   resolve(request: BashExecRequest): BashExecSpec {
     return {
       command: request.command,
       workdir: request.workdir ?? '/stub',
       timeoutMs: request.timeoutMs ?? 1000,
+      stdoutMaxBytes: request.stdoutMaxBytes ?? 64_000,
       ...request.signal ? { signal: request.signal } : {},
-      owner: request.owner,
       sandboxMode: request.sandboxMode,
     }
   }
 
-  async run(_spec: BashExecSpec): Promise<BashRunResult> {
+  async run(spec: BashExecSpec): Promise<BashRunResult> {
     return {
       exitCode: 0,
       signal: null,
       timedOut: false,
       aborted: false,
-      timeoutMs: 1000,
+      timeoutMs: spec.timeoutMs,
       stdout: { text: 'ok', truncated: false },
       stderr: { text: '', truncated: false },
     }
   }
 
-  start(spec: BashExecSpec): BashTask {
-    const task: BashTask = {
-      id: BashTaskId(`stub-${this.tasks.size + 1}`),
+  start(): BashProcess {
+    const proc: BashProcess = {
       status: 'running',
       exitCode: null,
       signal: null,
       done: Promise.resolve(),
+      readOutput: (): BashProcessRead => ({ delta: '', lossy: false }),
+      kill: (): boolean => {
+        if (proc.status !== 'running') return false
+        proc.status = 'killed'
+        return true
+      },
     }
-    this.tasks.set(task.id, task)
-    this.owners.set(task.id, spec.owner)
-    return task
+    return proc
   }
-
-  get(id: BashTaskId): BashTask | undefined {
-    return this.tasks.get(id)
-  }
-
-  ownerOf(id: BashTaskId): OwnerToken | undefined {
-    return this.owners.get(id)
-  }
-
-  list(): BashTask[] {
-    return [...this.tasks.values()]
-  }
-
-  readOutput(id: BashTaskId): BashTaskRead {
-    const task = this.tasks.get(id)
-    if (!task) throw new Error(`unknown bash task "${id}"`)
-    return { task, delta: '', lossy: false }
-  }
-
-  kill(id: BashTaskId): boolean {
-    const task = this.tasks.get(id)
-    if (!task) throw new Error(`unknown bash task "${id}"`)
-    if (task.status !== 'running') return false
-    task.status = 'killed'
-    return true
-  }
-
-  /** Expose the protected notifier for tests. */
-  fire(task: BashTask): void {
-    this.notifyTaskDone(task)
-  }
-}
-
-async function setup() {
-  const ctx = new Context()
-  await ctx.plugin(StubExecutor)
-  // ctx.bash resolves to the registered implementation.
-  const bash = ctx.bash as StubExecutor
-  return { ctx, bash }
 }
 
 describe('BashExecutor service seam', () => {
-  it('registers as ctx.bash and serves the abstract API', async () => {
-    const { bash } = await setup()
-    const task = bash.start(bash.resolve({ command: 'sleep 1' }))
-    expect(bash.get(task.id)).toBe(task)
-    expect(bash.list()).toEqual([task])
-    expect(bash.kill(task.id)).toBe(true)
-    expect(bash.kill(task.id)).toBe(false)
-    const result = await bash.run(bash.resolve({ command: 'true' }))
-    expect(result.exitCode).toBe(0)
-  })
-
-  it('reports no default sandbox mode (composition truth: the base never confines)', async () => {
-    const { bash } = await setup()
-    expect(bash.sandboxMode).toBeUndefined()
-  })
-
-  it('onTaskDone delivers completions to registered listeners', async () => {
-    const { bash } = await setup()
-    const seen: string[] = []
-    bash.onTaskDone(task => void seen.push(task.id))
-    const task = bash.start(bash.resolve({ command: 'x' }))
-    bash.fire(task)
-    expect(seen).toEqual([task.id])
-  })
-
-  it('onTaskDone disposer unsubscribes the listener', async () => {
-    const { bash } = await setup()
-    const listener = vi.fn()
-    const dispose = bash.onTaskDone(listener)
-    dispose()
-    bash.fire(bash.start(bash.resolve({ command: 'x' })))
-    expect(listener).not.toHaveBeenCalled()
-  })
-
-  it('listeners registered from a fiber are removed on dispose (HMR safety)', async () => {
-    const { ctx, bash } = await setup()
-    const listener = vi.fn()
-    const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      inner.bash.onTaskDone(listener)
-    }, { inject: ['bash'] }))
-    bash.fire(bash.start(bash.resolve({ command: 'one' })))
-    expect(listener).toHaveBeenCalledTimes(1)
-
-    await fiber.dispose()
-    bash.fire(bash.start(bash.resolve({ command: 'two' })))
-    expect(listener).toHaveBeenCalledTimes(1)
-  })
-
-  it('silences listeners once the service fiber is disposed', async () => {
+  it('a concrete subclass registers as ctx.bash and serves the abstract API', async () => {
     const ctx = new Context()
-    const fiber = await ctx.plugin(Object.assign(async (inner: Context) => {
-      await inner.plugin(StubExecutor)
-    }, {}))
-    const bash = ctx.bash as StubExecutor
-    const listener = vi.fn()
-    bash.onTaskDone(listener)
-    const task = bash.start(bash.resolve({ command: 'x' }))
+    await ctx.plugin(StubExecutor)
+    const spec = ctx.bash.resolve({ command: 'echo hi' })
+    expect(spec).toEqual({ command: 'echo hi', workdir: '/stub', timeoutMs: 1000, stdoutMaxBytes: 64_000, sandboxMode: undefined })
 
-    await fiber.dispose()
-    bash.fire(task)
-    expect(listener).not.toHaveBeenCalled()
+    const result = await ctx.bash.run(spec)
+    expect(result.exitCode).toBe(0)
+    expect(result.stdout.text).toBe('ok')
+
+    const proc = ctx.bash.start(spec)
+    expect(proc.status).toBe('running')
+    expect(proc.readOutput()).toEqual({ delta: '', lossy: false })
+    expect(proc.kill()).toBe(true)
+    expect(proc.kill()).toBe(false) // already settled → no-op
+    await proc.done
+  })
+
+  it('reports no default sandbox mode from the task-free base seam', async () => {
+    const ctx = new Context()
+    await ctx.plugin(StubExecutor)
+    expect(ctx.bash.sandboxMode).toBeUndefined()
+  })
+
+  it('loading a second implementation throws (one bash service per context — cordis standard)', async () => {
+    const ctx = new Context()
+    await ctx.plugin(StubExecutor)
+    class SecondExecutor extends StubExecutor {}
+    await expect(ctx.plugin(SecondExecutor)).rejects.toThrow(/service "bash" has been registered/)
   })
 })
diff --git a/packages/bash/bash/tsconfig.json b/packages/bash/bash/tsconfig.json
index 13d297a292..b9ab094edb 100644
--- a/packages/bash/bash/tsconfig.json
+++ b/packages/bash/bash/tsconfig.json
@@ -14,9 +14,6 @@
     {
       "path": "../../../vendor/cordis"
     },
-    {
-      "path": "../../util/brand"
-    },
     {
       "path": "../../sandbox/sandbox"
     },
diff --git a/packages/bash/tool-bash/README.md b/packages/bash/tool-bash/README.md
index 2c7fe345ad..d5e65f1a39 100644
--- a/packages/bash/tool-bash/README.md
+++ b/packages/bash/tool-bash/README.md
@@ -1,12 +1,12 @@
 # @deepseek-ai/dsh-tool-bash
 
-The model-facing bash tools — `bash`, `bash_output`, `bash_kill` — registered over the `ctx.bash` executor seam (`@deepseek-ai/dsh-bash`). This package owns schema and text shaping while process concerns stay behind the seam. Executor facts can change rendered results, and a sandboxing executor activates the escalation fields, without moving those presentation rules into the backend.
+The model-facing `bash` tool registered over the `ctx.bash` executor seam. Foreground execution stays behind that seam; a background process handle is registered with the generic `ctx.tasks` runtime and controlled through `task_output`, `task_list`, and `task_kill` from `@deepseek-ai/dsh-tool-tasks`.
 
 Requires a loaded executor implementation (e.g. `@deepseek-ai/dsh-bash-local`); the plugin stays pending until `ctx.bash` exists (`inject: ['tools', 'bash', 'systemPrompt']`).
 
-The package root exposes only the Cordis plugin contract (`name`, `inject`, `apply`); result rendering remains an implementation detail covered by same-package tests.
+The package root exposes only the Cordis plugin contract (`name`, `inject`, `Config`, `apply`); result rendering and background-process adaptation remain implementation details covered by same-package tests.
 
-The plugin also contributes the `tool:bash` prompt section (order 105) — the cross-call habit the per-tool descriptions cannot carry: check the `[exit code: N]` marker on every result and investigate failures before moving on. A sandboxing executor changes the `bash` schema and result markers but adds no mode statement or switch notice; see [Per-session mode](#per-session-mode-switching).
+The plugin also contributes the `tool:bash` prompt section (order 105): check the `[exit code: N]` marker on every result and investigate failures before moving on.
 
 ## Tools
 
@@ -24,31 +24,40 @@ The plugin also contributes the `tool:bash` prompt section (order 105) — the c
 
 `command`, `workdir`, and `timeoutMs` are resolved against the executor's config defaults via `ctx.bash.resolve()` before execution, so the executor seam (`BashExecSpec`) receives explicit `workdir`/`timeoutMs` values. The workdir default is applied in the tool layer (from the calling agent's `session.header.cwd`) BEFORE `resolve()` — the per-session cwd must come from `exec.agent`, since N sessions share one executor; only when no session cwd is available does the executor fall back to its own config / `process.cwd()`.
 
+### Managed shell environment
+
+Every foreground and background model bash call receives a newly collected trusted `DSH_*` environment. `DSH_HOME` is the absolute Harness home resolved by [`@deepseek-ai/dsh-home`](../../util/home/README.md) (`dshHome` config, then ambient `$DSH_HOME`, then `~/.dsh`) and `DSH_SHELL=1` identifies the managed child. Agent calls additionally receive `DSH_SESSION_ID=agent.session.header.id`; when the active persistence seam locates a JSONL artifact they also receive `DSH_SESSION_JSONL=<absolute target path>`. The JSONL path is a location hint: it may not exist before the first flush or contain the current buffered turn, and it is not an authorization credential.
+
+`ctx.bashEnv` owns collection. Other plugins can register an effect-scoped contributor with a stable name, declared keys/descriptions, and `resolve(execution: ToolExecution)`; duplicate ownership and undeclared runtime keys fail loudly, while `list()` enumerates declarations without executing providers. Harness built-ins reserve `DSH_HOME`, `DSH_SHELL`, and `DSH_SESSION_ID`; tool-bash's persistence translator owns `DSH_SESSION_JSONL` by reading the backend-neutral `sessionPersistence.locate()` seam.
+
+```ts
+import type { Context } from 'cordis'
+import type {} from '@deepseek-ai/dsh-tool-bash'
+
+export const inject = ['bashEnv']
+
+export function apply(ctx: Context): void {
+  ctx.bashEnv.register({
+    name: 'deployment-region',
+    variables: { DSH_DEPLOYMENT_REGION: { description: 'Current deployment region.' } },
+    resolve: execution => execution.agent === undefined ? {} : { DSH_DEPLOYMENT_REGION: 'cn-north' },
+  })
+}
+```
+
+The overlay is computed from the current `ToolExecution` and passed through the dedicated `BashExecRequest.dshEnv` channel. The local executor removes all inherited `DSH_*` before merging that snapshot, so nested harnesses and concurrent parent/child agents cannot leak stale identities. `process.env` is never modified. The tool description teaches the generic `$DSH_*` convention rather than naming persistence-specific variables or adding a permanent system-prompt section.
+
 Result text contains stdout, an optional `[stderr]` section, then applicable sandbox-denial, timeout, signal, exit-code, and truncation markers. Timeout is reported independently of final exit status; nonzero exit remains a model-interpreted result rather than `isError`. Truncation links a safe complete spill file or reports it unavailable. Only infrastructure failures such as spawn errors and aborts produce `isError`.
 
-### `bash_output`
-
-`task_id` → output produced **since the previous `bash_output` call** plus a status line (`running` / `completed, exit code: N` / `killed`). A settled task classified as a sandbox denial carries the same `[sandbox: file access denied under <mode> mode]` marker on every read that sees it (denials are only classifiable once the whole stderr has been collected). Reads that lost data to buffer bounds say so and point at the full-output spill file when one is safely available, otherwise `(unavailable)`.
-
-### `bash_kill`
-
-`task_id` → ask the executor to kill the background task. The concrete executor decides how to signal or stop the process; killing an already-finished task is a reported no-op, and unknown ids are errors.
-
-### Task ownership (cross-session isolation)
-
-The executor stores the spawning session id as the task's owner. `bash_output` and `bash_kill` reject a caller with a different session id; agent-less tasks remain unowned, while agent-less calls cannot access owned tasks. Storing ownership on the task prevents predictable global ids from crossing ACP sessions and preserves the fence across tool-plugin reloads. Completion notices remain effect-scoped and may be missed during a reload gap.
+When `run_in_background` is true, this plugin preflights `ctx.tasks.start()` before spawning, registers the calling agent as owner, and adapts the returned `BashProcess` handle into generic cancel/done/incremental-output hooks. The task runtime owns ids, cross-session isolation, completion notices, waiting, and disposal cleanup; this plugin only maps bash exit/sandbox facts into task output and outcome detail. `enableRunInBackground: false` removes the parameter and rejects a forced background call at execution time.
 
 ## UI presentation
 
-UI presentation is tool-owned through `presentCall` and `presentResult`. Foreground `bash` uses a terminal card whose title is the exact command and whose optional description is separate; cwd follows an explicit `workdir`—resolved by the bridge against the session when relative—or the session cwd. Its result carries raw output plus exit or signal data, and clients without terminal support receive a bridge-derived fenced console fallback. Background runs, spawn failures, `bash_output`, and `bash_kill` use generic cards. Presenters are pure and replay-safe; malformed older arguments fall back to generic rendering. See [`dsh-tools`](../../core/tools/) and [`dsh-acp`](../../ui/acp/) for card semantics.
-
-## Background completion notices
-
-When a task finishes, the plugin resolves its owner token to a live agent and injects a durable completion notice. If the owner no longer exists, the notice is dropped. Injection affects the next request but does not wake an idle agent, so the model must poll with `bash_output` when it needs completion promptly.
+The tool owns its `presentCall`/`presentResult` render intent. A foreground call is a terminal card carrying command, description, cwd, raw output, and parsed exit status. A background start is a generic execute card because it returns only a task id; the generic `task_*` tools own their own cards. These presenters are pure and replay-safe.
 
 ## The tool builds its request from named args only
 
-The seam supports trusted-plugin `stdin` and `env`, but the model-facing tool does not. It builds requests only from its declared arguments, signal, and owner; extra model keys are ignored. Shell syntax already provides equivalent command-level behavior, while the local executor's credential scrub protects ambient secrets. See the [stdin/env RFC](../../../docs/rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md).
+The `BashExecRequest` seam carries optional `stdoutMaxBytes`, `stdin`, ordinary `env`, and managed `dshEnv`, used by trusted in-process plugins and this tool's environment registry. The model-facing tool exposes none of `stdoutMaxBytes`, `stdin`, or `env`: it builds requests from named command/workdir/timeout/signal/sandbox fields plus the registry-collected `dshEnv`. Extra model keys are ignored and cannot replace managed values. Shell syntax provides equivalent command-level behavior, while the local executor scrubs ambient credentials and stale `DSH_*` values. See the [stdin/env RFC](../../../docs/rfc/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-surface.md).
 
 ## Permissions and escalation
 
@@ -76,7 +85,7 @@ Check the [exit code: N] marker on every bash result; investigate failures befor
 
 ### Tool schemas
 
-**What the model sees**: The model sees the generated [`bash`, `bash_output`, and `bash_kill` schemas](../../../docs/tool-catalog.md#deepseek-aidsh-tool-bash). `sandbox_permissions` and `justification` augment `bash` only when the mounted executor advertises sandboxing. Agent-scoped tool restrictions can remove the definitions for that agent.
+**What the model sees**: The model sees the generated [`bash` schema](../../../docs/tool-catalog.md#deepseek-aidsh-tool-bash). `run_in_background` appears only when this producer enables it; `sandbox_permissions` and `justification` appear only when the mounted executor advertises sandboxing. Agent-scoped tool restrictions can remove the definition for that agent.
 
 **Token effect**: Fixed schema cost on every request where the tools are visible; sandbox support adds the escalation fields and its conditional description paragraph.
 
@@ -88,19 +97,18 @@ Check the [exit code: N] marker on every bash result; investigate failures befor
 
 ### Background task context and results
 
-**What the model sees**: Start returns exactly `started background task <taskId>`. Completion injects exactly `background bash task <taskId> finished <status>. Read its output with bash_output.` Reads return only the data-dependent delta or `(no new output)`, optionally `[some output was dropped from memory; full output: <paths-or-(unavailable)>]`, then exactly one of `[status: running]`, `[status: killed]`, `[status: killed by <signal>]`, or `[status: completed, exit code: <exitCode>]`. Kill returns `killed background task <taskId>` or `task <taskId> had already finished`.
+**What the model sees**: Start returns exactly `started background task <taskId>`. This producer supplies incremental process output, optional `[some output was dropped from memory; full output: <paths-or-(unavailable)>]`, sandbox facts, and terminal detail such as `exit code: <exitCode>` or `signal: <signal>` to the generic task runtime. [`dsh-tool-tasks`](../../tasks/tool-tasks/README.md) owns the visible status line, completion notice, listing, and cancellation response.
 
-**Token effect**: Start and status text is small; deltas are data-dependent. The completion notice and every tool result are retained until compaction, but polling does not repeat already-delivered output.
+**Token effect**: The start acknowledgement is small and retained; collected output is data-dependent and bounded by the executor's stream buffers. Consuming reads do not repeat prior output.
 
 ### Tool errors
 
-**What the model sees**: Validation and policy failures are normalized as `Error: <message>`. This package's stable messages are `invalid command: expected a non-empty string`, `invalid description: expected a non-empty string`, `invalid timeoutMs: expected a positive number, got <value>`, `invalid escalation: sandbox_permissions requires a justification`, `invalid escalation: justification is only valid together with sandbox_permissions`, `invalid justification: expected a non-empty sentence`, `invalid task_id: expected a string, got <value>`, `task <taskId> belongs to another session`, `sandbox_permissions is not available in this composition (no sandboxing executor to escalate)`, `sandbox escalation to "<mode>" is not strictly wider than this call's current "<mode>" mode`, the approval-availability/rejection/cancellation variants, and `command aborted`.
+**What the model sees**: Validation and policy failures are normalized as `Error: <message>`. This package's stable messages are `invalid command: expected a non-empty string`, `invalid description: expected a non-empty string`, `invalid timeoutMs: expected a positive number, got <value>`, `invalid escalation: sandbox_permissions requires a justification`, `invalid escalation: justification is only valid together with sandbox_permissions`, `invalid justification: expected a non-empty sentence`, `background execution is disabled for this bash tool`, `background tasks unavailable: load @deepseek-ai/dsh-tasks and @deepseek-ai/dsh-tool-tasks`, `sandbox_permissions is not available in this composition (no sandboxing executor to escalate)`, `sandbox escalation to "<mode>" is not strictly wider than this call's current "<mode>" mode`, the approval-availability/rejection/cancellation variants, and `command aborted`.
 
 **Token effect**: Only the failing call adds these retained tokens; a rejected escalation does not add command output because the command does not run.
 
 ## Known Limitations and Deferred Work
 
 - **Replay exit pills parse from result text** — output whose final line happens to be exactly `[exit code: N]` / `[killed by signal: …]` shows a wrong pill on session replay; a display-only known residual.
-- **The bash tools opt out of `timeout-policy` budgets** — `bash` keeps the executor-owned `BASH_TIMEOUT` path and `bash_output`/`bash_kill` declare no budget, per [the tool-call timeout-policy RFC](../../../docs/rfc/implemented/architecture/2026-07-07-tool-call-timeout-policy.md).
-- **Completion notices do not wake an idle agent** — they become durable context for the next request; a caller needing progress now must poll `bash_output` or send another message.
-- **Tasks started outside an agent have no ownership fence** — their predictable ids are readable and killable by any caller; only agent-started tasks carry a session owner token.
+- **The `bash` tool opts out of `timeout-policy` budgets** — it keeps the executor-owned `BASH_TIMEOUT` path, per [the tool-call timeout-policy RFC](../../../docs/rfc/implemented/architecture/2026-07-07-tool-call-timeout-policy.md).
+- **Background processes have no executor timeout** — callers must use `task_kill`, or rely on owner/service disposal, when work no longer matters.
diff --git a/packages/bash/tool-bash/package.json b/packages/bash/tool-bash/package.json
index 4e23a928a7..816be1ef88 100644
--- a/packages/bash/tool-bash/package.json
+++ b/packages/bash/tool-bash/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@deepseek-ai/dsh-tool-bash",
-  "description": "Model-facing bash tools (bash, bash_output, bash_kill) over the DeepSeek Harness bash executor seam",
+  "description": "Model-facing bash tool with optional generic background-task and sandbox-escalation support",
   "version": "0.0.1",
   "private": true,
   "type": "module",
@@ -25,24 +25,34 @@
     "@deepseek-ai/dsh-agent": "^0.0.1",
     "@deepseek-ai/dsh-user-approval": "^0.0.1",
     "@deepseek-ai/dsh-bash": "^0.0.1",
+    "@deepseek-ai/dsh-home": "^0.0.1",
     "@deepseek-ai/dsh-llm": "^0.0.1",
+    "@deepseek-ai/dsh-session-persistence": "^0.0.1",
     "@deepseek-ai/dsh-sandbox": "^0.0.1",
     "@deepseek-ai/dsh-system-prompt": "^0.0.1",
+    "@deepseek-ai/dsh-tasks": "^0.0.1",
     "@deepseek-ai/dsh-tools": "^0.0.1",
     "cordis": "^4.0.0-rc.7"
   },
+  "dependencies": {
+    "schemastery": "^3.18.0"
+  },
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-user-approval": "workspace:^",
     "@deepseek-ai/dsh-bash": "workspace:^",
     "@deepseek-ai/dsh-bash-local": "workspace:^",
-    "@deepseek-ai/dsh-bash-sandbox": "workspace:^",
+    "@deepseek-ai/dsh-home": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-sandbox": "workspace:^",
-    "@deepseek-ai/dsh-sandbox-local": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
+    "@deepseek-ai/dsh-session-persistence": "workspace:^",
+    "@deepseek-ai/dsh-session-persistence-jsonl": "workspace:^",
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-tasks": "workspace:^",
+    "@deepseek-ai/dsh-tool-tasks": "workspace:^",
     "@deepseek-ai/dsh-tools": "workspace:^",
     "cordis": "^4.0.0-rc.7"
   }
diff --git a/packages/bash/tool-bash/src/background.ts b/packages/bash/tool-bash/src/background.ts
new file mode 100644
index 0000000000..6ba67cdfde
--- /dev/null
+++ b/packages/bash/tool-bash/src/background.ts
@@ -0,0 +1,27 @@
+/**
+ * Generic-task adaptation for background bash process handles.
+ *
+ * @module @deepseek-ai/dsh-tool-bash/background
+ */
+
+import type { BashProcess } from '@deepseek-ai/dsh-bash'
+
+/**
+ * Map a settled background process onto the generic task-outcome vocabulary:
+ * `killed` stays `killed` (detail: the signal when one is known), everything
+ * else is `completed` with the exit code as detail. A nonzero command exit is
+ * reported, not failed, exactly like the foreground rendering.
+ * @param proc - the settled process handle.
+ * @returns the outcome for the `ctx.tasks` registration.
+ */
+export function processOutcome(proc: BashProcess): { status: 'completed' | 'killed'; detail: string } {
+  // TODO(background-infrastructure-outcome): widen BashProcess with an explicit
+  // infrastructure-failure outcome, then map spawn failures and
+  // sandbox.runnerFailed to task `failed`. The current seam aliases a spawn
+  // failure with a signal-less kill and a runner failure with an ordinary
+  // wrapper exit; real nonzero command exits must remain `completed`.
+  if (proc.status === 'killed') {
+    return { status: 'killed', detail: proc.signal !== null ? `signal: ${proc.signal}` : 'killed before exit' }
+  }
+  return { status: 'completed', detail: `exit code: ${proc.exitCode ?? 0}` }
+}
diff --git a/packages/bash/tool-bash/src/index.ts b/packages/bash/tool-bash/src/index.ts
index ac4716b43f..cad92c9276 100644
--- a/packages/bash/tool-bash/src/index.ts
+++ b/packages/bash/tool-bash/src/index.ts
@@ -1,36 +1,221 @@
 /**
- * Model-facing `bash`, `bash_output`, and `bash_kill` tools over the executor
- * seam. Background tasks are fenced by owning session, completion injects a
- * durable notice, and confining executors add one-shot approval-based escalation.
- * Notices do not wake idle agents. Ownership is stored with the executor task so
- * it survives this plugin's reload; per-call authority is escalation grant,
- * session override, then executor default. See the package README for the tool contract.
+ * Model-facing `bash` tool over the `ctx.bash` executor seam. Background calls
+ * register process handles with `ctx.tasks`; their work uses task cancellation
+ * rather than the tool-call signal after an id is returned.
+ *
+ * TODO(permissions): deployment policy belongs in `tools/pre-execute` and
+ * sandboxing executors; see docs/architecture.md § Extending The Harness.
  * @module @deepseek-ai/dsh-tool-bash
  */
 
-import type { Context } from 'cordis'
+import { Service, type Context } from 'cordis'
+import z from 'schemastery'
 import { isAbsolute, resolve as resolvePath } from 'node:path'
 import { defineTool } from '@deepseek-ai/dsh-tools'
 import type { GenericCallView, TerminalCallView, ToolExecution, ToolResult, ToolResultView } from '@deepseek-ai/dsh-tools'
 import type { Agent } from '@deepseek-ai/dsh-agent'
+import type {} from '@deepseek-ai/dsh-session-persistence'
 import { assertNever } from '@deepseek-ai/dsh-llm'
 import type {} from '@deepseek-ai/dsh-system-prompt'
-// Side-effect type import: declaration-merges `ctx.approval`, consumed
-// opportunistically by the escalation gate (`ctx.get('approval')` — the seam
-// stays optional at runtime, same pattern as dsh-tools' ask routing).
+import type {} from '@deepseek-ai/dsh-tasks'
 import type {} from '@deepseek-ai/dsh-user-approval'
 import type { SandboxMode } from '@deepseek-ai/dsh-sandbox'
-import { BashTaskId, OwnerToken, effectiveSandboxMode } from '@deepseek-ai/dsh-bash'
-import type { BashTask } from '@deepseek-ai/dsh-bash'
-import { parseExitStatus, renderResult } from './render.ts'
+import { DSH_ENV_PREFIX, effectiveSandboxMode } from '@deepseek-ai/dsh-bash'
+import type { DshEnvironment, DshEnvironmentKey } from '@deepseek-ai/dsh-bash'
+import { DSH_HOME_ENV, resolveDshHome } from '@deepseek-ai/dsh-home'
+import { processOutcome } from './background.ts'
+import { parseExitStatus, renderProcessRead, renderResult } from './render.ts'
+
+declare module 'cordis' {
+  interface Context {
+    bashEnv: BashEnvRegistry
+  }
+}
 
 export const name = 'tool-bash'
 export const inject = ['tools', 'bash', 'systemPrompt']
 
+/** Configuration for the bash tool and its managed child environment. */
+export interface Config {
+  /** Expose `run_in_background` (default true); disabled calls are also rejected. */
+  enableRunInBackground?: boolean
+  /** DeepSeek Harness home directory exposed as `DSH_HOME`; defaults to `$DSH_HOME` or `~/.dsh`. */
+  dshHome?: string
+}
+
+/** Runtime configuration schema for the bash tool plugin. */
+export const Config: z<Config> = z.object({
+  enableRunInBackground: z.boolean().default(true),
+  dshHome: z.string(),
+})
+
+/** Model-visible metadata for one managed `DSH_*` environment variable. */
+export interface BashEnvVariable {
+  /** Concise description of the environment fact represented by the variable. */
+  description: string
+}
+
 /**
- * Validate value constraints absent from SchemaSpec: non-empty strings, a
- * positive finite timeout, and paired escalation mode and justification.
+ * A plugin contribution to the managed environment of each model bash call.
+ * Declared keys make ownership conflicts detectable before the first command;
+ * `resolve` computes only the values available for the current execution.
  */
+export interface BashEnvContributor {
+  /** Stable contributor name used in diagnostics and duplicate detection. */
+  name: string
+  /** Complete set of `DSH_*` keys this contributor may return. */
+  variables: Readonly<Record<DshEnvironmentKey, BashEnvVariable>>
+  /**
+   * Resolve this contributor's available values for one tool execution.
+   * @param execution - the bash tool execution and its optional calling agent.
+   * @returns a partial map containing only keys declared in {@link variables}.
+   */
+  resolve(execution: ToolExecution): Readonly<Partial<Record<DshEnvironmentKey, string>>>
+}
+
+/** An enumerable declaration returned by {@link BashEnvRegistry.list}. */
+export interface BashEnvVariableInfo extends BashEnvVariable {
+  /** Contributor that owns the variable. */
+  contributor: string
+  /** Declared `DSH_*` environment variable name. */
+  key: DshEnvironmentKey
+}
+
+const DSH_SHELL_KEY = `${DSH_ENV_PREFIX}SHELL` as const
+const DSH_SESSION_ID_KEY = `${DSH_ENV_PREFIX}SESSION_ID` as const
+const DSH_SESSION_JSONL_KEY = `${DSH_ENV_PREFIX}SESSION_JSONL` as const
+const RESERVED_BASH_ENV_KEYS = new Set<DshEnvironmentKey>([
+  DSH_HOME_ENV,
+  DSH_SHELL_KEY,
+  DSH_SESSION_ID_KEY,
+])
+const BASH_ENV_KEY_SUFFIX = /^[A-Z][A-Z0-9_]*$/
+
+/**
+ * Registry (`ctx.bashEnv`) for trusted, per-execution `DSH_*` variables.
+ * The namespace is rebuilt for every model bash call: ambient `DSH_*` values
+ * are discarded by the executor, then the registry's current snapshot is
+ * injected. Built-in shell facts remain owned by the registry itself while
+ * plugins can register additional, enumerable facts with effect-scoped
+ * disposal.
+ */
+export class BashEnvRegistry extends Service {
+  private readonly contributors = new Map<string, BashEnvContributor>()
+  private readonly keyOwners = new Map<DshEnvironmentKey, string>()
+  private readonly dshHome: string
+
+  /**
+   * Create and install the `ctx.bashEnv` service.
+   * @param ctx - Cordis context that owns the service and registrations.
+   * @param config - home-directory configuration for the built-in variables.
+   */
+  constructor(ctx: Context, config: Config = {}) {
+    super(ctx, 'bashEnv')
+    this.dshHome = resolveDshHome(config.dshHome)
+  }
+
+  /**
+   * Register one environment contributor. Names and keys are unique; built-in
+   * keys are reserved. Registration is disposed with the calling plugin fiber.
+   * @param contributor - declared key ownership and per-execution resolver.
+   * @returns the disposer that unregisters the contribution.
+   */
+  register(contributor: BashEnvContributor): () => void {
+    const dispose = this.ctx.effect(function* (this: BashEnvRegistry) {
+      if (contributor.name.trim().length === 0) {
+        throw new Error('bash env contributor name must be non-empty')
+      }
+      if (this.contributors.has(contributor.name)) {
+        throw new Error(`bash env contributor "${contributor.name}" is already registered`)
+      }
+
+      const variables = Object.entries(contributor.variables) as [DshEnvironmentKey, BashEnvVariable][]
+      for (const [key, variable] of variables) {
+        if (!key.startsWith(DSH_ENV_PREFIX)
+          || !BASH_ENV_KEY_SUFFIX.test(key.slice(DSH_ENV_PREFIX.length))) {
+          throw new Error(`bash env contributor "${contributor.name}" declared invalid key "${key}"`)
+        }
+        if (RESERVED_BASH_ENV_KEYS.has(key)) {
+          throw new Error(`bash env contributor "${contributor.name}" cannot own reserved key "${key}"`)
+        }
+        if (variable.description.trim().length === 0) {
+          throw new Error(`bash env contributor "${contributor.name}" must describe "${key}"`)
+        }
+        const owner = this.keyOwners.get(key)
+        if (owner !== undefined) {
+          throw new Error(`bash env key "${key}" is already owned by contributor "${owner}"; contributor "${contributor.name}" cannot also own it`)
+        }
+      }
+
+      this.contributors.set(contributor.name, contributor)
+      for (const [key] of variables) this.keyOwners.set(key, contributor.name)
+      yield () => {
+        this.contributors.delete(contributor.name)
+        for (const [key] of variables) this.keyOwners.delete(key)
+      }
+    }.bind(this), 'bashEnv.register()')
+    return () => void dispose()
+  }
+
+  /**
+   * Build the trusted `DSH_*` snapshot for one bash tool execution.
+   * @param execution - the current tool execution.
+   * @returns an immutable environment overlay containing built-ins and current contributions.
+   */
+  collect(execution: ToolExecution): DshEnvironment {
+    const values: Record<DshEnvironmentKey, string> = {
+      [DSH_HOME_ENV]: this.dshHome,
+      [DSH_SHELL_KEY]: '1',
+    }
+    if (execution.agent !== undefined) {
+      values[DSH_SESSION_ID_KEY] = execution.agent.session.header.id
+    }
+
+    for (const contributor of [...this.contributors.values()].sort((left, right) => left.name.localeCompare(right.name))) {
+      const resolved = contributor.resolve(execution)
+      for (const [rawKey, value] of Object.entries(resolved)) {
+        const key = rawKey as DshEnvironmentKey
+        if (!Object.hasOwn(contributor.variables, key)) {
+          throw new Error(`bash env contributor "${contributor.name}" returned undeclared key "${key}"`)
+        }
+        if (typeof value !== 'string') {
+          throw new Error(`bash env contributor "${contributor.name}" returned a non-string value for "${key}"`)
+        }
+        values[key] = value
+      }
+    }
+
+    return Object.freeze(Object.fromEntries(Object.entries(values).sort(([left], [right]) => left.localeCompare(right))))
+  }
+
+  // TODO(bash-env-list-builtins): Include registry-owned built-ins before diagnostics,
+  // prompt, or UI code treats list() as an exhaustive environment catalog.
+  /**
+   * Enumerate plugin-contributed variables without executing their resolvers.
+   * @returns declarations sorted by environment variable name.
+   */
+  list(): BashEnvVariableInfo[] {
+    return [...this.contributors.values()]
+      .flatMap(contributor => Object.entries(contributor.variables).map(([key, variable]) => ({
+        contributor: contributor.name,
+        description: variable.description,
+        key: key as DshEnvironmentKey,
+      })))
+      .sort((left, right) => left.key.localeCompare(right.key))
+  }
+}
+
+/** Parsed tool args; execute validates value constraints absent from SchemaSpec. */
+interface BashToolArgs {
+  command: string
+  description: string
+  timeoutMs?: number
+  workdir?: string
+  run_in_background?: boolean
+  sandbox_permissions?: string
+  justification?: string
+}
+
 function validateBashArgs(args: BashToolArgs): void {
   if (args.command.trim().length === 0) {
     throw new Error('invalid command: expected a non-empty string')
@@ -52,59 +237,24 @@ function validateBashArgs(args: BashToolArgs): void {
   }
 }
 
-/**
- * Reject an empty `task_id`; SchemaSpec already validates type and presence.
- */
-function validateTaskId(value: string): BashTaskId {
-  if (value.length === 0) {
-    throw new Error(`invalid task_id: expected a string, got ${JSON.stringify(value)}`)
-  }
-  return BashTaskId(value)
-}
-
-/**
- * Validated bash arguments. Escalation fields are advertised only when the
- * mounted executor reports a confining mode.
- */
-interface BashToolArgs {
-  command: string
-  description: string
-  timeoutMs?: number
-  workdir?: string
-  run_in_background?: boolean
-  sandbox_permissions?: string
-  justification?: string
-}
-
-/**
- * Strictly wider modes for each effective mode. Execution checks this table
- * because the schema is global while the effective mode is per call.
- */
 const WIDER_MODES: Record<string, readonly SandboxMode[]> = {
   'read-only': ['workspace-write', 'danger-full-access'],
   'workspace-write': ['danger-full-access'],
 }
 
-/**
- * All possible escalation targets. Advertise the global set because a session
- * override may be narrower than the executor default; execution rejects a
- * target that is not wider for that call.
- */
 const ESCALATION_TARGETS: readonly SandboxMode[] = ['workspace-write', 'danger-full-access']
 
-/**
- * The bash tool's byte-stable base description. Escalation guidance is added
- * only when the mounted executor can honor it, as the one exception to the
- * ordinary no-retry guidance.
- */
-function bashDescription(escalationModes: readonly SandboxMode[]): string {
+function bashDescription(backgroundEnabled: boolean, escalationModes: readonly SandboxMode[]): string {
+  const background = backgroundEnabled
+    ? 'Set `run_in_background: true` for long-running commands: the call returns a task id immediately; read its output with `task_output` and stop it with `task_kill`.'
+    : 'Background execution is not available; long-running commands must finish within the timeout.'
   const base = 'Execute a bash command (`bash -c`) and return its stdout/stderr. '
     + 'Each call runs in a fresh shell: no state (cwd, variables, functions) persists between calls — '
     + 'pass `workdir` instead of using `cd`. Non-zero exits are reported as `[exit code: N]`. '
-    + 'Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way (a background task reports the same marker via bash_output once it has finished). '
+    + `Current harness environment facts are exposed through managed \`$${DSH_ENV_PREFIX}*\` variables; inspect them when needed. `
+    + 'Commands may run under a file sandbox; a blocked file operation is reported as `[sandbox: file access denied under <mode> mode]` — a policy denial, not a bug in the command; do not retry another way. '
     + 'Long output is truncated to its tail; the full output is saved to a file whose path is reported when available. '
-    + 'Set `run_in_background: true` for long-running commands: the call returns a task id immediately; '
-    + 'poll it with `bash_output` and stop it with `bash_kill`.'
+    + background
   if (escalationModes.length === 0) return base
   return base + ' Attempting a command the sandbox may deny is safe and expected: run it and read the '
     + 'marker rather than assuming the denial. When a command is denied and a wider mode would let it '
@@ -119,16 +269,14 @@ function bashDescription(escalationModes: readonly SandboxMode[]): string {
     + 'it — but it does not forbid attempting or escalating other commands later.'
 }
 
-// Pure tool-owned presentation used for both live events and replay.
-
 /**
  * Present foreground calls as terminals and background starts as generic cards.
+ * The command remains the title on both paths; foreground cwd is passed through
+ * for the bridge to resolve, while background descriptions remain card content.
  */
 type BashCallArgs = { command: string; description: string; workdir?: string; run_in_background?: boolean }
 
 function presentBashCall(args: BashCallArgs): GenericCallView | TerminalCallView {
-  // A background start is not an interactive terminal — a generic execute card
-  // with the command as rawInput and the description as a content block.
   if (args.run_in_background === true) {
     return {
       card: 'generic',
@@ -138,7 +286,6 @@ function presentBashCall(args: BashCallArgs): GenericCallView | TerminalCallView
       content: [{ type: 'text', text: args.description }],
     }
   }
-  // A foreground run is a terminal; an explicit workdir supplies its cwd.
   return {
     card: 'terminal',
     title: args.command,
@@ -156,21 +303,13 @@ function presentBashResult(args: unknown, result: ToolResult): ToolResultView |
   if (block === undefined || block.type !== 'text') return undefined
   const raw = block.text
   const isBackground = typeof args === 'object' && args !== null && (args as { run_in_background?: unknown }).run_in_background === true
-  // A background ack or an errored run is not a real terminal exit: render the
-  // fenced ```console fallback as generic content (no exit pill).
+  // Background acknowledgements and errors have no terminal exit status.
   if (isBackground || result.isError) {
     return { card: 'generic', content: [{ type: 'text', text: `\`\`\`console\n${raw.replace(/\n+$/, '')}\n\`\`\`` }] }
   }
-  // A finished foreground run supplies raw output and parsed exit status.
-  // The bridge derives the no-capability fenced fallback from `output`.
   return { card: 'terminal', output: raw, ...parseExitStatus(raw) }
 }
 
-/** Pending-state presentation for `bash_output`/`bash_kill` (background-task tools). */
-function presentTaskCall(verb: string, args: { task_id: string }): GenericCallView {
-  return { card: 'generic', title: `${verb} background task ${args.task_id}`, kind: 'execute', rawInput: args.task_id }
-}
-
 /**
  * Resolve an explicit workdir first, making a relative one session-cwd-relative;
  * otherwise use the session cwd and leave executor defaulting as the fallback.
@@ -184,86 +323,33 @@ function resolveWorkdir(modelWorkdir: string | undefined, exec: { agent?: Agent
   return modelWorkdir
 }
 
-/** Status line for background task reads. */
-function statusLine(task: BashTask): string {
-  switch (task.status) {
-    case 'running': return '[status: running]'
-    case 'killed': return `[status: killed${task.signal !== null ? ` by ${task.signal}` : ''}]`
-    case 'completed': return `[status: completed, exit code: ${task.exitCode ?? 0}]`
-  }
-}
-
-export function apply(ctx: Context): void {
-  // Cross-call guidance belongs in the prompt rather than one tool description.
-  ctx.systemPrompt.section({
-    name: 'tool:bash',
-    order: 105,
-    text: 'Check the [exit code: N] marker on every bash result; investigate failures before moving on.',
+export function apply(ctx: Context, config: Config = {}): void {
+  const bashEnv = new BashEnvRegistry(ctx, config)
+  bashEnv.register({
+    name: 'session-persistence',
+    variables: {
+      [DSH_SESSION_JSONL_KEY]: {
+        description: 'Absolute target path of the current session JSONL when the active persistence backend provides one.',
+      },
+    },
+    resolve(execution) {
+      const agent = execution.agent
+      if (agent === undefined) return {}
+      const location = ctx.get('sessionPersistence')?.locate(agent.session.header)
+      return location?.kind === 'jsonl' ? { [DSH_SESSION_JSONL_KEY]: location.path } : {}
+    },
   })
-
-  /**
-   * Return the canonical session-header id used by ACP and persistence as the
-   * task owner, or undefined for a non-agent caller.
-   */
-  const callerToken = (exec: { agent?: Agent }): OwnerToken | undefined =>
-    exec.agent ? OwnerToken(exec.agent.session.header.id) : undefined
-
-  /**
-   * Reject access when a task has a different session owner. Unowned tasks are
-   * allowed; unknown ids still fail in the subsequent read or kill.
-   */
-  const assertTaskAccess = (taskId: BashTaskId, exec: { agent?: Agent }): void => {
-    const owner = ctx.bash.ownerOf(taskId)
-    if (owner !== undefined && owner !== callerToken(exec)) {
-      throw new Error(`task ${taskId} belongs to another session`)
-    }
-  }
-
-  // Completion runs on the bash fiber, so use topology-independent lookup and
-  // match the executor's stored session-owner token to a live agent.
-  ctx.bash.onTaskDone((task) => {
-    const ownerToken = ctx.bash.ownerOf(task.id)
-    if (ownerToken === undefined) return
-    const agent = ctx.get('agents')?.list().find(a => OwnerToken(a.session.header.id) === ownerToken)
-    if (!agent) return
-    try {
-      agent.inject(
-        [{ type: 'text', text: `background bash task ${task.id} finished ${statusLine(task)}. Read its output with bash_output.` }],
-        { source: { kind: 'plugin', plugin: 'tool-bash' } },
-      )
-    } catch (error: unknown) {
-      // The one expected failure: the agent was disposed between task completion and this
-      // injection (ReactLoopAgent.inject throws `agent "<id>" is disposed`).
-      if (error instanceof Error && error.message.includes('is disposed')) return
-      throw error
-    }
-  })
-
-  // The escalation surface exists whenever the mounted executor confines.
-  // Advertise the closed target vocabulary globally, then enforce strict
-  // widening against each call's effective session mode.
+  const backgroundEnabled = config.enableRunInBackground ?? true
   const defaultMode = ctx.bash.sandboxMode
   const escalationModes: readonly SandboxMode[] = defaultMode === undefined ? [] : ESCALATION_TARGETS
 
-  /**
-   * Return the calling session's folded standing mode. Approval outranks this
-   * value and the executor default applies when it is absent; non-sandboxing
-   * and agent-less calls have no override.
-   */
   const sessionOverride = (exec: ToolExecution): SandboxMode | undefined =>
     defaultMode === undefined || exec.agent === undefined ? undefined : effectiveSandboxMode(exec.agent.session.events)
 
-  /**
-   * Request one-shot escalation before execution. Missing approval context,
-   * rejection, cancellation, and unavailable answers throw without running the
-   * command; the optional seam is resolved per call through `ctx.get`.
-   */
   const approveEscalation = async (mode: string, justification: string, exec: ToolExecution): Promise<SandboxMode> => {
-    // Reject an unadvertised escalation before prompting for a nonexistent sandbox.
     if (escalationModes.length === 0) {
       throw new Error('sandbox_permissions is not available in this composition (no sandboxing executor to escalate)')
     }
-    // Reject sandbox widening against the call's effective mode before requesting approval.
     const effectiveMode = (sessionOverride(exec) ?? defaultMode) as SandboxMode
     if (!(WIDER_MODES[effectiveMode] ?? []).includes(mode as SandboxMode)) {
       throw new Error(`sandbox escalation to "${mode}" is not strictly wider than this call's current "${effectiveMode}" mode`)
@@ -279,13 +365,10 @@ export function apply(ctx: Context): void {
       agent: exec.agent,
       toolName: 'bash',
       callId: exec.callId,
-      // Self-contained for the audit trail: approval/asked stores this
-      // reason, and the target mode is part of the grant's identity.
       reason: `escalate sandbox to ${mode}: ${justification}`,
       ...exec.signal ? { signal: exec.signal } : {},
     })
     switch (outcome) {
-      // Schema validation pins the vocabulary; the per-call check proves widening.
       case 'allowed-once': return mode as SandboxMode
       case 'rejected': throw new Error(`the user rejected escalating this command to "${mode}"`)
       case 'cancelled': throw new Error(`approval for escalating to "${mode}" was cancelled`)
@@ -294,9 +377,16 @@ export function apply(ctx: Context): void {
     }
   }
 
+  // Cross-call guidance belongs in the prompt rather than one-call schema prose.
+  ctx.systemPrompt.section({
+    name: 'tool:bash',
+    order: 105,
+    text: 'Check the [exit code: N] marker on every bash result; investigate failures before moving on.',
+  })
+
   ctx.tools.register(defineTool({
     name: 'bash',
-    description: bashDescription(escalationModes),
+    description: bashDescription(backgroundEnabled, escalationModes),
     parameters: {
       command: { type: 'string', required: true, description: 'The bash command to execute.' },
       description: {
@@ -308,104 +398,71 @@ export function apply(ctx: Context): void {
       },
       timeoutMs: { type: 'number', description: 'Timeout in milliseconds. The executor applies its configured default and cap, and kills the command on expiry.' },
       workdir: { type: 'string', description: 'Working directory for this command. Defaults to the session workspace; a relative path is resolved against it.' },
-      run_in_background: { type: 'boolean', description: 'Run in the background and return a task id immediately. No timeout applies.' },
+      ...backgroundEnabled ? {
+        run_in_background: { type: 'boolean' as const, description: 'Run in the background and return a task id immediately (collect with task_output, stop with task_kill). No timeout applies.' },
+      } : {},
       ...escalationModes.length > 0 ? {
         sandbox_permissions: {
           type: 'string' as const,
           enum: [...escalationModes],
-          description: 'The wider sandbox mode this command needs. Only valid as a one-shot retry '
-            + 'of a command the sandbox just denied; requires justification and user approval.',
+          description: 'The wider sandbox mode this command needs. Only valid as a one-shot retry of a command the sandbox just denied; requires justification and user approval.',
         },
         justification: {
           type: 'string' as const,
-          description: 'Required with sandbox_permissions: one sentence for the user explaining '
-            + 'why this exact command needs the wider access.',
+          description: 'Required with sandbox_permissions: one sentence for the user explaining why this exact command needs the wider access.',
         },
       } : {},
     },
     async execute(args: BashToolArgs, exec) {
       validateBashArgs(args)
-      // `description` is display/logging metadata only. Escalation approval
-      // completes before execution; grant > session override > executor default.
+      // Description is display metadata; workdir defaults to the caller's session.
       const sandboxMode = args.sandbox_permissions !== undefined && args.justification !== undefined
         ? await approveEscalation(args.sandbox_permissions, args.justification, exec)
         : sessionOverride(exec)
-      // Default the workdir to the calling agent's session cwd so each ACP
-      // session runs in its own workspace (see resolveWorkdir); an explicit
-      // model workdir still wins.
       const workdir = resolveWorkdir(args.workdir, exec)
+      const dshEnv = bashEnv.collect(exec)
       const request = {
         command: args.command,
         ...workdir !== undefined ? { workdir } : {},
         ...args.timeoutMs !== undefined ? { timeoutMs: args.timeoutMs } : {},
-        ...exec.signal ? { signal: exec.signal } : {},
+        dshEnv,
         ...sandboxMode !== undefined ? { sandboxMode } : {},
       }
       if (args.run_in_background === true) {
-        // Store the session owner on the task for bash_output/bash_kill isolation.
-        const task = ctx.bash.start(ctx.bash.resolve({ ...request, owner: callerToken(exec) }))
-        return [{ type: 'text', text: `started background task ${task.id}` }]
+        // Undeclared keys are allowed, so schema omission also needs enforcement.
+        if (!backgroundEnabled) {
+          throw new Error('run_in_background is disabled for this deployment (enableRunInBackground: false)')
+        }
+        const tasks = ctx.get('tasks')
+        if (tasks === undefined) {
+          throw new Error('background tasks unavailable: load @deepseek-ai/dsh-tasks and @deepseek-ai/dsh-tool-tasks')
+        }
+        // Reject pre-start cancellation; returned tasks use their own lifecycle.
+        if (exec.signal?.aborted) throw new Error('command aborted')
+        // Task preflight finishes before the starter can spawn a process.
+        const id = tasks.start({
+          kind: 'bash',
+          label: args.command,
+          ...exec.agent ? { owner: exec.agent } : {},
+          run: () => {
+            const proc = ctx.bash.start(ctx.bash.resolve(request))
+            return {
+              cancel: () => void proc.kill(),
+              done: proc.done.then(() => processOutcome(proc)),
+              readOutput: () => renderProcessRead(proc.readOutput(), proc.sandbox, escalationModes),
+            }
+          },
+        })
+        return [{ type: 'text', text: `started background task ${id}` }]
       }
-      const result = await ctx.bash.run(ctx.bash.resolve(request))
+      const result = await ctx.bash.run(ctx.bash.resolve({
+        ...request,
+        ...exec.signal ? { signal: exec.signal } : {},
+      }))
       if (result.aborted) throw new Error('command aborted')
       return [{ type: 'text', text: renderResult(result, escalationModes) }]
     },
     presentCall: presentBashCall,
     presentResult: presentBashResult,
   }))
-
-  ctx.tools.register(defineTool({
-    name: 'bash_output',
-    description: 'Read new output from a background bash task started with `bash` + `run_in_background`. '
-      + 'Returns only output produced since the previous bash_output call, plus the task status. '
-      + 'Tasks keep running while you do other work; poll again later for more output.',
-    parameters: {
-      task_id: { type: 'string', required: true, description: 'Task id returned by the bash tool.' },
-    },
-    // execute is synchronous (registry reads + string shaping) but the
-    // ToolDefinition contract wants a Promise — hence resolve(), not async.
-    execute(args, exec) {
-      const id = validateTaskId(args.task_id)
-      assertTaskAccess(id, exec)
-      const read = ctx.bash.readOutput(id)
-      let text = read.delta.length > 0 ? read.delta : '(no new output)'
-      if (read.lossy) {
-        const paths = [read.stdoutSpillPath, read.stderrSpillPath].filter((p): p is string => p !== undefined)
-        const fullOutput = paths.length > 0 ? paths.join(', ') : '(unavailable)'
-        text += `\n[some output was dropped from memory; full output: ${fullOutput}]`
-      }
-      text += `\n${statusLine(read.task)}`
-      if (read.task.sandbox?.runnerFailed) {
-        // Background settlement carries the runner-failure fact that a
-        // foreground call exposes as SANDBOX_UNAVAILABLE.
-        text += `\n[sandbox: the sandbox runner itself failed under ${read.task.sandbox.mode} mode — the command did not run; this is a sandbox problem, not a command failure]`
-      } else if (read.task.sandbox?.denied) {
-        // Mirrors the foreground result marker (and its same-turn escalation hint).
-        text += `\n[sandbox: file access denied under ${read.task.sandbox.mode} mode]`
-        if (escalationModes.length > 0) {
-          text += '\n[sandbox: escalation available — retry this exact command once with sandbox_permissions (the narrowest wider mode that suffices) + justification; the approval prompt asks the user]'
-        }
-      }
-      return Promise.resolve([{ type: 'text', text }])
-    },
-    presentCall: args => presentTaskCall('Read output from', args),
-  }))
-
-  ctx.tools.register(defineTool({
-    name: 'bash_kill',
-    description: 'Ask the executor to kill a running background bash task by task id.',
-    parameters: {
-      task_id: { type: 'string', required: true, description: 'Task id returned by the bash tool.' },
-    },
-    execute(args, exec) {
-      const id = validateTaskId(args.task_id)
-      assertTaskAccess(id, exec)
-      const killed = ctx.bash.kill(id)
-      return Promise.resolve([{
-        type: 'text',
-        text: killed ? `killed background task ${id}` : `task ${id} had already finished`,
-      }])
-    },
-    presentCall: args => presentTaskCall('Kill', args),
-  }))
 }
diff --git a/packages/bash/tool-bash/src/render.ts b/packages/bash/tool-bash/src/render.ts
index 924861bb1e..2ce36af593 100644
--- a/packages/bash/tool-bash/src/render.ts
+++ b/packages/bash/tool-bash/src/render.ts
@@ -4,7 +4,7 @@
  * @module @deepseek-ai/dsh-tool-bash/render
  */
 
-import type { BashRunResult, CollectedOutput } from '@deepseek-ai/dsh-bash'
+import type { BashProcessRead, BashRunResult, BashSandboxInfo, CollectedOutput } from '@deepseek-ai/dsh-bash'
 import type { SandboxMode } from '@deepseek-ai/dsh-sandbox'
 
 /** Append the truncation notice (with the full-output spill path) to a stream's text. */
@@ -15,7 +15,7 @@ function streamText(output: CollectedOutput): string {
 
 /**
  * Shape one finished run into the text the model sees: stdout, then a marked
- * stderr section, then exit-status markers. Non-zero exits are REPORTED, not
+ * stderr section, then exit-status markers. Non-zero exits are reported, not
  * errored — the model decides how to react; only infrastructure failures
  * (spawn errors, aborts) surface as isError results.
  * @param result - the completed foreground run from the executor.
@@ -40,23 +40,15 @@ export function renderResult(
   if (body.length === 0) body = '(no output)'
 
   const markers: string[] = []
-  // The sandbox marker precedes the exit-status markers so `[exit code: N]`
-  // stays the LAST line (exitStatus() anchors its parse there). Denial is a
-  // reported fact like timeout: the model decides how to react.
+  // Keep the exit marker last because parseExitStatus anchors there.
   if (result.sandbox?.denied) {
     markers.push(`[sandbox: file access denied under ${result.sandbox.mode} mode]`)
-    // The same-turn nudge lives at the decision point: only when this
-    // composition advertises the fields (a lever is never hinted that the
-    // schema does not offer), and inside the sandbox marker family so the
-    // exit-code marker stays the last line.
+    // Hint only when the composition exposes escalation, before the final exit marker.
     if (escalationModes.length > 0) {
       markers.push('[sandbox: escalation available — retry this exact command once with sandbox_permissions (the narrowest wider mode that suffices) + justification; the approval prompt asks the user]')
     }
   }
-  // Timeout is reported independently of how the process actually ended: a
-  // command can trap SIGTERM and exit 0 after our timer fired (e.g.
-  // `trap "exit 0" TERM; sleep 60`), giving timedOut:true / exitCode:0 /
-  // signal:null — the model must still see that the command was cut short.
+  // A command may trap SIGTERM and exit 0 after timeout; still report interruption.
   if (result.timedOut) markers.push(`[timed out after ${result.timeoutMs}ms]`)
   if (result.signal !== null) {
     markers.push(`[killed by signal: ${result.signal}]`)
@@ -69,6 +61,38 @@ export function renderResult(
   return body + markers.join('\n')
 }
 
+/**
+ * Shape one background-process read into the `task_output` delta the model
+ * sees: the incremental delta, plus the lossy-read notice (with full-stream
+ * spill paths) when in-memory truncation dropped unread bytes. Empty-delta
+ * rendering (`(no new output)`) is the generic control surface's job.
+ * @param read - one incremental read from the process handle.
+ * @param sandbox - settled sandbox facts, when this was a confined process.
+ * @param escalationModes - escalation targets advertised by this composition.
+ * @returns the delta text with any loss or sandbox notice appended.
+ */
+export function renderProcessRead(
+  read: BashProcessRead,
+  sandbox?: BashSandboxInfo,
+  escalationModes: readonly SandboxMode[] = [],
+): string {
+  const notices: string[] = []
+  if (read.lossy) {
+    const paths = [read.stdoutSpillPath, read.stderrSpillPath].filter((path): path is string => path !== undefined)
+    notices.push(`[some output was dropped from memory; full output: ${paths.length > 0 ? paths.join(', ') : '(unavailable)'}]`)
+  }
+  if (sandbox?.runnerFailed) {
+    notices.push(`[sandbox: the sandbox runner itself failed under ${sandbox.mode} mode — the command did not run; this is a sandbox problem, not a command failure]`)
+  } else if (sandbox?.denied) {
+    notices.push(`[sandbox: file access denied under ${sandbox.mode} mode]`)
+    if (escalationModes.length > 0) {
+      notices.push('[sandbox: escalation available — retry this exact command once with sandbox_permissions (the narrowest wider mode that suffices) + justification; the approval prompt asks the user]')
+    }
+  }
+  if (notices.length === 0) return read.delta
+  return `${read.delta}${read.delta.length > 0 && !read.delta.endsWith('\n') ? '\n' : ''}${notices.join('\n')}`
+}
+
 /**
  * Recover the structured exit status from a rendered {@link renderResult}
  * string — the inverse of the status markers it appends. A killed marker
diff --git a/packages/bash/tool-bash/tests/bash-env.spec.ts b/packages/bash/tool-bash/tests/bash-env.spec.ts
new file mode 100644
index 0000000000..03d29b572b
--- /dev/null
+++ b/packages/bash/tool-bash/tests/bash-env.spec.ts
@@ -0,0 +1,190 @@
+import { homedir } from 'node:os'
+import { join, resolve } from 'node:path'
+import { afterEach, describe, expect, it, vi } from 'vitest'
+import { Context } from 'cordis'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import type { Agent } from '@deepseek-ai/dsh-agent'
+import type { ToolExecution } from '@deepseek-ai/dsh-tools'
+import { BashEnvRegistry } from '@deepseek-ai/dsh-tool-bash'
+
+afterEach(() => vi.unstubAllEnvs())
+
+function execution(sessionId?: string): ToolExecution {
+  return {
+    token: Symbol('bash-env-test') as ToolExecution['token'],
+    callId: CallId('bash-env-call'),
+    name: 'bash',
+    arguments: { command: 'true' },
+    ...(sessionId === undefined
+      ? {}
+      : { agent: { session: { header: { version: 0, id: sessionId, createdAt: 0 } } } as Agent }),
+  }
+}
+
+describe('BashEnvRegistry', () => {
+  it('collects unconditional shell facts and the current agent session id', () => {
+    const ctx = new Context()
+    const registry = new BashEnvRegistry(ctx, { dshHome: './test-dsh-home' })
+
+    expect(registry.collect(execution())).toEqual({
+      DSH_HOME: resolve('./test-dsh-home'),
+      DSH_SHELL: '1',
+    })
+    expect(registry.collect(execution('session-a'))).toEqual({
+      DSH_HOME: resolve('./test-dsh-home'),
+      DSH_SESSION_ID: 'session-a',
+      DSH_SHELL: '1',
+    })
+  })
+
+  it('resolves DSH_HOME from the ambient override or the user-home default', () => {
+    vi.stubEnv('DSH_HOME', './ambient-dsh-home')
+    const fromEnvironment = new BashEnvRegistry(new Context())
+    expect(fromEnvironment.collect(execution()).DSH_HOME).toBe(resolve('./ambient-dsh-home'))
+
+    vi.stubEnv('DSH_HOME', undefined)
+    const fromDefault = new BashEnvRegistry(new Context())
+    expect(fromDefault.collect(execution()).DSH_HOME).toBe(join(homedir(), '.dsh'))
+  })
+
+  it('collects declared contributor variables and omits unavailable values', () => {
+    const ctx = new Context()
+    const registry = new BashEnvRegistry(ctx, { dshHome: './test-dsh-home' })
+    registry.register({
+      name: 'optional-session-fact',
+      variables: {
+        DSH_SESSION_OPTIONAL: { description: 'Optional session-scoped test fact.' },
+      },
+      resolve: exec => exec.agent === undefined ? {} : { DSH_SESSION_OPTIONAL: exec.agent.session.header.id },
+    })
+    registry.register({
+      name: 'always-available-fact',
+      variables: {
+        DSH_ALWAYS_AVAILABLE: { description: 'Always-available test fact.' },
+      },
+      resolve: () => ({ DSH_ALWAYS_AVAILABLE: 'yes' }),
+    })
+
+    expect(registry.collect(execution())).not.toHaveProperty('DSH_SESSION_OPTIONAL')
+    expect(registry.collect(execution()).DSH_ALWAYS_AVAILABLE).toBe('yes')
+    expect(registry.collect(execution('session-b')).DSH_SESSION_OPTIONAL).toBe('session-b')
+    expect(registry.list()).toEqual([
+      {
+        contributor: 'always-available-fact',
+        description: 'Always-available test fact.',
+        key: 'DSH_ALWAYS_AVAILABLE',
+      },
+      {
+        contributor: 'optional-session-fact',
+        description: 'Optional session-scoped test fact.',
+        key: 'DSH_SESSION_OPTIONAL',
+      },
+    ])
+  })
+
+  it('rejects duplicate variable ownership at registration time', () => {
+    const ctx = new Context()
+    const registry = new BashEnvRegistry(ctx, { dshHome: './test-dsh-home' })
+    registry.register({
+      name: 'first',
+      variables: { DSH_SHARED: { description: 'First owner.' } },
+      resolve: () => ({ DSH_SHARED: 'first' }),
+    })
+
+    expect(() => registry.register({
+      name: 'second',
+      variables: { DSH_SHARED: { description: 'Second owner.' } },
+      resolve: () => ({ DSH_SHARED: 'second' }),
+    })).toThrow(/DSH_SHARED.*first.*second|DSH_SHARED.*second.*first/)
+  })
+
+  it('rejects duplicate contributor names and malformed declarations', () => {
+    const registry = new BashEnvRegistry(new Context(), { dshHome: './test-dsh-home' })
+    registry.register({
+      name: 'declared',
+      variables: { DSH_DECLARED: { description: 'Declared fact.' } },
+      resolve: () => ({}),
+    })
+
+    expect(() => registry.register({
+      name: 'declared',
+      variables: { DSH_ANOTHER: { description: 'Another fact.' } },
+      resolve: () => ({}),
+    })).toThrow(/already registered/)
+    expect(() => registry.register({
+      name: ' ',
+      variables: { DSH_BLANK_NAME: { description: 'Blank owner.' } },
+      resolve: () => ({}),
+    })).toThrow(/name must be non-empty/)
+    expect(() => registry.register({
+      name: 'invalid-key',
+      variables: { dsh_invalid: { description: 'Invalid key.' } } as unknown as Record<'DSH_INVALID', { description: string }>,
+      resolve: () => ({}),
+    })).toThrow(/invalid key/)
+    expect(() => registry.register({
+      name: 'reserved-key',
+      variables: { DSH_HOME: { description: 'Reserved key.' } },
+      resolve: () => ({}),
+    })).toThrow(/reserved key/)
+    expect(() => registry.register({
+      name: 'blank-description',
+      variables: { DSH_BLANK_DESCRIPTION: { description: ' ' } },
+      resolve: () => ({}),
+    })).toThrow(/must describe/)
+  })
+
+  it('rejects undeclared variables returned by a contributor', () => {
+    const ctx = new Context()
+    const registry = new BashEnvRegistry(ctx, { dshHome: './test-dsh-home' })
+    registry.register({
+      name: 'drifted-provider',
+      variables: { DSH_DECLARED: { description: 'Declared fact.' } },
+      resolve: () => ({ DSH_UNDECLARED: 'bad' }),
+    })
+
+    expect(() => registry.collect(execution())).toThrow(/drifted-provider.*DSH_UNDECLARED/)
+  })
+
+  it('rejects non-string values returned by a contributor', () => {
+    const registry = new BashEnvRegistry(new Context(), { dshHome: './test-dsh-home' })
+    registry.register({
+      name: 'wrong-value-type',
+      variables: { DSH_STRING: { description: 'String fact.' } },
+      resolve: () => ({ DSH_STRING: 42 }) as unknown as Record<'DSH_STRING', string>,
+    })
+
+    expect(() => registry.collect(execution())).toThrow(/wrong-value-type.*non-string.*DSH_STRING/)
+  })
+
+  it('removes an effect-scoped contributor when its plugin is disposed', async () => {
+    const ctx = new Context()
+    const registry = new BashEnvRegistry(ctx, { dshHome: './test-dsh-home' })
+    const fiber = await ctx.plugin({
+      inject: ['bashEnv'],
+      apply(inner: Context) {
+        inner.bashEnv.register({
+          name: 'temporary',
+          variables: { DSH_TEMPORARY: { description: 'Temporary fact.' } },
+          resolve: () => ({ DSH_TEMPORARY: 'present' }),
+        })
+      },
+    })
+
+    expect(registry.collect(execution()).DSH_TEMPORARY).toBe('present')
+    await fiber.dispose()
+    expect(registry.collect(execution())).not.toHaveProperty('DSH_TEMPORARY')
+  })
+
+  it('returns an explicit contributor disposer', () => {
+    const registry = new BashEnvRegistry(new Context(), { dshHome: './test-dsh-home' })
+    const dispose = registry.register({
+      name: 'explicit-disposal',
+      variables: { DSH_EXPLICIT_DISPOSAL: { description: 'Explicitly disposed fact.' } },
+      resolve: () => ({ DSH_EXPLICIT_DISPOSAL: 'present' }),
+    })
+
+    expect(registry.collect(execution()).DSH_EXPLICIT_DISPOSAL).toBe('present')
+    dispose()
+    expect(registry.collect(execution())).not.toHaveProperty('DSH_EXPLICIT_DISPOSAL')
+  })
+})
diff --git a/packages/bash/tool-bash/tests/integration.spec.ts b/packages/bash/tool-bash/tests/integration.spec.ts
index a22e86bd99..1f9055c67a 100644
--- a/packages/bash/tool-bash/tests/integration.spec.ts
+++ b/packages/bash/tool-bash/tests/integration.spec.ts
@@ -1,36 +1,44 @@
-import { describe, expect, it } from 'vitest'
+import { afterEach, describe, expect, it, vi } from 'vitest'
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import type { SessionEvent } from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import { existsSync, mkdtempSync, readFileSync, rmSync } from 'node:fs'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { SessionId, type SessionEvent } from '@deepseek-ai/dsh-session'
+import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
+import { AgentId } from '@deepseek-ai/dsh-agent'
 import AgentLoop, { ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
+import TaskService from '@deepseek-ai/dsh-tasks'
+import * as ToolTasks from '@deepseek-ai/dsh-tool-tasks'
 import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
-import { BashTaskId } from '@deepseek-ai/dsh-bash'
 import * as ToolBash from '@deepseek-ai/dsh-tool-bash'
 import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
 
 /**
  * Full-loop integration: a scripted mock model drives the REAL bash tool
  * through the agent loop, exercising the same seams a live model would
- * (tool/call + tool/result session events, agent.inject notifications).
+ * (tool/call + tool/result session events, the generic `ctx.tasks` runtime,
+ * agent.inject completion notices).
  */
-async function harness(adapter: MockAdapter) {
+async function harness(adapter: MockAdapter, sessionRoot?: string, dshHome?: string) {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
+  if (sessionRoot !== undefined) await ctx.plugin(SessionPersistenceJsonl, { root: sessionRoot })
   await ctx.plugin(AgentLoop, { agents: [] })
+  await ctx.plugin(TaskService)
+  await ctx.plugin(ToolTasks)
   await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
-  await ctx.plugin(ToolBash)
+  await ctx.plugin(ToolBash, dshHome === undefined ? {} : { dshHome })
   ctx.llm.registerAdapter(['mock'], adapter)
   return ctx
 }
 
+const dirs: string[] = []
+afterEach(() => {
+  vi.unstubAllEnvs()
+  for (const dir of dirs.splice(0)) rmSync(dir, { recursive: true, force: true })
+})
+
 function waitForIdle(ctx: Context, agent: ReactLoopAgent): Promise<void> {
   return new Promise((resolve) => {
     const dispose = ctx.on('agent/status', (subject, status) => {
@@ -67,14 +75,57 @@ function resultText(event: SessionEvent): string {
     .join('')
 }
 
+/** Poll until `predicate` holds (background settlement races turn end). */
+async function pollUntil(predicate: () => boolean, timeoutMs = 5_000): Promise<void> {
+  const deadline = Date.now() + timeoutMs
+  while (Date.now() < deadline) {
+    if (predicate()) return
+    await new Promise(resolve => setTimeout(resolve, 20))
+  }
+  throw new Error(`condition not met within ${timeoutMs}ms`)
+}
+
 describe('bash tool through the agent loop', () => {
+  it('first-turn bash receives session identity before the lazy JSONL file materializes', async () => {
+    const root = mkdtempSync(join(tmpdir(), 'dsh-bash-session-env-'))
+    dirs.push(root)
+    const dshHome = join(root, 'dsh-home')
+    vi.stubEnv('DSH_STALE_PARENT', 'stale')
+    const adapter = new MockAdapter([
+      toolCallResponse('call-1', 'bash', {
+        command: 'printf \'%s\\n%s\\n%s\\n%s\\n%s\\n\' "$DSH_HOME" "$DSH_SHELL" "$DSH_SESSION_ID" "$DSH_SESSION_JSONL" "${DSH_STALE_PARENT-unset}"; if [ -e "$DSH_SESSION_JSONL" ]; then printf \'present\\n\'; else printf \'absent\\n\'; fi',
+        description: 'inspect session environment',
+      }),
+      textResponse('Session environment inspected.'),
+    ])
+    const ctx = await harness(adapter, root, dshHome)
+    const handle = await ctx.agents.create({
+      agentId: AgentId('session-env'),
+      sessionId: SessionId('session-env-id'),
+      agentOptions: { provider: 'mock', model: 'mock' },
+    })
+    const agent = handle.agent as ReactLoopAgent
+    const location = ctx.sessionPersistence.locate(agent.session.header)
+    expect(location?.kind).toBe('jsonl')
+
+    agent.send([{ type: 'text', text: 'inspect the current session' }])
+    await waitForIdle(ctx, agent)
+
+    const result = findEvent(events(agent), 'tool/result')
+    expect(resultText(result)).toBe(`${dshHome}\n1\nsession-env-id\n${location?.path}\nunset\nabsent\n`)
+    expect(existsSync(location!.path)).toBe(true)
+    const header = JSON.parse(readFileSync(location!.path, 'utf8').split('\n')[0]!) as { type: string; id: string }
+    expect(header).toMatchObject({ type: 'session', id: 'session-env-id' })
+    await handle.dispose()
+  })
+
   it('foreground: model calls bash, sees the result, replies', async () => {
     const adapter = new MockAdapter([
       toolCallResponse('call-1', 'bash', { command: 'echo integration-ok', description: 'test command' }, 'Running it.'),
       textResponse('The command printed integration-ok.'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('it-fg'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('it-fg'), { provider: 'mock', model: 'mock' })
 
     agent.send([{ type: 'text', text: 'run echo integration-ok' }])
     await waitForIdle(ctx, agent)
@@ -106,7 +157,7 @@ describe('bash tool through the agent loop', () => {
       textResponse('It failed with code 9.'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('it-exit'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('it-exit'), { provider: 'mock', model: 'mock' })
 
     agent.send([{ type: 'text', text: 'run exit 9' }])
     await waitForIdle(ctx, agent)
@@ -116,46 +167,41 @@ describe('bash tool through the agent loop', () => {
     expect(resultText(toolResult)).toContain('[exit code: 9]')
   })
 
-  it('background: start → poll → completion notice lands as context/message', async () => {
+  it('background: start ack → completion notice as context/message → task_output collects it', async () => {
+    // The task id is deterministic (a fresh TaskService counts per kind from 1),
+    // so the script can name `bash-1` without threading a generated id.
     const adapter = new MockAdapter([
       toolCallResponse('call-1', 'bash', { command: 'echo bg-ok', description: 'test command', run_in_background: true }),
-      // Each harness owns a fresh BashLocal service, whose first task id is
-      // deterministically bash-1. Keep the scripted call faithful to what the
-      // model sent; tool arguments are immutable once execution policy begins.
-      toolCallResponse('call-2', 'bash_output', { task_id: 'bash-1' }, undefined),
+      textResponse('Started it in the background.'),
+      toolCallResponse('call-2', 'task_output', { task_id: 'bash-1' }),
       textResponse('Background task finished.'),
     ])
-    let taskId = ''
-
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('it-bg'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('it-bg'), { provider: 'mock', model: 'mock' })
 
-    // Capture the generated id so the deterministic fixture is checked against
-    // the real executor instead of silently assuming it.
-    ctx.on('session/event', (_session, event) => {
-      if (event.type === 'tool/result' && taskId === '') {
-        const match = /task (bash-\d+)/.exec(resultText(event))
-        if (match) taskId = match[1]!
-      }
-    })
     agent.send([{ type: 'text', text: 'run echo bg-ok in the background' }])
     await waitForIdle(ctx, agent)
 
-    expect(taskId).toBe('bash-1')
+    const firstResult = findEvent(events(agent), 'tool/result')
+    expect(firstResult.data.isError).toBe(false)
+    expect(resultText(firstResult)).toBe('started background task bash-1')
 
-    // Wait for the background task itself (completion may race turn end).
-    const task = ctx.bash.get(BashTaskId(taskId))
-    if (!task) throw new Error(`task ${taskId} not registered`)
-    await task.done
-
-    const log = events(agent)
-    const firstResult = findEvent(log, 'tool/result')
-    expect(resultText(firstResult)).toBe(`started background task ${taskId}`)
-
-    const notice = findEvent(log, 'context/message')
+    // The task settles on its own; the tool-tasks notice listener injects a
+    // durable context/message into the owning agent's session (settlement may
+    // race turn end, so poll for it).
+    await pollUntil(() => events(agent).some(event => event.type === 'context/message'))
+    const notice = findEvent(events(agent), 'context/message')
     expect(notice.data.content.some(
-      block => block.type === 'text' && block.text.includes(`background bash task ${taskId} finished`),
+      block => block.type === 'text' && block.text.includes('background task bash-1 (bash: echo bg-ok) finished'),
     )).toBe(true)
-    expect(notice.data.source).toEqual({ kind: 'plugin', plugin: 'tool-bash' })
+    expect(notice.data.source).toEqual({ kind: 'plugin', plugin: 'tool-tasks' })
+
+    // The next turn collects the output through the generic task tool.
+    agent.send([{ type: 'text', text: 'collect it' }])
+    await waitForIdle(ctx, agent)
+    const readResult = findEvent(events(agent), 'tool/result', 'last')
+    expect(readResult.data.isError).toBe(false)
+    expect(resultText(readResult)).toContain('bg-ok')
+    expect(resultText(readResult)).toContain('[status: completed, exit code: 0]')
   })
 })
diff --git a/packages/bash/tool-bash/tests/tools.spec.ts b/packages/bash/tool-bash/tests/tools.spec.ts
index d2cbc643a4..c19846ab13 100644
--- a/packages/bash/tool-bash/tests/tools.spec.ts
+++ b/packages/bash/tool-bash/tests/tools.spec.ts
@@ -1,40 +1,29 @@
-import { chmodSync, mkdirSync, mkdtempSync } from 'node:fs'
+import { mkdtempSync } from 'node:fs'
 import { tmpdir } from 'node:os'
 import { join } from 'node:path'
 import { describe, expect, it, vi } from 'vitest'
 import { Context } from 'cordis'
 import { CallId } from '@deepseek-ai/dsh-llm'
-import { BashExecutor, BashTaskId, setSandboxMode } from '@deepseek-ai/dsh-bash'
-import type { BashExecRequest, BashExecSpec, BashRunResult, BashTask, BashTaskRead, OwnerToken } from '@deepseek-ai/dsh-bash'
-import { Session, SessionId } from '@deepseek-ai/dsh-session'
+import { BashExecutor } from '@deepseek-ai/dsh-bash'
+import type { BashExecRequest, BashExecSpec, BashProcess, BashProcessRead, BashRunResult } from '@deepseek-ai/dsh-bash'
 import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
 import ToolRegistry from '@deepseek-ai/dsh-tools'
 import AgentRegistry from '@deepseek-ai/dsh-agent'
 import type { Agent } from '@deepseek-ai/dsh-agent'
-import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
-import { SandboxBashExecutor } from '@deepseek-ai/dsh-bash-sandbox'
-import { SandboxProvider } from '@deepseek-ai/dsh-sandbox'
-import type { ConfinedArgv } from '@deepseek-ai/dsh-sandbox'
-import { LocalSandboxProvider } from '@deepseek-ai/dsh-sandbox-local'
+import SessionStore from '@deepseek-ai/dsh-session'
+import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
+import TaskService from '@deepseek-ai/dsh-tasks'
+import * as ToolTasks from '@deepseek-ai/dsh-tool-tasks'
 import ApprovalService from '@deepseek-ai/dsh-user-approval'
 import type { ApprovalOutcome } from '@deepseek-ai/dsh-user-approval'
+import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
 import * as ToolBash from '@deepseek-ai/dsh-tool-bash'
-import { renderResult } from '../src/render.ts'
+import { processOutcome } from '../src/background.ts'
+import { renderProcessRead, renderResult } from '../src/render.ts'
 
 const spillDir = mkdtempSync(join(tmpdir(), 'dsh-tool-bash-spec-'))
 
-// Pure-config passthrough runner (same knob the snapshot tier uses): skips the
-// profile args up to `--` and execs the command unconfined — deterministic
-// without a host bwrap.
-const PASSTHROUGH_RUNNER = ['bash', '-c', 'while [ "$1" != "--" ]; do shift; done; shift; exec "$@"', 'passthrough-runner']
-const PASSTHROUGH_RUNNER_CONFIG = {
-  runnerCommand: PASSTHROUGH_RUNNER,
-  // The script has no pre-exec failure path; the provider still requires an
-  // explicit dialect so a future script change cannot silently turn runner
-  // failure into an ordinary command result.
-  runnerFailureSignatures: ['passthrough-runner: profile rejected'],
-}
-
+/** Foreground-only harness: no task runtime (backgrounding fails loud here). */
 async function setup() {
   const ctx = new Context()
   await ctx.plugin(SystemPrompt)
@@ -46,35 +35,40 @@ async function setup() {
   return ctx
 }
 
+/** Full harness: the generic task runtime + its control surface, then the bash tool. */
+async function setupWithTasks() {
+  const ctx = new Context()
+  await ctx.plugin(SystemPrompt)
+  await ctx.plugin(ToolRegistry)
+  await ctx.plugin(AgentRegistry)
+  await ctx.plugin(TaskService)
+  await ctx.plugin(ToolTasks)
+  await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000, graceMs: 200 })
+  ;(ctx.bash as LocalBashExecutor).internals = { spillDir }
+  await ctx.plugin(ToolBash)
+  return ctx
+}
+
 /**
- * Build a fake {@link Agent} whose session token is `sessionId`, REGISTER it in
- * `ctx.agents` (the completion-notice path finds the owning agent by scanning
- * the registry for a matching `session.header.id`), and return it. The returned
- * agent is also passed to `execute` as `exec.agent` so it owns the spawned task.
- * The registration disposer is tracked so {@link unregisterFakeAgents} can drop
- * it (simulating the owning session disconnecting before a task completes).
+ * Build a fake {@link Agent} whose session token is `sessionId`, give it a
+ * dedicated lifecycle fiber for `Agent.ctx`, and register it in `ctx.agents`.
+ * The agent id is deliberately different from the session token so a
+ * wrong-field ownership match fails the test.
  */
-const fakeAgentDisposers = new Map<Context, (() => Promise<void> | void)[]>()
-function registerFakeAgent(ctx: Context, sessionId: string, inject: (...args: unknown[]) => void): Agent {
-  // A config agent has distinct registry (`agent.id`) and owner (`session.header.id`) tokens.
-  // Keeping them unequal makes notice lookup by the wrong field fail instead of passing by chance.
-  const agent = { id: `agent-${sessionId}`, inject, session: { header: { version: 0, id: sessionId, createdAt: 0 } } } as unknown as Agent
-  const dispose = ctx.agents.register(agent)
-  const list = fakeAgentDisposers.get(ctx) ?? []
-  list.push(dispose)
-  fakeAgentDisposers.set(ctx, list)
+function registerFakeAgent(ctx: Context, sessionId: string, inject: (...args: unknown[]) => void = () => {}): Agent {
+  const scopeFiber = ctx.plugin(() => {})
+  const agent = {
+    id: `agent-${sessionId}`,
+    ctx: scopeFiber.ctx,
+    inject,
+    session: { header: { version: 0, id: sessionId, createdAt: 0 } },
+  } as unknown as Agent
+  ctx.agents.register(agent)
   return agent
 }
-
-/** Unregister every fake agent in this ctx (simulate the owning session disconnecting). */
-function unregisterFakeAgents(ctx: Context): void {
-  for (const dispose of fakeAgentDisposers.get(ctx) ?? []) void dispose()
-  fakeAgentDisposers.delete(ctx)
-}
-
 let callCounter = 0
-function call(ctx: Context, name: string, args: unknown) {
-  return ctx.tools.execute({ callId: CallId(`call-${++callCounter}`), name, arguments: args })
+function call(ctx: Context, name: string, args: unknown, agent?: Agent) {
+  return ctx.tools.execute({ callId: CallId(`call-${++callCounter}`), name, arguments: args, ...agent ? { agent } : {} })
 }
 
 function text(result: { content: { type: string; text?: string }[] }): string {
@@ -98,56 +92,110 @@ async function callUntilText(
   throw new Error(`${name} output did not include ${JSON.stringify(expected)}; last text was ${JSON.stringify(last !== undefined ? text(last) : '')}`)
 }
 
-abstract class TestBashExecutor extends BashExecutor {
+class RecordingSandboxExecutor extends BashExecutor {
+  readonly modes: Array<string | undefined> = []
+
+  override get sandboxMode() {
+    return 'read-only' as const
+  }
+
   resolve(request: BashExecRequest): BashExecSpec {
     return {
       command: request.command,
       workdir: request.workdir ?? process.cwd(),
-      timeoutMs: request.timeoutMs ?? 0,
+      stdoutMaxBytes: request.stdoutMaxBytes ?? 64_000,
+      timeoutMs: request.timeoutMs ?? 1000,
       ...request.signal ? { signal: request.signal } : {},
-      owner: request.owner,
-      sandboxMode: request.sandboxMode,
+      sandboxMode: request.sandboxMode ?? 'read-only',
+    }
+  }
+
+  run(spec: BashExecSpec): Promise<BashRunResult> {
+    this.modes.push(spec.sandboxMode)
+    return Promise.resolve({
+      exitCode: 0,
+      signal: null,
+      timedOut: false,
+      aborted: false,
+      timeoutMs: spec.timeoutMs,
+      stdout: { text: 'ok', truncated: false },
+      stderr: { text: '', truncated: false },
+      sandbox: { mode: spec.sandboxMode ?? 'read-only', denied: false },
+    })
+  }
+
+  start(spec: BashExecSpec): BashProcess {
+    this.modes.push(spec.sandboxMode)
+    return {
+      status: 'completed',
+      exitCode: 0,
+      signal: null,
+      done: Promise.resolve(),
+      sandbox: { mode: spec.sandboxMode ?? 'read-only', denied: false },
+      readOutput: () => ({ delta: '', lossy: false }),
+      kill: () => false,
     }
   }
 }
 
-class LossyReadBashExecutor extends TestBashExecutor {
-  private readonly task: BashTask = {
-    id: BashTaskId('bash-lossy'),
-    status: 'running',
-    exitCode: null,
-    signal: null,
-    done: Promise.resolve(),
+/** Test executor that records whether the background start boundary was crossed. */
+class CountingStartExecutor extends BashExecutor {
+  starts = 0
+
+  resolve(request: BashExecRequest): BashExecSpec {
+    return {
+      command: request.command,
+      workdir: request.workdir ?? '/x',
+      timeoutMs: request.timeoutMs ?? 0,
+      stdoutMaxBytes: request.stdoutMaxBytes ?? 64_000,
+      sandboxMode: request.sandboxMode,
+    }
   }
 
-  run(): Promise<BashRunResult> {
-    return Promise.reject(new Error('not used'))
-  }
+  run(): Promise<BashRunResult> { return Promise.reject(new Error('unused')) }
 
-  start(): BashTask {
-    return this.task
+  start(): BashProcess {
+    this.starts += 1
+    return {
+      status: 'completed',
+      exitCode: 0,
+      signal: null,
+      done: Promise.resolve(),
+      readOutput: () => ({ delta: '', lossy: false }),
+      kill: () => false,
+    }
   }
+}
 
-  get(id: BashTaskId): BashTask | undefined {
-    return id === this.task.id ? this.task : undefined
-  }
+async function setupSandboxed(withApproval = false) {
+  const ctx = new Context()
+  await ctx.plugin(SystemPrompt)
+  await ctx.plugin(ToolRegistry)
+  await ctx.plugin(AgentRegistry)
+  await ctx.plugin(TaskService)
+  await ctx.plugin(ToolTasks)
+  await ctx.plugin(RecordingSandboxExecutor)
+  if (withApproval) await ctx.plugin(ApprovalService)
+  await ctx.plugin(ToolBash)
+  return { ctx, bash: ctx.bash as RecordingSandboxExecutor }
+}
 
-  ownerOf(): OwnerToken | undefined {
-    return undefined
-  }
-
-  list(): BashTask[] {
-    return [this.task]
-  }
-
-  readOutput(id: BashTaskId): BashTaskRead {
-    if (id !== this.task.id) throw new Error(`unknown bash task "${id}"`)
-    return { task: this.task, delta: 'tail', lossy: true }
-  }
-
-  kill(): boolean {
-    return false
-  }
+function sandboxAgent(mode?: 'read-only' | 'workspace-write' | 'danger-full-access', ctx?: Context): Agent {
+  const events: Array<{ type: string; data?: Record<string, unknown> }> = [{ type: 'turn/start' }]
+  if (mode !== undefined) events.push({ type: 'bash/sandbox-mode', data: { mode } })
+  return {
+    id: 'sandbox-agent',
+    ...ctx === undefined ? {} : { ctx: ctx.plugin(() => {}).ctx },
+    session: {
+      header: { version: 0, id: 'sandbox-session', createdAt: 0 },
+      events,
+      append: (type: string, data: Record<string, unknown>) => {
+        const event = { type, data }
+        events.push(event)
+        return event
+      },
+    },
+  } as unknown as Agent
 }
 
 describe('bash tool', () => {
@@ -222,7 +270,7 @@ describe('bash tool', () => {
     expect(text(result)).toMatch(/ENOENT/)
   })
 
-  it('surfaces aborts as isError', async () => {
+  it('surfaces foreground aborts as isError', async () => {
     const ctx = await setup()
     const controller = new AbortController()
     const pending = ctx.tools.execute({
@@ -237,7 +285,7 @@ describe('bash tool', () => {
     expect(text(result)).toMatch(/aborted/)
   })
 
-  // Type and required-key violations are now rejected by the harness
+  // Type and required-key violations are rejected by the harness
   // (defineTool validates against the SchemaSpec — the arg-validation RFC) before execute.
   it.each([
     [{}, /missing required property "command"/],
@@ -275,15 +323,18 @@ describe('bash tool', () => {
     expect(text(result)).toContain('tool execution arguments must be losslessly JSON-serializable')
   })
 
-  it('registers all three schemas in the system prompt assembly', async () => {
+  it('registers the bash schema with run_in_background exposed by default', async () => {
     const ctx = await setup()
-    const names = ctx.tools.schemas().map(schema => schema.name)
-    expect(names).toEqual(['bash', 'bash_output', 'bash_kill'])
-    const bashSchema = ctx.tools.schemas()[0]!
+    const schemas = ctx.tools.schemas()
+    expect(schemas.map(schema => schema.name)).toEqual(['bash'])
+    const bashSchema = schemas[0]!
     expect(bashSchema.parameters).toMatchObject({
       type: 'object',
       required: ['command', 'description'],
     })
+    expect(Object.keys(bashSchema.parameters.properties as Record<string, unknown>))
+      .toContain('run_in_background')
+    expect(bashSchema.description).toContain('task_output')
   })
 
   it('contributes the exit-code habit as its prompt section (guidance the descriptions cannot carry)', async () => {
@@ -308,7 +359,7 @@ describe('bash tool', () => {
     await ctx.plugin(ToolRegistry)
     await ctx.plugin(LocalBashExecutor, {})
     const fiber = await ctx.plugin(ToolBash)
-    expect(ctx.tools.schemas()).toHaveLength(3)
+    expect(ctx.tools.schemas()).toHaveLength(1)
     expect((await ctx.systemPrompt.assemble()).sections.map(s => s.name)).toEqual(['harness:identity', 'deployment:persona', 'tool:bash'])
     await fiber.dispose()
     expect(ctx.tools.schemas()).toHaveLength(0)
@@ -325,338 +376,352 @@ describe('bash tool', () => {
     expect(ctx.tools.schemas()).toHaveLength(0)
     await ctx.plugin(LocalBashExecutor, {})
     await new Promise(resolve => setTimeout(resolve, 0))
-    expect(ctx.tools.schemas()).toHaveLength(3)
+    expect(ctx.tools.schemas()).toHaveLength(1)
+  })
+
+  it('applies the built-in background default when apply() receives a bare config', async () => {
+    // Bypasses the schemastery defaults on purpose: apply() must stand on its
+    // own `?? true` fallback when embedded programmatically without the schema.
+    const ctx = new Context()
+    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(LocalBashExecutor, {})
+    ToolBash.apply(ctx, {})
+    const schema = ctx.tools.schemas()[0]!
+    expect(Object.keys(schema.parameters.properties as Record<string, unknown>))
+      .toContain('run_in_background')
   })
 })
 
-describe('background tools', () => {
-  it('bash with run_in_background returns a task id immediately', async () => {
-    const ctx = await setup()
-    const result = await call(ctx, 'bash', { command: 'sleep 0.2; echo bg-done', description: 'test command', run_in_background: true })
-    expect(result.isError).toBe(false)
-    expect(text(result)).toMatch(/^started background task bash-\d+$/)
+describe('background execution through the task runtime', () => {
+  it('run_in_background acks with the task id, readable through the REAL task_output tool', async () => {
+    const ctx = await setupWithTasks()
+    const started = await call(ctx, 'bash', { command: 'echo bg-ok', description: 'test command', run_in_background: true })
+    expect(started.isError).toBe(false)
+    expect(text(started)).toBe('started background task bash-1')
+
+    const read = await callUntilText(ctx, 'task_output', { task_id: 'bash-1' }, 'bg-ok')
+    expect(text(read)).toContain('bg-ok')
+    // A later read reports the terminal outcome in the generic status line.
+    const final = await callUntilText(ctx, 'task_output', { task_id: 'bash-1' }, '[status: completed, exit code: 0]')
+    expect(final.isError).toBe(false)
   })
 
-  it('bash_output polls incrementally and reports status', async () => {
-    const ctx = await setup()
-    const started = await call(ctx, 'bash', { command: 'echo first; sleep 1; echo second', description: 'test command', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
+  it('a running background task is killable through the REAL task_kill tool', async () => {
+    const ctx = await setupWithTasks()
+    await call(ctx, 'bash', { command: 'sleep 60', description: 'test command', run_in_background: true })
 
-    const first = await callUntilText(ctx, 'bash_output', { task_id: id }, 'first')
-    expect(text(first)).toContain('first')
-    expect(text(first)).toContain('[status: running]')
-
-    await ctx.bash.get(id)!.done
-    const second = await call(ctx, 'bash_output', { task_id: id })
-    expect(text(second)).toContain('second')
-    expect(text(second)).not.toContain('first')
-    expect(text(second)).toContain('[status: completed, exit code: 0]')
-
-    const third = await call(ctx, 'bash_output', { task_id: id })
-    expect(text(third)).toContain('(no new output)')
+    const killed = await call(ctx, 'task_kill', { task_id: 'bash-1' })
+    expect(text(killed)).toBe('requested cancellation of task bash-1')
+    // The cancel reached the process handle; the task settles as killed with
+    // the signal detail mapped by processOutcome.
+    const final = await call(ctx, 'task_output', { task_id: 'bash-1', wait: true })
+    expect(text(final)).toContain('[status: killed, signal: SIGTERM]')
   })
 
-  it('bash_output flags lossy reads with spill paths', async () => {
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(LocalBashExecutor, { maxOutputBytes: 100, graceMs: 200 })
-    ;(ctx.bash as LocalBashExecutor).internals = { spillDir }
-    await ctx.plugin(ToolBash)
+  it('a self-signal background exit is reported as killed through the REAL task_output tool', async () => {
+    const ctx = await setupWithTasks()
+    await call(ctx, 'bash', { command: 'kill -TERM $$', description: 'test command', run_in_background: true })
 
-    const started = await call(ctx, 'bash', { command: 'for i in $(seq 1 200); do printf "line-%04d\\n" $i; done', description: 'test command', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    await ctx.bash.get(id)!.done
-    const read = await call(ctx, 'bash_output', { task_id: id })
-    expect(text(read)).toContain('[some output was dropped from memory; full output: ')
+    const final = await call(ctx, 'task_output', { task_id: 'bash-1', wait: true })
+    expect(text(final)).toContain('[status: killed, signal: SIGTERM]')
   })
 
-  it('bash_output reports unavailable when a lossy read has no safe spill path', async () => {
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(LossyReadBashExecutor)
-    await ctx.plugin(ToolBash)
+  it('a background task started by an agent is registered with that agent as owner', async () => {
+    // The producer must forward exec.agent as the task owner.
+    const ctx = await setupWithTasks()
+    const agent = registerFakeAgent(ctx, 'sess-owner')
+    const started = await call(ctx, 'bash', { command: 'sleep 60', description: 'test command', run_in_background: true }, agent)
+    expect(text(started)).toBe('started background task bash-1')
 
-    const read = await call(ctx, 'bash_output', { task_id: 'bash-lossy' })
-    expect(text(read)).toBe('tail\n[some output was dropped from memory; full output: (unavailable)]\n[status: running]')
+    const anon = await call(ctx, 'task_output', { task_id: 'bash-1' })
+    expect(anon.isError).toBe(true)
+    expect(text(anon)).toMatch(/belongs to another session/)
+
+    const killed = await call(ctx, 'task_kill', { task_id: 'bash-1' }, agent)
+    expect(killed.isError).toBe(false)
+    await call(ctx, 'task_output', { task_id: 'bash-1', wait: true }, agent) // await settlement — no orphan
   })
 
-  it('bash_kill stops a running task; repeat reports already-finished', async () => {
-    const ctx = await setup()
-    const started = await call(ctx, 'bash', { command: 'sleep 60', description: 'test command', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-
-    const killed = await call(ctx, 'bash_kill', { task_id: id })
-    expect(text(killed)).toBe(`killed background task ${id}`)
-    await ctx.bash.get(id)!.done
-
-    const again = await call(ctx, 'bash_kill', { task_id: id })
-    expect(text(again)).toBe(`task ${id} had already finished`)
-
-    const status = await call(ctx, 'bash_output', { task_id: id })
-    expect(text(status)).toContain('[status: killed by SIGTERM]')
-  })
-
-  it('unknown task ids are isError for both tools', async () => {
-    const ctx = await setup()
-    const read = await call(ctx, 'bash_output', { task_id: 'bash-999' })
-    expect(read.isError).toBe(true)
-    expect(text(read)).toMatch(/unknown bash task/)
-    const kill = await call(ctx, 'bash_kill', { task_id: 'bash-999' })
-    expect(kill.isError).toBe(true)
-  })
-
-  it.each([
-    ['bash_output', {}, /missing required property "task_id"/],
-    ['bash_output', { task_id: 9 }, /"task_id" must be a string/],
-    ['bash_kill', { task_id: '' }, /invalid task_id/],
-  ])('%s rejects invalid task_id %j', async (tool, args, pattern) => {
-    const ctx = await setup()
-    const result = await call(ctx, tool, args)
+  it('fails loud when the task runtime is not loaded', async () => {
+    const ctx = await setup() // no TaskService / ToolTasks
+    const result = await call(ctx, 'bash', { command: 'sleep 60', description: 'test command', run_in_background: true })
     expect(result.isError).toBe(true)
-    expect(text(result)).toMatch(pattern)
+    expect(text(result)).toContain('background tasks unavailable: load @deepseek-ai/dsh-tasks and @deepseek-ai/dsh-tool-tasks')
   })
 
-  it('injects a completion notice into the owning agent (found via the registry by session token)', async () => {
-    const ctx = await setup()
-    const inject = vi.fn()
-    // Notices look up the agent in ctx.agents by session token, so passing it to execute is not
-    // enough: the fake must be registered with a matching `session.header.id`.
-    const agent = registerFakeAgent(ctx, 'bg', inject)
+  it('a pre-aborted call refuses to start: isError, no process spawned', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(AgentRegistry)
+    await ctx.plugin(TaskService)
+    await ctx.plugin(ToolTasks)
+    await ctx.plugin(CountingStartExecutor)
+    await ctx.plugin(ToolBash)
 
-    const started = await ctx.tools.execute({
-      callId: CallId('call-bg'),
+    const controller = new AbortController()
+    controller.abort()
+    const result = await ctx.tools.execute({
+      callId: CallId('call-pre-aborted'),
       name: 'bash',
-      arguments: { command: 'true', description: 'test command', run_in_background: true },
-      agent,
+      arguments: { command: 'sleep 60', description: 'test command', run_in_background: true },
+      signal: controller.signal,
     })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    await ctx.bash.get(id)!.done
-
-    expect(inject).toHaveBeenCalledTimes(1)
-    const [content, options] = inject.mock.calls[0] as [
-      { type: string; text: string }[],
-      { source: { kind: string; plugin: string } },
-    ]
-    expect(content[0]!.text).toContain(`background bash task ${id} finished`)
-    expect(content[0]!.text).toContain('bash_output')
-    expect(options.source).toEqual({ kind: 'plugin', plugin: 'tool-bash' })
+    expect(result.isError).toBe(true)
+    expect(text(result)).toContain('command aborted')
+    expect((ctx.bash as CountingStartExecutor).starts).toBe(0)
   })
 
-  it('swallows ONLY the disposed-agent inject error', async () => {
-    const ctx = await setup()
-    const agent = registerFakeAgent(ctx, 'bg', () => { throw new Error('agent "x" is disposed') })
+  it('never spawns the process when tasks.start preflight throws (no orphan, by construction)', async () => {
+    // With no control surface, task preflight fails before the executor can spawn.
+    const ctx = new Context()
+    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(AgentRegistry)
+    await ctx.plugin(TaskService)
+    await ctx.plugin(CountingStartExecutor)
+    await ctx.plugin(ToolBash)
 
-    const started = await ctx.tools.execute({
-      callId: CallId('call-bg2'),
-      name: 'bash',
-      arguments: { command: 'true', description: 'test command', run_in_background: true },
-      agent,
-    })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    await expect(ctx.bash.get(id)!.done).resolves.toBeUndefined()
+    const result = await call(ctx, 'bash', { command: 'sleep 60', description: 'test command', run_in_background: true })
+    expect(result.isError).toBe(true)
+    expect(text(result)).toContain('no control surface is attached')
+    // Declare-then-execute: the failed preflight means no process ever ran.
+    expect((ctx.bash as CountingStartExecutor).starts).toBe(0)
   })
 
-  it('rethrows a non-disposed inject failure (not blindly swallowed)', async () => {
-    const ctx = await setup()
-    // A real bug in inject (not the benign disposed race) must surface — the
-    // base-class notifier contains it (logs, does not reject task.done), but
-    // the listener itself must have thrown rather than silently eaten it.
-    const errorSpy = vi.spyOn(console, 'error').mockImplementation(() => undefined)
-    try {
-      const agent = registerFakeAgent(ctx, 'bg', () => { throw new Error('unexpected inject bug') })
+  it('enableRunInBackground: false removes the parameter and flips the description', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(LocalBashExecutor, {})
+    await ctx.plugin(ToolBash, { enableRunInBackground: false })
 
-      const started = await ctx.tools.execute({
-        callId: CallId('call-bg3'),
-        name: 'bash',
-        arguments: { command: 'true', description: 'test command', run_in_background: true },
-        agent,
-      })
-      const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-      await ctx.bash.get(id)!.done
-      // notifyTaskDone caught and logged the rethrown error.
-      expect(errorSpy).toHaveBeenCalled()
-      const logged = errorSpy.mock.calls.flat().some(arg => arg instanceof Error && arg.message === 'unexpected inject bug')
-      expect(logged).toBe(true)
-    } finally {
-      errorSpy.mockRestore()
+    const schema = ctx.tools.schemas().find(s => s.name === 'bash')!
+    expect(Object.keys(schema.parameters.properties as Record<string, unknown>))
+      .toEqual(['command', 'description', 'timeoutMs', 'workdir'])
+    expect(schema.description).toContain('Background execution is not available')
+    expect(schema.description).not.toContain('run_in_background')
+    // The registry-held definition agrees (schema and capability never disagree).
+    const parameters = ctx.tools.get('bash')!.parameters as { properties: Record<string, unknown> }
+    expect('run_in_background' in parameters.properties).toBe(false)
+
+    // Schema omission is advertising; execution must also enforce the opt-out.
+    const forced = await call(ctx, 'bash', { command: 'echo hi', description: 'test command', run_in_background: true })
+    expect(forced.isError).toBe(true)
+    expect(text(forced)).toContain('run_in_background is disabled for this deployment')
+    const foreground = await call(ctx, 'bash', { command: 'echo hi', description: 'test command' })
+    expect(foreground.isError).toBe(false)
+  })
+})
+
+describe('sandbox escalation through the generic task producer', () => {
+  const escalate = {
+    command: 'true',
+    description: 'test escalation',
+    sandbox_permissions: 'workspace-write',
+    justification: 'the command needs workspace writes',
+  }
+
+  it('advertises the sandbox fields and validates their pairing', async () => {
+    const { ctx } = await setupSandboxed()
+    const schema = ctx.tools.schemas().find(item => item.name === 'bash')!
+    const properties = schema.parameters.properties as Record<string, { enum?: string[] }>
+    expect(properties['sandbox_permissions']?.enum).toEqual(['workspace-write', 'danger-full-access'])
+    expect(schema.description).toContain('approval prompt')
+
+    for (const args of [
+      { command: 'true', description: 'd', sandbox_permissions: 'workspace-write' },
+      { command: 'true', description: 'd', justification: 'why' },
+      { command: 'true', description: 'd', sandbox_permissions: 'workspace-write', justification: ' ' },
+    ]) {
+      expect((await call(ctx, 'bash', args)).isError).toBe(true)
     }
   })
 
-  it('drops the notice cleanly when the owning agent is gone from the registry by completion', async () => {
-    // Host-scoped bash tasks can outlive a per-session agent after an ACP disconnect. The task
-    // retains its owner token, but with no matching live agent the notice is dropped without error.
-    const ctx = await setup()
-    const inject = vi.fn()
-    const agent = registerFakeAgent(ctx, 'bg', inject)
-    const started = await ctx.tools.execute({
-      callId: CallId('call-bg4'),
-      name: 'bash',
-      arguments: { command: 'true', description: 'test command', run_in_background: true },
-      agent,
+  it('rejects injected escalation without a sandbox and non-widening escalation without prompting', async () => {
+    const plain = await setup()
+    expect(text(await call(plain, 'bash', escalate))).toContain('not available in this composition')
+
+    const { ctx } = await setupSandboxed(true)
+    const prompted = vi.fn()
+    ctx.on('approval/request', () => { prompted(); return Promise.resolve<ApprovalOutcome>('allowed-once') })
+    const result = await call(ctx, 'bash', { ...escalate, sandbox_permissions: 'workspace-write' }, sandboxAgent('workspace-write'))
+    expect(text(result)).toContain('not strictly wider')
+    expect(prompted).not.toHaveBeenCalled()
+
+    const malformed = sandboxAgent()
+    ;(malformed.session.events as unknown as Array<{ type: string; data: { mode: string } }>).push({
+      type: 'bash/sandbox-mode',
+      data: { mode: 'unknown-mode' },
     })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    // Unregister the agent BEFORE the task completes (simulate disconnect).
-    unregisterFakeAgents(ctx)
-    await expect(ctx.bash.get(id)!.done).resolves.toBeUndefined()
-    expect(inject).not.toHaveBeenCalled()
+    expect(text(await call(ctx, 'bash', escalate, malformed))).toContain('not strictly wider')
   })
 
-  it('does not notify when no agent owned the task', async () => {
-    const ctx = await setup()
-    const started = await call(ctx, 'bash', { command: 'true', description: 'test command', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    await expect(ctx.bash.get(id)!.done).resolves.toBeUndefined()
+  it('fails closed when approval cannot be routed', async () => {
+    const withoutService = await setupSandboxed()
+    expect(text(await call(withoutService.ctx, 'bash', escalate, sandboxAgent()))).toContain('no approval service')
+
+    const withService = await setupSandboxed(true)
+    expect(text(await call(withService.ctx, 'bash', escalate))).toContain('no agent to route')
+    expect(text(await call(withService.ctx, 'bash', escalate, sandboxAgent()))).toContain('no approval channel')
+  })
+
+  it.each([
+    ['rejected', 'user rejected'],
+    ['cancelled', 'was cancelled'],
+  ] as const)('maps an approval %s to its distinct failure', async (outcome, message) => {
+    const { ctx, bash } = await setupSandboxed(true)
+    ctx.on('approval/request', () => Promise.resolve<ApprovalOutcome>(outcome))
+    const result = await call(ctx, 'bash', escalate, sandboxAgent())
+    expect(text(result)).toContain(message)
+    expect(bash.modes).toEqual([])
+  })
+
+  it('runs a granted foreground or background call under the approved mode', async () => {
+    const { ctx, bash } = await setupSandboxed(true)
+    ctx.on('approval/request', () => Promise.resolve<ApprovalOutcome>('allowed-once'))
+    const agent = sandboxAgent(undefined, ctx)
+    ctx.agents.register(agent)
+    const foreground = await ctx.tools.execute({
+      callId: CallId('sandbox-signal'),
+      name: 'bash',
+      arguments: escalate,
+      agent,
+      signal: new AbortController().signal,
+    })
+    expect(foreground.isError).toBe(false)
+    const background = await call(ctx, 'bash', { ...escalate, run_in_background: true }, agent)
+    expect(text(background)).toBe('started background task bash-1')
+    expect(bash.modes).toEqual(['workspace-write', 'workspace-write'])
+  })
+
+  it('uses the session override for ordinary calls and evaluates widening against it', async () => {
+    const { ctx, bash } = await setupSandboxed(true)
+    const agent = sandboxAgent('workspace-write')
+    await call(ctx, 'bash', { command: 'true', description: 'ordinary' }, agent)
+    ctx.on('approval/request', () => Promise.resolve<ApprovalOutcome>('allowed-once'))
+    await call(ctx, 'bash', { ...escalate, sandbox_permissions: 'danger-full-access' }, agent)
+    expect(bash.modes).toEqual(['workspace-write', 'danger-full-access'])
+  })
+
+  it('keeps the exhaustiveness backstop for a rogue approval implementation', async () => {
+    const { ctx } = await setupSandboxed(true)
+    ctx.approval.request = () => Promise.resolve('rogue' as ApprovalOutcome)
+    const result = await call(ctx, 'bash', escalate, sandboxAgent())
+    expect(text(result)).toContain('unreachable variant in ApprovalOutcome')
   })
 })
 
-describe('background task ownership (cross-session isolation)', () => {
-  /** Run a tool on behalf of a specific agent (sets exec.agent). */
-  function callAs(ctx: Context, agent: import('@deepseek-ai/dsh-agent').Agent | undefined, name: string, args: unknown) {
-    return ctx.tools.execute({ callId: CallId(`own-${++callCounter}`), name, arguments: args, ...agent ? { agent } : {} })
+describe('renderProcessRead', () => {
+  const base: BashProcessRead = { delta: 'out\n', lossy: false }
+
+  it('returns the delta verbatim for a lossless read', () => {
+    expect(renderProcessRead(base)).toBe('out\n')
+    expect(renderProcessRead({ delta: '', lossy: false })).toBe('')
+  })
+
+  it('appends the loss notice with the available spill paths', () => {
+    expect(renderProcessRead({ ...base, lossy: true, stdoutSpillPath: '/spill/out.log' }))
+      .toBe('out\n[some output was dropped from memory; full output: /spill/out.log]')
+    expect(renderProcessRead({ ...base, lossy: true, stdoutSpillPath: '/spill/out.log', stderrSpillPath: '/spill/err.log' }))
+      .toBe('out\n[some output was dropped from memory; full output: /spill/out.log, /spill/err.log]')
+  })
+
+  it('reports (unavailable) when a lossy read has no safe spill path', () => {
+    expect(renderProcessRead({ ...base, lossy: true }))
+      .toBe('out\n[some output was dropped from memory; full output: (unavailable)]')
+  })
+
+  it('an empty lossy delta is the notice alone', () => {
+    expect(renderProcessRead({ delta: '', lossy: true, stderrSpillPath: '/spill/err.log' }))
+      .toBe('[some output was dropped from memory; full output: /spill/err.log]')
+  })
+
+  it('inserts the separating newline only when the delta lacks one', () => {
+    expect(renderProcessRead({ delta: 'tail', lossy: true }))
+      .toBe('tail\n[some output was dropped from memory; full output: (unavailable)]')
+    expect(renderProcessRead({ delta: 'tail\n', lossy: true }))
+      .toBe('tail\n[some output was dropped from memory; full output: (unavailable)]')
+  })
+
+  it('appends settled sandbox denial and runner-failure facts', () => {
+    expect(renderProcessRead(base, { mode: 'read-only', denied: true }, ['workspace-write']))
+      .toContain('[sandbox: escalation available')
+    expect(renderProcessRead({ delta: 'tail', lossy: false }, { mode: 'read-only', denied: true }))
+      .toBe('tail\n[sandbox: file access denied under read-only mode]')
+    const runner = renderProcessRead(
+      { delta: '', lossy: false },
+      { mode: 'workspace-write', denied: true, runnerFailed: true },
+      ['danger-full-access'],
+    )
+    expect(runner).toContain('sandbox runner itself failed under workspace-write mode')
+    expect(runner).not.toContain('file access denied')
+  })
+})
+
+describe('processOutcome', () => {
+  function settled(over: Partial<BashProcess>): BashProcess {
+    return {
+      status: 'completed',
+      exitCode: 0,
+      signal: null,
+      done: Promise.resolve(),
+      readOutput: () => ({ delta: '', lossy: false }),
+      kill: () => false,
+      ...over,
+    }
   }
-  // Ownership uses `session.header.id`, not object identity. Distinct ids keep the isolation tests
-  // from passing accidentally because every fake produced the same owner token.
-  const fakeAgent = (sessionId: string) =>
-    ({ inject: () => undefined, session: { header: { version: 0, id: sessionId, createdAt: 0 } } }) as unknown as import('@deepseek-ai/dsh-agent').Agent
 
-  it('rejects bash_output/bash_kill for a task owned by a DIFFERENT session token', async () => {
-    const ctx = await setup()
-    const a = fakeAgent('sess-a')
-    const b = fakeAgent('sess-b')
-    // Agent A starts a long-running background task.
-    const started = await callAs(ctx, a, 'bash', { command: 'sleep 60', description: 'bg', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-
-    // Agent B (a different session token) cannot read or kill A's task.
-    const readByB = await callAs(ctx, b, 'bash_output', { task_id: id })
-    expect(readByB.isError).toBe(true)
-    expect(text(readByB)).toMatch(/belongs to another session/)
-    const killByB = await callAs(ctx, b, 'bash_kill', { task_id: id })
-    expect(killByB.isError).toBe(true)
-    expect(text(killByB)).toMatch(/belongs to another session/)
-
-    // The task is still running (B's kill did nothing) — A can still kill it.
-    const killByA = await callAs(ctx, a, 'bash_kill', { task_id: id })
-    expect(killByA.isError).toBe(false)
-    expect(text(killByA)).toBe(`killed background task ${id}`)
+  it('maps a signal-killed process to killed with the signal detail', () => {
+    expect(processOutcome(settled({ status: 'killed', signal: 'SIGTERM' })))
+      .toEqual({ status: 'killed', detail: 'signal: SIGTERM' })
   })
 
-  it('a DIFFERENT Agent object with the SAME session token may access the task (ownership is by token, not object identity)', async () => {
-    // Ownership fences by session.header.id, NOT Agent object identity. Two
-    // distinct Agent objects sharing one session token (e.g. an agent re-created
-    // on the same session) are the SAME owner.
-    const ctx = await setup()
-    const a1 = fakeAgent('sess-shared')
-    const a2 = fakeAgent('sess-shared') // distinct object, same token
-    const started = await callAs(ctx, a1, 'bash', { command: 'sleep 60', description: 'bg', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    const readByA2 = await callAs(ctx, a2, 'bash_output', { task_id: id })
-    expect(readByA2.isError).toBe(false)
-    await callAs(ctx, a1, 'bash_kill', { task_id: id }) // cleanup
+  it('maps a killed process without a recorded signal (kill raced exit / spawn failure)', () => {
+    expect(processOutcome(settled({ status: 'killed', exitCode: null })))
+      .toEqual({ status: 'killed', detail: 'killed before exit' })
   })
 
-  it('the no-agent (non-loop) caller cannot access an owned task', async () => {
-    const ctx = await setup()
-    const a = fakeAgent('sess-a')
-    const started = await callAs(ctx, a, 'bash', { command: 'sleep 60', description: 'bg', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    // A call with no exec.agent has no token → cannot prove ownership of an owned task.
-    const read = await callAs(ctx, undefined, 'bash_output', { task_id: id })
-    expect(read.isError).toBe(true)
-    expect(text(read)).toMatch(/belongs to another session/)
-    await callAs(ctx, a, 'bash_kill', { task_id: id }) // cleanup
+  it('maps a completed process to its exit code', () => {
+    expect(processOutcome(settled({ exitCode: 3 })))
+      .toEqual({ status: 'completed', detail: 'exit code: 3' })
   })
 
-  it('an UNOWNED task (started with no agent) is accessible to anyone', async () => {
-    const ctx = await setup()
-    // Started by a non-loop caller (no exec.agent) → no owner token recorded.
-    const started = await callAs(ctx, undefined, 'bash', { command: 'sleep 60', description: 'bg', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    // Any agent (and the no-agent caller) may read/kill it.
-    const read = await callAs(ctx, fakeAgent('sess-x'), 'bash_output', { task_id: id })
-    expect(read.isError).toBe(false)
-    const killed = await callAs(ctx, undefined, 'bash_kill', { task_id: id })
-    expect(killed.isError).toBe(false)
-  })
-
-  it('the owner can still access its task AFTER it completes (owner token persists on the task)', async () => {
-    const ctx = await setup()
-    const a = fakeAgent('sess-a')
-    const b = fakeAgent('sess-b')
-    const started = await callAs(ctx, a, 'bash', { command: 'echo done', description: 'bg', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    await ctx.bash.get(id)!.done
-    // Completion does NOT clear ownership: B is still rejected, A still allowed.
-    const readByB = await callAs(ctx, b, 'bash_output', { task_id: id })
-    expect(readByB.isError).toBe(true)
-    expect(text(readByB)).toMatch(/belongs to another session/)
-    const readByA = await callAs(ctx, a, 'bash_output', { task_id: id })
-    expect(readByA.isError).toBe(false)
-  })
-
-  it('ownership SURVIVES an independent tool-bash HMR reload (token lives on the executor)', async () => {
-    // The executor task owns the token, so reloading only tool-bash preserves ownership. A
-    // plugin-local map would lose it and incorrectly expose the task to agent B.
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000, graceMs: 200 })
-    ;(ctx.bash as LocalBashExecutor).internals = { spillDir }
-    const fiber = await ctx.plugin(ToolBash)
-
-    const a = fakeAgent('sess-a')
-    const b = fakeAgent('sess-b')
-    const started = await callAs(ctx, a, 'bash', { command: 'sleep 60', description: 'bg', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    // Before reload: B is rejected (A owns it).
-    expect((await callAs(ctx, b, 'bash_output', { task_id: id })).isError).toBe(true)
-
-    // Reload ONLY tool-bash; the executor and its running task (with its owner
-    // token) survive.
-    await fiber.dispose()
-    await ctx.plugin(ToolBash)
-    expect(ctx.bash.get(id)?.status).toBe('running')
-    expect(ctx.bash.ownerOf(id)).toBe('sess-a')
-
-    // After reload, ownership is INTACT → B is STILL rejected.
-    expect((await callAs(ctx, b, 'bash_output', { task_id: id })).isError).toBe(true)
-    await callAs(ctx, a, 'bash_kill', { task_id: id }) // cleanup
+  it('defensively reads a null exit code as 0 (handle shapes from other executors)', () => {
+    expect(processOutcome(settled({ exitCode: null })))
+      .toEqual({ status: 'completed', detail: 'exit code: 0' })
   })
 })
 
 describe('session-cwd routing (per-session workdir)', () => {
-  function callAs(ctx: Context, agent: import('@deepseek-ai/dsh-agent').Agent | undefined, args: unknown) {
-    return ctx.tools.execute({ callId: CallId(`cwd-${++callCounter}`), name: 'bash', arguments: args, ...agent ? { agent } : {} })
-  }
   // An agent whose session header carries a cwd (what session/new records).
   const agentInCwd = (cwd: string) =>
-    ({ inject: () => undefined, session: { header: { version: 0, id: 'c', createdAt: 0, cwd } } }) as unknown as import('@deepseek-ai/dsh-agent').Agent
+    ({ inject: () => undefined, session: { header: { version: 0, id: 'c', createdAt: 0, cwd } } }) as unknown as Agent
 
   it('defaults bash to the agent\'s session cwd (not the server launch dir)', async () => {
     const ctx = await setup()
-    const result = await callAs(ctx, agentInCwd('/tmp'), { command: 'pwd', description: 'pwd' })
+    const result = await call(ctx, 'bash', { command: 'pwd', description: 'pwd' }, agentInCwd('/tmp'))
     expect(text(result).trim()).toMatch(/\/tmp$/)
   })
 
   it('an explicit absolute workdir overrides the session cwd', async () => {
     const ctx = await setup()
-    const result = await callAs(ctx, agentInCwd('/'), { command: 'pwd', description: 'pwd', workdir: '/tmp' })
+    const result = await call(ctx, 'bash', { command: 'pwd', description: 'pwd', workdir: '/tmp' }, agentInCwd('/'))
     expect(text(result).trim()).toMatch(/\/tmp$/)
   })
 
   it('a relative workdir is resolved against the session cwd', async () => {
     const ctx = await setup()
     // session cwd /usr + relative 'bin' → /usr/bin
-    const result = await callAs(ctx, agentInCwd('/usr'), { command: 'pwd', description: 'pwd', workdir: 'bin' })
+    const result = await call(ctx, 'bash', { command: 'pwd', description: 'pwd', workdir: 'bin' }, agentInCwd('/usr'))
     expect(text(result).trim()).toMatch(/\/usr\/bin$/)
   })
 
   it('two sessions with different cwds each run bash in their own dir', async () => {
     const ctx = await setup()
-    const inUsr = await callAs(ctx, agentInCwd('/usr'), { command: 'pwd', description: 'pwd' })
-    const inTmp = await callAs(ctx, agentInCwd('/tmp'), { command: 'pwd', description: 'pwd' })
+    const inUsr = await call(ctx, 'bash', { command: 'pwd', description: 'pwd' }, agentInCwd('/usr'))
+    const inTmp = await call(ctx, 'bash', { command: 'pwd', description: 'pwd' }, agentInCwd('/tmp'))
     expect(text(inUsr).trim()).toMatch(/\/usr$/)
     expect(text(inTmp).trim()).toMatch(/\/tmp$/)
   })
@@ -718,34 +783,22 @@ describe('renderResult', () => {
     expect(renderResult({ ...base, stdout: { text: 'tail', truncated: true } }))
       .toBe('tail\n[output truncated; full output: (unavailable)]')
   })
-})
 
-describe('status lines', () => {
-  it('reports kills without a recorded signal (executor raced process exit)', async () => {
-    const ctx = await setup()
-    const started = await call(ctx, 'bash', { command: 'sleep 60', description: 'test command', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    const task = ctx.bash.get(id)!
-
-    await call(ctx, 'bash_kill', { task_id: id })
-    await task.done
-    // Simulate the variant where the close event carried no signal.
-    task.signal = null
-    const read = await call(ctx, 'bash_output', { task_id: id })
-    expect(text(read)).toContain('[status: killed]')
-  })
-
-  it('reports completed tasks with a null exit code as exit 0', async () => {
-    const ctx = await setup()
-    const started = await call(ctx, 'bash', { command: 'true', description: 'test command', run_in_background: true })
-    const id = BashTaskId(/task (bash-\d+)/.exec(text(started))![1]!)
-    const task = ctx.bash.get(id)!
-    await task.done
-    // Defensive: completed tasks always carry an exit code in practice; the
-    // ?? 0 fallback covers task shapes from other executor implementations.
-    task.exitCode = null
-    const read = await call(ctx, 'bash_output', { task_id: id })
-    expect(text(read)).toContain('[status: completed, exit code: 0]')
+  it('reports sandbox denials before exit status and hints only when escalation is advertised', () => {
+    const result: BashRunResult = {
+      exitCode: 1,
+      signal: null,
+      timedOut: false,
+      aborted: false,
+      timeoutMs: 1000,
+      stdout: { text: '', truncated: false },
+      stderr: { text: 'denied', truncated: false },
+      sandbox: { mode: 'read-only', denied: true },
+    }
+    expect(renderResult(result)).toMatch(/denied under read-only mode\]\n\[exit code: 1\]$/)
+    expect(renderResult(result, ['workspace-write'])).toContain('[sandbox: escalation available')
+    expect(renderResult({ ...result, sandbox: { mode: 'read-only', denied: false } }, ['workspace-write']))
+      .not.toContain('[sandbox:')
   })
 })
 
@@ -871,14 +924,6 @@ describe('tool-owned UI presentation (presentCall / presentResult)', () => {
     })).toBeUndefined()
   })
 
-  it('bash_output / bash_kill presentCall: a readable task-scoped title, task id as rawInput', async () => {
-    const ctx = await setup()
-    expect(ctx.tools.get('bash_output')!.presentCall!({ task_id: 'bash-3' }))
-      .toEqual({ card: 'generic', title: 'Read output from background task bash-3', kind: 'execute', rawInput: 'bash-3' })
-    expect(ctx.tools.get('bash_kill')!.presentCall!({ task_id: 'bash-3' }))
-      .toEqual({ card: 'generic', title: 'Kill background task bash-3', kind: 'execute', rawInput: 'bash-3' })
-  })
-
   it('presentCall validates softly: malformed args (missing required description) return undefined, never throw', async () => {
     const ctx = await setup()
     // `defineTool` soft-validates replayed logged args before presentation. Invalid shapes return
@@ -888,11 +933,20 @@ describe('tool-owned UI presentation (presentCall / presentResult)', () => {
 })
 
 describe('the model-facing bash tool builds its request from named args only (no {...args} forward)', () => {
+  const recordingDshHome = join(spillDir, 'dsh-home')
+
   /**
-   * Records requests passed to `resolve()` so tests can prove the model-facing tool forwards only
-   * named arguments. It intentionally exposes neither `stdin` nor `env`; this catches a future
-   * `...args` spread into the post-scrub env merge. The credential scrub remains the security
-   * boundary; see the bash stdin/env RFC. Foreground `run()` is canned and `start()` is unused.
+   * Records every {@link BashExecRequest} the consumer hands to `resolve()`, so a
+   * test can assert what the model-facing tool DID and DID NOT forward. The `bash`
+   * tool does not expose trusted-plugin fields (`stdoutMaxBytes`, `stdin`, or
+   * `env`) as parameters, so it must build its request from named args only and
+   * never spread unknown tool-call keys into it. This guard's job is to catch a
+   * future refactor that blindly forwards `...args` — which would silently thread
+   * model input into the post-scrub `env` merge or per-run capture budget — NOT
+   * to defend a trust boundary
+   * (the credential scrub in dsh-bash-local is the security control; see the
+   * bash-stdin-env RFC). Foreground `run()` returns a canned result; `start()`
+   * hands back an already-settled fake handle so the task registration completes.
    */
   class RecordingBashExecutor extends BashExecutor {
     readonly requests: BashExecRequest[] = []
@@ -902,10 +956,11 @@ describe('the model-facing bash tool builds its request from named args only (no
         command: request.command,
         workdir: request.workdir ?? process.cwd(),
         timeoutMs: request.timeoutMs ?? 0,
+        stdoutMaxBytes: request.stdoutMaxBytes ?? 64_000,
         ...request.signal ? { signal: request.signal } : {},
         ...request.stdin !== undefined ? { stdin: request.stdin } : {},
         ...request.env !== undefined ? { env: request.env } : {},
-        owner: request.owner,
+        ...request.dshEnv !== undefined ? { dshEnv: request.dshEnv } : {},
         sandboxMode: request.sandboxMode,
       }
     }
@@ -915,25 +970,139 @@ describe('the model-facing bash tool builds its request from named args only (no
         stdout: { text: 'ok', truncated: false }, stderr: { text: '', truncated: false },
       })
     }
-    start(): BashTask { throw new Error('unused') }
-    get(): BashTask | undefined { return undefined }
-    ownerOf(): OwnerToken | undefined { return undefined }
-    list(): BashTask[] { return [] }
-    readOutput(): BashTaskRead { throw new Error('unused') }
-    kill(): boolean { return false }
+    start(): BashProcess {
+      return {
+        status: 'completed',
+        exitCode: 0,
+        signal: null,
+        done: Promise.resolve(),
+        readOutput: () => ({ delta: '', lossy: false }),
+        kill: () => false,
+      }
+    }
   }
 
-  async function setupRecording() {
+  async function setupRecording(withJsonl = false) {
     const ctx = new Context()
     await ctx.plugin(SystemPrompt)
     await ctx.plugin(ToolRegistry)
     await ctx.plugin(AgentRegistry)
+    if (withJsonl) {
+      await ctx.plugin(SessionStore)
+      await ctx.plugin(SessionPersistenceJsonl, { root: join(spillDir, 'jsonl') })
+    }
+    await ctx.plugin(TaskService)
+    await ctx.plugin(ToolTasks)
     await ctx.plugin(RecordingBashExecutor)
-    await ctx.plugin(ToolBash)
+    await ctx.plugin(ToolBash, { dshHome: recordingDshHome })
     return { ctx, bash: ctx.bash as RecordingBashExecutor }
   }
 
-  it('does not forward env/stdin even when the model includes them as extra arguments', async () => {
+  it('describes the managed harness environment namespace to the model', async () => {
+    const { ctx } = await setupRecording()
+    const description = ctx.tools.get('bash')?.description ?? ''
+    expect(description).toContain('$DSH_*')
+    expect(description).not.toContain('DSH_SESSION_JSONL')
+  })
+
+  it('injects the session id and JSONL target path into a foreground request', async () => {
+    const { ctx, bash } = await setupRecording(true)
+    const agent = registerFakeAgent(ctx, 'request-fg', () => undefined)
+    const path = ctx.sessionPersistence.locate(agent.session.header)?.path
+
+    await ctx.tools.execute({
+      callId: CallId('session-env-fg'),
+      name: 'bash',
+      arguments: { command: 'true', description: 'run command' },
+      agent,
+    })
+
+    expect(bash.requests[0]?.dshEnv).toEqual({
+      DSH_HOME: recordingDshHome,
+      DSH_SESSION_ID: 'request-fg',
+      DSH_SESSION_JSONL: path,
+      DSH_SHELL: '1',
+    })
+  })
+
+  it('injects the same trusted variables into a background request without forwarding model env', async () => {
+    const { ctx, bash } = await setupRecording(true)
+    const agent = registerFakeAgent(ctx, 'request-bg', () => undefined)
+    const path = ctx.sessionPersistence.locate(agent.session.header)?.path
+
+    await ctx.tools.execute({
+      callId: CallId('session-env-bg'),
+      name: 'bash',
+      arguments: {
+        command: 'sleep 1',
+        description: 'run command',
+        run_in_background: true,
+        env: { DSH_SESSION_ID: 'spoofed', DSH_SESSION_JSONL: '/tmp/spoofed' },
+      },
+      agent,
+    })
+
+    expect(bash.requests[0]?.env).toBeUndefined()
+    expect(bash.requests[0]?.dshEnv).toEqual({
+      DSH_HOME: recordingDshHome,
+      DSH_SESSION_ID: 'request-bg',
+      DSH_SESSION_JSONL: path,
+      DSH_SHELL: '1',
+    })
+  })
+
+  it('injects built-ins and the stable session id when no JSONL locator is available', async () => {
+    const { ctx, bash } = await setupRecording()
+    const agent = registerFakeAgent(ctx, 'request-id-only', () => undefined)
+    const ambient = process.env.DSH_SESSION_ID
+
+    await ctx.tools.execute({
+      callId: CallId('session-env-id-only'),
+      name: 'bash',
+      arguments: { command: 'true', description: 'run command' },
+      agent,
+    })
+
+    expect(bash.requests[0]?.dshEnv).toEqual({
+      DSH_HOME: recordingDshHome,
+      DSH_SESSION_ID: 'request-id-only',
+      DSH_SHELL: '1',
+    })
+    expect(process.env.DSH_SESSION_ID).toBe(ambient)
+  })
+
+  it('keeps parent and child agent session environments isolated', async () => {
+    const { ctx, bash } = await setupRecording(true)
+    const parent = registerFakeAgent(ctx, 'request-parent', () => undefined)
+    const child = registerFakeAgent(ctx, 'request-child', () => undefined)
+
+    for (const [callId, agent] of [['parent', parent], ['child', child]] as const) {
+      await ctx.tools.execute({
+        callId: CallId(`session-env-${callId}`),
+        name: 'bash',
+        arguments: { command: 'true', description: 'run command' },
+        agent,
+      })
+    }
+
+    expect(bash.requests.map(request => request.dshEnv)).toEqual([
+      {
+        DSH_HOME: recordingDshHome,
+        DSH_SESSION_ID: 'request-parent',
+        DSH_SESSION_JSONL: ctx.sessionPersistence.locate(parent.session.header)?.path,
+        DSH_SHELL: '1',
+      },
+      {
+        DSH_HOME: recordingDshHome,
+        DSH_SESSION_ID: 'request-child',
+        DSH_SESSION_JSONL: ctx.sessionPersistence.locate(child.session.header)?.path,
+        DSH_SHELL: '1',
+      },
+    ])
+    expect(bash.requests[0]?.dshEnv?.DSH_SESSION_JSONL).not.toBe(bash.requests[1]?.dshEnv?.DSH_SESSION_JSONL)
+  })
+
+  it('does not forward trusted-only fields even when the model includes them as extra arguments', async () => {
     const { ctx, bash } = await setupRecording()
     // Unknown `env` and `stdin` keys are ignored by the schema and named request construction.
     // This preserves the request shape; it is not a security boundary because shell syntax can
@@ -946,6 +1115,7 @@ describe('the model-facing bash tool builds its request from named args only (no
         description: 'echo',
         env: { SNEAKY_API_KEY: 'leak' },
         stdin: 'malicious payload',
+        stdoutMaxBytes: 999_999,
       },
     })
     expect(bash.requests).toHaveLength(1)
@@ -953,13 +1123,12 @@ describe('the model-facing bash tool builds its request from named args only (no
     expect(request.command).toBe('echo hi')
     expect('env' in request).toBe(false)
     expect('stdin' in request).toBe(false)
+    expect('stdoutMaxBytes' in request).toBe(false)
   })
 
-  it('a background bash call likewise carries no env/stdin', async () => {
+  it('a background bash call likewise carries no trusted-only fields', async () => {
     const { ctx, bash } = await setupRecording()
-    // start() throws in this recorder, but resolve() runs first and records the
-    // request — which is all this no-forward assertion needs.
-    await ctx.tools.execute({
+    const result = await ctx.tools.execute({
       callId: CallId('no-forward-2'),
       name: 'bash',
       arguments: {
@@ -968,510 +1137,18 @@ describe('the model-facing bash tool builds its request from named args only (no
         run_in_background: true,
         env: { TOKEN: 'leak' },
         stdin: 'x',
+        stdoutMaxBytes: 999_999,
       },
     })
+    // The call really went down the background path (the recorder sees the real
+    // request the consumer built, so the absent env/stdin below is a real
+    // negative, not a recorder that drops everything).
+    expect(text(result)).toBe('started background task bash-1')
     expect(bash.requests).toHaveLength(1)
     const request = bash.requests[0]!
+    expect(request.command).toBe('sleep 1')
     expect('env' in request).toBe(false)
     expect('stdin' in request).toBe(false)
-    // The owner token IS set on a background call (the isolation fence) — proving
-    // the recorder sees the real request the consumer built, so the absent
-    // env/stdin above is a real negative, not a recorder that drops everything.
-    expect('owner' in request).toBe(true)
+    expect('stdoutMaxBytes' in request).toBe(false)
   })
 })
-
-describe('sandbox rendering', () => {
-  const sandboxResult = (denied: boolean, exitCode: number): BashRunResult => ({
-    exitCode,
-    signal: null,
-    timedOut: false,
-    aborted: false,
-    timeoutMs: 1000,
-    stdout: { text: '', truncated: false },
-    stderr: { text: denied ? 'bash: /x: Read-only file system' : 'boom', truncated: false },
-    sandbox: { mode: 'read-only', denied },
-  })
-
-  it('renders a denial marker BEFORE the exit-code marker (the $-anchored parse survives)', () => {
-    const text = renderResult(sandboxResult(true, 1))
-    expect(text).toMatch(/\[sandbox: file access denied under read-only mode\]\n\[exit code: 1\]$/)
-  })
-
-  it('appends the same-turn escalation hint to a denial exactly when the fields are advertised', () => {
-    const hinted = renderResult(sandboxResult(true, 1), ['workspace-write', 'danger-full-access'])
-    expect(hinted).toMatch(
-      /denied under read-only mode\]\n\[sandbox: escalation available — retry this exact command once with sandbox_permissions [^\n]+\]\n\[exit code: 1\]$/, // eslint-disable-line @stylistic/max-len -- the hint sentence is pinned verbatim
-    )
-    // Default (no advertisement): no hint — a lever the schema does not offer is never suggested.
-    expect(renderResult(sandboxResult(true, 1))).not.toContain('escalation available')
-    // A non-denied result never hints, advertised or not.
-    expect(renderResult(sandboxResult(false, 2), ['danger-full-access'])).not.toContain('escalation available')
-  })
-
-  it('renders no sandbox marker for a plain failure under a sandboxed mode', () => {
-    expect(renderResult(sandboxResult(false, 2))).not.toContain('[sandbox:')
-  })
-
-  it('bash_output reports a settled background denial with the same marker', async () => {
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
-    await ctx.plugin(LocalSandboxProvider, PASSTHROUGH_RUNNER_CONFIG)
-    await ctx.plugin(SandboxBashExecutor, { graceMs: 200 })
-    const bash = ctx.bash as SandboxBashExecutor
-    bash.internals = { spillDir }
-    await ctx.plugin(ToolBash)
-    const started = await call(ctx, 'bash', { command: 'echo "x: Permission denied" >&2; exit 1', description: 'test command', run_in_background: true })
-    const id = text(started).match(/started background task (bash-\d+)/)![1]
-    await bash.list().find(task => task.id === id)!.done
-    const read = await call(ctx, 'bash_output', { task_id: id })
-    expect(text(read)).toMatch(
-      /\[status: completed, exit code: 1\]\n\[sandbox: file access denied under read-only mode\]\n\[sandbox: escalation available[^\n]+\]$/,
-    )
-  })
-
-  it('a settled background denial renders no escalation hint without a confining executor (defensive arm)', async () => {
-    // Structurally near-unreachable through the real stack — every confining
-    // default advertises the static target set — but the read path guards
-    // it anyway: an executor that reports no sandboxMode (fields never
-    // advertised) whose task nonetheless carries denial facts must render
-    // the marker without suggesting a lever the schema does not offer.
-    class FactsOnlyExecutor extends TestBashExecutor {
-      private readonly task: BashTask = {
-        id: BashTaskId('bash-facts'),
-        status: 'completed',
-        exitCode: 1,
-        signal: null,
-        done: Promise.resolve(),
-        sandbox: { mode: 'read-only', denied: true },
-      }
-
-      run(): Promise<BashRunResult> { return Promise.reject(new Error('not used')) }
-      start(): BashTask { return this.task }
-      get(id: string): BashTask | undefined { return id === this.task.id ? this.task : undefined }
-      list(): BashTask[] { return [this.task] }
-      kill(): boolean { return false }
-      ownerOf(): OwnerToken | undefined { return undefined }
-      readOutput(): BashTaskRead {
-        return { task: this.task, delta: '', lossy: false }
-      }
-    }
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
-    await ctx.plugin(FactsOnlyExecutor)
-    await ctx.plugin(ToolBash)
-    const read = await call(ctx, 'bash_output', { task_id: 'bash-facts' })
-    expect(text(read)).toMatch(/\[sandbox: file access denied under read-only mode\]$/)
-    expect(text(read)).not.toContain('escalation available')
-  })
-
-  it('bash_output reports a settled background RUNNER failure as a sandbox problem, outranking the denial marker', async () => {
-    // A provider whose wrap carries a runner-failure signature: the settled
-    // task's stderr matching it means the sandbox itself broke and the
-    // command never ran — even though the same stderr also carries denial
-    // words (a runner's error text may contain them).
-    class FakeProvider extends SandboxProvider {
-      confine(argv: readonly string[]): ConfinedArgv {
-        return { argv: [...argv], enforcement: 'full', denialSignatures: ['permission denied'], runnerFailureSignatures: ['fake-runner: '] }
-      }
-    }
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
-    await ctx.plugin(FakeProvider)
-    await ctx.plugin(SandboxBashExecutor, { graceMs: 200 })
-    const bash = ctx.bash as SandboxBashExecutor
-    bash.internals = { spillDir }
-    await ctx.plugin(ToolBash)
-    const started = await call(ctx, 'bash', { command: 'echo "fake-runner: cannot open rule path: /x: Permission denied" >&2; exit 125', description: 'test command', run_in_background: true })
-    const id = text(started).match(/started background task (bash-\d+)/)![1]
-    await bash.list().find(task => task.id === id)!.done
-    const read = await call(ctx, 'bash_output', { task_id: id })
-    expect(text(read)).toMatch(/\[sandbox: the sandbox runner itself failed under read-only mode — the command did not run; /)
-    expect(text(read)).toMatch(/this is a sandbox problem, not a command failure\]$/)
-    expect(text(read)).not.toContain('file access denied')
-  })
-
-  it('classifies an executable configured runner that refuses its profile before the command runs', async () => {
-    const signature = 'custom-runner-rejected'
-    const ctx = new Context()
-    await ctx.plugin(LocalSandboxProvider, {
-      runnerCommand: ['bash', '-c', `printf '${signature}\\n' >&2; exit 125`, 'custom-runner'],
-      runnerFailureSignatures: [signature],
-    })
-    await ctx.plugin(SandboxBashExecutor, { graceMs: 200 })
-    const bash = ctx.bash as SandboxBashExecutor
-    bash.internals = { spillDir }
-
-    await expect(bash.run(bash.resolve({ command: 'echo command-must-not-run' })))
-      .rejects.toMatchObject({ code: 'SANDBOX_UNAVAILABLE' })
-
-    const task = bash.start(bash.resolve({ command: 'echo command-must-not-run' }))
-    await task.done
-    expect(task.sandbox).toEqual({ mode: 'read-only', denied: false, enforcement: 'full', runnerFailed: true })
-  })
-
-  it('reports a real denial end-to-end through the shipping sandbox executor', async () => {
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
-    await ctx.plugin(LocalSandboxProvider, PASSTHROUGH_RUNNER_CONFIG)
-    await ctx.plugin(SandboxBashExecutor, { graceMs: 200 })
-    const bash = ctx.bash as SandboxBashExecutor
-    bash.internals = { spillDir }
-    await ctx.plugin(ToolBash)
-    const lockedDir = join(mkdtempSync(join(tmpdir(), 'dsh-tool-bash-denied-')), 'locked')
-    mkdirSync(lockedDir)
-    chmodSync(lockedDir, 0o555)
-    const result = await call(ctx, 'bash', { command: `echo x > ${lockedDir}/f`, description: 'Write into a locked directory' })
-    expect(result.isError).toBe(false)
-    expect(text(result)).toMatch(
-      /denied under read-only mode\]\n\[sandbox: escalation available[^\n]+\]\n\[exit code: \d+\]$/,
-    )
-  })
-})
-
-describe('sandbox escalation (sandbox_permissions / justification)', () => {
-  /** Compose the real sandbox stack (passthrough runner) at a given default mode. */
-  async function setupSandboxed(mode?: 'read-only' | 'workspace-write' | 'danger-full-access', opts: { approval?: boolean; policy?: 'ask' | 'never' } = {}) {
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
-    await ctx.plugin(LocalSandboxProvider, PASSTHROUGH_RUNNER_CONFIG)
-    await ctx.plugin(SandboxBashExecutor, { graceMs: 200, ...mode !== undefined ? { mode } : {} })
-    const bash = ctx.bash as SandboxBashExecutor
-    bash.internals = { spillDir }
-    if (opts.approval === true) await ctx.plugin(ApprovalService, opts.policy !== undefined ? { policy: opts.policy } : {})
-    await ctx.plugin(ToolBash)
-    return { ctx, bash }
-  }
-
-  /** The registered bash tool's wire schema (what the model actually sees). */
-  function bashSchema(ctx: Context) {
-    const schema = ctx.tools.schemas().find(s => s.name === 'bash')
-    if (!schema) throw new Error('bash tool not registered')
-    return schema as unknown as { description: string; parameters: { properties: Record<string, { enum?: string[] }> } }
-  }
-
-  /**
-   * A fake agent whose session records appends — the approval audit surface.
-   * Seeded mid-turn: an escalating call always runs inside one, and request()
-   * enforces the enclosure.
-   */
-  function escalationAgent(events: Array<{ type: string; data: Record<string, unknown> }>): Agent {
-    return {
-      id: 'agent-esc',
-      session: {
-        header: { version: 0, id: 'sess-esc', createdAt: 0 },
-        events: [{ type: 'turn/start' }],
-        append: (type: string, data: Record<string, unknown>) => { events.push({ type, data }) },
-      },
-    } as unknown as Agent
-  }
-
-  let escCall = 0
-  function callAs(ctx: Context, agent: Agent | undefined, args: unknown) {
-    return ctx.tools.execute({ callId: CallId(`call-esc-${++escCall}`), name: 'bash', arguments: args, ...agent ? { agent } : {} })
-  }
-
-  const ESCALATE = { command: 'true', description: 'test escalation', sandbox_permissions: 'workspace-write', justification: 'the test needs it' }
-
-  it('advertises no escalation surface under a non-sandboxing executor', async () => {
-    const ctx = await setup()
-    expect(ctx.bash.sandboxMode).toBeUndefined()
-    const schema = bashSchema(ctx)
-    expect(schema.parameters.properties['sandbox_permissions']).toBeUndefined()
-    expect(schema.parameters.properties['justification']).toBeUndefined()
-    expect(schema.description).not.toContain('sanctioned exception')
-  })
-
-  it('advertises the full closed target vocabulary under any confining default', async () => {
-    // The enum is deliberately NOT default-relative: a session's effective
-    // mode is per-session and switchable, so every confining composition
-    // advertises every possible target — strict widening is checked at
-    // execution against the call's effective mode instead.
-    for (const mode of [undefined, 'workspace-write', 'danger-full-access'] as const) {
-      const { ctx } = await setupSandboxed(mode)
-      const schema = bashSchema(ctx)
-      expect(schema.parameters.properties['sandbox_permissions']?.enum).toEqual(['workspace-write', 'danger-full-access'])
-      expect(schema.parameters.properties['justification']).toBeDefined()
-      expect(schema.description).toContain('sanctioned exception')
-    }
-  })
-
-  it('a non-widening request fails at execution with its own text and prompts no one', async () => {
-    const { ctx } = await setupSandboxed('danger-full-access', { approval: true })
-    const consulted = vi.fn()
-    ctx.on('approval/request', (_req, next) => { consulted(); return next() })
-    const result = await callAs(ctx, escalationAgent([]), { command: 'true', description: 'd', sandbox_permissions: 'workspace-write', justification: 'already wider' })
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('not strictly wider than this call\'s current "danger-full-access" mode')
-    expect(consulted).not.toHaveBeenCalled()
-  })
-
-  it('rejects sandbox_permissions without a justification, and vice versa, and a blank justification', async () => {
-    const { ctx } = await setupSandboxed()
-    const missing = await callAs(ctx, undefined, { command: 'true', description: 'd', sandbox_permissions: 'workspace-write' })
-    expect(missing.isError).toBe(true)
-    expect(text(missing)).toContain('sandbox_permissions requires a justification')
-    const orphan = await callAs(ctx, undefined, { command: 'true', description: 'd', justification: 'why not' })
-    expect(orphan.isError).toBe(true)
-    expect(text(orphan)).toContain('only valid together with sandbox_permissions')
-    const blank = await callAs(ctx, undefined, { command: 'true', description: 'd', sandbox_permissions: 'workspace-write', justification: '   ' })
-    expect(blank.isError).toBe(true)
-    expect(text(blank)).toContain('expected a non-empty sentence')
-  })
-
-  it('the schema enum rejects a mode outside the target vocabulary before execute (registry-level, any caller)', async () => {
-    const { ctx } = await setupSandboxed()
-    const result = await callAs(ctx, undefined, { command: 'true', description: 'd', sandbox_permissions: 'read-only', justification: 'narrow' })
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('must be one of')
-  })
-
-  it('rejects an unadvertised sandbox_permissions injection under a non-sandboxing executor', async () => {
-    const ctx = await setup()
-    const result = await callAs(ctx, undefined, { command: 'true', description: 'd', sandbox_permissions: 'workspace-write', justification: 'sneaky' })
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('not available in this composition')
-  })
-
-  it('fails closed with its own text when no approval service is composed', async () => {
-    const { ctx } = await setupSandboxed()
-    const result = await callAs(ctx, escalationAgent([]), ESCALATE)
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('no approval service is composed')
-  })
-
-  it('fails closed with its own text for an agent-less escalating call', async () => {
-    const { ctx } = await setupSandboxed('read-only', { approval: true })
-    const result = await callAs(ctx, undefined, ESCALATE)
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('no agent to route it through')
-  })
-
-  it('fails closed with its own text when the service has no answerer', async () => {
-    const { ctx } = await setupSandboxed('read-only', { approval: true })
-    const result = await callAs(ctx, escalationAgent([]), ESCALATE)
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('no approval channel is available')
-  })
-
-  it('a grant runs THAT call under the wider mode — the denial marker names it — and lands the audit pair', async () => {
-    const { ctx } = await setupSandboxed('read-only', { approval: true })
-    ctx.on('approval/request', () => Promise.resolve<ApprovalOutcome>('allowed-once'))
-    const events: Array<{ type: string; data: Record<string, unknown> }> = []
-    // A real unix denial under the passthrough runner: the marker's mode can
-    // only say workspace-write if the override actually rode the spec.
-    const lockedDir = join(mkdtempSync(join(tmpdir(), 'dsh-esc-denied-')), 'locked')
-    mkdirSync(lockedDir)
-    chmodSync(lockedDir, 0o555)
-    const result = await callAs(ctx, escalationAgent(events), {
-      command: `echo x > ${lockedDir}/f`,
-      description: 'write into a locked directory',
-      sandbox_permissions: 'workspace-write',
-      justification: 'must write outside the workspace',
-    })
-    expect(result.isError).toBe(false)
-    expect(text(result)).toMatch(/\[sandbox: file access denied under workspace-write mode\]/)
-    expect(events.map(e => e.type)).toEqual(['approval/asked', 'approval/decided'])
-    expect(events[0]?.data['toolName']).toBe('bash')
-    expect(events[0]?.data['reason']).toBe('escalate sandbox to workspace-write: must write outside the workspace')
-    expect(events[1]?.data['outcome']).toBe('allowed-once')
-  })
-
-  it('a granted background start settles with the wider mode\'s facts', async () => {
-    const { ctx, bash } = await setupSandboxed('read-only', { approval: true })
-    ctx.on('approval/request', () => Promise.resolve<ApprovalOutcome>('allowed-once'))
-    const started = await callAs(ctx, escalationAgent([]), { ...ESCALATE, run_in_background: true })
-    expect(started.isError).toBe(false)
-    const id = text(started).match(/started background task (bash-\d+)/)?.[1]
-    const task = bash.list().find(t => t.id === id)
-    if (!task) throw new Error('escalated task not tracked')
-    await task.done
-    expect(task.sandbox).toMatchObject({ mode: 'workspace-write', denied: false })
-  })
-
-  it('a rejection denies with the user-said-no text and runs nothing', async () => {
-    const { ctx } = await setupSandboxed('read-only', { approval: true })
-    ctx.on('approval/request', () => Promise.resolve<ApprovalOutcome>('rejected'))
-    // A live (non-aborted) signal rides the execution: the gate threads it
-    // into the approval request so a turn cancellation can withdraw the ask.
-    const result = await ctx.tools.execute({
-      callId: CallId(`call-esc-${++escCall}`),
-      name: 'bash',
-      arguments: ESCALATE,
-      agent: escalationAgent([]),
-      signal: new AbortController().signal,
-    })
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('the user rejected escalating this command to "workspace-write"')
-  })
-
-  it('a cancellation denies with the cancelled text', async () => {
-    const { ctx } = await setupSandboxed('read-only', { approval: true })
-    ctx.on('approval/request', () => Promise.resolve<ApprovalOutcome>('cancelled'))
-    const result = await callAs(ctx, escalationAgent([]), ESCALATE)
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('approval for escalating to "workspace-write" was cancelled')
-  })
-
-  it('a rogue approval stand-in returning a non-vocabulary outcome hits the exhaustiveness backstop', async () => {
-    const { ctx } = await setupSandboxed()
-    ctx.provide('approval', { request: () => Promise.resolve('yolo') } as unknown as InstanceType<typeof ApprovalService>)
-    const result = await callAs(ctx, escalationAgent([]), ESCALATE)
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('unreachable')
-  })
-
-  it('a never policy rejects an escalation deterministically without consulting any answerer', async () => {
-    // The live-session e.md case: the model requests escalation against a
-    // 'never' session — the prepend gate answers rejected before any
-    // interactive answerer, the fail-closed text is the ordinary rejection
-    // wording, and the audit pair still lands.
-    const { ctx } = await setupSandboxed('read-only', { approval: true, policy: 'never' })
-    const consulted = vi.fn()
-    ctx.on('approval/request', (_req, next) => { consulted(); return next() })
-    const events: Array<{ type: string; data: Record<string, unknown> }> = []
-    const result = await callAs(ctx, escalationAgent(events), ESCALATE)
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('the user rejected escalating this command to "workspace-write"')
-    expect(consulted).not.toHaveBeenCalled()
-    expect(events.map(e => e.type)).toEqual(['approval/asked', 'approval/decided'])
-    expect(events[1]?.data).toMatchObject({ outcome: 'rejected' })
-  })
-
-  it('a plain call under a sandboxing executor never consults approval', async () => {
-    const { ctx } = await setupSandboxed('read-only', { approval: true })
-    const asked = vi.fn()
-    ctx.on('approval/request', (_req, next) => { asked(); return next() })
-    const result = await callAs(ctx, escalationAgent([]), { command: 'echo plain', description: 'plain run' })
-    expect(result.isError).toBe(false)
-    expect(text(result)).toContain('plain')
-    expect(asked).not.toHaveBeenCalled()
-  })
-})
-
-describe('per-session sandbox mode (the bash/sandbox-mode fold)', () => {
-  /** Compose the real sandbox stack (passthrough runner) at a given default mode. */
-  async function setupModal(mode: 'read-only' | 'workspace-write' | 'danger-full-access' = 'read-only', opts: { approval?: boolean } = {}) {
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
-    await ctx.plugin(LocalSandboxProvider, PASSTHROUGH_RUNNER_CONFIG)
-    await ctx.plugin(SandboxBashExecutor, { graceMs: 200, mode })
-    ;(ctx.bash as SandboxBashExecutor).internals = { spillDir }
-    if (opts.approval === true) await ctx.plugin(ApprovalService)
-    await ctx.plugin(ToolBash)
-    return ctx
-  }
-
-  /**
-   * An agent stand-in over a REAL Session — the stamping folds real events;
-   * the opened turn satisfies approval's enclosure precondition on escalating
-   * calls.
-   */
-  function sessionAgent(id: string): { agent: Agent; session: Session; injected: string[] } {
-    const session = new Session(SessionId(id))
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    const injected: string[] = []
-    const agent = {
-      id,
-      session,
-      inject: (content: { type: string; text: string }[]) => { injected.push(content[0]?.text ?? '') },
-    } as unknown as Agent
-    return { agent, session, injected }
-  }
-
-  let modeCall = 0
-  const callAs = (ctx: Context, agent: Agent | undefined, args: unknown) =>
-    ctx.tools.execute({ callId: CallId(`call-mode-${++modeCall}`), name: 'bash', arguments: args, ...agent ? { agent } : {} })
-
-
-  it('stamps calls with grant > session override > nothing (executor default)', async () => {
-    const ctx = await setupModal('read-only', { approval: true })
-    ctx.on('approval/request', () => Promise.resolve<ApprovalOutcome>('allowed-once'))
-    const seen: (string | undefined)[] = []
-    const original = ctx.bash.resolve.bind(ctx.bash)
-    vi.spyOn(ctx.bash, 'resolve').mockImplementation((req) => {
-      seen.push(req.sandboxMode)
-      return original(req)
-    })
-    const { agent, session } = sessionAgent('sess-stamp-1')
-    const run = { command: 'true', description: 'stamp probe' }
-    await callAs(ctx, agent, run)                       // no override yet
-    setSandboxMode(session, 'workspace-write')
-    await callAs(ctx, agent, run)                       // standing override
-    await callAs(ctx, undefined, run)                   // agent-less caller: no session to fold
-    await callAs(ctx, agent, { ...run, sandbox_permissions: 'danger-full-access', justification: 'grant outranks override' })
-    expect(seen).toEqual([undefined, 'workspace-write', undefined, 'danger-full-access'])
-  })
-
-  it('escalates relative to the session effective mode, not the executor default (narrower override)', async () => {
-    // With a workspace-write default and read-only override, escalation must return to
-    // workspace-write. The static target vocabulary exposes it, and validation compares it with
-    // the call's effective override rather than a default-relative ladder.
-    const ctx = await setupModal('workspace-write', { approval: true })
-    ctx.on('approval/request', () => Promise.resolve<ApprovalOutcome>('allowed-once'))
-    const seen: (string | undefined)[] = []
-    const original = ctx.bash.resolve.bind(ctx.bash)
-    vi.spyOn(ctx.bash, 'resolve').mockImplementation((req) => {
-      seen.push(req.sandboxMode)
-      return original(req)
-    })
-    const { agent, session } = sessionAgent('sess-esc-narrow')
-    setSandboxMode(session, 'read-only')
-    const result = await callAs(ctx, agent, { command: 'true', description: 'd', sandbox_permissions: 'workspace-write', justification: 'the override is narrower than the default' })
-    expect(result.isError).toBe(false)
-    expect(seen).toEqual(['workspace-write'])
-  })
-
-  it('a danger-full-access default still offers the lever to a narrower-switched session', async () => {
-    // Under the default-relative ladder these fields VANISHED (nothing is
-    // wider than the default), stranding a read-only-overridden session
-    // with no escalation path at all.
-    const ctx = await setupModal('danger-full-access', { approval: true })
-    ctx.on('approval/request', () => Promise.resolve<ApprovalOutcome>('allowed-once'))
-    const schema = ctx.tools.schemas().find(t => t.name === 'bash') as unknown as { parameters: { properties: Record<string, { enum?: string[] }> } }
-    expect(schema.parameters.properties['sandbox_permissions']?.enum).toEqual(['workspace-write', 'danger-full-access'])
-    const { agent, session } = sessionAgent('sess-esc-dfa')
-    setSandboxMode(session, 'read-only')
-    const result = await callAs(ctx, agent, { command: 'true', description: 'd', sandbox_permissions: 'workspace-write', justification: 'confined by override under a wide default' })
-    expect(result.isError).toBe(false)
-  })
-
-  it('rejects a non-widening request against the OVERRIDDEN effective mode without prompting', async () => {
-    const ctx = await setupModal('read-only', { approval: true })
-    const consulted = vi.fn()
-    ctx.on('approval/request', (_req, next) => { consulted(); return next() })
-    const { agent, session } = sessionAgent('sess-esc-nonwide')
-    setSandboxMode(session, 'danger-full-access')
-    const result = await callAs(ctx, agent, { command: 'true', description: 'd', sandbox_permissions: 'workspace-write', justification: 'already wider via override' })
-    expect(result.isError).toBe(true)
-    expect(text(result)).toContain('not strictly wider than this call\'s current "danger-full-access" mode')
-    expect(consulted).not.toHaveBeenCalled()
-  })
-
-  it('never stamps an override under a non-sandboxing executor (nothing honors it)', async () => {
-    const ctx = await setup()
-    const seen: (string | undefined)[] = []
-    const original = ctx.bash.resolve.bind(ctx.bash)
-    vi.spyOn(ctx.bash, 'resolve').mockImplementation((req) => {
-      seen.push(req.sandboxMode)
-      return original(req)
-    })
-    const { agent, session } = sessionAgent('sess-stamp-2')
-    setSandboxMode(session, 'danger-full-access')
-    await callAs(ctx, agent, { command: 'true', description: 'plain probe' })
-    expect(seen).toEqual([undefined])
-  })
-
-})
diff --git a/packages/bash/tool-bash/tsconfig.json b/packages/bash/tool-bash/tsconfig.json
index c4d738c7dd..407e78ebd8 100644
--- a/packages/bash/tool-bash/tsconfig.json
+++ b/packages/bash/tool-bash/tsconfig.json
@@ -14,6 +14,9 @@
     {
       "path": "../../../vendor/cordis"
     },
+    {
+      "path": "../../../vendor/schemastery"
+    },
     {
       "path": "../../llm/llm"
     },
@@ -23,9 +26,18 @@
     {
       "path": "../../core/agent"
     },
+    {
+      "path": "../../session-persistence/session-persistence"
+    },
     {
       "path": "../../bash/bash"
     },
+    {
+      "path": "../../util/home"
+    },
+    {
+      "path": "../../tasks/tasks"
+    },
     {
       "path": "../../core/system-prompt"
     },
diff --git a/packages/compact/README.md b/packages/compact/README.md
index 08c3ddd707..b5f5987571 100644
--- a/packages/compact/README.md
+++ b/packages/compact/README.md
@@ -5,7 +5,7 @@ A three-package capability seam (see [capability seams](../../docs/rfc/implement
 | Package | Role | ctx key |
 |---|---|---|
 | `compact/` | Abstract compaction seam (interface + `compact/*` events + `CompactionResult`) | `ctx.compact` |
-| `compact-basic/` | A backend: chars-per-token estimation (`charsPerToken`, default 4) + token-budget retention + `llm.stream()` summarization | (registers `ctx.compact`) |
+| `compact-basic/` | A backend: `ctx.tokenMeter` pressure + token-budget retention + `llm.stream()` summarization | (registers `ctx.compact`) |
 | `tool-compact/` (deferred) | Model-facing `/compact` tool over `ctx.compact` | (registers on `ctx.tools`) |
 
-The interface lives at `compact/compact/`, the backend at `compact/compact-basic/`. Unlike the bash seam, it depends on `dsh-session` and `dsh-llm` — its verbs are defined over a `Session` and its output is the `ContentBlock` vocabulary, so the contract cannot be expressed without naming them. That deviation from the "interface depends only on cordis" guidance is intentional and recorded in the [compaction capability-seam RFC](../../docs/rfc/implemented/feature/2026-06-18-compaction-capability-seam.md). A tokenizer- or template-based backend would replace `compact-basic` without touching the interface or the tool.
+The interface lives at `compact/compact/`, the backend at `compact/compact-basic/`. Unlike the bash seam, it depends on `dsh-session` and `dsh-llm` — its verbs are defined over a `Session` and its output is the `ContentBlock` vocabulary, so the contract cannot be expressed without naming them. That deviation from the "interface depends only on cordis" guidance is intentional and recorded in the [compaction capability-seam RFC](../../docs/rfc/implemented/feature/2026-06-18-compaction-capability-seam.md). Token measurement is a reusable LLM-family service rather than a `CompactService` method; a template- or model-backed compactor can replace `compact-basic` without changing the meter or callers.
diff --git a/packages/compact/compact-basic/README.md b/packages/compact/compact-basic/README.md
index d9c0f472a8..937120eebe 100644
--- a/packages/compact/compact-basic/README.md
+++ b/packages/compact/compact-basic/README.md
@@ -1,6 +1,6 @@
 # @deepseek-ai/dsh-compact-basic
 
-The **basic compaction backend**: a `BasicCompactService` implementing the `@deepseek-ai/dsh-compact` seam with a chars-per-token heuristic (the `charsPerToken` config, default 4), token-budget retention, and summarization as a direct one-shot `ctx.llm.stream()` call (interceptable at `llm/stream`).
+The **basic compaction backend**: a `BasicCompactService` implementing the `@deepseek-ai/dsh-compact` seam with reusable `ctx.tokenMeter` pressure, token-budget retention, and summarization as a direct one-shot `ctx.llm.stream()` call (interceptable at `llm/stream`).
 
 This is the implementation tier of the compaction capability — see the [interface package](../compact/README.md) for the seam and the [capability-seam RFC](../../../docs/rfc/implemented/feature/2026-06-18-compaction-capability-seam.md) for the design.
 
@@ -8,49 +8,43 @@ This is the implementation tier of the compaction capability — see the [interf
 
 This backend owns the compaction policy:
 
-- **Estimation** — a configurable characters-per-token heuristic counts the current session prefix supplied to pre-step, derived history, and system prompt, matching the next request rather than stale logged prefix state.
-- **Retention** — compact the oldest whole surface units while preserving a recent tail and balanced tool-call/result cuts. Turn boundaries do not protect old steps inside a runaway turn. An open indivisible tail declines until it closes; a single unit larger than the budget remains out of scope.
+- **Measurement** — the singleton `ctx.tokenMeter` prices the provisional request envelope and current surface at one consumed-log revision. The current prompt and prefix override their logged values; the pre-step boundary reuses logged tools and call config.
+- **Retention** — compact the oldest whole surface units while preserving a recent tail and balanced tool-call/result cuts through the [`dsh-compact` boundary helpers](../compact/README.md#tool-pairing-boundaries). Turn boundaries do not protect old steps inside a runaway turn. An open indivisible tail declines until it closes; a single unit larger than the budget remains out of scope.
 - **Convergence** — retry head-checkpoint compaction up to `compactionRetries`; reject a summary that does not shrink its source, and throw if retries cannot return below threshold.
-- **Summarization** — a direct `llm/stream` call uses the configured model and cap without running the loop-only `agent/request` seam. The input transcript preserves non-text blocks as tagged placeholders; only returned text enters the checkpoint, excluding reasoning and tool calls that would leak private reasoning or create an orphaned call.
+- **Summarization** — a direct `llm/stream` call uses the configured provider/model pair and cap, falling back to the latest logged request target and then the agent target, without running the loop-only `agent/request` seam. The input transcript preserves non-text blocks as tagged placeholders; only returned text enters the checkpoint, excluding reasoning and tool calls that would leak private reasoning or create an orphaned call.
 - **Framing** — the replacement user message marks established checkpoint context with `<compacted-summary>` tags. The raw summary remains on the provenance event, and later automatic cycles merge the prior checkpoint.
-- **Lifecycle** — `compactRegion()` records its start, summary, replacement, and end. The serial `agent/pre-step` listener checks pressure before every step, outside an open step, so a tool-heavy turn remains compactable and the loop derives history once after mutation.
+- **Lifecycle** — `compactRegion()` requires its agent to own the exact target session and rejects mismatch before resolution or mutation; a valid call records its start, summary, replacement, and end. The serial `agent/pre-step` listener checks pressure before every step, outside an open step, so a tool-heavy turn remains compactable and the loop derives history once after mutation.
 - **Failure handling** — an unmatched `compact/start` is an inert crash marker because no replacement landed. Recoverable failure records an error end and leaves the surface unchanged.
 
-`estimateContentTokens()` and `summarize()` are overridable hooks: a tokenizer-based or template-based backend can subclass `BasicCompactService` and override just those, reusing the retention walk and surface plumbing. `summarize()` returns the summary blocks together with the call envelope it actually used (`{ summary, model, maxTokens? }`) — the caller logs that envelope on the `compact/summary` provenance event, so an overriding backend reports its own envelope honestly.
+`summarize()` is the sole subclass hook. A template- or remote-summarizer subclass can override it while pressure, retention, provenance, shrink validation, and shadowed-token accounting stay on `ctx.tokenMeter`. The hook returns the summary blocks together with the call envelope it used (`{ summary, provider, model, maxTokens? }`), which is logged on `compact/summary`.
 
 ## Config (`BasicCompactConfig`)
 
-Every knob is **required** except `auto` — there is no concrete data yet to justify default thresholds/budgets, so a consumer states each value explicitly rather than inherit a guessed default. `auto` alone defaults to `true`.
+Every setting is optional. The pressure and retention policy applies to the token meter's single context window. Unrecognized top-level keys are rejected.
 
 | Key | Required | Meaning |
 |---|---|---|
-| `contextWindow` | yes | Context window size in tokens. |
-| `thresholdRatio` | yes | Compact when estimated usage exceeds this fraction of the window. |
-| `retainTokens` | yes | Tokens of recent context to keep intact. |
-| `summarizationModel` | yes | Model for summarization (`''` → use the agent's model). |
-| `maxTokens` | yes | Provider generation cap for the summarization call; may include reasoning tokens. |
-| `compactionRetries` | yes | Extra compaction attempts after the first if the compacted surface remains over threshold. |
-| `auto` | no (default `true`) | Register the `agent/pre-step` auto-compaction listener. Set `false` for manual-only. |
-| `charsPerToken` | no (default `4`) | Token-estimator text density (estimated tokens = chars / `charsPerToken`; may be fractional). The default suits English text; CJK-heavy deployments should set ~1-2 or the estimate undershoots several-fold and compaction fires too late. |
+| `thresholdRatio` | no (default `0.8`) | Compact at `floor(contextWindow × ratio)`. |
+| `retainTokens` | no (default `floor(contextWindow × 0.16)`) | Recent surface budget kept verbatim; must be below the threshold. |
+| `summarizationProvider` | no (default `''`) | Set together with `summarizationModel`; an empty pair resolves the latest logged request target, then the `AgentOptions` pair. |
+| `summarizationModel` | no (default `''`) | Set together with `summarizationProvider`; an empty pair resolves the latest logged request target, then the `AgentOptions` pair. |
+| `maxTokens` | no (default `8192`) | Provider generation cap for the summarization call; may include reasoning tokens. |
+| `compactionRetries` | no (default `1`) | Extra attempts after the first when pressure remains above threshold. |
+| `auto` | no (default `true`) | Register the `agent/pre-step` automatic listener. Set `false` for manual-only. |
 
 ## Usage
 
 ```ts
 import type { Context } from 'cordis'
 import { BasicCompactService } from '@deepseek-ai/dsh-compact-basic'
+import TokenMeterService from '@deepseek-ai/dsh-token-meter'
 
 export const name = 'compact-basic'
 export const inject = ['llm']
 
 export function apply(ctx: Context): void {
-  ctx.plugin(BasicCompactService, {
-    contextWindow: 128000,
-    thresholdRatio: 0.8,
-    retainTokens: 20480,
-    summarizationModel: '',
-    maxTokens: 8192,
-    compactionRetries: 1,
-  })
+  ctx.plugin(TokenMeterService)
+  ctx.plugin(BasicCompactService)
 }
 ```
 
@@ -122,8 +116,8 @@ Rules:
 
 ## Known Limitations and Deferred Work
 
-- **Token estimation is the chars/`charsPerToken` heuristic** — a marked TODO schedules replacing it with an exact count (a real tokenizer, or provider `usage` fed back) so thresholds track the model's actual budget.
-- **`estimatePressure()` does not count the request's `tools` field** — pressure is underestimated by the size of the serialized tool schemas the request also carries.
+- **Pre-step sees a provisional request envelope** — the current prompt and prefix are exact, but routing and tool changes made later in `agent/request` are not logged yet. A router-only agent with no provisional provider/model pair skips that check.
+- **Meter accuracy follows the fixed heuristic** — missing reusable provider usage falls back to character count plus structural overhead rather than exact tokenization.
 - **`compactRegion` requires an open turn** — a manual call on a fully-closed session throws ("no open turn") rather than compacting.
 - **Summarization failure fails closed with full, over-budget history** — including truncation at the summarization `maxTokens`, which hidden reasoning tokens can consume; the auto path logs a warning and proceeds.
 - **The summarization call has no transcript-snapshot coverage** — `dsh-llm-replay` derives calls from `assistant/chunk` events, so this chunk-less direct `ctx.llm.stream()` call cannot replay (named deferred replay infrastructure in [the seam RFC](../../../docs/rfc/implemented/feature/2026-06-18-compaction-capability-seam.md)).
diff --git a/packages/compact/compact-basic/package.json b/packages/compact/compact-basic/package.json
index 32852a060f..a0ee2b4036 100644
--- a/packages/compact/compact-basic/package.json
+++ b/packages/compact/compact-basic/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@deepseek-ai/dsh-compact-basic",
-  "description": "Basic compaction backend (chars-per-token estimation + token-budget retention + llm.generate() summarization) for the DeepSeek Harness",
+  "description": "Token-meter-driven compaction policy and LLM summarization backend for the DeepSeek Harness",
   "version": "0.0.1",
   "private": true,
   "type": "module",
@@ -26,16 +26,23 @@
     "@deepseek-ai/dsh-compact": "^0.0.1",
     "@deepseek-ai/dsh-llm": "^0.0.1",
     "@deepseek-ai/dsh-session": "^0.0.1",
+    "@deepseek-ai/dsh-token-meter": "^0.0.1",
     "cordis": "^4.0.0-rc.7"
   },
+  "dependencies": {
+    "schemastery": "^3.18.0"
+  },
   "devDependencies": {
+    "@cordisjs/plugin-include": "workspace:^",
+    "@cordisjs/plugin-loader": "workspace:^",
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-compact": "workspace:^",
     "@deepseek-ai/dsh-invariants": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
-    "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-token-meter": "workspace:^",
     "@deepseek-ai/dsh-tools": "workspace:^",
     "cordis": "^4.0.0-rc.7"
   }
diff --git a/packages/compact/compact-basic/src/automatic.ts b/packages/compact/compact-basic/src/automatic.ts
new file mode 100644
index 0000000000..edb317d270
--- /dev/null
+++ b/packages/compact/compact-basic/src/automatic.ts
@@ -0,0 +1,52 @@
+/**
+ * Automatic pre-step pressure listener for compact-basic.
+ *
+ * @module @deepseek-ai/dsh-compact-basic/automatic
+ */
+
+import type { Context } from 'cordis'
+import type { CompactionResult } from '@deepseek-ai/dsh-compact'
+import type { Message } from '@deepseek-ai/dsh-llm'
+import type { Agent } from '@deepseek-ai/dsh-agent'
+
+interface AutomaticCompactor {
+  compactIfNeeded(
+    agent: Agent,
+    fullSystemPrompt: string,
+    sessionPrefix: readonly Message[],
+    signal: AbortSignal,
+  ): Promise<CompactionResult | null>
+}
+
+/**
+ * Register the implementation-owned automatic compaction listener.
+ * @param ctx - context owning the listener effect and logger.
+ * @param service - compactor whose public methods remain dynamically dispatched.
+ */
+export function registerAutomaticCompaction(
+  ctx: Context,
+  service: AutomaticCompactor,
+): void {
+  ctx.on('agent/pre-step', async (
+    agent: Agent,
+    _turn: number,
+    _step: number,
+    fullSystemPrompt: string,
+    sessionPrefix: readonly Message[],
+    signal: AbortSignal,
+  ) => {
+    try {
+      const result = await service.compactIfNeeded(agent, fullSystemPrompt, sessionPrefix, signal)
+      if (result !== null) {
+        ctx.logger.info(
+          `compaction: shadowed ${result.shadowedSeqs.length} surface nodes `
+          + `(seqs ${result.shadowedRange.start}-${result.shadowedRange.end}, `
+          + `~${result.shadowedTokenCount} tokens)`,
+        )
+      }
+    } catch (error: unknown) {
+      const message = error instanceof Error ? error.message : String(error)
+      ctx.logger.warn(`compaction failed: ${message}; proceeding with full history`)
+    }
+  })
+}
diff --git a/packages/compact/compact-basic/src/config.ts b/packages/compact/compact-basic/src/config.ts
new file mode 100644
index 0000000000..30241987a4
--- /dev/null
+++ b/packages/compact/compact-basic/src/config.ts
@@ -0,0 +1,107 @@
+/**
+ * Runtime defaulting and policy validation for compact-basic.
+ *
+ * @module @deepseek-ai/dsh-compact-basic/config
+ */
+
+import { deepFreeze } from '@deepseek-ai/dsh-llm'
+import type { TokenMeterService } from '@deepseek-ai/dsh-token-meter'
+import type { BasicCompactConfig, ResolvedConfig } from './types.ts'
+
+/** Default request-pressure fraction of the token meter's context window. */
+const DEFAULT_THRESHOLD_RATIO = 0.8
+
+/** Default verbatim-tail fraction of the token meter's context window. */
+const DEFAULT_RETAIN_RATIO = 0.16
+
+/** Complete public configuration key set. */
+const BASIC_COMPACT_CONFIG_KEYS: ReadonlySet<string> = new Set([
+  'thresholdRatio',
+  'retainTokens',
+  'summarizationProvider',
+  'summarizationModel',
+  'maxTokens',
+  'compactionRetries',
+  'auto',
+])
+
+/** Reject stale or misspelled keys before defaults can hide them. */
+function validateConfigKeys(config: BasicCompactConfig): void {
+  for (const key of Object.keys(config)) {
+    if (!BASIC_COMPACT_CONFIG_KEYS.has(key)) {
+      throw new Error(
+        `BasicCompactConfig: unknown key "${key}" `
+        + '(allowed: thresholdRatio, retainTokens, summarizationProvider, summarizationModel, maxTokens, compactionRetries, auto)',
+      )
+    }
+  }
+}
+
+/**
+ * Resolve defaults and validate the service-wide compaction policy.
+ * @param config - raw compact-basic configuration.
+ * @param tokenMeter - token meter supplying the context capacity.
+ * @returns a detached deeply immutable configuration.
+ */
+export function resolveConfig(
+  config: BasicCompactConfig = {},
+  tokenMeter: TokenMeterService,
+): ResolvedConfig {
+  validateConfigKeys(config)
+  const thresholdRatio = config.thresholdRatio ?? DEFAULT_THRESHOLD_RATIO
+  const retainTokens = config.retainTokens
+    ?? Math.floor(tokenMeter.contextWindow * DEFAULT_RETAIN_RATIO)
+  const resolved: ResolvedConfig = {
+    thresholdRatio,
+    retainTokens,
+    summarizationProvider: config.summarizationProvider ?? '',
+    summarizationModel: config.summarizationModel ?? '',
+    maxTokens: config.maxTokens ?? 8192,
+    compactionRetries: config.compactionRetries ?? 1,
+    auto: config.auto ?? true,
+  }
+
+  assertRatio('thresholdRatio', resolved.thresholdRatio)
+  assertNonNegativeInteger('retainTokens', resolved.retainTokens)
+  const thresholdTokens = Math.floor(tokenMeter.contextWindow * resolved.thresholdRatio)
+  if (resolved.retainTokens >= thresholdTokens) {
+    throw new Error(
+      `BasicCompactConfig: retainTokens (${resolved.retainTokens}) must be less than threshold tokens ${thresholdTokens}`,
+    )
+  }
+  assertPositiveInteger('maxTokens', resolved.maxTokens)
+  assertNonNegativeInteger('compactionRetries', resolved.compactionRetries)
+  if (typeof resolved.summarizationProvider !== 'string') {
+    throw new Error('BasicCompactConfig: summarizationProvider must be a string')
+  }
+  if (typeof resolved.summarizationModel !== 'string') {
+    throw new Error('BasicCompactConfig: summarizationModel must be a string')
+  }
+  if ((resolved.summarizationProvider.length === 0) !== (resolved.summarizationModel.length === 0)) {
+    throw new Error(
+      'BasicCompactConfig: summarizationProvider and summarizationModel must both be set or both be empty',
+    )
+  }
+  if (typeof resolved.auto !== 'boolean') {
+    throw new Error('BasicCompactConfig: auto must be a boolean')
+  }
+  return deepFreeze(resolved)
+}
+
+function assertPositiveInteger(name: string, value: number): void {
+  if (!Number.isInteger(value) || value <= 0) {
+    throw new Error(`BasicCompactConfig: ${name} (${value}) must be a positive integer`)
+  }
+}
+
+function assertNonNegativeInteger(name: string, value: number): void {
+  if (!Number.isInteger(value) || value < 0) {
+    throw new Error(`BasicCompactConfig: ${name} (${value}) must be a non-negative integer`)
+  }
+}
+
+function assertRatio(name: string, value: number): void {
+  if (typeof value !== 'number' || !Number.isFinite(value) || value <= 0 || value > 1) {
+    throw new Error(`BasicCompactConfig: ${name} (${value}) must be a number in (0, 1]`)
+  }
+}
diff --git a/packages/compact/compact-basic/src/index.ts b/packages/compact/compact-basic/src/index.ts
index bf349f7f73..60452581c5 100644
--- a/packages/compact/compact-basic/src/index.ts
+++ b/packages/compact/compact-basic/src/index.ts
@@ -1,287 +1,118 @@
 /**
- * Basic compaction backend. It estimates request pressure, retains a recent
- * tool-balanced surface tail, summarizes the older head through a one-shot model
- * call, and replaces that head with one checkpoint. Auto-compaction runs before
- * every step so a growing turn can compact its earlier closed steps.
+ * Basic replay-aware compaction backend.
+ *
  * @module @deepseek-ai/dsh-compact-basic
  */
 
 import { Context } from 'cordis'
-import { CompactService, renderTranscript } from '@deepseek-ai/dsh-compact'
+import z from 'schemastery'
+import { CompactService } from '@deepseek-ai/dsh-compact'
 import type { CompactionResult } from '@deepseek-ai/dsh-compact'
-import { BlockAssembler } from '@deepseek-ai/dsh-llm'
-import type { ContentBlock, FinishReason, GenerateOptions, Message } from '@deepseek-ai/dsh-llm'
-import type { Session, SessionEvent } from '@deepseek-ai/dsh-session'
-import { isToolPairingBalanced } from '@deepseek-ai/dsh-session'
+import { canonicalHeader } from '@deepseek-ai/dsh-session'
+import type { EpochHeader, Session } from '@deepseek-ai/dsh-session'
+import type { ContentBlock, Message } from '@deepseek-ai/dsh-llm'
 import type { Agent } from '@deepseek-ai/dsh-agent'
-import type { BasicCompactConfig, ResolvedConfig } from './types.ts'
-import { resolveConfig } from './types.ts'
+import { registerAutomaticCompaction } from './automatic.ts'
+import { resolveConfig } from './config.ts'
+import { compactSurfaceRegion, selectCompactableRange } from './region.ts'
+import { summarizeWithLlm } from './summarizer.ts'
+import type {
+  BasicCompactConfig,
+  ResolvedConfig,
+} from './types.ts'
 
-export type { BasicCompactConfig, ResolvedConfig } from './types.ts'
-export { resolveConfig } from './types.ts'
+export { resolveConfig } from './config.ts'
+export type {
+  BasicCompactConfig,
+  ResolvedConfig,
+} from './types.ts'
 
-/** Per-block structural overhead for JSON framing / type tag. */
-const BLOCK_OVERHEAD = 4
-
-/** Role-field framing overhead added per message in {@link BasicCompactService.estimateTokens}. */
-const ROLE_OVERHEAD = 4
-
-/** Tags wrapping the structured summary inside the landed checkpoint node. */
-const SUMMARY_OPEN_TAG = '<compacted-summary>'
-const SUMMARY_CLOSE_TAG = '</compacted-summary>'
-
-/**
- * Fixed summary structure for resumable checkpoints. A tagged prior checkpoint
- * is merged with newer history instead of copied forward verbatim.
- */
-const SUMMARIZE_SYSTEM_PROMPT = [
-  'You are a compaction engine for an AI coding assistant. Condense the conversation transcript into a structured checkpoint that lets another model resume the work with no loss of essential context.',
-  '',
-  'Output EXACTLY the Markdown structure below: keep every section, in order. Use terse bullets, not prose paragraphs. Write "(none)" for an empty section — never drop a section.',
-  '',
-  '## Primary Request and Intent',
-  "- [the user's original and evolving goals; quote verbatim where the exact wording matters]",
-  '',
-  '## Key Technical Concepts',
-  '- [technologies, frameworks, patterns, and conventions in play]',
-  '',
-  '## Files and Code',
-  '- [exact path: why it matters, key changes or snippets]',
-  '',
-  '## Errors and Fixes',
-  '- [error: how it was resolved, plus any related user feedback]',
-  '',
-  '## Pending Tasks',
-  '- [explicitly requested work not yet completed]',
-  '',
-  '## Current Work',
-  '- [precisely what was in progress at this checkpoint]',
-  '',
-  '## Next Step',
-  '- [the single next action, directly in line with the most recent request, or "(none)"]',
-  '',
-  '## Critical Context',
-  '- [decisions and their rationale, constraints, user preferences, open questions, data needed to continue]',
-  '',
-  'Rules:',
-  '- Preserve exact file paths, commands, error strings, identifiers, and function signatures.',
-  '- Capture user feedback and explicit instructions faithfully, especially corrections.',
-  '- Do NOT mention this summarization process or that the context was compacted.',
-  `- If the transcript already contains a ${SUMMARY_OPEN_TAG} block, it is a PRIOR checkpoint. Do not copy it forward verbatim: preserve still-true facts, drop stale ones, and merge newer information into a single consolidated summary under the same structure.`,
-].join('\n')
-
-/** Framing that makes a landed summary established context rather than a new request. */
-const CHECKPOINT_PREAMBLE =
-  'This is an automatically generated checkpoint condensing an earlier span of the conversation to free up context. Treat the captured context as established background and build on it without restating it. Continue the task directly from the messages that follow, without acknowledging this checkpoint.'
-
-/**
- * Map a terminal summary failure to an error. A max-token finish is rejected
- * because committing an incomplete checkpoint would shadow the full history.
- */
-function finishError(finish: FinishReason): Error | undefined {
-  switch (finish.kind) {
-    case 'error': {
-      const error = new Error(finish.message) as Error & { code?: string }
-      if (finish.code !== undefined) error.code = finish.code
-      return error
-    }
-    case 'aborted': {
-      const error = new Error('summarization stream aborted') as Error & { code?: string }
-      error.code = 'ABORTED'
-      return error
-    }
-    case 'max-tokens': {
-      const error = new Error('summarization truncated at the token cap (incomplete checkpoint)') as Error & { code?: string }
-      error.code = 'MAX_TOKENS'
-      return error
-    }
-    default:
-      return undefined
+/** Resolve the latest actual routed provider/model, then the complete agent fallback pair. */
+function effectiveTarget(agent: Agent): { provider: string; model: string } | undefined {
+  const latest = agent.session.requestHeader()?.config
+  if (latest !== undefined) return { provider: latest.provider, model: latest.model }
+  const { provider, model } = agent.options
+  if (provider === undefined || provider.length === 0 || model === undefined || model.length === 0) {
+    return undefined
   }
+  return { provider, model }
 }
 
 /**
- * Basic, dependency-light compaction backend: estimates the surface's token
- * footprint, summarizes the stale prefix through the model, and shadows it
- * behind a durable checkpoint. Every threshold/budget knob is required config
- * ({@link BasicCompactConfig}); the estimator's text density is the
- * `charsPerToken` knob.
+ * Build the provisional pre-step request envelope. Prompt and prefix are exact;
+ * tools and non-model call config come from the latest logged request because
+ * later request middleware has not run yet.
+ */
+function provisionalHeader(
+  target: { provider: string; model: string },
+  session: Session,
+  fullSystemPrompt: string,
+  sessionPrefix: readonly Message[],
+): EpochHeader {
+  const latest = session.requestHeader()
+  return canonicalHeader({
+    config: latest === undefined ? target : { ...latest.config, ...target },
+    ...fullSystemPrompt.length === 0 ? {} : { system: fullSystemPrompt },
+    ...latest?.tools === undefined ? {} : { tools: latest.tools },
+    ...sessionPrefix.length === 0 ? {} : { messagePrefix: [...sessionPrefix] },
+  })
+}
+
+/**
+ * Dependency-light compaction backend using `ctx.tokenMeter` for pressure,
+ * retention, provenance, and summary-convergence pricing.
+ *
+ * `summarize()` is the sole subclass customization hook; the replay and durable
+ * mutation strategy stays fixed so every pricing decision uses the singleton
+ * token meter.
  */
 export class BasicCompactService extends CompactService {
-  static inject = ['llm']
+  static inject = ['llm', 'tokenMeter']
 
-  /** Resolved configuration (`auto` defaulted). */
+  static Config: z<BasicCompactConfig> = z.object({
+    thresholdRatio: z.number().default(0.8),
+    retainTokens: z.number().step(1),
+    summarizationProvider: z.string().default(''),
+    summarizationModel: z.string().default(''),
+    maxTokens: z.number().step(1).min(1).default(8192),
+    compactionRetries: z.number().step(1).min(0).default(1),
+    auto: z.boolean().default(true),
+  })
+
+  /** Resolved and validated compaction configuration. */
   readonly config: ResolvedConfig
 
-  constructor(ctx: Context, config: BasicCompactConfig) {
+  constructor(ctx: Context, config: BasicCompactConfig = {}) {
     super(ctx)
-    this.config = resolveConfig(config)
-
-    if (this.config.auto) {
-      // Check before every step so a single growing turn can compact earlier closed steps.
-      // This serial pre-step seam mutates the surface outside the pending step.
-      ctx.on('agent/pre-step', async (agent: Agent, _turn: number, _step: number, fullSystemPrompt: string, sessionPrefix: readonly Message[], signal: AbortSignal) => {
-        try {
-          const result = await this.compactIfNeeded(agent, fullSystemPrompt, sessionPrefix, signal)
-          if (result) {
-            const after = this.estimatePressure(agent.session, fullSystemPrompt, sessionPrefix)
-            ctx.logger.info(
-              `compaction: shadowed ${result.shadowedSeqs.length} surface nodes ` +
-              `(seqs ${result.shadowedRange.start}-${result.shadowedRange.end}, ` +
-              `~${result.shadowedTokenCount} tokens) ` +
-              `→ ${after} estimated tokens after compaction`,
-            )
-          }
-        } catch (error: unknown) {
-          // A failed compaction must not prevent the model call — the surface is
-          // untouched on failure, so the loop derives the full history and the
-          // call proceeds.
-          const msg = error instanceof Error ? error.message : String(error)
-          ctx.logger.warn(`compaction failed: ${msg}; proceeding with full history`)
-        }
-      })
-    }
-  }
-
-  // ---- Token estimation (overridable hooks) ----
-
-  // TODO: chars/charsPerToken is a coarse heuristic. Replace with an exact
-  // count — a real tokenizer, or the provider's post-response `usage` (input
-  // tokens) fed back as a correction — so threshold decisions match the
-  // model's actual budget.
-  /**
-   * Estimate the token count of content blocks — chars divided by the
-   * `charsPerToken` config, with per-block overhead. Override in a subclass to
-   * plug in a real tokenizer.
-   *
-   * @param blocks - the blocks to estimate; `tool-result` blocks recurse into
-   *   their nested content, and unknown (merge-extended) types fall back to
-   *   their JSON-stringified length.
-   * @returns the estimated token count.
-   */
-  estimateContentTokens(blocks: readonly ContentBlock[]): number {
-    const { charsPerToken } = this.config
-    let tokens = 0
-    for (const block of blocks) {
-      switch (block.type) {
-        case 'text':
-        case 'reasoning':
-          tokens += Math.ceil(block.text.length / charsPerToken) + BLOCK_OVERHEAD
-          break
-        case 'tool-call':
-          tokens += Math.ceil(block.name.length / charsPerToken)
-            + Math.ceil(block.arguments.length / charsPerToken)
-            + BLOCK_OVERHEAD
-          break
-        case 'tool-result':
-          tokens += this.estimateContentTokens(block.content) + BLOCK_OVERHEAD
-          break
-        default:
-          // Unknown block types (merge-extensible ContentBlockMap):
-          // estimate conservatively via JSON stringify.
-          tokens += BLOCK_OVERHEAD + Math.ceil(JSON.stringify(block).length / charsPerToken)
-      }
-    }
-    return tokens
+    this.config = resolveConfig(config, ctx.tokenMeter)
+    if (this.config.auto) registerAutomaticCompaction(ctx, this)
   }
 
   /**
-   * Estimate token count for a single session event. Returns 0 for non-message
-   * event types (boundaries, chunks, usage, errors, compact markers).
-   *
-   * @param event - any session event; only the message-bearing types carry
-   *   content to count.
-   * @returns the estimated token count of the event's content, or 0 for a
-   *   non-message event.
-   */
-  estimateEventTokens(event: SessionEvent): number {
-    switch (event.type) {
-      case 'user/message':
-      case 'assistant/message':
-      case 'context/message':
-      case 'steering/message':
-      case 'tool/result':
-        return this.estimateContentTokens(event.data.content)
-      default:
-        return 0
-    }
-  }
-
-  /**
-   * Estimate total tokens across a list of messages plus optional system prompt.
-   *
-   * @param messages - the derived conversation messages; each adds a fixed
-   *   role-framing overhead on top of its content estimate.
-   * @param systemPrompt - counted at chars / `charsPerToken` when provided.
-   * @returns the estimated token footprint of the whole request.
-   */
-  estimateTokens(messages: readonly Message[], systemPrompt?: string): number {
-    let total = 0
-    for (const msg of messages) {
-      total += this.estimateContentTokens(msg.content)
-      total += ROLE_OVERHEAD
-    }
-    if (systemPrompt) total += Math.ceil(systemPrompt.length / this.config.charsPerToken)
-    return total
-  }
-
-  /**
-   * Summarize through a direct one-shot `ctx.llm.stream()` call, not an agent
-   * step or `agent/request` dispatch. Failure finishes and truncated summaries
-   * reject; the signal is forwarded and only text reaches the checkpoint.
-   *
-   * @param text - plain-text rendering of the conversation region to condense.
-   * @param agent - supplies the fallback model and the session id stamped on
-   *   the call; throws when neither it nor the config names a model.
-   * @param signal - optional abort signal, forwarded into the model call.
-   * @returns the text-only summary blocks plus the call envelope used
-   *   (`model`, and `maxTokens` when the summarizer has a cap).
+   * Summarize a rendered region through a direct one-shot `ctx.llm.stream()`
+   * call. Override this sole hook for a template or remote summarizer.
+   * @param text - plain-text conversation region to condense.
+   * @param agent - supplies routed-model history, fallback model, and session id.
+   * @param signal - optional cancellation forwarded to the adapter.
+   * @returns safe text summary blocks and exact auxiliary-call provenance.
    */
   async summarize(
-    text: string, agent: Agent, signal?: AbortSignal,
-  ): Promise<{ summary: ContentBlock[]; model: string; maxTokens?: number }> {
-    const assembler = new BlockAssembler()
-    const options: GenerateOptions = {
-      model: this.config.summarizationModel || agent.options.model || '',
-      messages: [{
-        role: 'user',
-        content: [{ type: 'text', text: `Summarize this conversation history:\n\n${text}\n\nSummary:` }],
-      }],
-      system: SUMMARIZE_SYSTEM_PROMPT,
-      maxTokens: this.config.maxTokens,
-      sessionId: agent.session.id,
-    }
-    // exactOptionalPropertyTypes: only set `signal` when present — assigning
-    // `undefined` to an optional `signal?: AbortSignal` is a type error.
-    if (signal) options.signal = signal
-    if (!options.model) {
-      throw new Error('no model available for summarization: set BasicCompactConfig.summarizationModel or AgentOptions.model')
-    }
-    for await (const chunk of this.ctx.llm.stream(options)) {
-      assembler.push(chunk)
-    }
-
-    const error = finishError(assembler.finish)
-    if (error) throw error
-
-    const summary = this._textOnly(assembler.message().content)
-    if (!summary.some(block => block.type === 'text' && block.text.trim().length > 0)) {
-      throw new Error('summarization produced no text summary content')
-    }
-
-    // config.maxTokens is required and validated positive, so this backend's
-    // envelope always carries the cap; the return type's optionality exists
-    // for overriding subclasses whose summarizer has none.
-    return { summary, model: options.model, maxTokens: this.config.maxTokens }
+    text: string,
+    agent: Agent,
+    signal?: AbortSignal,
+  ): Promise<{ summary: ContentBlock[]; provider: string; model: string; maxTokens?: number }> {
+    return summarizeWithLlm(this.ctx, this.config, text, agent, signal)
   }
 
-  // ---- Core API (implements the abstract contract) ----
-
   /**
-   * The sole pressure gate: count the next request's prefix, derived history,
-   * and system prompt. Above threshold, retain a recent tool-balanced tail and
-   * compact the head, reconsolidating any prior automatic checkpoint. Returns
-   * `null` when no safe or necessary range exists.
+   * Check replayed pressure for the provisional pre-step envelope and compact
+   * a tool-balanced head until it falls below the service-wide threshold.
+   * A genuinely model-less router-first step skips this provisional check.
+   * @param agent - agent whose session and provisional provider/model are measured.
+   * @param fullSystemPrompt - current assembled system prompt override.
+   * @param sessionPrefix - current request-only prefix override.
+   * @param signal - live step cancellation signal forwarded to summarization.
+   * @returns the latest compaction result, or `null` when no check/work applies.
    */
   override async compactIfNeeded(
     agent: Agent,
@@ -289,47 +120,45 @@ export class BasicCompactService extends CompactService {
     sessionPrefix: readonly Message[],
     signal: AbortSignal,
   ): Promise<CompactionResult | null> {
-    const session = agent.session
-    const threshold = Math.floor(this.config.contextWindow * this.config.thresholdRatio)
-    let result: CompactionResult | null = null
-    for (let attempt = 0; attempt <= this.config.compactionRetries; attempt++) {
-      const totalTokens = this.estimatePressure(session, fullSystemPrompt, sessionPrefix)
-      if (totalTokens < threshold) return result
+    const target = effectiveTarget(agent)
+    if (target === undefined) return null
+    const meter = this.ctx.tokenMeter
+    const requestHeader = provisionalHeader(target, agent.session, fullSystemPrompt, sessionPrefix)
+    const threshold = Math.floor(meter.contextWindow * this.config.thresholdRatio)
+    let measurement = meter.measure(agent.session, requestHeader)
+    if (measurement.totalTokens < threshold) return null
 
-      const range = this._compactableRange(session)
+    let result: CompactionResult | null = null
+    for (let attempt = 0; attempt <= this.config.compactionRetries; attempt += 1) {
+      const range = selectCompactableRange(agent.session, measurement, this.config.retainTokens)
       if (range === null) {
-        /* v8 ignore else -- defensive for non-standard subclass mutations; the concrete replace keeps a compactable head checkpoint. */
+        /* v8 ignore else -- concrete replacement preserves a compactable checkpoint; subclass hooks cannot mutate it. */
         if (result === null) return null
-        /* v8 ignore next -- paired with the ignored defensive branch above. */
+        /* v8 ignore next -- paired with the defensive post-success branch above. */
         break
       }
-
-      result = await this.compactRegion(session, range.start, range.end, agent, signal)
+      result = await this.compactRegion(agent.session, range.start, range.end, agent, signal)
+      measurement = meter.measure(agent.session, requestHeader)
+      if (measurement.totalTokens < threshold) return result
     }
 
-    const totalTokens = this.estimatePressure(session, fullSystemPrompt, sessionPrefix)
-    if (totalTokens < threshold) return result
-
     throw new Error(
       `compaction still above threshold after ${this.config.compactionRetries + 1} compaction attempts `
-      + `(${totalTokens} estimated tokens >= threshold ${threshold})`,
+      + `(${measurement.totalTokens} estimated tokens >= threshold ${threshold})`,
     )
   }
 
   /**
-   * Estimated token pressure of the NEXT request: the session prefix
-   * (`EpochHeader.messagePrefix` — request-only messages the loop sends in
-   * front of the derived history, composed before the pre-step seam and
-   * handed to the gate), the derived history, and the system prompt.
-   * @param session - the session whose next request is being estimated.
-   * @param fullSystemPrompt - the assembled system prompt (counts toward pressure).
-   * @param sessionPrefix - the instance's composed session prefix (counts toward pressure).
-   * @returns the estimated token total the next request will carry.
+   * Compact one inclusive positional surface range using the effective
+   * token meter for all retention and shrink pricing. Reject an agent that does
+   * not own the exact target before any mutation.
+   * @param session - session whose surface is mutated; must equal `agent.session`.
+   * @param start - inclusive first surface-node seq.
+   * @param end - inclusive last surface-node seq.
+   * @param agent - owner of the target session, used by the summarizer.
+   * @param signal - optional summarization cancellation signal.
+   * @returns the successful durable compaction result.
    */
-  estimatePressure(session: Session, fullSystemPrompt: string, sessionPrefix: readonly Message[]): number {
-    return this.estimateTokens([...sessionPrefix, ...session.deriveMessages()], fullSystemPrompt)
-  }
-
   override async compactRegion(
     session: Session,
     start: number,
@@ -337,215 +166,13 @@ export class BasicCompactService extends CompactService {
     agent: Agent,
     signal?: AbortSignal,
   ): Promise<CompactionResult> {
-    // Resolve by surface position: a newer replacement seq may occupy an older slot.
-    const nodes = session.surface.nodes
-    const startIdx = nodes.findIndex(n => n.seq === start)
-    const endIdx = nodes.findIndex(n => n.seq === end)
-    if (startIdx === -1) throw new Error(`compactRegion: start seq ${start} not found in surface`)
-    if (endIdx === -1) throw new Error(`compactRegion: end seq ${end} not found in surface`)
-    if (startIdx > endIdx) {
-      throw new Error(`compactRegion: start seq ${start} (position ${startIdx}) is after end seq ${end} (position ${endIdx}) on the surface`)
+    if (session !== agent.session) {
+      throw new Error('compactRegion: agent.session must be the exact target session')
     }
-
-    // Both range edges must preserve assistant tool-call/result pairing.
-    const events = session.events
-    if (!isToolPairingBalanced(nodes, events, start)) {
-      throw new Error(`compactRegion: start seq ${start} is not a balanced boundary (would split a step's tool-call/result pair)`)
-    }
-    // The cut after `end` is named by `end`'s surface successor, or `null` when
-    // `end` is the tail.
-    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-    const afterEnd: number | null = nodes[endIdx]!.next
-    if (!isToolPairingBalanced(nodes, events, afterEnd)) {
-      throw new Error(`compactRegion: end seq ${end} is not a balanced boundary (would split a step, or the step is still open)`)
-    }
-
-    if (this._isCompactionInProgress(session)) {
-      throw new Error('compaction already in progress')
-    }
-
-    // Compaction's events (compact/* and the replacement user/message) must be turn-enclosed:
-    // the session-log contract rejects any plugin event appended outside an open turn.
-    const openTurn = this._openTurn(session)
-    if (openTurn === null) {
-      throw new Error('compactRegion: no open turn — compaction events must be enclosed in a turn')
-    }
-    // Slice the ordered surface nodes [startIdx, endIdx] inclusive — the
-    // shadowed range is positional, so this is the set the replace op covers.
-    const shadowedSeqs = nodes.slice(startIdx, endIdx + 1).map(n => n.seq)
-
-    // --- Acquire lock ---
-    const startEvent = session.append('compact/start', { turn: openTurn })
-
-    try {
-      // --- Extract text and summarize ---
-      const text = renderTranscript(session.events, shadowedSeqs)
-      const { summary, model, maxTokens } = await this.summarize(text, agent, signal)
-
-      // Estimate token count of the shadowed content for provenance.
-      let shadowedTokenCount = 0
-      for (const seq of shadowedSeqs) {
-        // seq comes from a surface node — always a valid log index by construction.
-        // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-        shadowedTokenCount += this.estimateEventTokens(session.events[seq]!)
-      }
-      const framedSummary = this._frameSummary(summary)
-      const framedSummaryTokenCount = this.estimateContentTokens(framedSummary)
-      if (framedSummaryTokenCount >= shadowedTokenCount) {
-        throw new Error(
-          `summary is not smaller than the shadowed content (${framedSummaryTokenCount} estimated framed tokens >= ${shadowedTokenCount})`,
-        )
-      }
-      // --- Provenance record (log-only) ---
-      const summaryEvent = session.append('compact/summary', {
-        summary,
-        shadowedRange: { start, end },
-        shadowedSeqs,
-        shadowedTokenCount,
-        model,
-        ...maxTokens !== undefined ? { maxTokens } : {},
-      })
-
-      // --- Surface replacement --- The user/message directly shadows all compacted surface
-      // nodes with a single replace op.
-      session.append('user/message', {
-        content: framedSummary,
-        source: { kind: 'plugin', plugin: 'compact' },
-      }, {
-        surfaceOp: { op: 'replace', start, end },
-        sourceEventSeqs: [startEvent.seq, summaryEvent.seq, ...shadowedSeqs],
-      })
-
-      // --- Release lock (log-only) ---
-      // Appended LAST so the lock brackets the WHOLE operation: a crash between
-      // compact/start and here leaves a detectable orphaned lock (a compact/start
-      // with no matching compact/end) rather than a compact/end that falsely
-      // claims compaction finished before the surface replacement landed.
-      const endEvent = session.append('compact/end', { turn: openTurn })
-
-      return {
-        startSeq: startEvent.seq,
-        summarySeq: summaryEvent.seq,
-        endSeq: endEvent.seq,
-        summary,
-        shadowedRange: { start, end },
-        shadowedSeqs,
-        shadowedTokenCount,
-      }
-    } catch (error: unknown) {
-      // Always release the lock — append compact/end with the error so a
-      // wedged lock is impossible.
-      const msg = error instanceof Error ? error.message : String(error)
-      session.append('compact/end', { turn: openTurn, error: msg })
-      throw error
-    }
-  }
-
-  // ---- Internal helpers ----
-
-  /**
-   * Frame the raw summary blocks into the content that lands on the surface:
-   * a checkpoint preamble (so a resuming model reads it as a checkpoint, not a
-   * fresh user request) followed by the summary wrapped in
-   * {@link SUMMARY_OPEN_TAG}/{@link SUMMARY_CLOSE_TAG}. The tags make a prior
-   * checkpoint detectable in the transcript on the next compaction cycle, which
-   * triggers the merge rule in the summarization prompt. The raw, unframed
-   * `summary` is preserved separately on the `compact/summary` provenance event.
-   */
-  private _frameSummary(summary: readonly ContentBlock[]): ContentBlock[] {
-    return [
-      { type: 'text', text: `${CHECKPOINT_PREAMBLE}\n\n${SUMMARY_OPEN_TAG}` },
-      ...summary,
-      { type: 'text', text: SUMMARY_CLOSE_TAG },
-    ]
-  }
-
-  /**
-   * Whether a compaction is currently in progress for `session` — an unmatched `compact/start`
-   * (no later `compact/end`) WITHIN the current turn.
-   */
-  private _isCompactionInProgress(session: Session): boolean {
-    const events = session.events
-    for (let i = events.length - 1; i >= 0; i--) {
-      // Index bounded by i >= 0 and i < events.length — never undefined.
-      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-      const e = events[i]!
-      if (e.type === 'compact/start') return true
-      if (e.type === 'compact/end') break
-      // A turn/end bounds the scan: anything before it belongs to a prior
-      // (closed) turn and cannot be an in-progress compaction of THIS turn.
-      if (e.type === 'turn/end') break
-    }
-    return false
-  }
-
-  /** Resolve the next head-anchored compactable surface range, or `null`. */
-  private _compactableRange(session: Session): { start: number; end: number } | null {
-    const nodes = session.surface.nodes
-    if (nodes.length === 0) return null
-
-    const events = session.events
-    const retainBudget = this.config.retainTokens
-
-    // Walk tail→head summing per-node token estimates. `keepFromIdx` is the
-    // index of the OLDEST node we retain verbatim; everything strictly older
-    // (`[0, keepFromIdx - 1]`) is the compactable range.
-    let accumulated = 0
-    let keepFromIdx = nodes.length // nothing retained yet
-    for (let i = nodes.length - 1; i >= 0; i--) {
-      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-      const node = nodes[i]!
-      const event = events[node.seq]
-      /* v8 ignore next -- node.seq is a surface-node seq, always a valid log index by construction */
-      if (event) accumulated += this.estimateEventTokens(event)
-      keepFromIdx = i
-      if (accumulated >= retainBudget) break
-    }
-
-    // The whole surface fits the retain budget — nothing to compact.
-    if (keepFromIdx === 0) return null
-
-    // Round the cutoff to a tool-pairing boundary: if the cut before `nodes[keepFromIdx]` is
-    // unbalanced (an unanswered tool-call sits before it — i.e. it is mid-step), extend the
-    // retained side head-ward until the cut is balanced, so the compacted range ends without
-    // splitting an assistant↔result pair.
-    while (keepFromIdx > 0) {
-      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-      if (isToolPairingBalanced(nodes, events, nodes[keepFromIdx]!.seq)) break
-      keepFromIdx -= 1
-    }
-    if (keepFromIdx === 0) return null
-
-    // The compacted range is [head … keepFromIdx - 1], anchored at the head.
-    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-    const firstSeq = nodes[0]!.seq
-    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-    const cutoffSeq = nodes[keepFromIdx - 1]!.seq
-    return { start: firstSeq, end: cutoffSeq }
-  }
-
-  /** Keep only text; checkpoints cannot contain reasoning or orphan tool calls. */
-  private _textOnly(blocks: readonly ContentBlock[]): ContentBlock[] {
-    return blocks.filter((block): block is Extract<ContentBlock, { type: 'text' }> => block.type === 'text')
-  }
-
-  /**
-   * The turn number of the currently OPEN turn — a `turn/start` not yet
-   * followed by its `turn/end` — or `null` if the session has no open turn.
-   *
-   * Compaction's events must be enclosed in a turn, so scanning back from the
-   * tail: a `turn/start` means that turn is open (return it); a `turn/end` means
-   * the most recent turn already closed (return null). The whole compaction
-   * sequence (compact/start … compact/end) is stamped with this turn.
-   */
-  private _openTurn(session: Session): number | null {
-    for (let i = session.events.length - 1; i >= 0; i--) {
-      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-      const e = session.events[i]!
-      if (e.type === 'turn/start') return e.data.turn
-      if (e.type === 'turn/end') return null
-    }
-    return null
+    return compactSurfaceRegion({
+      meter: this.ctx.tokenMeter,
+      summarize: (text, owner, abort) => this.summarize(text, owner, abort),
+    }, session, start, end, agent, signal)
   }
 }
 
diff --git a/packages/compact/compact-basic/src/region.ts b/packages/compact/compact-basic/src/region.ts
new file mode 100644
index 0000000000..ac1ee260b1
--- /dev/null
+++ b/packages/compact/compact-basic/src/region.ts
@@ -0,0 +1,197 @@
+/**
+ * Surface retention selection and the log-recorded compaction transaction.
+ *
+ * @module @deepseek-ai/dsh-compact-basic/region
+ */
+
+import {
+  renderTranscript,
+  toolPairingBalancedAfter,
+  toolPairingBalancedBefore,
+} from '@deepseek-ai/dsh-compact'
+import type { CompactionResult } from '@deepseek-ai/dsh-compact'
+import type { TokenMeasurement, TokenMeterService } from '@deepseek-ai/dsh-token-meter'
+import type { Session, SessionEvent } from '@deepseek-ai/dsh-session'
+import type { Agent } from '@deepseek-ai/dsh-agent'
+import { frameSummary } from './summarizer.ts'
+import type { SummaryResult } from './summarizer.ts'
+
+interface RegionDependencies {
+  readonly meter: TokenMeterService
+  summarize(text: string, agent: Agent, signal?: AbortSignal): Promise<SummaryResult>
+}
+
+/**
+ * Resolve the next head-anchored range while retaining a priced recent tail
+ * and never splitting an assistant tool-call/result pair.
+ * @param session - session supplying authoritative current surface positions.
+ * @param measurement - unified pressure and surface measurement from the conversation meter.
+ * @param retainTokens - minimum recent tail budget retained verbatim.
+ * @returns the inclusive positional seq range to compact, or `null`.
+ */
+export function selectCompactableRange(
+  session: Session,
+  measurement: TokenMeasurement,
+  retainTokens: number,
+): { start: number; end: number } | null {
+  const pricedNodes = measurement.nodes
+  if (pricedNodes.length === 0) return null
+
+  const surfaceNodes = session.surface.nodes
+  if (surfaceNodes.length !== pricedNodes.length
+    || surfaceNodes.some((seq, index) => seq !== pricedNodes[index]?.seq)) {
+    throw new Error('compaction: token-meter surface does not match the current session surface')
+  }
+
+  let accumulated = 0
+  let keepFromIdx = pricedNodes.length
+  for (let index = pricedNodes.length - 1; index >= 0; index -= 1) {
+    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+    accumulated += pricedNodes[index]!.tokens
+    keepFromIdx = index
+    if (accumulated >= retainTokens) break
+  }
+  if (keepFromIdx === 0) return null
+
+  while (keepFromIdx > 0) {
+    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+    if (toolPairingBalancedBefore(session, surfaceNodes[keepFromIdx]!)) break
+    keepFromIdx -= 1
+  }
+  if (keepFromIdx === 0) return null
+
+  // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+  const first = surfaceNodes[0]!
+  // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+  const cutoff = surfaceNodes[keepFromIdx - 1]!
+  return { start: first, end: cutoff }
+}
+
+/**
+ * Validate and compact one positional surface span.
+ * @param dependencies - conversation meter and dynamically dispatched summarizer hook.
+ * @param session - session whose surface is mutated.
+ * @param start - inclusive first surface-node seq.
+ * @param end - inclusive last surface-node seq.
+ * @param agent - agent used by the summarizer.
+ * @param signal - optional summarization cancellation signal.
+ * @returns the successful durable compaction result.
+ */
+export async function compactSurfaceRegion(
+  dependencies: RegionDependencies,
+  session: Session,
+  start: number,
+  end: number,
+  agent: Agent,
+  signal?: AbortSignal,
+): Promise<CompactionResult> {
+  const nodes = session.surface.nodes
+  const startIdx = nodes.indexOf(start)
+  const endIdx = nodes.indexOf(end)
+  if (startIdx === -1) throw new Error(`compactRegion: start seq ${start} not found in surface`)
+  if (endIdx === -1) throw new Error(`compactRegion: end seq ${end} not found in surface`)
+  if (startIdx > endIdx) {
+    throw new Error(
+      `compactRegion: start seq ${start} (position ${startIdx}) is after end seq ${end} (position ${endIdx}) on the surface`,
+    )
+  }
+  // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+  if (!toolPairingBalancedBefore(session, nodes[startIdx]!)) {
+    throw new Error(`compactRegion: start seq ${start} is not a balanced boundary (would split a step's tool-call/result pair)`)
+  }
+  // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+  if (!toolPairingBalancedAfter(session, nodes[endIdx]!)) {
+    throw new Error(`compactRegion: end seq ${end} is not a balanced boundary (would split a step, or the step is still open)`)
+  }
+
+  const tail = inspectTurnTail(session.events)
+  if (tail.compactionInProgress) throw new Error('compaction already in progress')
+  if (tail.turn === null) {
+    throw new Error('compactRegion: no open turn — compaction events must be enclosed in a turn')
+  }
+
+  const shadowedSeqs = nodes.slice(startIdx, endIdx + 1)
+  const startEvent = session.append('compact/start', { turn: tail.turn })
+  try {
+    // Capture after the lock event so any later durable append, including a
+    // log-only one, invalidates the async selection before replacement.
+    const lockedMeasurement = dependencies.meter.measure(session)
+    const selected = lockedMeasurement.nodes.slice(startIdx, endIdx + 1)
+    if (selected.length !== shadowedSeqs.length
+      || selected.some((node, index) => node.seq !== shadowedSeqs[index])) {
+      throw new Error('compaction: selected surface changed before summarization began')
+    }
+    const shadowedTokenCount = selected.reduce((total, node) => total + node.tokens, 0)
+    const text = renderTranscript(session.events, shadowedSeqs)
+    const { summary, provider, model, maxTokens } = await dependencies.summarize(text, agent, signal)
+
+    const currentMeasurement = dependencies.meter.measure(session)
+    if (currentMeasurement.logRevision !== lockedMeasurement.logRevision) {
+      throw new Error('compaction: session log changed during summarization')
+    }
+    const framedSummary = frameSummary(summary)
+    const framedSummaryTokenCount = dependencies.meter.estimateMessage({
+      role: 'user',
+      content: framedSummary,
+    })
+    if (framedSummaryTokenCount >= shadowedTokenCount) {
+      throw new Error(
+        `summary is not smaller than the shadowed content (${framedSummaryTokenCount} estimated framed tokens >= ${shadowedTokenCount})`,
+      )
+    }
+
+    const summaryEvent = session.append('compact/summary', {
+      summary,
+      shadowedRange: { start, end },
+      shadowedSeqs,
+      shadowedTokenCount,
+      provider,
+      model,
+      ...maxTokens === undefined ? {} : { maxTokens },
+    })
+    session.append('user/message', {
+      content: framedSummary,
+      source: { kind: 'plugin', plugin: 'compact' },
+    }, {
+      surfaceOp: { op: 'replace', start, end },
+      sourceEventSeqs: [startEvent.seq, summaryEvent.seq, ...shadowedSeqs],
+    })
+    const endEvent = session.append('compact/end', { turn: tail.turn })
+    return {
+      startSeq: startEvent.seq,
+      summarySeq: summaryEvent.seq,
+      endSeq: endEvent.seq,
+      summary,
+      shadowedRange: { start, end },
+      shadowedSeqs,
+      shadowedTokenCount,
+    }
+  } catch (error: unknown) {
+    const message = error instanceof Error ? error.message : String(error)
+    session.append('compact/end', { turn: tail.turn, error: message })
+    throw error
+  }
+}
+
+/** Inspect the current turn boundary and latest compaction bracket once. */
+function inspectTurnTail(
+  events: readonly SessionEvent[],
+): { turn: number | null; compactionInProgress: boolean } {
+  let compactionInProgress = false
+  let compactionStateKnown = false
+  for (let index = events.length - 1; index >= 0; index -= 1) {
+    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+    const event = events[index]!
+    if (!compactionStateKnown) {
+      if (event.type === 'compact/start') {
+        compactionInProgress = true
+        compactionStateKnown = true
+      } else if (event.type === 'compact/end') {
+        compactionStateKnown = true
+      }
+    }
+    if (event.type === 'turn/start') return { turn: event.data.turn, compactionInProgress }
+    if (event.type === 'turn/end') return { turn: null, compactionInProgress }
+  }
+  return { turn: null, compactionInProgress }
+}
diff --git a/packages/compact/compact-basic/src/summarizer.ts b/packages/compact/compact-basic/src/summarizer.ts
new file mode 100644
index 0000000000..c814c7ecc8
--- /dev/null
+++ b/packages/compact/compact-basic/src/summarizer.ts
@@ -0,0 +1,169 @@
+/**
+ * Default one-shot summarization and durable checkpoint framing.
+ *
+ * @module @deepseek-ai/dsh-compact-basic/summarizer
+ */
+
+import type { Context } from 'cordis'
+import { BlockAssembler } from '@deepseek-ai/dsh-llm'
+import type { ContentBlock, FinishReason, GenerateOptions } from '@deepseek-ai/dsh-llm'
+import type { Agent } from '@deepseek-ai/dsh-agent'
+import type { ResolvedConfig } from './types.ts'
+
+/** Tags wrapping the structured summary inside the landed checkpoint node. */
+const SUMMARY_OPEN_TAG = '<compacted-summary>'
+const SUMMARY_CLOSE_TAG = '</compacted-summary>'
+
+/** Fixed structure required from the auxiliary summarization call. */
+const SUMMARIZE_SYSTEM_PROMPT = [
+  'You are a compaction engine for an AI coding assistant. Condense the conversation transcript into a structured checkpoint that lets another model resume the work with no loss of essential context.',
+  '',
+  'Output EXACTLY the Markdown structure below: keep every section, in order. Use terse bullets, not prose paragraphs. Write "(none)" for an empty section — never drop a section.',
+  '',
+  '## Primary Request and Intent',
+  "- [the user's original and evolving goals; quote verbatim where the exact wording matters]",
+  '',
+  '## Key Technical Concepts',
+  '- [technologies, frameworks, patterns, and conventions in play]',
+  '',
+  '## Files and Code',
+  '- [exact path: why it matters, key changes or snippets]',
+  '',
+  '## Errors and Fixes',
+  '- [error: how it was resolved, plus any related user feedback]',
+  '',
+  '## Pending Tasks',
+  '- [explicitly requested work not yet completed]',
+  '',
+  '## Current Work',
+  '- [precisely what was in progress at this checkpoint]',
+  '',
+  '## Next Step',
+  '- [the single next action, directly in line with the most recent request, or "(none)"]',
+  '',
+  '## Critical Context',
+  '- [decisions and their rationale, constraints, user preferences, open questions, data needed to continue]',
+  '',
+  'Rules:',
+  '- Preserve exact file paths, commands, error strings, identifiers, and function signatures.',
+  '- Capture user feedback and explicit instructions faithfully, especially corrections.',
+  '- Do NOT mention this summarization process or that the context was compacted.',
+  `- If the transcript already contains a ${SUMMARY_OPEN_TAG} block, it is a PRIOR checkpoint. Do not copy it forward verbatim: preserve still-true facts, drop stale ones, and merge newer information into a single consolidated summary under the same structure.`,
+].join('\n')
+
+/** Framing that makes the replacement user message established context. */
+const CHECKPOINT_PREAMBLE =
+  'This is an automatically generated checkpoint condensing an earlier span of the conversation to free up context. Treat the captured context as established background and build on it without restating it. Continue the task directly from the messages that follow, without acknowledging this checkpoint.'
+
+/** Safe summary content plus the exact auxiliary call envelope recorded in provenance. */
+export interface SummaryResult {
+  summary: ContentBlock[]
+  provider: string
+  model: string
+  maxTokens?: number
+}
+
+/**
+ * Run the default direct `ctx.llm.stream()` summarization call.
+ * @param ctx - context providing the LLM service.
+ * @param config - resolved backend configuration.
+ * @param text - rendered transcript region to summarize.
+ * @param agent - supplies routed-model history, fallback model, and session id.
+ * @param signal - optional cancellation forwarded to the adapter.
+ * @returns safe text-only summary blocks and exact call provenance.
+ */
+export async function summarizeWithLlm(
+  ctx: Context,
+  config: ResolvedConfig,
+  text: string,
+  agent: Agent,
+  signal?: AbortSignal,
+): Promise<SummaryResult> {
+  const latest = agent.session.requestHeader()?.config
+  const configured = config.summarizationProvider.length === 0
+    ? undefined
+    : { provider: config.summarizationProvider, model: config.summarizationModel }
+  const agentTarget = agent.options.provider !== undefined
+    && agent.options.provider.length > 0
+    && agent.options.model !== undefined
+    && agent.options.model.length > 0
+    ? { provider: agent.options.provider, model: agent.options.model }
+    : undefined
+  const target = configured ?? latest ?? agentTarget
+  if (target === undefined) {
+    throw new Error(
+      'no provider/model available for summarization: set both BasicCompactConfig summarization fields, route one request, or set both AgentOptions fields',
+    )
+  }
+
+  const assembler = new BlockAssembler()
+  const options: GenerateOptions = {
+    provider: target.provider,
+    model: target.model,
+    messages: [{
+      role: 'user',
+      content: [{ type: 'text', text: `Summarize this conversation history:\n\n${text}\n\nSummary:` }],
+    }],
+    system: SUMMARIZE_SYSTEM_PROMPT,
+    maxTokens: config.maxTokens,
+    sessionId: agent.session.id,
+    ...signal === undefined ? {} : { signal },
+  }
+  for await (const chunk of ctx.llm.stream(options)) assembler.push(chunk)
+  const error = finishError(assembler.finish)
+  if (error !== undefined) throw error
+
+  const summary = textOnly(assembler.message().content)
+  if (!summary.some(block => block.text.trim().length > 0)) {
+    throw new Error('summarization produced no text summary content')
+  }
+  return {
+    summary,
+    provider: target.provider,
+    model: target.model,
+    maxTokens: config.maxTokens,
+  }
+}
+
+/**
+ * Wrap raw summary blocks in the durable checkpoint framing.
+ * @param summary - safe text-only model output.
+ * @returns content for the synthesized replacement user message.
+ */
+export function frameSummary(summary: readonly ContentBlock[]): ContentBlock[] {
+  return [
+    { type: 'text', text: `${CHECKPOINT_PREAMBLE}\n\n${SUMMARY_OPEN_TAG}` },
+    ...summary,
+    { type: 'text', text: SUMMARY_CLOSE_TAG },
+  ]
+}
+
+/** Map a terminal summarization finish to its fail-closed error. */
+function finishError(finish: FinishReason): Error | undefined {
+  switch (finish.kind) {
+    case 'error': {
+      const error = new Error(finish.message) as Error & { code?: string }
+      if (finish.code !== undefined) error.code = finish.code
+      return error
+    }
+    case 'aborted': {
+      const error = new Error('summarization stream aborted') as Error & { code?: string }
+      error.code = 'ABORTED'
+      return error
+    }
+    case 'max-tokens': {
+      const error = new Error('summarization truncated at the token cap (incomplete checkpoint)') as Error & { code?: string }
+      error.code = 'MAX_TOKENS'
+      return error
+    }
+    default:
+      return undefined
+  }
+}
+
+/** Keep only text blocks before synthesizing a user message. */
+function textOnly(
+  blocks: readonly ContentBlock[],
+): Array<Extract<ContentBlock, { type: 'text' }>> {
+  return blocks.filter((block): block is Extract<ContentBlock, { type: 'text' }> => block.type === 'text')
+}
diff --git a/packages/compact/compact-basic/src/types.ts b/packages/compact/compact-basic/src/types.ts
index 05169286d7..8145ce164e 100644
--- a/packages/compact/compact-basic/src/types.ts
+++ b/packages/compact/compact-basic/src/types.ts
@@ -1,94 +1,34 @@
 /**
- * Configuration vocabulary for the basic compaction backend.
- *
- * Every tunable lives here, in the implementation — the abstract contract
- * (`@deepseek-ai/dsh-compact`) carries no config, because thresholds and
- * retention policy are HOW decisions a different backend would make
- * differently.
+ * Configuration vocabulary for the replay-aware basic compaction backend.
  *
  * @module @deepseek-ai/dsh-compact-basic/types
  */
 
-/**
- * Backend configuration. Every knob is REQUIRED except `auto` and
- * `charsPerToken`: there is no concrete data yet to justify default
- * thresholds/budgets, so a consumer must state each value explicitly rather
- * than inherit a guessed default. `auto` alone defaults to `true`
- * (auto-compaction is the intended posture), and `charsPerToken` defaults to
- * the English-text heuristic its estimator was calibrated on.
- */
+/** Basic compaction configuration; every common field has a deployment default. */
 export interface BasicCompactConfig {
-  /** Context window size in tokens. */
-  contextWindow: number
-  /** Compact when estimated token usage exceeds this fraction of context window. */
-  thresholdRatio: number
-  /** Number of tokens of recent context to retain during compaction. */
-  retainTokens: number
-  /** Model to use for summarization (`''` — uses the agent's model). */
-  summarizationModel: string
-  /** Provider generation cap for the summarization call. */
-  maxTokens: number
-  /** Extra compaction attempts when the first compacted surface is still over threshold. */
-  compactionRetries: number
-  /** Enable automatic compaction on the `agent/pre-step` seam (default true). */
+  /** Compact at this fraction of the token meter's context window. Defaults to `0.8`. */
+  thresholdRatio?: number
+  /** Recent surface tokens retained verbatim. Defaults to `floor(contextWindow * 0.16)`. */
+  retainTokens?: number
+  /** Summary provider; `''` resolves the latest routed pair, then the agent pair. Defaults to `''`. */
+  summarizationProvider?: string
+  /** Summary model; `''` resolves the latest routed pair, then the agent pair. Defaults to `''`. */
+  summarizationModel?: string
+  /** Provider generation cap for summarization. Defaults to `8192`. */
+  maxTokens?: number
+  /** Extra attempts after the first compaction when pressure remains above threshold. Defaults to `1`. */
+  compactionRetries?: number
+  /** Enable the automatic `agent/pre-step` pressure listener. Defaults to `true`. */
   auto?: boolean
-  /**
-   * Text density for the token estimator: estimated tokens = chars /
-   * `charsPerToken`. Defaults to 4 (typical English text). A CJK-heavy
-   * deployment should set ~1-2 — CJK runs at roughly 1-2 chars per token, so
-   * the default UNDERestimates several-fold and compaction fires far too late.
-   * May be fractional.
-   */
-  charsPerToken?: number
 }
 
-/** Resolved config with `auto` and `charsPerToken` defaulted. */
-export type ResolvedConfig = Required<BasicCompactConfig>
-
-/**
- * Default `auto`/`charsPerToken` when unset and reject nonsensical numeric knobs.
- *
- * @param config - the raw, unresolved backend config.
- * @returns the validated config with `auto` and `charsPerToken` defaulted.
- */
-export function resolveConfig(config: BasicCompactConfig): ResolvedConfig {
-  const resolved: ResolvedConfig = { auto: true, charsPerToken: 4, ...config }
-
-  assertPositiveInteger('contextWindow', resolved.contextWindow)
-  assertRatio('thresholdRatio', resolved.thresholdRatio)
-  assertNonNegativeInteger('retainTokens', resolved.retainTokens)
-  assertPositiveInteger('maxTokens', resolved.maxTokens)
-  assertNonNegativeInteger('compactionRetries', resolved.compactionRetries)
-  assertPositiveFinite('charsPerToken', resolved.charsPerToken)
-  if (typeof resolved.summarizationModel !== 'string') {
-    throw new Error('BasicCompactConfig: summarizationModel must be a string.')
-  }
-  if (typeof resolved.auto !== 'boolean') {
-    throw new Error('BasicCompactConfig: auto must be a boolean.')
-  }
-  return resolved
-}
-
-function assertPositiveInteger(name: string, value: number): void {
-  if (!Number.isInteger(value) || value <= 0) {
-    throw new Error(`BasicCompactConfig: ${name} (${value}) must be a positive integer.`)
-  }
-}
-
-function assertNonNegativeInteger(name: string, value: number): void {
-  if (!Number.isInteger(value) || value < 0) {
-    throw new Error(`BasicCompactConfig: ${name} (${value}) must be a non-negative integer.`)
-  }
-}
-
-function assertPositiveFinite(name: string, value: number): void {
-  if (typeof value !== 'number' || !Number.isFinite(value) || value <= 0) {
-    throw new Error(`BasicCompactConfig: ${name} (${value}) must be a positive finite number.`)
-  }
-}
-
-function assertRatio(name: string, value: number): void {
-  if (typeof value !== 'number' || !Number.isFinite(value) || value <= 0 || value > 1) {
-    throw new Error(`BasicCompactConfig: ${name} (${value}) must be a number in (0, 1].`)
-  }
+/** Validated and detached compaction configuration. */
+export interface ResolvedConfig {
+  readonly thresholdRatio: number
+  readonly retainTokens: number
+  readonly summarizationProvider: string
+  readonly summarizationModel: string
+  readonly maxTokens: number
+  readonly compactionRetries: number
+  readonly auto: boolean
 }
diff --git a/packages/compact/compact-basic/tests/compact-basic.spec.ts b/packages/compact/compact-basic/tests/compact-basic.spec.ts
index 8c5cc183c1..2d315293ed 100644
--- a/packages/compact/compact-basic/tests/compact-basic.spec.ts
+++ b/packages/compact/compact-basic/tests/compact-basic.spec.ts
@@ -1,1682 +1,790 @@
-import { describe, expect, it } from 'vitest'
+import { describe, expect, it, vi } from 'vitest'
 import { Context } from 'cordis'
-import { BasicCompactService } from '@deepseek-ai/dsh-compact-basic'
+import BasicCompactService, { resolveConfig } from '@deepseek-ai/dsh-compact-basic'
 import type { BasicCompactConfig } from '@deepseek-ai/dsh-compact-basic'
+import { selectCompactableRange } from '@deepseek-ai/dsh-compact-basic/src/region.ts'
+import type { CompactionResult } from '@deepseek-ai/dsh-compact'
+import LlmService, { CallId, LlmAdapter } from '@deepseek-ai/dsh-llm'
 import type { ContentBlock, GenerateOptions, Message, StreamChunk } from '@deepseek-ai/dsh-llm'
-import { CallId, LlmAdapter, LlmService } from '@deepseek-ai/dsh-llm'
-import SessionStore, { Session, SessionId } from '@deepseek-ai/dsh-session'
-import type { SessionEvent, SurfaceEvent } from '@deepseek-ai/dsh-session'
-import * as Invariants from '@deepseek-ai/dsh-invariants'
+import { Session, SessionId } from '@deepseek-ai/dsh-session'
+import TokenMeterService from '@deepseek-ai/dsh-token-meter'
 import type { Agent } from '@deepseek-ai/dsh-agent'
 
-/** A never-aborted signal for the required `compactIfNeeded`/listener arg. */
 const SIGNAL = new AbortController().signal
+const MODEL = 'test-model'
 
-/**
- * Baseline config with every required knob set. `BasicCompactConfig` has no
- * defaults for the numeric/model knobs (only `auto` defaults), so each test
- * builds a complete config via `cfg()` and overrides only the knob under test.
- */
-const TEST_CONFIG: BasicCompactConfig = {
-  contextWindow: 128000,
-  thresholdRatio: 0.8,
-  retainTokens: 20480,
-  summarizationModel: '',
-  maxTokens: 8192,
-  compactionRetries: 1,
-}
-
-/** A complete config with `overrides` applied over the baseline. */
-function cfg(overrides: Partial<BasicCompactConfig> = {}): BasicCompactConfig {
-  return { ...TEST_CONFIG, ...overrides }
-}
-
-/** Long enough that the real checkpoint preamble is smaller than two fixture messages. */
-const LONG_FIXTURE_TEXT = ' Detailed fixture context that makes framed checkpoint compaction genuinely shrinking.'.repeat(20)
-
-/**
- * A BasicCompactService with summarize() stubbed (no real model call) and a
- * predictable token estimate, for deterministic unit tests of the algorithm.
- */
-class TestCompactService extends BasicCompactService {
-  private readonly summaryOutputs = new WeakSet<readonly ContentBlock[]>()
-  /** Boundary/unit tests use tiny fixtures; keep framing from dominating them unless a test opts out. */
-  estimateFramedSummariesCheaply = true
-  /** Track calls to summarize for test assertions. */
-  summarizeCalls: { text: string; model: string }[] = []
-  /** The fixed summary to return. */
-  mockSummary: ContentBlock[] = [{ type: 'text', text: 'Test summary of compacted content.' }]
-  /** Per-call summaries; when set, each summarize() call shifts one value. */
-  mockSummaryQueue: ContentBlock[][] = []
-  /** If set, summarize() throws this error. */
-  summarizeError: Error | null = null
-
-  override estimateContentTokens(blocks: readonly ContentBlock[]): number {
-    if (this.summaryOutputs.has(blocks)) return blocks.length * 2
-    if (this.estimateFramedSummariesCheaply && isFramedCheckpoint(blocks)) return blocks.length * 2
-    // 10 tokens per block — predictable for retention/threshold math.
-    return blocks.length * 10
-  }
-
-  override async summarize(text: string, agent: Agent): Promise<{ summary: ContentBlock[]; model: string; maxTokens?: number }> {
-    const model = this.config.summarizationModel || agent.options.model || ''
-    this.summarizeCalls.push({ text, model })
-    if (this.summarizeError) throw this.summarizeError
-    const summary = this.mockSummaryQueue.shift() ?? this.mockSummary
-    this.summaryOutputs.add(summary)
-    return { summary, model }
-  }
-}
-
-function isFramedCheckpoint(blocks: readonly ContentBlock[]): boolean {
-  const first = blocks[0]
-  const last = blocks[blocks.length - 1]
-  return first?.type === 'text'
-    && first.text.includes('<compacted-summary>')
-    && last?.type === 'text'
-    && last.text === '</compacted-summary>'
-}
-
-/** Create a test service with a throwaway context (auto disabled — no model). */
-function createTestService(overrides: Partial<BasicCompactConfig> = {}): TestCompactService {
-  return new TestCompactService(new Context(), cfg({ auto: false, ...overrides }))
-}
-
-/** Build closed turns plus an open compaction turn unless `leaveOpen` is false. */
-function multiTurnSession(turns: number, messagesPerTurn: number = 2, opts: { leaveOpen?: boolean } = {}): Session {
-  const leaveOpen = opts.leaveOpen ?? true
-  const s = new Session(SessionId('test'))
-  for (let t = 1; t <= turns; t++) {
-    s.append('turn/start', { turn: t, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: t, step: 1 })
-    for (let m = 0; m < messagesPerTurn; m++) {
-      s.append('user/message', {
-        content: [{ type: 'text', text: `turn ${t} user message ${m + 1}.${LONG_FIXTURE_TEXT}` }],
-        source: { kind: 'user' },
-      }, { surfaceOp: 'append' })
-      s.append('assistant/message', {
-        turn: t, step: 1,
-        content: [{ type: 'text', text: `turn ${t} assistant response ${m + 1}.${LONG_FIXTURE_TEXT}` }],
-      }, { surfaceOp: 'append' })
-    }
-    s.append('step/end', { turn: t, step: 1 })
-    s.append('turn/end', { turn: t, reason: { kind: 'completed' } })
-  }
-  // Open one more turn so compaction's events are turn-enclosed, as they are
-  // when the loop runs the auto-compaction listener mid-turn.
-  if (leaveOpen) {
-    s.append('turn/start', { turn: turns + 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-  }
-  return s
-}
-
-/** Build a session with tool calls for richer extraction tests. */
-function sessionWithTools(): Session {
-  const s = new Session(SessionId('tools'))
-  s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-  s.append('step/start', { turn: 1, step: 1 })
-  s.append('user/message', {
-    content: [{ type: 'text', text: 'read file x' }],
-    source: { kind: 'user' },
-  }, { surfaceOp: 'append' })
-  s.append('assistant/message', {
-    turn: 1, step: 1,
-    content: [
-      { type: 'text', text: 'Let me read that file.' },
-      { type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{"command":"cat x"}' },
-    ],
-  }, { surfaceOp: 'append' })
-  s.append('tool/call', { turn: 1, step: 1, callId: CallId('c1'), name: 'bash', arguments: '{"command":"cat x"}' })
-  s.append('tool/result', {
-    turn: 1, step: 1, callId: CallId('c1'),
-    content: [{ type: 'text', text: 'hello world' }],
-    isError: false,
-  }, { surfaceOp: 'append' })
-  s.append('assistant/message', {
-    turn: 1, step: 1,
-    content: [{ type: 'text', text: 'The file contains: hello world' }],
-  }, { surfaceOp: 'append' })
-  s.append('step/end', { turn: 1, step: 1 })
-  s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-  // Open a trailing turn so compaction's events are turn-enclosed (as they are
-  // when the loop runs the auto-compaction listener mid-turn).
-  s.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
-  return s
-}
-
-/**
- * Build a session of `turns` turns, each a SINGLE step containing an
- * assistant/message that issues a tool-call plus its tool/result — the real
- * multi-node-step shape (a step is two surface nodes: the assistant and the
- * result). Each turn is preceded by a user/message. Used to exercise
- * step-alignment: a region boundary must not fall between the assistant and its
- * result.
- */
-function toolTurnSession(turns: number): Session {
-  const s = new Session(SessionId('tools-multi'))
-  for (let t = 1; t <= turns; t++) {
-    s.append('turn/start', { turn: t, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('user/message', {
-      content: [{ type: 'text', text: `turn ${t} request` }],
-      source: { kind: 'user' },
-    }, { surfaceOp: 'append' })
-    s.append('step/start', { turn: t, step: 1 })
-    s.append('assistant/message', {
-      turn: t, step: 1,
-      content: [
-        { type: 'text', text: `turn ${t} calling tool` },
-        { type: 'tool-call', id: CallId(`c${t}`), name: 'bash', arguments: '{"command":"ls"}' },
-      ],
-    }, { surfaceOp: 'append' })
-    s.append('tool/call', { turn: t, step: 1, callId: CallId(`c${t}`), name: 'bash', arguments: '{"command":"ls"}' })
-    s.append('tool/result', {
-      turn: t, step: 1, callId: CallId(`c${t}`),
-      content: [{ type: 'text', text: `turn ${t} output` }],
-      isError: false,
-    }, { surfaceOp: 'append' })
-    s.append('step/end', { turn: t, step: 1 })
-    s.append('turn/end', { turn: t, reason: { kind: 'completed' } })
-  }
-  // Open a trailing turn so compaction's events are turn-enclosed.
-  s.append('turn/start', { turn: turns + 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-  return s
-}
-
-/**
- * Assert the derived transcript has NO orphaned tool-result: every
- * `tool-result` block's `toolCallId` must be matched by a preceding `tool-call`
- * block in an earlier (assistant) message. A dangling tool-result is exactly
- * what splitting a step at compaction produces, and every provider rejects it.
- */
-function expectNoOrphanToolResults(messages: Message[]): void {
-  const seenCallIds = new Set<string>()
-  for (const msg of messages) {
-    for (const block of msg.content) {
-      if (block.type === 'tool-call') seenCallIds.add(block.id)
-      if (block.type === 'tool-result') {
-        expect(seenCallIds.has(block.toolCallId),
-          `orphaned tool-result for callId ${block.toolCallId} (no preceding tool-call)`).toBe(true)
-      }
-    }
-  }
-}
-
-describe('BasicCompactService step-alignment (never split a tool-call/result pair)', () => {
-  it('compactIfNeeded rounds the retained boundary head-ward to keep a whole step (no orphaned tool-result)', async () => {
-    // Retain the recent tail while the older assistant/result pairs compact as
-    // whole units; no boundary may orphan a result.
-    const svc = createTestService({ contextWindow: 280, thresholdRatio: 0.5, retainTokens: 55 })
-    const session = toolTurnSession(3)
-
-    const result = await compactIfNeeded(svc, session, '', 'm', SIGNAL)
-    expect(result).not.toBeNull()
-    expect(result!.shadowedSeqs.length).toBeGreaterThan(0)
-    // No dangling tool-result: every compacted/retained step stayed whole.
-    expectNoOrphanToolResults(session.deriveMessages())
-    // The most-recent step's result is retained verbatim (still on the surface).
-    const lastResultSeq = session.events.findLast(e => e.type === 'tool/result')!.seq
-    expect(result!.shadowedSeqs).not.toContain(lastResultSeq)
-  })
-
-  it('compactIfNeeded returns null when the only compactable region is an un-splittable single step', async () => {
-    // The only candidate cut is inside one assistant/result pair; with no safe
-    // compactable prefix, decline rather than split it.
-    const s = new Session(SessionId('one-step'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('assistant/message', {
-      turn: 1, step: 1,
-      content: [{ type: 'text', text: 'calling' }, { type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{}' }],
-    }, { surfaceOp: 'append' })
-    s.append('tool/call', { turn: 1, step: 1, callId: CallId('c1'), name: 'bash', arguments: '{}' })
-    s.append('tool/result', { turn: 1, step: 1, callId: CallId('c1'), content: [{ type: 'text', text: 'out' }], isError: false }, { surfaceOp: 'append' })
-    s.append('step/end', { turn: 1, step: 1 })
-    // Turn stays open.
-
-    const svc = createTestService({ contextWindow: 100, thresholdRatio: 0.1, retainTokens: 5 })
-    const result = await compactIfNeeded(svc, s, '', 'm', SIGNAL)
-    expect(result).toBeNull()
-    expect(s.events.some(e => e.type === 'compact/start')).toBe(false)
-  })
-
-  it('compactRegion rejects a start that splits a step (unbalanced boundary)', async () => {
-    const svc = createTestService()
-    const session = toolTurnSession(1)
-    const nodes = session.surface.nodes // [user, asst(tool-call), result]
-    const userSeq = nodes[0]!.seq
-    const resultSeq = nodes[2]!.seq
-    // start = the tool/result: its issuing assistant precedes it IN THE SAME STEP,
-    // so starting here would orphan that assistant's tool-call. end is fine (user).
-    await expect(compactRegion(svc, session, resultSeq, resultSeq, 'm'))
-      .rejects.toThrow(/start seq .* is not a balanced boundary/)
-    expect(userSeq).toBeLessThan(resultSeq) // sanity: ordering as expected
-  })
-
-  it('compactRegion rejects an end that splits a step (unbalanced boundary)', async () => {
-    const svc = createTestService()
-    const session = toolTurnSession(1)
-    const nodes = session.surface.nodes
-    const userSeq = nodes[0]!.seq
-    const asstSeq = nodes[1]!.seq
-    // end = the assistant/message: its tool/result follows IN THE SAME STEP, so
-    // ending here would strand that result. start is fine (the pre-step user).
-    await expect(compactRegion(svc, session, userSeq, asstSeq, 'm'))
-      .rejects.toThrow(/end seq .* is not a balanced boundary/)
-  })
-
-  it('compactRegion rejects an end inside an open tail step', async () => {
-    const svc = createTestService()
-    const s = new Session(SessionId('open-tail'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('user/message', { content: [{ type: 'text', text: 'go' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('assistant/message', {
-      turn: 1, step: 1,
-      content: [{ type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{}' }],
-    }, { surfaceOp: 'append' })
-    const nodes = s.surface.nodes // [user, asst]
-    const userSeq = nodes[0]!.seq
-    const asstSeq = nodes[1]!.seq
-    await expect(compactRegion(svc, s, userSeq, asstSeq, 'm'))
-      .rejects.toThrow(/end seq .* is not a balanced boundary/)
-  })
-
-  it('compactRegion accepts step-aligned boundaries (pre-step user → last result of a closed step)', async () => {
-    const svc = createTestService()
-    const session = toolTurnSession(2)
-    const nodes = session.surface.nodes // [user1, asst1, res1, user2, asst2, res2]
-    const startSeq = nodes[0]!.seq // pre-step user1 (free boundary)
-    const endSeq = nodes[2]!.seq   // res1 = last node of turn 1's closed step
-    const result = await compactRegion(svc, session, startSeq, endSeq, 'm')
-    expect(result.shadowedRange).toEqual({ start: startSeq, end: endSeq })
-    expectNoOrphanToolResults(session.deriveMessages())
-  })
-
-  it('compactRegion accepts a single inter-step node (start === end on a pre-step user/message)', async () => {
-    const svc = createTestService()
-    const session = toolTurnSession(1)
-    const nodes = session.surface.nodes
-    const userSeq = nodes[0]!.seq // pre-step user: free boundary both ways
-    const result = await compactRegion(svc, session, userSeq, userSeq, 'm')
-    expect(result.shadowedRange).toEqual({ start: userSeq, end: userSeq })
-  })
-
-  it('compactRegion accepts an injection-turn context node (no step at all)', async () => {
-    const svc = createTestService()
-    const s = new Session(SessionId('inject'))
-    // An idle inject(): turn/start → context/message, NO step. A later turn is
-    // open so compaction's events are turn-enclosed.
-    s.append('turn/start', { turn: 1, trigger: { kind: 'injection', source: { kind: 'user' } } })
-    s.append('context/message', { content: [{ type: 'text', text: 'ctx' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    s.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
-    const nodes = s.surface.nodes
-    const ctxSeq = nodes[0]!.seq
-    const result = await compactRegion(svc, s, ctxSeq, ctxSeq, 'm')
-    expect(result.shadowedRange).toEqual({ start: ctxSeq, end: ctxSeq })
-  })
-})
-
-describe('BasicCompactService.estimateEventTokens', () => {
-  it('returns 0 for non-message events (boundary, chunk, step/end, tool/call)', () => {
-    const svc = createTestService()
-    expect(svc.estimateEventTokens({ type: 'turn/start', seq: 0, time: 1, data: { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } } })).toBe(0)
-    expect(svc.estimateEventTokens({ type: 'step/start', seq: 1, time: 2, data: { turn: 1, step: 1 } })).toBe(0)
-    expect(svc.estimateEventTokens({ type: 'assistant/chunk', seq: 2, time: 3, data: { turn: 1, step: 1, chunk: { type: 'text-delta', index: 0, text: 'h' } } })).toBe(0)
-    expect(svc.estimateEventTokens({ type: 'step/end', seq: 3, time: 4, data: { turn: 1, step: 1 } })).toBe(0)
-    expect(svc.estimateEventTokens({ type: 'tool/call', seq: 4, time: 5, data: { turn: 1, step: 1, callId: CallId('c1'), name: 'read', arguments: '{}' } })).toBe(0)
-  })
-
-  it('returns estimate for message-producing events', () => {
-    const svc = createTestService()
-    const userEvent: SessionEvent = { type: 'user/message', seq: 0, time: 1, data: { content: [{ type: 'text', text: 'hello' }], source: { kind: 'user' } } }
-    expect(svc.estimateEventTokens(userEvent)).toBe(10)
-
-    const asstEvent: SessionEvent = { type: 'assistant/message', seq: 1, time: 2, data: { turn: 1, step: 1, content: [{ type: 'text', text: 'a' }, { type: 'text', text: 'b' }] } }
-    expect(svc.estimateEventTokens(asstEvent)).toBe(20)
-
-    const toolEvent: SessionEvent = { type: 'tool/result', seq: 2, time: 3, data: { turn: 1, step: 1, callId: CallId('c1'), content: [{ type: 'text', text: 'output' }], isError: false } }
-    expect(svc.estimateEventTokens(toolEvent)).toBe(10)
-  })
-})
-
-describe('BasicCompactService.estimateTokens', () => {
-  it('sums token estimates across messages', () => {
-    const svc = createTestService()
-    const messages: Message[] = [
-      { role: 'user', content: [{ type: 'text', text: 'hello' }] },
-      { role: 'assistant', content: [{ type: 'text', text: 'hi' }, { type: 'text', text: 'there' }] },
-    ]
-    // 1 block * 10 + 4 (role) + 2 blocks * 10 + 4 (role) = 10 + 4 + 20 + 4 = 38
-    expect(svc.estimateTokens(messages)).toBe(38)
-  })
-
-  it('includes system prompt in the estimate', () => {
-    const svc = createTestService()
-    const messages: Message[] = [
-      { role: 'user', content: [{ type: 'text', text: 'hi' }] },
-    ]
-    const systemPrompt = 'You are a helpful assistant.'
-    // 1 block * 10 + 4 (role) + ceil(28/4) = 10 + 4 + 7 = 21
-    expect(svc.estimateTokens(messages, systemPrompt)).toBe(21)
-  })
-})
-
-describe('BasicCompactService.compactRegion', () => {
-  it('shadows surface nodes and inserts a summary via user/message', async () => {
-    const svc = createTestService()
-    const session = multiTurnSession(3, 1) // 3 turns, 2 surface nodes each = 6 nodes
-
-    const nodes = session.surface.nodes
-    expect(nodes.length).toBe(6)
-
-    const firstSeq = nodes[0]!.seq
-    const secondSeq = nodes[1]!.seq
-    const result = await compactRegion(svc, session, firstSeq, secondSeq, 'test-model')
-
-    expect(result.shadowedSeqs).toEqual([firstSeq, secondSeq])
-    expect(result.shadowedRange.start).toBe(firstSeq)
-    expect(result.shadowedRange.end).toBe(secondSeq)
-    expect(result.summary).toEqual(svc.mockSummary)
-
-    const events = session.events
-    const startEvent = events.findLast(e => e.type === 'compact/start')
-    const summaryEvent = events.findLast(e => e.type === 'compact/summary')
-    const endEvent = events.findLast(e => e.type === 'compact/end')
-    expect(startEvent).toBeDefined()
-    expect(summaryEvent).toBeDefined()
-    expect(endEvent).toBeDefined()
-    // The provenance record carries the summarize call's envelope, so "which
-    // model wrote this summary" is answerable from the log alone.
-    expect(summaryEvent?.type === 'compact/summary' && summaryEvent.data.model).toBe('test-model')
-
-    // compact/* events are log-only — no surfaceOp (type system enforces this).
-    const startRaw = startEvent as unknown as { surfaceOp?: unknown }
-    expect(startRaw.surfaceOp).toBeUndefined()
-
-    // The user/message carries the replace surfaceOp.
-    const userMsg = events.findLast(e => e.type === 'user/message')!
-    const surfaceUserMsg = userMsg as SurfaceEvent
-    expect(surfaceUserMsg.surfaceOp).toEqual({ op: 'replace', start: firstSeq, end: secondSeq })
-    expect(surfaceUserMsg.sourceEventSeqs).toContain(startEvent!.seq)
-    expect(surfaceUserMsg.sourceEventSeqs).toContain(summaryEvent!.seq)
-    expect(surfaceUserMsg.sourceEventSeqs).toContain(firstSeq)
-    expect(surfaceUserMsg.sourceEventSeqs).toContain(secondSeq)
-    // compact/end is appended AFTER the replacement (the lock brackets the whole
-    // op), so the replacement cannot reference it — sourceEventSeqs may only
-    // reference earlier seqs.
-    expect(surfaceUserMsg.sourceEventSeqs).not.toContain(endEvent!.seq)
-    expect(endEvent!.seq).toBeGreaterThan(userMsg.seq)
-
-    // Surface now has: summary user/message + retained 4 nodes = 5 nodes.
-    const newNodes = session.surface.nodes
-    expect(newNodes.length).toBe(5)
-    expect(newNodes[0]!.seq).toBe(userMsg.seq)
-
-    // deriveMessages() produces the framed summary as a user-role message:
-    // a checkpoint preamble + tag-wrapped summary blocks.
-    const derived = session.deriveMessages()
-    expect(derived.length).toBe(5)
-    expect(derived[0]!.role).toBe('user')
-    const framed = derived[0]!.content
-    expect(framed[0]).toMatchObject({ type: 'text' })
-    expect((framed[0] as { text: string }).text).toContain('<compacted-summary>')
-    expect(framed).toContainEqual(svc.mockSummary[0])
-    expect((framed[framed.length - 1] as { text: string }).text).toBe('</compacted-summary>')
-  })
-
-  it('throws when start or end are not surface nodes', async () => {
-    const svc = createTestService()
-    const session = multiTurnSession(1, 1)
-    await expect(compactRegion(svc, session, 999, 1000, 'm'))
-      .rejects.toThrow(/start seq 999 not found in surface/)
-  })
-
-  it('throws when start is positioned after end on the surface', async () => {
-    const svc = createTestService()
-    const session = multiTurnSession(2, 1)
-    const nodes = session.surface.nodes
-    await expect(compactRegion(svc, session, nodes[1]!.seq, nodes[0]!.seq, 'm'))
-      .rejects.toThrow(/is after end seq .* on the surface/)
-  })
-
-  it('throws when compaction is already in progress', async () => {
-    const svc = createTestService()
-    const session = multiTurnSession(2, 1)
-    const nodes = session.surface.nodes
-    session.append('compact/start', { turn: 2 })
-    await expect(compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'm'))
-      .rejects.toThrow(/compaction already in progress/)
-  })
-
-  it('appends compact/end with error on summarize failure', async () => {
-    const svc = createTestService()
-    svc.summarizeError = new Error('model unavailable')
-    const session = multiTurnSession(2, 1)
-    const nodes = session.surface.nodes
-
-    await expect(compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'm'))
-      .rejects.toThrow('model unavailable')
-
-    const endEvent = session.events.findLast(e => e.type === 'compact/end')
-    expect(endEvent).toBeDefined()
-    // multiTurnSession(2,…) closes turns 1-2 and leaves turn 3 open; compaction
-    // stamps the open turn.
-    expect(endEvent!.data).toMatchObject({ turn: 3, error: 'model unavailable' })
-
-    // No replace-op user/message was appended (summarize failed).
-    const userMsgsAfter = session.events.filter(e => e.type === 'user/message')
-    const replaceMsgs = userMsgsAfter.filter((e) => {
-      const se = e as unknown as { surfaceOp?: unknown }
-      return se.surfaceOp !== undefined && typeof se.surfaceOp !== 'string'
-    })
-    expect(replaceMsgs.length).toBe(0)
-  })
-
-  it('extracts conversation text for summarization', async () => {
-    const svc = createTestService()
-    const session = multiTurnSession(1, 2)
-    const nodes = session.surface.nodes
-
-    await compactRegion(svc, session, nodes[0]!.seq, nodes[nodes.length - 1]!.seq, 'm')
-
-    expect(svc.summarizeCalls.length).toBe(1)
-    const { text, model } = svc.summarizeCalls[0]!
-    expect(model).toBe('m')
-    expect(text).toContain('User: turn 1 user message 1')
-    expect(text).toContain('Assistant: turn 1 assistant response 1')
-  })
-
-  it('frames the landed summary with a checkpoint preamble and tags, keeping raw provenance', async () => {
-    const svc = createTestService()
-    svc.mockSummary = [{ type: 'text', text: 'STRUCTURED SUMMARY' }]
-    const session = multiTurnSession(3, 1)
-    const nodes = session.surface.nodes
-
-    const result = await compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'm')
-
-    // Provenance (compact/summary) carries the RAW, unframed summary.
-    expect(result.summary).toEqual([{ type: 'text', text: 'STRUCTURED SUMMARY' }])
-    const summaryEvent = session.events.findLast(e => e.type === 'compact/summary')!
-    expect(summaryEvent.data).toMatchObject({ summary: [{ type: 'text', text: 'STRUCTURED SUMMARY' }] })
-
-    // The landed surface node is framed: preamble + tag-wrapped summary.
-    const landed = session.deriveMessages()[0]!.content
-    expect((landed[0] as { text: string }).text).toContain('checkpoint')
-    expect((landed[0] as { text: string }).text).toContain('<compacted-summary>')
-    expect(landed).toContainEqual({ type: 'text', text: 'STRUCTURED SUMMARY' })
-    expect((landed[landed.length - 1] as { text: string }).text).toBe('</compacted-summary>')
-  })
-
-  it('extracts tool-call and tool-result context', async () => {
-    const svc = createTestService()
-    const session = sessionWithTools()
-    const nodes = session.surface.nodes
-
-    const firstSeq = nodes[0]!.seq
-    const lastSeq = nodes[nodes.length - 1]!.seq
-    await compactRegion(svc, session, firstSeq, lastSeq, 'm')
-
-    expect(svc.summarizeCalls.length).toBe(1)
-    const { text } = svc.summarizeCalls[0]!
-    expect(text).toContain('read file x')
-    expect(text).toContain('bash')
-    expect(text).toContain('Tool result')
-  })
-})
-
-describe('BasicCompactService.compactIfNeeded', () => {
-  it('returns null when tokens are under threshold', async () => {
-    const svc = createTestService({ contextWindow: 128000, thresholdRatio: 0.8 })
-    const session = multiTurnSession(1, 1)
-    expect(await compactIfNeeded(svc, session, '', 'm', SIGNAL)).toBeNull()
-  })
-
-  it('compacts when tokens exceed threshold', async () => {
-    const svc = createTestService({ contextWindow: 100, thresholdRatio: 0.5, retainTokens: 10 })
-    const session = multiTurnSession(3, 1) // 6 surface nodes, 10 tokens each = 60
-
-    const result = await compactIfNeeded(svc, session, '', 'm', SIGNAL)
-    expect(result).not.toBeNull()
-    expect(result!.shadowedSeqs.length).toBeGreaterThan(0)
-  })
-
-  it('counts the session prefix toward pressure (every request carries it in front of the history)', async () => {
-    const svc = createTestService({ contextWindow: 200, thresholdRatio: 0.5, retainTokens: 10 })
-    const session = multiTurnSession(3, 1) // 6 derived messages ≈ 84 estimated tokens — under the 100 threshold alone
-    expect(await compactIfNeeded(svc, session, '', 'm', SIGNAL)).toBeNull()
-
-    // The loop composes the agent/session-prefix product before the pre-step
-    // seam and hands it to the gate; it rides every request, so pressure must
-    // include it — the same history now crosses the threshold.
-    const sessionPrefix: Message[] = [
-      { role: 'user', content: [{ type: 'text', text: `opener one.${LONG_FIXTURE_TEXT}` }] },
-      { role: 'user', content: [{ type: 'text', text: `opener two.${LONG_FIXTURE_TEXT}` }] },
-    ]
-    const result = await compactIfNeeded(svc, session, '', 'm', SIGNAL, sessionPrefix)
-    expect(result).not.toBeNull()
-    // The prefix itself is NOT history: compaction shadowed surface nodes only.
-    expect(sessionPrefix).toHaveLength(2)
-  })
-
-  it('returns the first compaction result when a zero-retry pass converges after the loop', async () => {
-    // With compactionRetries=0 there is no next-loop threshold check after the
-    // first mutation, so the success path is the post-loop `return result`.
-    const svc = createTestService({
-      contextWindow: 100,
-      thresholdRatio: 0.7,
-      retainTokens: 10,
-      compactionRetries: 0,
-    })
-    const session = multiTurnSession(3, 1) // 6 derived messages = 84 estimated tokens.
-
-    const result = await compactIfNeeded(svc, session, '', 'm', SIGNAL)
-
-    expect(result).not.toBeNull()
-    expect(session.events.filter(e => e.type === 'compact/summary')).toHaveLength(1)
-    expect(svc.estimateTokens(session.deriveMessages(), '')).toBeLessThan(70)
-  })
-
-  it('walks tail→head and retains nodes within token budget', async () => {
-    const svc = createTestService({ contextWindow: 350, thresholdRatio: 0.2, retainTokens: 15 })
-    const session = multiTurnSession(5, 1) // 10 surface nodes = ~100 tokens
-
-    const result = await compactIfNeeded(svc, session, '', 'm', SIGNAL)
-    expect(result).not.toBeNull()
-    const nodes = session.surface.nodes
-    expect(result!.shadowedSeqs.length).toBeGreaterThan(0)
-    expect(result!.shadowedSeqs).not.toContain(nodes[nodes.length - 1]!.seq)
-  })
-
-  it('returns null when the whole surface fits the retain budget (over threshold by role/system overhead)', async () => {
-    // Role overhead pushes the request above its 48-token threshold, but the
-    // raw four-node retention walk remains below retainTokens=45, so all fit.
-    const svc = createTestService({ contextWindow: 480, thresholdRatio: 0.1, retainTokens: 45 })
-    const session = multiTurnSession(2, 1)
-    expect(await compactIfNeeded(svc, session, '', 'm', SIGNAL)).toBeNull()
-  })
-
-  it('compacts a runaway turn: its early CLOSED steps summarize while recent steps stay verbatim', async () => {
-    // Completed early steps of the open turn remain eligible; protecting the
-    // whole turn would make a runaway turn impossible to compact.
-    const svc = createTestService({ contextWindow: 800, thresholdRatio: 0.1, retainTokens: 25 })
-    const s = new Session(SessionId('runaway'))
-    // ONE open turn with 5 closed steps; each step is [asst(tool-call), result].
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('user/message', { content: [{ type: 'text', text: 'do a big multi-step task' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    for (let step = 1; step <= 5; step++) {
-      s.append('step/start', { turn: 1, step })
-      s.append('assistant/message', {
-        turn: 1, step,
-        content: [{ type: 'text', text: `step ${step}` }, { type: 'tool-call', id: CallId(`c${step}`), name: 'bash', arguments: '{}' }],
-      }, { surfaceOp: 'append' })
-      s.append('tool/call', { turn: 1, step, callId: CallId(`c${step}`), name: 'bash', arguments: '{}' })
-      s.append('tool/result', { turn: 1, step, callId: CallId(`c${step}`), content: [{ type: 'text', text: `out ${step}` }], isError: false }, { surfaceOp: 'append' })
-      s.append('step/end', { turn: 1, step })
-    }
-    // The turn stays OPEN (no turn/end) — the model is mid-turn, about to run
-    // step 6. Surface: user + 5×[asst, result] = 11 nodes.
-    const nodesBefore = s.surface.nodes.length
-    expect(nodesBefore).toBe(11)
-
-    const result = await compactIfNeeded(svc, s, '', 'm', SIGNAL)
-    expect(result).not.toBeNull()
-    // Early steps of the SAME open turn were shadowed (impossible under layer 2).
-    expect(result!.shadowedSeqs.length).toBeGreaterThan(0)
-    // The most-recent step's tool result is retained verbatim (still on surface).
-    const lastResultSeq = s.events.findLast(e => e.type === 'tool/result')!.seq
-    expect(result!.shadowedSeqs).not.toContain(lastResultSeq)
-    expect(s.surface.nodes.some(n => n.seq === lastResultSeq)).toBe(true)
-    // No orphaned tool-result survives (whole-step boundaries respected).
-    expectNoOrphanToolResults(s.deriveMessages())
-  })
-
-  it('returns null for an empty surface', async () => {
-    const svc = createTestService({ contextWindow: 100, thresholdRatio: 0.5, retainTokens: 10 })
-    const session = new Session(SessionId('empty'))
-    expect(await compactIfNeeded(svc, session, '', 'm', SIGNAL)).toBeNull()
-  })
-
-  it('compacts again after a prior summary node heads the surface (the summary stays eligible)', async () => {
-    // Head-anchored recompaction must include the previous summary and retained context.
-    const svc = createTestService({ contextWindow: 800, thresholdRatio: 0.1, retainTokens: 25 })
-    const s = multiTurnSession(4, 1) // turns 1-4 closed, turn 5 open (no surface yet)
-
-    const first = await compactIfNeeded(svc, s, '', 'm', SIGNAL)
-    expect(first).not.toBeNull()
-    // The summary node now heads the surface with a fresh high seq.
-    const summaryHeadSeq = s.surface.nodes[0]!.seq
-    const turn5StartSeq = s.events.filter(e => e.type === 'turn/start').at(-1)!.seq
-    expect(summaryHeadSeq).toBeGreaterThan(turn5StartSeq)
-
-    // Append a verbatim node in the open turn (a step's output), still over
-    // threshold, then compact again — the older summary + closed turns compact,
-    // the fresh nodes are retained.
-    s.append('step/start', { turn: 5, step: 1 })
-    s.append('user/message', { content: [{ type: 'text', text: 'turn 5 work' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('assistant/message', { turn: 5, step: 1, content: [{ type: 'text', text: 'reply 5' }] }, { surfaceOp: 'append' })
-    s.append('step/end', { turn: 5, step: 1 })
-
-    const second = await compactIfNeeded(svc, s, '', 'm', SIGNAL)
-    expect(second).not.toBeNull()
-    expect(second!.shadowedSeqs.length).toBeGreaterThan(0)
-    // The fresh open-turn nodes were NOT compacted.
-    const turn5UserSeq = s.events.find(e => e.type === 'user/message' && e.data.content.some(b => b.type === 'text' && b.text === 'turn 5 work'))!.seq
-    expect(second!.shadowedSeqs).not.toContain(turn5UserSeq)
-  })
-
-  it('re-compacts smaller summaries until the post-compaction surface drops below threshold', async () => {
-    const svc = createTestService({
-      contextWindow: 100,
-      thresholdRatio: 0.5,
-      retainTokens: 10,
-      compactionRetries: 2,
-    })
-    svc.estimateFramedSummariesCheaply = false
-    svc.mockSummaryQueue = [
-      Array.from({ length: 4 }, (_, index) => ({ type: 'text', text: `first ${index}` })),
-      [{ type: 'text', text: 'second' }],
-    ]
-    const session = multiTurnSession(4, 1)
-
-    const result = await compactIfNeeded(svc, session, '', 'm', SIGNAL)
-
-    expect(result).not.toBeNull()
-    expect(svc.summarizeCalls).toHaveLength(2)
-    expect(session.events.filter(e => e.type === 'compact/summary')).toHaveLength(2)
-    expect(svc.estimateTokens(session.deriveMessages(), '')).toBeLessThan(50)
-  })
-
-  it('throws after the configured re-compaction attempts still leave the surface above threshold', async () => {
-    const svc = createTestService({
-      contextWindow: 100,
-      thresholdRatio: 0.5,
-      retainTokens: 10,
-      compactionRetries: 1,
-    })
-    svc.estimateFramedSummariesCheaply = false
-    svc.mockSummaryQueue = [
-      Array.from({ length: 4 }, (_, index) => ({ type: 'text', text: `first ${index}` })),
-      Array.from({ length: 3 }, (_, index) => ({ type: 'text', text: `second ${index}` })),
-    ]
-    const session = multiTurnSession(4, 1)
-
-    await expect(compactIfNeeded(svc, session, '', 'm', SIGNAL))
-      .rejects.toThrow(/still above threshold after 2 compaction attempts/)
-    expect(svc.summarizeCalls).toHaveLength(2)
-  })
-})
-
-describe('BasicCompactService replay equivalence', () => {
-  it('produces identical deriveMessages() after seeding from compacted log', async () => {
-    const svc = createTestService()
-    const session = multiTurnSession(3, 1)
-    const nodes = session.surface.nodes
-
-    await compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'm')
-    const derived = session.deriveMessages()
-
-    const replayed = new Session(SessionId('replay'), [...session.events])
-    expect(replayed.deriveMessages()).toEqual(derived)
-  })
-})
-
-describe('BasicCompactService blocking (compaction in progress)', () => {
-  it('detects in-progress compaction from unmatched compact/start', async () => {
-    const svc = createTestService()
-    const session = multiTurnSession(1, 1)
-    session.append('compact/start', { turn: 1 })
-    const nodes = session.surface.nodes
-    // Whole step (user → assistant) is a step-aligned region, so the call reaches
-    // the in-progress check rather than being rejected for splitting a step.
-    await expect(compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'm'))
-      .rejects.toThrow(/compaction already in progress/)
-  })
-
-  it('allows compaction after compact/end is appended', async () => {
-    const svc = createTestService()
-    const session = multiTurnSession(2, 1)
-    const nodes = session.surface.nodes
-    session.append('compact/start', { turn: 1 })
-    session.append('compact/end', { turn: 1 })
-    const result = await compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'm')
-    expect(result).toBeDefined()
-  })
-
-  it('is not wedged by an orphaned compact/start from a prior (now-closed) turn', async () => {
-    // An orphaned start in a closed repaired turn is stale; only the current
-    // turn participates in the in-progress lock.
-    const svc = createTestService()
-    const s = new Session(SessionId('stale-lock'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('user/message', { content: [{ type: 'text', text: 'turn 1' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'reply 1' }] }, { surfaceOp: 'append' })
-    s.append('compact/start', { turn: 1 }) // ← orphaned: no matching compact/end
-    s.append('step/end', { turn: 1, step: 1 })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } }) // repair closed the turn
-    // A new open turn.
-    s.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
-    const nodes = s.surface.nodes
-
-    // The stale start is before the turn/end, so it is NOT seen as in-progress.
-    const result = await compactRegion(svc, s, nodes[0]!.seq, nodes[1]!.seq, 'm')
-    expect(result).toBeDefined()
-  })
-})
-
-describe('BasicCompactService token estimation (char/4 heuristic)', () => {
-  it('estimates text blocks with char/4 + overhead', () => {
-    const svc = new BasicCompactService(new Context(), cfg({ auto: false }))
-    // 'this is a somewhat longer text block' = 36 → ceil(36/4)+4 = 13; 'short' = 5 → 2+4 = 6
-    const blocks: ContentBlock[] = [
-      { type: 'text', text: 'this is a somewhat longer text block' },
-      { type: 'text', text: 'short' },
-    ]
-    expect(svc.estimateContentTokens(blocks)).toBe(19)
-  })
-
-  it('estimates reasoning blocks same as text', () => {
-    const svc = new BasicCompactService(new Context(), cfg({ auto: false }))
-    // 'thinking about this...' = 22 → ceil(22/4)+4 = 10
-    expect(svc.estimateContentTokens([{ type: 'reasoning', text: 'thinking about this...' }])).toBe(10)
-  })
-
-  it('estimates tool-call blocks from name + arguments', () => {
-    const svc = new BasicCompactService(new Context(), cfg({ auto: false }))
-    // 'bash' = 4 → 1; '{"command":"ls"}' = 16 → 4; + 4 overhead = 9
-    expect(svc.estimateContentTokens([
-      { type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{"command":"ls"}' },
-    ])).toBe(9)
-  })
-
-  it('estimates tool-result blocks recursively', () => {
-    const svc = new BasicCompactService(new Context(), cfg({ auto: false }))
-    // inner text 5 → 2+4 = 6; outer 6 + 4 overhead = 10
-    expect(svc.estimateContentTokens([
-      { type: 'tool-result', toolCallId: CallId('c1'), content: [{ type: 'text', text: 'hello' }], isError: false },
-    ])).toBe(10)
-  })
-
-  it('returns 0 for empty content blocks', () => {
-    const svc = new BasicCompactService(new Context(), cfg({ auto: false }))
-    expect(svc.estimateContentTokens([])).toBe(0)
-  })
-
-  it('honors a configured charsPerToken (fractional densities included)', () => {
-    // 'this is a somewhat longer text block' = 36 chars.
-    const blocks: ContentBlock[] = [{ type: 'text', text: 'this is a somewhat longer text block' }]
-    // charsPerToken 2: ceil(36/2)+4 = 22 — a CJK-density config doubles the estimate.
-    const dense = new BasicCompactService(new Context(), cfg({ auto: false, charsPerToken: 2 }))
-    expect(dense.estimateContentTokens(blocks)).toBe(22)
-    // Fractional density is legal: ceil(36/1.5)+4 = 28.
-    const fractional = new BasicCompactService(new Context(), cfg({ auto: false, charsPerToken: 1.5 }))
-    expect(fractional.estimateContentTokens(blocks)).toBe(28)
-    // The system-prompt term scales with the same knob: 36-char prompt at density 2 → ceil(36/2) = 18.
-    expect(dense.estimateTokens([], 'this is a somewhat longer text block')).toBe(18)
-  })
-})
-
-describe('BasicCompactService HMR safety', () => {
-  it('registers as ctx.compact', () => {
-    const ctx = new Context()
-    void new BasicCompactService(ctx, cfg({ auto: false }))
-    expect(ctx.compact).toBeDefined()
-    expect(ctx.compact).toBeInstanceOf(BasicCompactService)
-  })
-
-  it('disposing the plugin fiber unregisters ctx.compact', async () => {
-    // Mount through the real plugin fiber (the Loader path), then dispose it and confirm the
-    // service registration is torn down.
-    const ctx = new Context()
-    await ctx.plugin(LlmService)
-    const fiber = await ctx.plugin(BasicCompactService, cfg({ auto: false }))
-    expect(ctx.get('compact')).toBeInstanceOf(BasicCompactService)
-
-    await fiber.dispose()
-    expect(ctx.get('compact')).toBeUndefined()
-  })
-})
-
-describe('BasicCompactService config validation', () => {
-  it('rejects invalid numeric config values', () => {
-    expect(() => new BasicCompactService(new Context(), cfg({ auto: false, contextWindow: 0 })))
-      .toThrow(/contextWindow .* positive integer/)
-    expect(() => new BasicCompactService(new Context(), cfg({ auto: false, thresholdRatio: 0 }))).toThrow(/thresholdRatio .* \(0, 1\]/)
-    expect(() => new BasicCompactService(new Context(), cfg({ auto: false, thresholdRatio: 1.1 }))).toThrow(/thresholdRatio .* \(0, 1\]/)
-    expect(() => new BasicCompactService(new Context(), cfg({ auto: false, retainTokens: -1 })))
-      .toThrow(/retainTokens .* non-negative integer/)
-    expect(() => new BasicCompactService(new Context(), cfg({ auto: false, maxTokens: 0 }))).toThrow(/maxTokens .* positive integer/)
-    expect(() => new BasicCompactService(new Context(), cfg({ auto: false, compactionRetries: -1 })))
-      .toThrow(/compactionRetries .* non-negative integer/)
-    expect(() => new BasicCompactService(
-      new Context(), cfg({ auto: false, summarizationModel: 1 } as unknown as Partial<BasicCompactConfig>),
-    )).toThrow(/summarizationModel must be a string/)
-    expect(() => new BasicCompactService(new Context(), cfg({ auto: 'no' } as unknown as Partial<BasicCompactConfig>)))
-      .toThrow(/auto must be a boolean/)
-    expect(() => new BasicCompactService(new Context(), cfg({ auto: false, charsPerToken: 0 })))
-      .toThrow(/charsPerToken .* positive finite number/)
-    expect(() => new BasicCompactService(new Context(), cfg({ auto: false, charsPerToken: Number.NaN })))
-      .toThrow(/charsPerToken .* positive finite number/)
-  })
-
-  it('accepts a large retain budget because convergence is enforced dynamically', () => {
-    expect(() => new BasicCompactService(new Context(), cfg({
-      auto: false,
-      contextWindow: 1000,
-      thresholdRatio: 0.5,
-      retainTokens: 900,
-    }))).not.toThrow()
-  })
-
-  it('the default config is valid', () => {
-    expect(() => new BasicCompactService(new Context(), cfg({ auto: false }))).not.toThrow()
-  })
-})
-
-/** An adapter that emits a fixed summary text, for exercising the real summarize() path. */
-class ScriptedAdapter extends LlmAdapter {
-  lastOptions: GenerateOptions | null = null
-  constructor(private summaryText: string) {
-    super()
-  }
-
-  async * stream(options: GenerateOptions): AsyncIterable<StreamChunk> {
-    this.lastOptions = options
-    yield { type: 'block-start', index: 0, blockType: 'text' }
-    yield { type: 'text-delta', index: 0, text: this.summaryText }
-    yield { type: 'finish', reason: { kind: 'stop' } }
-  }
-}
-
-/** An adapter that emits arbitrary content blocks, preserving reasoning/text shape. */
-class BlocksAdapter extends LlmAdapter {
-  lastOptions: GenerateOptions | null = null
-  constructor(private blocks: readonly ContentBlock[]) {
-    super()
-  }
-
-  async * stream(options: GenerateOptions): AsyncIterable<StreamChunk> {
-    this.lastOptions = options
-    for (const [index, block] of this.blocks.entries()) {
-      yield { type: 'block-start', index, blockType: block.type }
-      switch (block.type) {
-        case 'text':
-          yield { type: 'text-delta', index, text: block.text }
-          break
-        case 'reasoning':
-          yield { type: 'reasoning-delta', index, text: block.text }
-          break
-        default:
-          yield { type: 'block-end', index, block }
-      }
-    }
-    yield { type: 'finish', reason: { kind: 'stop' } }
-  }
-}
-
-/** Wire a real LlmService + arbitrary-block adapter into a context. */
-async function ctxWithBlocks(blocks: readonly ContentBlock[], model = 'test-model'): Promise<{ ctx: Context; adapter: BlocksAdapter }> {
+function createContext(contextWindow = 1_000): Context {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  const adapter = new BlocksAdapter(blocks)
-  ctx.llm.registerAdapter([model], adapter)
-  return { ctx, adapter }
-}
-
-/** Wire a real LlmService + scripted adapter into a context. */
-async function ctxWithModel(summaryText: string, model = 'test-model'): Promise<{ ctx: Context; adapter: ScriptedAdapter }> {
-  const ctx = new Context()
-  await ctx.plugin(LlmService)
-  const adapter = new ScriptedAdapter(summaryText)
-  ctx.llm.registerAdapter([model], adapter)
-  return { ctx, adapter }
-}
-
-/** An adapter whose stream ends with a finish chunk of the given reason (no content). */
-class FinishOnlyAdapter extends LlmAdapter {
-  constructor(private reason: StreamChunk & { type: 'finish' }) {
-    super()
-  }
-
-  async * stream(): AsyncIterable<StreamChunk> {
-    yield this.reason
-  }
-}
-
-/** Wire a real LlmService + finish-only adapter into a context. */
-async function ctxWithFinish(reason: (StreamChunk & { type: 'finish' })['reason'], model = 'test-model'): Promise<Context> {
-  const ctx = new Context()
-  await ctx.plugin(LlmService)
-  ctx.llm.registerAdapter([model], new FinishOnlyAdapter({ type: 'finish', reason }))
+  void new TokenMeterService(ctx, { contextWindow })
   return ctx
 }
 
-/** A minimal Agent stub carrying just session + options (enough for the listeners). */
-function stubAgent(session: Session, model?: string): Agent {
-  return { session, options: { model } } as unknown as Agent
+function agent(session: Session, model?: string): Agent {
+  return { session, options: model === undefined ? {} : { provider: model, model } } as Agent
 }
 
-function compactIfNeeded(
-  svc: BasicCompactService,
-  session: Session,
-  fullSystemPrompt: string,
-  model: string,
-  signal: AbortSignal,
-  sessionPrefix: readonly Message[] = [],
-) {
-  return svc.compactIfNeeded(stubAgent(session, model), fullSystemPrompt, sessionPrefix, signal)
-}
-
-function compactRegion(
-  svc: BasicCompactService,
-  session: Session,
-  start: number,
-  end: number,
-  model: string,
-  signal?: AbortSignal,
-) {
-  return svc.compactRegion(session, start, end, stubAgent(session, model), signal)
-}
-
-function summarize(svc: BasicCompactService, text: string, model: string) {
-  return svc.summarize(text, stubAgent(new Session(SessionId('summary')), model))
-}
-
-describe('BasicCompactService.summarize (real ctx.llm.stream)', () => {
-  it('summarizes via the registered adapter and returns its content', async () => {
-    const { ctx, adapter } = await ctxWithModel('SUMMARY TEXT')
-    const svc = new BasicCompactService(ctx, cfg({ auto: false, maxTokens: 512 }))
-
-    const { summary, model, maxTokens } = await summarize(svc, 'User: hi\n\nAssistant: hello', 'test-model')
-    expect(summary).toEqual([{ type: 'text', text: 'SUMMARY TEXT' }])
-    // The returned envelope reports what the call actually used — the caller
-    // logs it on compact/summary (the reconstructability RFC).
-    expect(model).toBe('test-model')
-    expect(maxTokens).toBe(512)
-    // The fixed system prompt and maxTokens flow through.
-    expect(adapter.lastOptions!.system).toContain('compaction engine')
-    expect(adapter.lastOptions!.system).toContain('## Next Step')
-    expect(adapter.lastOptions!.maxTokens).toBe(512)
-    expect(adapter.lastOptions!.sessionId).toBe(SessionId('summary'))
-    expect(adapter.lastOptions!.messages[0]!.content[0]).toMatchObject({ type: 'text' })
-  })
-
-  it('uses maxTokens as the summarization provider cap', async () => {
-    const { ctx, adapter } = await ctxWithModel('SUMMARY TEXT')
-    const svc = new BasicCompactService(ctx, cfg({
-      auto: false,
-      maxTokens: 50,
-    }))
-
-    await summarize(svc, 'User: hi', 'test-model')
-
-    expect(adapter.lastOptions!.maxTokens).toBe(50)
-  })
-
-  it('keeps only text blocks in the stored summary (drops reasoning and tool-call)', async () => {
-    const { ctx } = await ctxWithBlocks([
-      { type: 'reasoning', text: 'private chain of thought' },
-      { type: 'text', text: 'PUBLIC SUMMARY' },
-      // A model reply can carry a tool-call; it must not survive into the
-      // synthesized user/message summary as an orphaned call.
-      { type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{}' },
-    ])
-    const svc = new BasicCompactService(ctx, cfg({ auto: false }))
-
-    const { summary } = await summarize(svc, 'User: hi', 'test-model')
-
-    expect(summary).toEqual([{ type: 'text', text: 'PUBLIC SUMMARY' }])
-  })
-
-  it('throws when no text block remains after filtering', async () => {
-    const { ctx } = await ctxWithBlocks([{ type: 'reasoning', text: 'private only' }])
-    const svc = new BasicCompactService(ctx, cfg({ auto: false }))
-
-    await expect(summarize(svc, 'User: hi', 'test-model')).rejects.toThrow(/no text summary content/)
-  })
-
-  it('throws when no model is provided', async () => {
-    const { ctx } = await ctxWithModel('x')
-    const svc = new BasicCompactService(ctx, cfg({ auto: false }))
-    await expect(summarize(svc, 'text', '')).rejects.toThrow(/no model available/)
-  })
-
-  it('rethrows when the stream ends with a finish-error chunk', async () => {
-    const ctx = await ctxWithFinish({ kind: 'error', message: 'provider 401', code: 'UNAUTHORIZED' })
-    const svc = new BasicCompactService(ctx, cfg({ auto: false }))
-    await expect(summarize(svc, 'text', 'test-model')).rejects.toMatchObject({ message: 'provider 401', code: 'UNAUTHORIZED' })
-  })
-
-  it('rethrows a finish-error chunk without a code (code stays undefined)', async () => {
-    const ctx = await ctxWithFinish({ kind: 'error', message: 'opaque failure' })
-    const svc = new BasicCompactService(ctx, cfg({ auto: false }))
-    const error = await summarize(svc, 'text', 'test-model').then(() => null, (e: unknown) => e as Error & { code?: string })
-    expect(error?.message).toBe('opaque failure')
-    expect(error?.code).toBeUndefined()
-  })
-
-  it('rethrows when the stream ends with a finish-aborted chunk', async () => {
-    const ctx = await ctxWithFinish({ kind: 'aborted' })
-    const svc = new BasicCompactService(ctx, cfg({ auto: false }))
-    await expect(summarize(svc, 'text', 'test-model')).rejects.toMatchObject({ message: 'summarization stream aborted', code: 'ABORTED' })
-  })
-
-  it('fails closed on a max-tokens finish (an incomplete checkpoint must not commit)', async () => {
-    const ctx = await ctxWithFinish({ kind: 'max-tokens' })
-    const svc = new BasicCompactService(ctx, cfg({ auto: false }))
-    await expect(summarize(svc, 'text', 'test-model')).rejects.toMatchObject({ code: 'MAX_TOKENS' })
-  })
-
-  it('compactRegion leaves the surface intact when summarization hits max-tokens', async () => {
-    const ctx = await ctxWithFinish({ kind: 'max-tokens' })
-    const svc = new BasicCompactService(ctx, cfg({ auto: false }))
-    const session = multiTurnSession(2, 1)
-    const before = [...session.surface.nodes]
-    const nodes = session.surface.nodes
-
-    await expect(compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'test-model'))
-      .rejects.toMatchObject({ code: 'MAX_TOKENS' })
-
-    // No replacement landed — the surface is byte-identical, and the lock was
-    // released with the error (compact/end carries it).
-    expect(session.surface.nodes).toEqual(before)
-    const endEvent = session.events.findLast(e => e.type === 'compact/end')!
-    const endData = endEvent.data as { error?: string }
-    expect(endData.error).toContain('truncated')
-  })
-
-  it('compactRegion uses the real summarizer end-to-end', async () => {
-    const { ctx } = await ctxWithModel('CONDENSED')
-    const svc = new BasicCompactService(ctx, cfg({ auto: false }))
-    const session = multiTurnSession(2, 1)
-    const nodes = session.surface.nodes
-
-    const result = await compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'test-model')
-    expect(result.summary).toEqual([{ type: 'text', text: 'CONDENSED' }])
-    // The raw summary is wrapped in the checkpoint framing on the surface.
-    expect(session.deriveMessages()[0]!.content).toContainEqual({ type: 'text', text: 'CONDENSED' })
-  })
-
-  it('rejects a summary that is not smaller than the shadowed content', async () => {
-    const svc = createTestService({ auto: false })
-    const session = multiTurnSession(2, 1)
-    const nodes = session.surface.nodes
-    svc.mockSummary = Array.from({ length: 20 }, (_, index) => ({ type: 'text', text: `large ${index}` }))
-
-    await expect(compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'm'))
-      .rejects.toThrow(/summary is not smaller than the shadowed content/)
-    expect(session.events.some(e => e.type === 'compact/summary')).toBe(false)
-  })
-
-  it('rejects when the framed checkpoint is not smaller than the shadowed content', async () => {
-    const svc = createTestService({ auto: false })
-    svc.estimateFramedSummariesCheaply = false
-    const session = new Session(SessionId('framed-nonshrinking'))
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    session.append('step/start', { turn: 1, step: 1 })
-    session.append('user/message', { content: [{ type: 'text', text: 'tiny user' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    session.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'tiny assistant' }] }, { surfaceOp: 'append' })
-    session.append('step/end', { turn: 1, step: 1 })
-    session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    session.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
-    const before = [...session.surface.nodes]
-    const nodes = session.surface.nodes
-
-    await expect(compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'm'))
-      .rejects.toThrow(/summary is not smaller than the shadowed content/)
-    expect(session.events.some(e => e.type === 'compact/summary')).toBe(false)
-    expect(session.surface.nodes).toEqual(before)
-  })
-})
-
-describe('BasicCompactService auto-compaction (agent/pre-step listener)', () => {
-  /** Fire the agent/pre-step serial checkpoint as the loop does. */
-  function firePreStep(ctx: Context, agent: Agent, step: number, fullSystemPrompt: string): Promise<unknown> {
-    return ctx.serial('agent/pre-step', agent, 1, step, fullSystemPrompt, [], SIGNAL)
-  }
-
-  it('compacts (mutating the surface) when over threshold', async () => {
-    const { ctx } = await ctxWithModel('SUMMARY')
-    void new BasicCompactService(ctx, cfg({ contextWindow: 200, thresholdRatio: 0.5, retainTokens: 20 }))
-    const session = multiTurnSession(5, 1) // 10 surface nodes
-    const agent = stubAgent(session, 'test-model')
-    const before = session.surface.nodes.length
-
-    await firePreStep(ctx, agent, 1, '')
-
-    // The surface shrank in place, and a summary checkpoint landed.
-    expect(session.surface.nodes.length).toBeLessThan(before)
-    expect(session.events.some(e => e.type === 'compact/summary')).toBe(true)
-    // The re-derived head message is the framed summary checkpoint.
-    expect(session.deriveMessages()[0]!.content).toContainEqual({ type: 'text', text: 'SUMMARY' })
-  })
-
-  it('logs compaction details when auto-compaction returns a converged result', async () => {
-    const ctx = new Context()
-    const infos: string[] = []
-    ctx.logger.info = ((msg: string) => void infos.push(msg)) as typeof ctx.logger.info
-    void new TestCompactService(ctx, cfg({
-      contextWindow: 100,
-      thresholdRatio: 0.7,
-      retainTokens: 10,
-      compactionRetries: 0,
-    }))
-    const session = multiTurnSession(3, 1)
-    const agent = stubAgent(session, 'test-model')
-
-    await firePreStep(ctx, agent, 1, '')
-
-    expect(session.events.filter(e => e.type === 'compact/summary')).toHaveLength(1)
-    expect(infos.some(msg => msg.includes('compaction: shadowed'))).toBe(true)
-    expect(infos.some(msg => msg.includes('estimated tokens after compaction'))).toBe(true)
-  })
-
-  it('compacts mid-turn on steps after the first (the surface grows within a turn)', async () => {
-    const { ctx } = await ctxWithModel('SUMMARY')
-    void new BasicCompactService(ctx, cfg({ contextWindow: 100, thresholdRatio: 0.5, retainTokens: 10 }))
-    const session = multiTurnSession(3, 1) // over the 0.5 threshold
-    const agent = stubAgent(session, 'test-model')
-
-    // A step-2 checkpoint (a tool-heavy turn's later step) must still compact —
-    // the surface accumulated assistant/message + tool/result nodes since step 1.
-    await firePreStep(ctx, agent, 2, '')
-    expect(session.events.some(e => e.type === 'compact/start')).toBe(true)
-  })
-
-  it('does nothing when under threshold', async () => {
-    const { ctx } = await ctxWithModel('SUMMARY')
-    void new BasicCompactService(ctx, cfg({ contextWindow: 128000, thresholdRatio: 0.8 }))
-    const session = multiTurnSession(1, 1)
-    const agent = stubAgent(session, 'test-model')
-
-    await firePreStep(ctx, agent, 1, '')
-    expect(session.events.some(e => e.type === 'compact/start')).toBe(false)
-  })
-
-  it('leaves the surface intact when compaction fails (summarize rejects)', async () => {
-    // No adapter registered for this model → summarize() rejects → caught, the
-    // surface is untouched (the loop derives the full history).
-    const ctx = new Context()
-    await ctx.plugin(LlmService)
-    void new BasicCompactService(ctx, cfg({ contextWindow: 300, thresholdRatio: 0.1, retainTokens: 10 }))
-    const session = multiTurnSession(3, 1)
-    const agent = stubAgent(session, 'missing-model')
-    const before = session.surface.nodes.length
-
-    await firePreStep(ctx, agent, 1, '')
-    // No summary landed; the surface is unchanged.
-    expect(session.events.some(e => e.type === 'compact/summary')).toBe(false)
-    expect(session.surface.nodes.length).toBe(before)
-  })
-
-  it('does not register the listener when auto is false', async () => {
-    const { ctx } = await ctxWithModel('SUMMARY')
-    void new BasicCompactService(ctx, cfg({ auto: false, contextWindow: 100, thresholdRatio: 0.1, retainTokens: 5 }))
-    const session = multiTurnSession(3, 1)
-    const agent = stubAgent(session, 'test-model')
-
-    await firePreStep(ctx, agent, 1, '')
-    expect(session.events.some(e => e.type === 'compact/start')).toBe(false)
-  })
-
-  it('summarization is interceptable at llm/stream (model routing for direct calls)', async () => {
-    const { ctx, adapter } = await ctxWithModel('ROUTED SUMMARY', 'routed-model')
-    // One-shot summaries bypass agent/request but remain mutable at llm/stream;
-    // adapter selection happens after the waterfall rewrite.
-    ctx.on('llm/stream', (options, next) => {
-      options.model = 'routed-model'
-      return next()
-    })
-    void new BasicCompactService(ctx, cfg({ contextWindow: 200, thresholdRatio: 0.5, retainTokens: 20 }))
-    const session = multiTurnSession(5, 1)
-    const agent = stubAgent(session, 'agent-model')
-
-    await ctx.serial('agent/pre-step', agent, 1, 1, '', [], SIGNAL)
-
-    expect(adapter.lastOptions?.model).toBe('routed-model')
-    expect(session.events.some(e => e.type === 'compact/summary')).toBe(true)
-    expect(session.deriveMessages()[0]!.content).toContainEqual({ type: 'text', text: 'ROUTED SUMMARY' })
-  })
-
-  it('removes the auto pre-step listener when the plugin fiber is disposed', async () => {
-    const { ctx } = await ctxWithModel('SUMMARY')
-    const fiber = await ctx.plugin(BasicCompactService, cfg({
-      contextWindow: 200,
-      thresholdRatio: 0.5,
-      retainTokens: 20,
-    }))
-    const session = multiTurnSession(5, 1)
-    const agent = stubAgent(session, 'test-model')
-
-    await fiber.dispose()
-    await firePreStep(ctx, agent, 1, '')
-
-    expect(session.events.some(e => e.type === 'compact/start')).toBe(false)
-    expect(ctx.get('compact')).toBeUndefined()
-  })
-})
-
-describe('BasicCompactService transcript rendering (delegated to dsh-compact)', () => {
-  it('renders reasoning, context, and steering messages', async () => {
-    const svc = createTestService()
-    const s = new Session(SessionId('rich'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('context/message', {
-      content: [{ type: 'text', text: 'project context here' }],
-      source: { kind: 'user' },
-    }, { surfaceOp: 'append' })
-    s.append('assistant/message', {
-      turn: 1, step: 1,
-      content: [{ type: 'reasoning', text: 'thinking hard' }, { type: 'text', text: 'answer' }],
-    }, { surfaceOp: 'append' })
-    s.append('steering/message', {
-      turn: 1,
-      content: [{ type: 'text', text: 'steer this way' }],
-      source: { kind: 'user' },
-    }, { surfaceOp: 'append' })
-    s.append('step/end', { turn: 1, step: 1 })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    s.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
-
-    const nodes = s.surface.nodes
-    await compactRegion(svc, s, nodes[0]!.seq, nodes[nodes.length - 1]!.seq, 'm')
-
-    const { text } = svc.summarizeCalls[0]!
-    expect(text).toContain('[Context: project context here]')
-    expect(text).toContain('[reasoning: thinking hard]')
-    expect(text).toContain('[Steering: steer this way]')
-  })
-
-  it('labels tool errors distinctly from tool results', async () => {
-    const svc = createTestService()
-    const s = new Session(SessionId('toolerr'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('user/message', { content: [{ type: 'text', text: 'run it' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('assistant/message', {
-      turn: 1, step: 1,
-      content: [{ type: 'tool-call', id: CallId('c9'), name: 'bash', arguments: '{}' }],
-    }, { surfaceOp: 'append' })
-    s.append('tool/call', { turn: 1, step: 1, callId: CallId('c9'), name: 'bash', arguments: '{}' })
-    s.append('tool/result', {
-      turn: 1, step: 1, callId: CallId('c9'),
-      content: [{ type: 'text', text: 'boom failure' }],
-      isError: true,
-    }, { surfaceOp: 'append' })
-    s.append('step/end', { turn: 1, step: 1 })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    s.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
-
-    const nodes = s.surface.nodes
-    await compactRegion(svc, s, nodes[0]!.seq, nodes[nodes.length - 1]!.seq, 'm')
-    expect(svc.summarizeCalls[0]!.text).toContain('Tool error (call c9): boom failure')
-  })
-})
-
-describe('BasicCompactService edge cases', () => {
-  it('renders bare and nested tool-result placeholders and unknown blocks', async () => {
-    const svc = createTestService()
-    const s = new Session(SessionId('toolresult'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    // assistant/message carrying a nested tool-result block, an unknown block,
-    // and the tool-call that the following tool/result answers (so the surface
-    // is tool-pairing balanced).
-    s.append('assistant/message', {
-      turn: 1, step: 1,
-      content: [
-        { type: 'tool-result', toolCallId: CallId('n1'), content: [{ type: 'chart', data: 'x' } as unknown as ContentBlock] },
-        { type: 'custom-widget', payload: 'x' } as unknown as ContentBlock,
-        { type: 'tool-call', id: CallId('b1'), name: 'bash', arguments: '{}' },
-      ],
-    }, { surfaceOp: 'append' })
-    // tool/result whose content is itself only non-text → bare '[tool-result]'.
-    s.append('tool/call', { turn: 1, step: 1, callId: CallId('b1'), name: 'bash', arguments: '{}' })
-    s.append('tool/result', {
-      turn: 1, step: 1, callId: CallId('b1'),
-      content: [{ type: 'tool-result', toolCallId: CallId('inner'), content: [] }],
-      isError: false,
-    }, { surfaceOp: 'append' })
-    s.append('step/end', { turn: 1, step: 1 })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    s.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
-
-    const nodes = s.surface.nodes
-    await compactRegion(svc, s, nodes[0]!.seq, nodes[nodes.length - 1]!.seq, 'm')
-    const { text } = svc.summarizeCalls[0]!
-    expect(text).toContain('[tool-result: [chart]]') // nested tool-result with content
-    expect(text).toContain('[custom-widget]') // unknown block placeholder
-    expect(text).toContain('Tool result (call b1): [tool-result]') // empty nested → bare placeholder
-  })
-
-  it('estimates unknown block types via JSON length (default branch)', () => {
-    const svc = new BasicCompactService(new Context(), cfg({ auto: false }))
-    // A block whose type is none of the known kinds — exercises the default arm.
-    const unknown = { type: 'custom-widget', payload: 'some data' } as unknown as ContentBlock
-    expect(svc.estimateContentTokens([unknown])).toBeGreaterThan(0)
-  })
-
-  it('auto-compaction reports bounded retry exhaustion after committing a smaller summary', async () => {
-    const { ctx } = await ctxWithModel('SUMMARY')
-    const warnings: string[] = []
-    ctx.logger.warn = ((msg: string) => void warnings.push(msg)) as typeof ctx.logger.warn
-    void new BasicCompactService(ctx, cfg({
-      contextWindow: 300,
-      thresholdRatio: 0.1,
-      retainTokens: 5,
-      compactionRetries: 0,
-    }))
-    const session = multiTurnSession(4, 1)
-    const agent = stubAgent(session, 'test-model')
-
-    await ctx.serial('agent/pre-step', agent, 1, 1, '', [], SIGNAL)
-    expect(session.events.some(e => e.type === 'compact/summary')).toBe(true)
-    // The surface was mutated; the head message is the framed summary checkpoint.
-    expect(session.deriveMessages()[0]!.content).toContainEqual({ type: 'text', text: 'SUMMARY' })
-    expect(warnings.some(w => w.includes('still above threshold after 1 compaction attempts'))).toBe(true)
-  })
-
-  it('rejects compaction when no turn is open (compaction events must be turn-enclosed)', async () => {
-    const svc = createTestService()
-    // A session whose only turn has CLOSED — scanning back from the tail hits
-    // turn/end before any turn/start, so there is no open turn to enclose
-    // compaction's compact/* + replacement events, which the log contract forbids.
-    const s = new Session(SessionId('noturn'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('user/message', { content: [{ type: 'text', text: 'orphan' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'reply' }] }, { surfaceOp: 'append' })
-    s.append('step/end', { turn: 1, step: 1 })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    const nodes = s.surface.nodes
-
-    await expect(compactRegion(svc, s, nodes[0]!.seq, nodes[1]!.seq, 'm'))
-      .rejects.toThrow(/no open turn/)
-    // The lock was never acquired — no compact/start landed.
-    expect(s.events.some(e => e.type === 'compact/start')).toBe(false)
-  })
-
-  it('rejects compaction on a session with no turn boundaries at all', async () => {
-    const svc = createTestService()
-    // No turn events whatsoever — the open-turn scan falls through to the end
-    // of the log and finds none, so compaction is rejected (its events have no
-    // turn to enclose them).
-    const s = new Session(SessionId('turnless'))
-    s.append('user/message', { content: [{ type: 'text', text: 'orphan' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    const nodes = s.surface.nodes
-
-    await expect(compactRegion(svc, s, nodes[0]!.seq, nodes[0]!.seq, 'm'))
-      .rejects.toThrow(/no open turn/)
-    expect(s.events.some(e => e.type === 'compact/start')).toBe(false)
-  })
-
-  it('compactIfNeeded returns null for empty surface even when over threshold', async () => {
-    const svc = createTestService({ contextWindow: 1000, thresholdRatio: 0.1, retainTokens: 5 })
-    const session = new Session(SessionId('empty-but-pressured'))
-    // No surface nodes, but a large system prompt pushes the estimate over threshold.
-    const bigPrompt = 'x'.repeat(800) // ceil(800/4) = 200 tokens >> threshold 100
-    expect(await compactIfNeeded(svc, session, bigPrompt, 'm', SIGNAL)).toBeNull()
-  })
-
-  it('compactRegion throws when end is not a surface node (start valid)', async () => {
-    const svc = createTestService()
-    const session = multiTurnSession(1, 1)
-    const nodes = session.surface.nodes
-    await expect(compactRegion(svc, session, nodes[0]!.seq, 9999, 'm'))
-      .rejects.toThrow(/end seq 9999 not found in surface/)
-  })
-
-  it('compactRegion stringifies a non-Error thrown by summarize', async () => {
-    const svc = createTestService()
-    // Throw a non-Error value to exercise the String(error) branch in the catch.
-    svc.summarizeError = 'plain string failure' as unknown as Error
-    const session = multiTurnSession(1, 1)
-    const nodes = session.surface.nodes
-
-    // Whole step (user → assistant): a step-aligned region that reaches summarize.
-    await expect(compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'm')).rejects.toBe('plain string failure')
-    const endEvent = session.events.findLast(e => e.type === 'compact/end')!
-    expect(endEvent.data).toMatchObject({ error: 'plain string failure' })
-  })
-
-  it('auto-compaction listener stringifies a non-Error and proceeds', async () => {
-    const { ctx } = await ctxWithModel('SUMMARY')
-    const warnings: string[] = []
-    ctx.logger.warn = ((msg: string) => void warnings.push(msg)) as typeof ctx.logger.warn
-    const svc = new TestCompactService(ctx, cfg({ contextWindow: 300, thresholdRatio: 0.1, retainTokens: 10 }))
-    svc.summarizeError = 'boom' as unknown as Error
-    const session = multiTurnSession(3, 1)
-    const agent = stubAgent(session, 'test-model')
-    const before = session.surface.nodes.length
-
-    await ctx.serial('agent/pre-step', agent, 1, 1, '', [], SIGNAL)
-    // The failure was swallowed; the surface is untouched and a warning logged.
-    expect(session.surface.nodes.length).toBe(before)
-    expect(session.events.some(e => e.type === 'compact/summary')).toBe(false)
-    expect(warnings.some(w => w.includes('compaction failed: boom'))).toBe(true)
-  })
-
-  it('auto-compaction listener takes the result-null branch (nothing to compact)', async () => {
-    const { ctx } = await ctxWithModel('SUMMARY')
-    // A large system prompt pushes the listener's estimate over threshold, but
-    // retainTokens is huge so compactIfNeeded walks everything and returns null.
-    // threshold = floor(2000*0.1) = 200; invariant: 5 + 150 = 155 ≤ 200.
-    const svc = new TestCompactService(ctx, cfg({ contextWindow: 2000, thresholdRatio: 0.1, retainTokens: 150 }))
-    const session = multiTurnSession(2, 1)
-    const agent = stubAgent(session, 'test-model')
-    const bigSystem = 'x'.repeat(900) // ceil(900/4)=225 > threshold 200
-
-    await ctx.serial('agent/pre-step', agent, 1, 1, bigSystem, [], SIGNAL)
-    expect(session.events.some(e => e.type === 'compact/start')).toBe(false)
-    expect(svc.summarizeCalls.length).toBe(0)
-  })
-
-  it('skips messages whose extracted text is empty across all kinds', async () => {
-    const svc = createTestService()
-    const s = new Session(SessionId('empties'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('user/message', { content: [{ type: 'text', text: '' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'reasoning', text: '' }] }, { surfaceOp: 'append' })
-    s.append('context/message', { content: [], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('steering/message', { turn: 1, content: [{ type: 'text', text: '' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('step/end', { turn: 1, step: 1 })
-    // Keep the log pairing-valid while the empty result covers the final message kind.
-    s.append('step/start', { turn: 1, step: 2 })
-    s.append('assistant/message', {
-      turn: 1, step: 2,
-      content: [{ type: 'tool-call', id: CallId('z1'), name: 'bash', arguments: '{}' }],
-    }, { surfaceOp: 'append' })
-    s.append('tool/call', { turn: 1, step: 2, callId: CallId('z1'), name: 'bash', arguments: '{}' })
-    s.append('tool/result', { turn: 1, step: 2, callId: CallId('z1'), content: [], isError: false }, { surfaceOp: 'append' })
-    s.append('step/end', { turn: 1, step: 2 })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    s.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
-
-    const nodes = s.surface.nodes
-    await compactRegion(svc, s, nodes[0]!.seq, nodes[nodes.length - 1]!.seq, 'm')
-    expect(svc.summarizeCalls[0]!.text).toBe('Assistant: [tool-call: bash({})]')
-  })
-
-  it('renders non-text blocks as type-tagged placeholders across all message kinds', async () => {
-    const svc = createTestService()
-    const s = new Session(SessionId('placeholders'))
-    // A plugin-added block type (merge-extensible ContentBlockMap) — the
-    // placeholder path must cover every message kind, not just assistant.
-    const chart = (id: string): ContentBlock => ({ type: 'chart', data: id } as unknown as ContentBlock)
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    // user/message with only a plugin-added block → '[chart]' placeholder.
-    s.append('user/message', { content: [chart('y')], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    // assistant/message with a plugin-added block AND the tool-call its
-    // tool/result answers (so the surface is tool-pairing balanced).
-    s.append('assistant/message', {
-      turn: 1, step: 1,
-      content: [
-        chart('z'),
-        { type: 'tool-call', id: CallId('e1'), name: 'bash', arguments: '{}' },
-      ],
-    }, { surfaceOp: 'append' })
-    // tool/result with a plugin-added block → '[chart]' placeholder.
-    s.append('tool/call', { turn: 1, step: 1, callId: CallId('e1'), name: 'bash', arguments: '{}' })
-    s.append('tool/result', { turn: 1, step: 1, callId: CallId('e1'), content: [chart('r')], isError: false }, { surfaceOp: 'append' })
-    // context/message and steering/message with plugin-added content.
-    s.append('context/message', { content: [chart('c')], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('steering/message', { turn: 1, content: [chart('s')], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('step/end', { turn: 1, step: 1 })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    s.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
-
-    const nodes = s.surface.nodes
-    await compactRegion(svc, s, nodes[0]!.seq, nodes[nodes.length - 1]!.seq, 'm')
-    const { text } = svc.summarizeCalls[0]!
-    // Every non-text block surfaces as a placeholder rather than being dropped.
-    expect(text).toContain('User: [chart]')
-    expect(text).toContain('Assistant: [chart]')
-    expect(text).toContain('Tool result (call e1): [chart]')
-    expect(text).toContain('[Context: [chart]]')
-    expect(text).toContain('[Steering: [chart]]')
-  })
-
-})
-
-describe('BasicCompactService positional range (surface seqs are not monotonic after a replace)', () => {
-  it('compacts a second region after the first replace lands a high-seq summary at the head position', async () => {
-    // Replacement makes surface seqs non-monotonic. The next region is a
-    // positional span even when startSeq > endSeq.
-    const svc = createTestService({ auto: false })
-    const session = multiTurnSession(4, 1)
-
-    // A replacement puts its high-seq summary at the surface head.
-    const nodes0 = session.surface.nodes
-    const first = await compactRegion(svc, session, nodes0[0]!.seq, nodes0[1]!.seq, 'm')
-
-    const nodes1 = session.surface.nodes
-    expect(nodes1[0]!.seq).toBeGreaterThanOrEqual(first.summarySeq)
-    expect(nodes1[0]!.seq).toBeGreaterThan(nodes1[1]!.seq)
-
-    const startSeq = nodes1[0]!.seq
-    const endSeq = nodes1[2]!.seq
-    expect(startSeq).toBeGreaterThan(endSeq)
-    const second = await compactRegion(svc, session, startSeq, endSeq, 'm')
-
-    // Selection follows surface positions, not sequence-number order.
-    expect(second.shadowedSeqs).toEqual([nodes1[0]!.seq, nodes1[1]!.seq, nodes1[2]!.seq])
-    const finalNodes = session.surface.nodes
-    expect(finalNodes[0]!.seq).toBeGreaterThanOrEqual(second.summarySeq)
-    expect(session.deriveMessages().length).toBe(finalNodes.length)
-  })
-
-  it('extracts the second-compaction transcript in surface order, not log-seq order', async () => {
-    const svc = createTestService({ auto: false })
-    const session = multiTurnSession(3, 1)
-
-    // Put a high-seq summary at the head; log order would place retained older nodes first.
-    const n0 = session.surface.nodes
-    await compactRegion(svc, session, n0[0]!.seq, n0[1]!.seq, 'm')
-
-    const n1 = session.surface.nodes
-    svc.summarizeCalls = []
-    await compactRegion(svc, session, n1[0]!.seq, n1[2]!.seq, 'm')
-
-    // Extraction must match surface and `deriveMessages()` order.
-    const { text } = svc.summarizeCalls[0]!
-    const checkpointIdx = text.indexOf('compacted-summary')
-    const olderIdx = text.indexOf('turn 2 user')
-    expect(checkpointIdx).toBeGreaterThanOrEqual(0)
-    expect(olderIdx).toBeGreaterThan(checkpointIdx)
-  })
-})
-
-describe('BasicCompactService llm inject (real plugin-load path)', () => {
-  it('declares llm in static inject so a sibling fiber can resolve ctx.llm', () => {
-    // summarize() reads ctx.llm; the inject lets the cordis ctx proxy resolve a sibling
-    // LlmService when this service is mounted as its own plugin fiber.
-    expect(BasicCompactService.inject).toContain('llm')
-  })
-
-  it('resolves ctx.llm and summarizes when mounted as a sibling plugin of LlmService', async () => {
-    const ctx = new Context()
-    await ctx.plugin(LlmService)
-    ctx.llm.registerAdapter(['test-model'], new ScriptedAdapter('CONDENSED'))
-    // Mount the service through its real plugin fiber (NOT new …(rootCtx)), so
-    // the sibling-fiber ctx.llm resolution actually exercises the inject.
-    const fiber = await ctx.plugin(BasicCompactService, cfg({ auto: false }))
-
-    const svc = ctx.compact as BasicCompactService
-    const session = multiTurnSession(2, 1)
-    const nodes = session.surface.nodes
-    const result = await compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'test-model')
-    expect(result.summary).toEqual([{ type: 'text', text: 'CONDENSED' }])
-
-    // Tear the fiber down so this test owns no leaked registration; the
-    // dedicated cleanup assertion lives in the "HMR safety" suite.
-    await fiber.dispose()
-    expect(ctx.get('compact')).toBeUndefined()
-  })
-})
-
-describe('BasicCompactService under the real invariants plugin', () => {
-  /**
-   * Drive compaction through a session whose `session/event` listeners include
-   * the real dev-mode invariants plugin (as a real app loads it via agent-core).
-   * The invariants throw on append, so a passing run proves the compaction
-   * sequence is contract-valid: every event is turn-enclosed, and the positional
-   * replace op is accepted even when the surface is no longer seq-ordered.
-   */
-  async function setup(): Promise<{ ctx: Context; session: Session; svc: BasicCompactService }> {
-    const ctx = new Context()
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(Invariants)
-    await ctx.plugin(LlmService)
-    ctx.llm.registerAdapter(['test-model'], new ScriptedAdapter('CONDENSED'))
-    await ctx.plugin(BasicCompactService, cfg({ auto: false }))
-    const session = ctx.sessions.create()
-    return { ctx, session, svc: ctx.compact as BasicCompactService }
-  }
-
-  /** Append one closed turn of [user, assistant] surface nodes via the store. */
-  function closedTurn(session: Session, turn: number): void {
+/** Closed two-message turns followed by one open turn for durable compaction events. */
+function conversation(turns = 4, text = 'fixture '.repeat(40).trim()): Session {
+  const session = new Session(SessionId(`conversation-${turns}`))
+  for (let turn = 1; turn <= turns; turn += 1) {
     session.append('turn/start', { turn, trigger: { kind: 'message', source: { kind: 'user' } } })
+    session.append('user/message', {
+      content: [{ type: 'text', text: `${text} user ${turn}` }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
     session.append('step/start', { turn, step: 1 })
-    session.append('user/message', { content: [{ type: 'text', text: `turn ${turn} user.${LONG_FIXTURE_TEXT}` }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    session.append('assistant/message', { turn, step: 1, content: [{ type: 'text', text: `turn ${turn} assistant.${LONG_FIXTURE_TEXT}` }] }, { surfaceOp: 'append' })
+    session.append('assistant/message', {
+      provenance: { provider: MODEL, model: MODEL },
+      turn,
+      step: 1,
+      content: [{ type: 'text', text: `${text} assistant ${turn}` }],
+    }, { surfaceOp: 'append' })
     session.append('step/end', { turn, step: 1 })
     session.append('turn/end', { turn, reason: { kind: 'completed' } })
   }
+  session.append('turn/start', {
+    turn: turns + 1,
+    trigger: { kind: 'message', source: { kind: 'user' } },
+  })
+  return session
+}
 
-  it('runs a turn-enclosed compaction whose positional replace the invariants accept', async () => {
-    const { session, svc } = await setup()
-    closedTurn(session, 1)
-    closedTurn(session, 2)
-    // Open turn 3, as the loop has when the auto-compaction listener fires.
-    session.append('turn/start', { turn: 3, trigger: { kind: 'message', source: { kind: 'user' } } })
+function toolConversation(): Session {
+  const session = new Session(SessionId('tools'))
+  for (let turn = 1; turn <= 3; turn += 1) {
+    const callId = CallId(`call-${turn}`)
+    session.append('turn/start', { turn, trigger: { kind: 'message', source: { kind: 'user' } } })
+    session.append('user/message', {
+      content: [{ type: 'text', text: `request ${turn} `.repeat(300) }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+    session.append('step/start', { turn, step: 1 })
+    session.append('assistant/message', {
+      provenance: { provider: MODEL, model: MODEL },
+      turn,
+      step: 1,
+      content: [
+        { type: 'text', text: `calling ${turn} `.repeat(300) },
+        { type: 'tool-call', id: callId, name: 'read', arguments: '{}' },
+      ],
+    }, { surfaceOp: 'append' })
+    session.append('tool/call', { turn, step: 1, callId, name: 'read', arguments: '{}' })
+    session.append('tool/result', {
+      turn,
+      step: 1,
+      callId,
+      content: [{ type: 'text', text: `result ${turn} `.repeat(300) }],
+      isError: false,
+    }, { surfaceOp: 'append' })
+    session.append('step/end', { turn, step: 1 })
+    session.append('turn/end', { turn, reason: { kind: 'completed' } })
+  }
+  session.append('turn/start', { turn: 4, trigger: { kind: 'message', source: { kind: 'user' } } })
+  return session
+}
 
-    const nodes = session.surface.nodes
-    // No invariant throws here: compact/* + the replacement are all in turn 3.
-    const result = await compactRegion(svc, session, nodes[0]!.seq, nodes[1]!.seq, 'test-model')
-    expect(result.shadowedSeqs.length).toBe(2)
-    expect(session.surface.nodes[0]!.seq).toBeGreaterThan(session.surface.nodes[1]!.seq)
+class TestCompactService extends BasicCompactService {
+  summary: ContentBlock[] = [{ type: 'text', text: 'small checkpoint' }]
+  summaryProvider = 'summary-provider'
+  summaryModel = 'summary-model'
+  error: unknown
+  mutateDuringSummary: (() => void) | undefined
+  calls: Array<{ text: string; signal: AbortSignal | undefined }> = []
+
+  override async summarize(
+    text: string,
+    _agent: Agent,
+    signal?: AbortSignal,
+  ): Promise<{ summary: ContentBlock[]; provider: string; model: string; maxTokens?: number }> {
+    this.calls.push({ text, signal })
+    this.mutateDuringSummary?.()
+    if (this.error !== undefined) throw this.error
+    return {
+      summary: this.summary,
+      provider: this.summaryProvider,
+      model: this.summaryModel,
+      maxTokens: 123,
+    }
+  }
+}
+
+function service(
+  config: BasicCompactConfig = { auto: false },
+  ctx = createContext(),
+): TestCompactService {
+  return new TestCompactService(ctx, config)
+}
+
+async function compactIfNeeded(
+  compact: BasicCompactService,
+  session: Session,
+  model: string | undefined = MODEL,
+  system = '',
+  prefix: readonly Message[] = [],
+): Promise<CompactionResult | null> {
+  return compact.compactIfNeeded(agent(session, model), system, prefix, SIGNAL)
+}
+
+describe('compact configuration and defaults', () => {
+  it('uses low-friction service-wide defaults', () => {
+    const ctx = createContext()
+    const resolved = resolveConfig({}, ctx.tokenMeter)
+
+    expect(resolved).toEqual({
+      thresholdRatio: 0.8,
+      retainTokens: 160,
+      summarizationProvider: '',
+      summarizationModel: '',
+      maxTokens: 8192,
+      compactionRetries: 1,
+      auto: true,
+    })
+    expect(Object.isFrozen(resolved)).toBe(true)
   })
 
-  it('accepts a second compaction over the non-monotonic surface left by the first', async () => {
-    const { session, svc } = await setup()
-    closedTurn(session, 1)
-    closedTurn(session, 2)
-    closedTurn(session, 3)
-    session.append('turn/start', { turn: 4, trigger: { kind: 'message', source: { kind: 'user' } } })
+  it('resolves threshold and retention overrides independently', () => {
+    const ctx = createContext()
+    const thresholdOnly = resolveConfig({
+      thresholdRatio: 0.5,
+    }, ctx.tokenMeter)
+    expect(thresholdOnly).toMatchObject({
+      thresholdRatio: 0.5,
+      retainTokens: 160,
+    })
 
-    const n0 = session.surface.nodes
-    await compactRegion(svc, session, n0[0]!.seq, n0[1]!.seq, 'test-model')
+    const retentionOnly = resolveConfig({
+      retainTokens: 70,
+    }, ctx.tokenMeter)
+    expect(retentionOnly).toMatchObject({
+      thresholdRatio: 0.8,
+      retainTokens: 70,
+    })
+  })
 
-    // Surface head now carries a higher seq than the older retained nodes. A
-    // second compaction spanning [head … a later closed-step end] must pass the
-    // invariants' positional replace check even though startSeq > endSeq.
-    const n1 = session.surface.nodes
-    expect(n1[0]!.seq).toBeGreaterThan(n1[2]!.seq)
-    const second = await compactRegion(svc, session, n1[0]!.seq, n1[2]!.seq, 'test-model')
-    expect(second.shadowedSeqs).toEqual([n1[0]!.seq, n1[1]!.seq, n1[2]!.seq])
+  it('validates common values and pressure-policy invariants', () => {
+    const ctx = createContext()
+    const bad = [
+      [{ maxTokens: 0 }, /maxTokens/],
+      [{ compactionRetries: -1 }, /compactionRetries/],
+      [{ auto: 'yes' }, /auto must be a boolean/],
+      [{ summarizationProvider: 1 }, /summarizationProvider must be a string/],
+      [{ summarizationModel: 1 }, /summarizationModel must be a string/],
+      [{ summarizationProvider: MODEL }, /must both be set or both be empty/],
+      [{ summarizationModel: MODEL }, /must both be set or both be empty/],
+      [{ thresholdRatio: 0 }, /number in \(0, 1\]/],
+      [{ thresholdRatio: 1.1 }, /number in \(0, 1\]/],
+      [{ retainTokens: -1 }, /non-negative integer/],
+      [{ thresholdRatio: 0.5, retainTokens: 500 }, /less than threshold/],
+      [{ models: { [MODEL]: { retainTokens: 10 } } }, /BasicCompactConfig: unknown key "models"/],
+      [{ thresholdRato: 0.5 }, /BasicCompactConfig: unknown key "thresholdRato"/],
+    ] as Array<[unknown, RegExp]>
+
+    for (const [config, pattern] of bad) {
+      expect(() => resolveConfig(config as BasicCompactConfig, ctx.tokenMeter)).toThrow(pattern)
+    }
+  })
+})
+
+describe('pressure measurement and retention', () => {
+  const compactConfig: BasicCompactConfig = {
+    auto: false,
+    thresholdRatio: 0.5,
+    retainTokens: 180,
+  }
+
+  it('skips the provisional check only when no routed or fallback model exists', async () => {
+    const compact = service(compactConfig)
+    const session = conversation()
+    expect(await compact.compactIfNeeded(agent(session), '', [], SIGNAL)).toBeNull()
+    expect(compact.calls).toHaveLength(0)
+  })
+
+  it('meters any routed model without profile resolution', async () => {
+    const compact = service(compactConfig)
+    await expect(compactIfNeeded(compact, conversation(), 'unlisted-model'))
+      .resolves.not.toBeNull()
+  })
+
+  it('does nothing below threshold and compacts a priced head above threshold', async () => {
+    const compact = service(compactConfig)
+    expect(await compactIfNeeded(compact, conversation(2))).toBeNull()
+
+    const session = conversation(4)
+    const result = await compactIfNeeded(compact, session)
+    expect(result).not.toBeNull()
+    expect(result?.shadowedSeqs.length).toBeGreaterThan(2)
+    expect(session.surface.nodes.length).toBeLessThan(8)
+  })
+
+  it('counts the current prompt and request prefix without putting either on the surface', async () => {
+    const compact = service({
+      auto: false,
+      thresholdRatio: 0.7,
+      retainTokens: 50,
+    })
+    const session = conversation(2, 'x'.repeat(200))
+    expect(await compactIfNeeded(compact, session)).toBeNull()
+
+    const prefix: Message[] = [{
+      role: 'user',
+      content: [{ type: 'text', text: 'p'.repeat(1_000) }],
+    }]
+    const result = await compactIfNeeded(compact, session, MODEL, 's'.repeat(1_000), prefix)
+    expect(result).not.toBeNull()
+    expect(prefix).toHaveLength(1)
+    expect(session.events.some(event => event.type === 'context/message')).toBe(false)
+  })
+
+  it('uses the latest logged routed model in the provisional request envelope', async () => {
+    const ctx = createContext()
+    const compact = service({
+      auto: false,
+      thresholdRatio: 0.5,
+      retainTokens: 180,
+    }, ctx)
+    const session = conversation(4)
+    session.append('request/header', {
+      header: { config: { provider: 'actual', model: 'actual' } },
+      reason: 'initial',
+    })
+    const measure = vi.spyOn(ctx.tokenMeter, 'measure')
+
+    const result = await compactIfNeeded(compact, session, 'fallback')
+    expect(result).not.toBeNull()
+    expect(measure.mock.calls[0]?.[1]?.config.provider).toBe('actual')
+    expect(measure.mock.calls[0]?.[1]?.config.model).toBe('actual')
+  })
+
+  it('declines when envelope pressure is high but the surface has no compactable range', async () => {
+    const compact = service(compactConfig)
+    const empty = new Session(SessionId('empty'))
+    expect(await compactIfNeeded(compact, empty, MODEL, 'x'.repeat(100_000))).toBeNull()
+
+    const retained = conversation(1)
+    expect(await compactIfNeeded(compact, retained, MODEL, 'x'.repeat(100_000))).toBeNull()
+  })
+
+  it('uses one unified measurement for each pressure-and-retention decision', async () => {
+    const ctx = createContext()
+    const compact = service(compactConfig, ctx)
+    const measure = vi.spyOn(ctx.tokenMeter, 'measure')
+    const stop = new Error('stop after first decision')
+    vi.spyOn(compact, 'compactRegion').mockRejectedValueOnce(stop)
+
+    await expect(compactIfNeeded(compact, conversation(4))).rejects.toBe(stop)
+    expect(measure).toHaveBeenCalledTimes(1)
+  })
+
+  it('bounds retries when a shrinking checkpoint remains above threshold', async () => {
+    const compact = service({
+      auto: false,
+      compactionRetries: 0,
+      thresholdRatio: 0.3,
+      retainTokens: 180,
+    })
+    compact.summary = Array.from({ length: 7 }, (_, index) => ({
+      type: 'text',
+      text: `summary ${index}`,
+    }))
+
+    await expect(compactIfNeeded(compact, conversation(4)))
+      .rejects.toThrow(/still above threshold after 1 compaction attempts/)
+  })
+
+  it('rounds a retention cut head-ward to preserve tool-call/result pairing', async () => {
+    const compact = service({
+      auto: false,
+      thresholdRatio: 0.8,
+      retainTokens: 80,
+    }, createContext(4_000))
+    const session = toolConversation()
+    const result = await compactIfNeeded(compact, session)
+    expect(result).not.toBeNull()
+
+    const messages = session.deriveMessages()
+    const calls = new Set<string>()
+    for (const message of messages) {
+      for (const block of message.content) {
+        if (block.type === 'tool-call') calls.add(block.id)
+        if (block.type === 'tool-result') expect(calls.has(block.toolCallId)).toBe(true)
+      }
+    }
+  })
+
+  it('rejects a priced surface that is not the current positional surface', () => {
+    const ctx = createContext()
+    const session = conversation(2)
+    const priced = ctx.tokenMeter.measure(session)
+    expect(() => selectCompactableRange(session, {
+      ...priced,
+      nodes: priced.nodes.slice(1),
+    }, 1)).toThrow(/does not match/)
+  })
+
+  it('declines when rounding a cut would consume the only tool pair', () => {
+    const ctx = createContext()
+    const session = new Session(SessionId('one-tool-pair'))
+    const callId = CallId('only')
+    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
+    session.append('step/start', { turn: 1, step: 1 })
+    session.append('assistant/message', {
+      provenance: { provider: MODEL, model: MODEL },
+      turn: 1,
+      step: 1,
+      content: [{ type: 'tool-call', id: callId, name: 'read', arguments: '{}' }],
+    }, { surfaceOp: 'append' })
+    session.append('tool/call', { turn: 1, step: 1, callId, name: 'read', arguments: '{}' })
+    session.append('tool/result', {
+      turn: 1,
+      step: 1,
+      callId,
+      content: [{ type: 'text', text: 'result' }],
+      isError: false,
+    }, { surfaceOp: 'append' })
+    session.append('step/end', { turn: 1, step: 1 })
+
+    const priced = ctx.tokenMeter.measure(session)
+    expect(selectCompactableRange(session, priced, 1)).toBeNull()
+  })
+})
+
+describe('compaction region transaction', () => {
+  it('rejects an agent that does not own the exact target session before mutation', async () => {
+    const compact = service()
+    const target = conversation(2)
+    const owner = conversation(1)
+    const targetEvents = [...target.events]
+    const ownerEvents = [...owner.events]
+    const nodes = target.surface.nodes
+
+    await expect(compact.compactRegion(
+      target,
+      nodes[0]!,
+      nodes[1]!,
+      agent(owner),
+    )).rejects.toThrow('compactRegion: agent.session must be the exact target session')
+
+    expect(target.events).toEqual(targetEvents)
+    expect(owner.events).toEqual(ownerEvents)
+    expect(compact.calls).toEqual([])
+  })
+
+  it('lands a framed, replayable checkpoint with exact pricing provenance', async () => {
+    const compact = service()
+    const session = conversation(3)
+    const before = session.surface.nodes
+    const result = await compact.compactRegion(
+      session,
+      before[0]!,
+      before[3]!,
+      agent(session, MODEL),
+      SIGNAL,
+    )
+
+    expect(result.shadowedSeqs).toEqual(before.slice(0, 4))
+    expect(result.shadowedTokenCount).toBeGreaterThan(0)
+    expect(compact.calls[0]).toMatchObject({ signal: SIGNAL })
+    expect(compact.calls[0]?.text).toContain('fixture user 1')
+    const summary = session.events.findLast(event => event.type === 'compact/summary')
+    expect(summary?.data).toMatchObject({
+      shadowedSeqs: result.shadowedSeqs,
+      shadowedTokenCount: result.shadowedTokenCount,
+      provider: 'summary-provider',
+      model: 'summary-model',
+      maxTokens: 123,
+    })
+    const head = session.deriveMessages()[0]!
+    expect(head.content[0]?.type).toBe('text')
+    expect(head.content[0]?.type === 'text' ? head.content[0].text : '').toContain('<compacted-summary>')
+    expect(head.content.at(-1)).toEqual({ type: 'text', text: '</compacted-summary>' })
+
+    const replay = new Session(SessionId('replay'), [...session.events])
+    expect(replay.deriveMessages()).toEqual(session.deriveMessages())
+  })
+
+  it.each([
+    ['start missing', 9_001, undefined, /start seq 9001 not found/],
+    ['end missing', undefined, 9_002, /end seq 9002 not found/],
+  ])('rejects %s', async (_label, startOverride, endOverride, pattern) => {
+    const compact = service()
+    const session = conversation(2)
+    const nodes = session.surface.nodes
+    await expect(compact.compactRegion(
+      session,
+      startOverride ?? nodes[0]!,
+      endOverride ?? nodes[1]!,
+      agent(session, MODEL),
+    )).rejects.toThrow(pattern)
+  })
+
+  it('rejects reversed and tool-unbalanced positional boundaries', async () => {
+    const compact = service()
+    const plain = conversation(2)
+    const nodes = plain.surface.nodes
+    await expect(compact.compactRegion(
+      plain,
+      nodes[2]!,
+      nodes[1]!,
+      agent(plain, MODEL),
+    )).rejects.toThrow(/is after end/)
+
+    const tools = toolConversation()
+    const toolNodes = tools.surface.nodes
+    await expect(compact.compactRegion(
+      tools,
+      toolNodes[2]!,
+      toolNodes[4]!,
+      agent(tools, MODEL),
+    )).rejects.toThrow(/start seq .* not a balanced boundary/)
+    await expect(compact.compactRegion(
+      tools,
+      toolNodes[0]!,
+      toolNodes[1]!,
+      agent(tools, MODEL),
+    )).rejects.toThrow(/end seq .* not a balanced boundary/)
+  })
+
+  it('requires an open turn and an idle compaction bracket', async () => {
+    const compact = service()
+    const closed = conversation(1)
+    closed.append('turn/end', { turn: 2, reason: { kind: 'completed' } })
+    const nodes = closed.surface.nodes
+    await expect(compact.compactRegion(
+      closed,
+      nodes[0]!,
+      nodes[1]!,
+      agent(closed, MODEL),
+    )).rejects.toThrow(/no open turn/)
+
+    const locked = conversation(1)
+    locked.append('compact/start', { turn: 2 })
+    const lockedNodes = locked.surface.nodes
+    await expect(compact.compactRegion(
+      locked,
+      lockedNodes[0]!,
+      lockedNodes[1]!,
+      agent(locked, MODEL),
+    )).rejects.toThrow(/already in progress/)
+  })
+
+  it('rejects a session with no turn boundary at all', async () => {
+    const compact = service()
+    const session = new Session(SessionId('turnless'))
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'orphan' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+    const node = session.surface.nodes[0]!
+
+    await expect(compact.compactRegion(
+      session,
+      node,
+      node,
+      agent(session, MODEL),
+    )).rejects.toThrow(/no open turn/)
+  })
+
+  it('rejects a meter snapshot that changed before summarization began', async () => {
+    const ctx = createContext()
+    const meter = ctx.tokenMeter
+    const original = meter.measure.bind(meter)
+    vi.spyOn(meter, 'measure').mockImplementationOnce((session) => {
+      const measurement = original(session)
+      return { ...measurement, nodes: measurement.nodes.slice(1) }
+    })
+    const compact = service({ auto: false }, ctx)
+    const session = conversation(2)
+    const nodes = session.surface.nodes
+
+    await expect(compact.compactRegion(
+      session,
+      nodes[0]!,
+      nodes[2]!,
+      agent(session, MODEL),
+    )).rejects.toThrow(/selected surface changed/)
+  })
+
+  it('records summarizer failures without mutating the surface', async () => {
+    const compact = service()
+    compact.error = new Error('summary unavailable')
+    const session = conversation(2)
+    const before = session.surface.nodes
+
+    await expect(compact.compactRegion(
+      session,
+      before[0]!,
+      before[2]!,
+      agent(session, MODEL),
+    )).rejects.toThrow('summary unavailable')
+    expect(session.surface.nodes).toEqual(before)
+    expect(session.events.findLast(event => event.type === 'compact/end')?.data)
+      .toMatchObject({ error: 'summary unavailable' })
+  })
+
+  it('stringifies non-Error failures in the durable end bracket', async () => {
+    const compact = service()
+    compact.error = 'plain failure'
+    const session = conversation(2)
+    const nodes = session.surface.nodes
+    await expect(compact.compactRegion(
+      session,
+      nodes[0]!,
+      nodes[2]!,
+      agent(session, MODEL),
+    )).rejects.toBe('plain failure')
+    expect(session.events.findLast(event => event.type === 'compact/end')?.data)
+      .toMatchObject({ error: 'plain failure' })
+  })
+
+  it('rejects concurrent durable appends before committing the replacement', async () => {
+    const compact = service()
+    const session = conversation(2)
+    compact.mutateDuringSummary = () => {
+      session.append('request/header', {
+        header: { config: { provider: MODEL, model: MODEL } },
+        reason: 'initial',
+      })
+    }
+    const nodes = session.surface.nodes
+
+    await expect(compact.compactRegion(
+      session,
+      nodes[0]!,
+      nodes[2]!,
+      agent(session, MODEL),
+    )).rejects.toThrow(/session log changed/)
+    expect(session.events.some(event => event.type === 'compact/summary')).toBe(false)
+  })
+
+  it('rejects a non-shrinking framed summary under the conversation meter', async () => {
+    const compact = service()
+    compact.summary = Array.from({ length: 100 }, (_, index) => ({
+      type: 'text',
+      text: `verbose ${index}`,
+    }))
+    const session = conversation(2)
+    const nodes = session.surface.nodes
+
+    await expect(compact.compactRegion(
+      session,
+      nodes[0]!,
+      nodes[2]!,
+      agent(session, MODEL),
+    )).rejects.toThrow(/summary is not smaller/)
+    expect(session.events.some(event => event.type === 'compact/summary')).toBe(false)
+  })
+
+  it('lets a model-independent custom summarizer compact without a conversation model', async () => {
+    const compact = service()
+    const session = conversation(1)
+    const nodes = session.surface.nodes
+    await expect(compact.compactRegion(
+      session,
+      nodes[0]!,
+      nodes[1]!,
+      agent(session),
+    )).resolves.toMatchObject({ shadowedSeqs: [nodes[0]!, nodes[1]!] })
+  })
+})
+
+class ScriptedAdapter extends LlmAdapter {
+  lastOptions: GenerateOptions | undefined
+
+  constructor(
+    private readonly blocks: readonly ContentBlock[],
+    private readonly finish: (StreamChunk & { type: 'finish' })['reason'] = { kind: 'stop' },
+  ) {
+    super()
+  }
+
+  override async * stream(options: GenerateOptions): AsyncIterable<StreamChunk> {
+    this.lastOptions = options
+    for (const [index, block] of this.blocks.entries()) {
+      yield { type: 'block-start', index, blockType: block.type }
+      if (block.type === 'text') {
+        yield { type: 'text-delta', index, text: block.text }
+      } else if (block.type === 'reasoning') {
+        yield { type: 'reasoning-delta', index, text: block.text }
+      } else {
+        yield { type: 'block-end', index, block }
+      }
+    }
+    yield { type: 'finish', reason: this.finish }
+  }
+}
+
+async function summarizerHarness(
+  blocks: readonly ContentBlock[],
+  finish?: (StreamChunk & { type: 'finish' })['reason'],
+  model = MODEL,
+  config: BasicCompactConfig = { auto: false },
+): Promise<{ ctx: Context; adapter: ScriptedAdapter; compact: BasicCompactService }> {
+  const ctx = new Context()
+  await ctx.plugin(LlmService)
+  void new TokenMeterService(ctx, { contextWindow: 1_000 })
+  const adapter = new ScriptedAdapter(blocks, finish)
+  ctx.llm.registerAdapter([model], adapter)
+  const compact = new BasicCompactService(ctx, config)
+  return { ctx, adapter, compact }
+}
+
+describe('default one-shot summarizer', () => {
+  it('uses configured model/default cap, forwards cancellation, and keeps only safe text', async () => {
+    const { adapter, compact } = await summarizerHarness([
+      { type: 'reasoning', text: 'private' },
+      { type: 'text', text: 'public summary' },
+      { type: 'tool-call', id: CallId('unexpected'), name: 'x', arguments: '{}' },
+    ], undefined, MODEL, {
+      auto: false,
+      summarizationProvider: MODEL,
+      summarizationModel: MODEL,
+      maxTokens: 321,
+    })
+    const session = conversation(1)
+    const output = await compact.summarize('transcript', agent(session, 'fallback'), SIGNAL)
+
+    expect(output).toEqual({
+      summary: [{ type: 'text', text: 'public summary' }],
+      provider: MODEL,
+      model: MODEL,
+      maxTokens: 321,
+    })
+    expect(adapter.lastOptions).toMatchObject({
+      provider: MODEL,
+      model: MODEL,
+      maxTokens: 321,
+      signal: SIGNAL,
+      sessionId: session.id,
+    })
+    expect(adapter.lastOptions?.system).toContain('## Primary Request and Intent')
+  })
+
+  it('resolves the latest routed provider/model before the AgentOptions pair', async () => {
+    const { adapter, compact } = await summarizerHarness([{ type: 'text', text: 'summary' }], undefined, 'routed')
+    const session = conversation(1)
+    session.append('request/header', {
+      header: { config: { provider: 'routed', model: 'routed' } },
+      reason: 'initial',
+    })
+    const output = await compact.summarize('history', agent(session, 'fallback'))
+    expect(output.provider).toBe('routed')
+    expect(output.model).toBe('routed')
+    expect(adapter.lastOptions?.provider).toBe('routed')
+    expect(adapter.lastOptions?.model).toBe('routed')
+  })
+
+  it('fails clearly when no complete summarization target can be resolved', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    void new TokenMeterService(ctx)
+    const compact = new BasicCompactService(ctx, { auto: false })
+    await expect(compact.summarize('history', agent(new Session(SessionId('model-less')))))
+      .rejects.toThrow(/no provider\/model available for summarization/)
+  })
+
+  it.each([
+    [{ kind: 'error', message: 'provider failed', code: 'PROVIDER' }, 'PROVIDER', /provider failed/],
+    [{ kind: 'error', message: 'opaque' }, undefined, /opaque/],
+    [{ kind: 'aborted' }, 'ABORTED', /aborted/],
+    [{ kind: 'max-tokens' }, 'MAX_TOKENS', /token cap/],
+  ] as Array<[(StreamChunk & { type: 'finish' })['reason'], string | undefined, RegExp]>) (
+    'rejects terminal finish %#',
+    async (finish, code, pattern) => {
+      const { compact } = await summarizerHarness([], finish)
+      let thrown: unknown
+      try {
+        await compact.summarize('history', agent(conversation(1), MODEL))
+      } catch (error: unknown) {
+        thrown = error
+      }
+      expect(thrown).toBeInstanceOf(Error)
+      expect((thrown as Error).message).toMatch(pattern)
+      expect((thrown as Error & { code?: string }).code).toBe(code)
+    },
+  )
+
+  it('rejects empty or reasoning-only successful output', async () => {
+    const { compact } = await summarizerHarness([{ type: 'reasoning', text: 'private' }])
+    await expect(compact.summarize('history', agent(conversation(1), MODEL)))
+      .rejects.toThrow(/no text summary content/)
+  })
+})
+
+describe('automatic listener and loader composition', () => {
+  function preStep(ctx: Context, owner: Agent): Promise<unknown> {
+    return ctx.serial('agent/pre-step', owner, 1, 1, '', [], SIGNAL)
+  }
+
+  it('compacts above threshold and remains idle below it', async () => {
+    const ctx = createContext()
+    const compact = new TestCompactService(ctx, {
+      thresholdRatio: 0.5,
+      retainTokens: 180,
+    })
+    const pressured = conversation(4)
+    await preStep(ctx, agent(pressured, MODEL))
+    expect(pressured.events.some(event => event.type === 'compact/summary')).toBe(true)
+
+    const small = conversation(1)
+    await preStep(ctx, agent(small, MODEL))
+    expect(small.events.some(event => event.type === 'compact/start')).toBe(false)
+    expect(compact.calls).toHaveLength(1)
+  })
+
+  it('warns and continues after operational failures, including non-Errors', async () => {
+    const ctx = createContext()
+    const warnings: string[] = []
+    ctx.logger.warn = ((message: string) => void warnings.push(message)) as typeof ctx.logger.warn
+    const compact = new TestCompactService(ctx, {
+      thresholdRatio: 0.5,
+      retainTokens: 180,
+    })
+    compact.error = 'temporary failure'
+    const session = conversation(4)
+
+    await expect(preStep(ctx, agent(session, MODEL))).resolves.toBeUndefined()
+    expect(warnings).toContainEqual(expect.stringContaining('temporary failure'))
+    expect(session.events.some(event => event.type === 'compact/summary')).toBe(false)
+  })
+
+  it('auto:false installs no listener', async () => {
+    const ctx = createContext()
+    void new TestCompactService(ctx, {
+      auto: false,
+      thresholdRatio: 0.5,
+      retainTokens: 180,
+    })
+    const session = conversation(4)
+    await preStep(ctx, agent(session, MODEL))
+    expect(session.events.some(event => event.type === 'compact/start')).toBe(false)
+  })
+
+  it('loads and disposes the real zero-config service stack', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    const meterFiber = await ctx.plugin(TokenMeterService)
+    const compactFiber = await ctx.plugin(BasicCompactService, { auto: false })
+
+    expect(ctx.tokenMeter.contextWindow).toBe(128_000)
+    expect(ctx.get('compact')).toBeInstanceOf(BasicCompactService)
+    await compactFiber.dispose()
+    expect(ctx.get('compact')).toBeUndefined()
+    await meterFiber.dispose()
+    expect(ctx.get('tokenMeter')).toBeUndefined()
+  })
+
+  it('removes its automatic listener with the plugin fiber', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    await ctx.plugin(TokenMeterService, { contextWindow: 1_000 })
+    const fiber = await ctx.plugin(TestCompactService, {
+      thresholdRatio: 0.5,
+      retainTokens: 180,
+    })
+    await fiber.dispose()
+
+    const session = conversation(4)
+    await preStep(ctx, agent(session, MODEL))
+    expect(session.events.some(event => event.type === 'compact/start')).toBe(false)
   })
 })
diff --git a/packages/compact/compact-basic/tests/compact-loop-repro.spec.ts b/packages/compact/compact-basic/tests/compact-loop-repro.spec.ts
index dbf8a3b737..d163d84804 100644
--- a/packages/compact/compact-basic/tests/compact-loop-repro.spec.ts
+++ b/packages/compact/compact-basic/tests/compact-loop-repro.spec.ts
@@ -1,16 +1,15 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
+import { toolPairingBalancedAfter, toolPairingBalancedBefore } from '@deepseek-ai/dsh-compact'
 import type { ContentBlock, GenerateOptions, StreamChunk } from '@deepseek-ai/dsh-llm'
 import { CallId, LlmAdapter } from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import { isToolPairingBalanced } from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry, { defineTool } from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import { defineTool } from '@deepseek-ai/dsh-tools'
+import { AgentId } from '@deepseek-ai/dsh-agent'
 import AgentLoop, { ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as Invariants from '@deepseek-ai/dsh-invariants'
 import { BasicCompactService } from '@deepseek-ai/dsh-compact-basic'
+import TokenMeterService from '@deepseek-ai/dsh-token-meter'
 import type { SurfaceEvent } from '@deepseek-ai/dsh-session'
 
 /**
@@ -20,15 +19,13 @@ import type { SurfaceEvent } from '@deepseek-ai/dsh-session'
  * surface-position semantics rather than raw-log scanning.
  */
 
-const TOKENS_PER_BLOCK = 10
-
 class ReproCompactService extends BasicCompactService {
-  override estimateContentTokens(blocks: readonly ContentBlock[]): number {
-    return blocks.length * TOKENS_PER_BLOCK
-  }
-
-  override async summarize(): Promise<{ summary: ContentBlock[]; model: string }> {
-    return { summary: [{ type: 'text', text: 'CHECKPOINT SUMMARY' }], model: 'stub' }
+  override async summarize(): Promise<{ summary: ContentBlock[]; provider: string; model: string }> {
+    return {
+      summary: [{ type: 'text', text: 'CHECKPOINT SUMMARY' }],
+      provider: 'mock',
+      model: 'stub',
+    }
   }
 }
 
@@ -60,13 +57,10 @@ class StepwiseToolAdapter extends LlmAdapter {
 
 async function harness(toolSteps: number): Promise<{ ctx: Context; compact: ReproCompactService }> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(Invariants)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
   await ctx.plugin(AgentLoop, { agents: [] })
+  await ctx.plugin(TokenMeterService, { contextWindow: 400 })
   ctx.llm.registerAdapter(['mock'], new StepwiseToolAdapter(toolSteps))
   ctx.tools.register(defineTool({
     name: 'work',
@@ -76,13 +70,12 @@ async function harness(toolSteps: number): Promise<{ ctx: Context; compact: Repr
       return [{ type: 'text', text: 'work result' }]
     },
   }))
-  // Tiny window so a couple of tool steps cross the threshold and compaction
-  // fires within the runaway turn.
+  // Small window so several tool steps cross the threshold and compaction
+  // fires within the runaway turn after enough history can shrink.
   const compact = new ReproCompactService(ctx, {
     auto: true,
-    contextWindow: 64,
     thresholdRatio: 0.5,
-    retainTokens: 20,
+    retainTokens: 50,
     summarizationModel: '',
     maxTokens: 8192,
     compactionRetries: 1,
@@ -105,7 +98,7 @@ describe('CBR-001: a real-loop checkpoint is a valid boundary on both sides', ()
   it('the head checkpoint the loop lands is a balanced cut on both sides', async () => {
     const { ctx } = await harness(8)
     try {
-      const agent = ctx.agentLoop.create(AgentId('repro'), { model: 'mock' })
+      const agent = ctx.agentLoop.create(AgentId('repro'), { provider: 'mock', model: 'mock' })
       agent.send([{ type: 'text', text: 'do a long multi-step task' }])
       await waitForIdle(ctx, agent)
 
@@ -122,12 +115,12 @@ describe('CBR-001: a real-loop checkpoint is a valid boundary on both sides', ()
       // its start and end cuts are balanced in surface order.
       const nodes = agent.session.surface.nodes
       for (const cp of checkpoints) {
-        const node = nodes.find(n => n.seq === cp.seq)
-        if (!node) continue // shadowed by a later checkpoint — no longer an edge.
-        expect(isToolPairingBalanced(nodes, events, node.seq),
-          `checkpoint seq ${node.seq} must be a balanced region START`).toBe(true)
-        expect(isToolPairingBalanced(nodes, events, node.next),
-          `checkpoint seq ${node.seq} must be a balanced region END`).toBe(true)
+        const index = nodes.indexOf(cp.seq)
+        if (index === -1) continue // shadowed by a later checkpoint — no longer an edge.
+        expect(toolPairingBalancedBefore(agent.session, cp.seq),
+          `checkpoint seq ${cp.seq} must be a balanced region START`).toBe(true)
+        expect(toolPairingBalancedAfter(agent.session, cp.seq),
+          `checkpoint seq ${cp.seq} must be a balanced region END`).toBe(true)
       }
     } finally {
       await ctx.fiber.dispose()
diff --git a/packages/compact/compact-basic/tests/loader-composition.spec.ts b/packages/compact/compact-basic/tests/loader-composition.spec.ts
new file mode 100644
index 0000000000..b13e8f8b67
--- /dev/null
+++ b/packages/compact/compact-basic/tests/loader-composition.spec.ts
@@ -0,0 +1,94 @@
+import { mkdtemp, rm, writeFile } from 'node:fs/promises'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { pathToFileURL } from 'node:url'
+import { afterEach, describe, expect, it } from 'vitest'
+import { Context } from 'cordis'
+import Loader from '@cordisjs/plugin-loader'
+import Include from '@cordisjs/plugin-include'
+import LlmService from '@deepseek-ai/dsh-llm'
+import TokenMeterService from '@deepseek-ai/dsh-token-meter'
+import BasicCompactService from '@deepseek-ai/dsh-compact-basic'
+
+let root: string | undefined
+let context: Context | undefined
+
+afterEach(async () => {
+  await context?.fiber.dispose()
+  context = undefined
+  if (root !== undefined) await rm(root, { recursive: true, force: true })
+  root = undefined
+})
+
+async function loadYaml(lines: readonly string[]): Promise<Context> {
+  root = await mkdtemp(join(tmpdir(), 'dsh-token-meter-loader-'))
+  const configPath = join(root, 'cordis.yml')
+  await writeFile(configPath, [...lines, ''].join('\n'))
+
+  context = new Context()
+  context.baseUrl = pathToFileURL(root).href + '/'
+  await context.plugin(Loader)
+  context.loader.builtins.include = Include
+  const modules = new Map<string, unknown>([
+    ['@deepseek-ai/dsh-llm', LlmService],
+    ['@deepseek-ai/dsh-token-meter', TokenMeterService],
+    ['@deepseek-ai/dsh-compact-basic', BasicCompactService],
+  ])
+  context.loader.internal = {
+    version: 'v2',
+    async import(specifier: string) {
+      if (!modules.has(specifier)) throw new Error(`unexpected Loader import: ${specifier}`)
+      return modules.get(specifier)
+    },
+  } as unknown as NonNullable<typeof context.loader.internal>
+  await context.loader.create({
+    name: 'cordis:include',
+    config: { path: pathToFileURL(configPath).href },
+  })
+  await context.loader.await()
+  return context
+}
+
+describe('real Loader composition', () => {
+  it('loads the flat token-meter and compact-basic YAML shape', async () => {
+    const loaded = await loadYaml([
+      "- name: '@deepseek-ai/dsh-llm'",
+      "- name: '@deepseek-ai/dsh-token-meter'",
+      '  config:',
+      '    contextWindow: 4096',
+      "- name: '@deepseek-ai/dsh-compact-basic'",
+      '  config:',
+      '    thresholdRatio: 0.5',
+      '    retainTokens: 512',
+      '    auto: false',
+    ])
+
+    const unloaded = [...loaded.loader.entries()]
+      .filter(entry => entry.fiber === undefined && !entry.disabled)
+      .map(entry => entry.options.name)
+    expect(unloaded).toEqual([])
+    expect(loaded.tokenMeter.contextWindow).toBe(4096)
+    expect(loaded.get('compact')).toBeInstanceOf(BasicCompactService)
+    expect((loaded.compact as BasicCompactService).config).toMatchObject({
+      thresholdRatio: 0.5,
+      retainTokens: 512,
+      auto: false,
+    })
+  })
+
+  it('rejects stale token-meter config after Schemastery normalization', async () => {
+    context = new Context()
+    await expect(context.plugin(TokenMeterService, {
+      models: { legacy: { contextWindow: 4096 } },
+    } as never)).rejects.toThrow(/TokenMeterConfig: unknown key "models"/)
+  })
+
+  it('rejects stale compact-basic config after Schemastery normalization', async () => {
+    context = new Context()
+    await context.plugin(LlmService)
+    await context.plugin(TokenMeterService)
+    await expect(context.plugin(BasicCompactService, {
+      models: { legacy: { thresholdRatio: 0.5 } },
+    } as never)).rejects.toThrow(/BasicCompactConfig: unknown key "models"/)
+  })
+})
diff --git a/packages/compact/compact-basic/tsconfig.json b/packages/compact/compact-basic/tsconfig.json
index 075c64cb61..0103ad82a8 100644
--- a/packages/compact/compact-basic/tsconfig.json
+++ b/packages/compact/compact-basic/tsconfig.json
@@ -8,7 +8,9 @@
   "references": [
     { "path": "../../../vendor/cosmokit" },
     { "path": "../../../vendor/cordis" },
+    { "path": "../../../vendor/schemastery" },
     { "path": "../../llm/llm" },
+    { "path": "../../llm/token-meter" },
     { "path": "../../core/session" },
     { "path": "../../core/agent" },
     { "path": "../compact" }
diff --git a/packages/compact/compact/README.md b/packages/compact/compact/README.md
index 9141012281..ec14947090 100644
--- a/packages/compact/compact/README.md
+++ b/packages/compact/compact/README.md
@@ -6,22 +6,28 @@ This package is the interface tier of the compaction capability, split so each c
 
 | Package | Role |
 |---|---|
-| `@deepseek-ai/dsh-compact` (this) | the interface: abstract service + `compact/*` events + `CompactionResult` + the shared transcript renderer (`renderTranscript`/`renderContentBlocks`) |
-| `@deepseek-ai/dsh-compact-basic` | a backend: chars-per-token estimation (`charsPerToken`, default 4) + token-budget retention + `llm.stream()` summarization |
+| `@deepseek-ai/dsh-compact` (this) | the interface: abstract service + `compact/*` events + `CompactionResult` + tool-pairing boundary helpers + the shared transcript renderer (`renderTranscript`/`renderContentBlocks`) |
+| `@deepseek-ai/dsh-compact-basic` | a backend: `ctx.tokenMeter` pressure + token-budget retention + `llm.stream()` summarization |
 | `@deepseek-ai/dsh-tool-compact` (deferred) | the model-facing `/compact` tool over `ctx.compact` |
 
 Unlike the bash seam, this interface depends on `@deepseek-ai/dsh-session` and `@deepseek-ai/dsh-llm` — the contract's verbs are defined over a `Session` and its output is the `ContentBlock` vocabulary, so they cannot be expressed without naming those packages. That deviation from the "interface depends only on cordis" guidance is intentional and recorded in the [compaction capability-seam RFC](../../../docs/rfc/implemented/feature/2026-06-18-compaction-capability-seam.md).
 
 ## Service API (`ctx.compact`)
 
-Both methods are **abstract** — the backend owns the entire strategy (token estimation, retention policy, event sequencing, summarization).
+Both methods are **abstract** — the backend owns trigger policy, retention, event sequencing, and summarization. Reusable request measurement is a separate service, [`ctx.tokenMeter`](../../llm/token-meter/README.md), rather than part of this interface.
 
 | Member | Semantics |
 |---|---|
 | `compactIfNeeded(agent, fullSystemPrompt, sessionPrefix, signal)` | Estimate the surface-derived history size; if over the backend's threshold, compact an older range via `compactRegion`, keeping recent context intact. Returns the `CompactionResult`, or `null` if nothing needed compacting. All parameters required — the loop's `agent/pre-step` checkpoint supplies the agent, assembled `fullSystemPrompt`, composed `sessionPrefix` (request-only messages every request carries but the derived history omits — the pressure estimate must count them), and turn `signal`. A backend's summarization request is a direct `ctx.llm.stream()` call (not a loop step), so per-call interception happens at `llm/stream`. |
-| `compactRegion(session, start, end, agent, signal?)` | Forcibly summarize surface nodes `[start, end]` (inclusive seqs) into a single replacement node. **Throws** if a compaction is already in progress, if `start`/`end` aren't surface nodes, or if `start` is positioned after `end` on the surface. The range is a SURFACE-POSITION span, not a numeric seq interval — after a prior replace lands a fresh high-seq summary node at the shadowed range's position, surface order no longer tracks seq order. |
+| `compactRegion(session, start, end, agent, signal?)` | Forcibly summarize surface nodes `[start, end]` (inclusive seqs) into a single replacement node. The agent must own the exact target (`session === agent.session`); a backend rejects mismatch before model resolution, lock acquisition, summarization, or log mutation. **Throws** if a compaction is already in progress, if `start`/`end` aren't surface nodes, or if `start` is positioned after `end` on the surface. The range is a SURFACE-POSITION span, not a numeric seq interval — after a prior replace lands a fresh high-seq summary node at the shadowed range's position, surface order no longer tracks seq order. |
 
-`compactIfNeeded` takes a required `signal`; `compactRegion`'s is optional. A backend that summarizes via `ctx.llm.stream()` **must** forward it into the call's `GenerateOptions.signal`, so an abort or fiber dispose tears down the in-flight summarization instead of leaving an orphaned model call running past the cancellation. The session being compacted comes from the agent context; the turn that the `compact/*` events belong to is recoverable from the log (the currently-open turn), so the backend stamps it from the log rather than trusting a caller-supplied value.
+`compactIfNeeded` takes a required `signal`; `compactRegion`'s is optional. A backend that summarizes via `ctx.llm.stream()` **must** forward it into the call's `GenerateOptions.signal`, so an abort or fiber dispose tears down the in-flight summarization instead of leaving an orphaned model call running past the cancellation. The turn that the `compact/*` events belong to is recoverable from the owned session's log (the currently-open turn), so the backend stamps it from the log rather than trusting a caller-supplied value.
+
+## Tool-pairing boundaries
+
+The interface exports `toolPairingBalancedBefore(session, seq)` and `toolPairingBalancedAfter(session, seq)` for snapping and validating compaction edges. A safe edge has no unanswered assistant tool call crossing it. Each helper validates that the event sequence is in the current surface and answers from balances cached per cut in surface order.
+
+The private per-session cache is keyed by `session.surface.replaceGeneration` and the processed surface-entry count. An unchanged generation extends the fold with unseen tail entries only; a log-only append with no new surface entry does no event reads, while a replacement generation rebuilds current membership and balances. Missing event seqs and a `tool/result` without a preceding open call reject as corrupt surface state.
 
 ## Surface contract
 
@@ -29,7 +35,7 @@ Both methods are **abstract** — the backend owns the entire strategy (token es
 
 1. appends `compact/start` (log-only) — acquires the lock,
 2. summarizes the range,
-3. appends `compact/summary` (log-only) — provenance: summary, range, shadowed seqs, token count,
+3. appends `compact/summary` (log-only) — provenance: summary, range, shadowed seqs, token count, and provider/model call envelope,
 4. appends a single `user/message` with `surfaceOp: { op: 'replace', start, end }` carrying the summary — **the only surface mutation**,
 5. appends `compact/end` (log-only) — releases the lock.
 
@@ -47,7 +53,7 @@ The `compact/*` events extend `SessionEventMap` (merge-extensible) via declarati
 
 ## Implementing a backend
 
-Subclass `CompactService`, implement `compactIfNeeded` and `compactRegion`, and load the subclass as a plugin — it registers as `ctx.compact`. A tokenizer-, template-, or model-backed implementation can live as a sibling package without changing callers.
+Subclass `CompactService`, implement `compactIfNeeded` and `compactRegion`, and load the subclass as a plugin — it registers as `ctx.compact`. A template- or model-backed implementation can live as a sibling package without changing callers or the shared token meter.
 
 ## Model Experience
 
diff --git a/packages/compact/compact/src/index.ts b/packages/compact/compact/src/index.ts
index f59e0dc328..7361d38ef3 100644
--- a/packages/compact/compact/src/index.ts
+++ b/packages/compact/compact/src/index.ts
@@ -14,6 +14,7 @@ import type { CompactionResult } from './types.ts'
 
 export type { CompactionResult } from './types.ts'
 export { renderContentBlocks, renderTranscript } from './render.ts'
+export { toolPairingBalancedAfter, toolPairingBalancedBefore } from './tool-pairing.ts'
 
 /** Minimal agent context compaction needs without depending on the agent package. */
 export interface CompactAgentContext {
@@ -28,10 +29,11 @@ declare module 'cordis' {
 }
 
 /**
- * Abstract compaction service. Implementations own token estimation, retention,
- * and summarization, but a successful run must replace the selected surface span
- * with one summary node and prevent concurrent compaction of the same session.
- * Load one implementation per context as `ctx.compact`.
+ * Abstract compaction service. Implementations own trigger policy, retention,
+ * and summarization, and may consume a separate measurement service. A
+ * successful run replaces the selected surface span with one summary node and
+ * prevents concurrent compaction of the same session. Load one implementation
+ * per context as `ctx.compact`.
  */
 export abstract class CompactService extends Service {
   constructor(ctx: Context) {
@@ -65,15 +67,19 @@ export abstract class CompactService extends Service {
    * `start` and `end` name an inclusive span by surface position, not numeric seq
    * order; replacements can make visible seqs non-monotonic. Both edges must be
    * balanced so assistant tool calls remain paired with their results. A model-
-   * backed implementation forwards cancellation and rejects active, missing,
-   * reversed, or unbalanced ranges.
+   * backed implementation forwards cancellation. The agent must own the exact
+   * target session object; implementations reject an ownership mismatch before
+   * model resolution, lock acquisition, summarization, or log mutation, and
+   * reject active, missing, reversed, or unbalanced ranges.
+   * Use {@link toolPairingBalancedBefore} and {@link toolPairingBalancedAfter}
+   * for the edge checks.
    *
-   * @param session - session to mutate.
+   * @param session - session to mutate; must be identical to `agent.session`.
    * @param start - first surface seq, inclusive.
    * @param end - last surface seq, inclusive.
-   * @param agent - summarizer context.
+   * @param agent - owner of the target session and summarizer context.
    * @param signal - optional cancellation; model-backed implementations must forward it.
-   * @throws when compaction is active or the range is missing, reversed, or unbalanced.
+   * @throws when the agent does not own `session`, compaction is active, or the range is missing, reversed, or unbalanced.
    * @returns the replaced range and summary.
    */
   abstract compactRegion(
diff --git a/packages/compact/compact/src/tool-pairing.ts b/packages/compact/compact/src/tool-pairing.ts
new file mode 100644
index 0000000000..56e7a0592d
--- /dev/null
+++ b/packages/compact/compact/src/tool-pairing.ts
@@ -0,0 +1,131 @@
+/**
+ * Tool-pairing balance over a session surface. Compaction changes surface
+ * positions, so safe cuts are derived from tool-call/result content in current
+ * surface order rather than step markers.
+ * @module @deepseek-ai/dsh-compact/tool-pairing
+ */
+
+import type { Session, SessionEvent } from '@deepseek-ai/dsh-session'
+
+/** Incremental balance state for one session surface generation. */
+interface BalanceCache {
+  /** Surface rewrite generation this state describes. */
+  generation: number
+  /**
+   * Balance of every surface cut in current order: a surface of N sequences has
+   * N + 1 cuts, entry `i` being the cut before sequence `i` and the final entry
+   * the cut after the surface tail.
+   */
+  cutBalanced: readonly boolean[]
+  /** Current surface position of each event seq, indexing {@link cutBalanced}. */
+  indexBySeq: Map<number, number>
+  /** In-progress tool-call count after the processed surface tail. */
+  inProgressToolCalls: number
+}
+
+const balanceCacheBySession = new WeakMap<Session, BalanceCache>()
+
+/** Return how one surface event changes the in-progress tool-call count. */
+function eventDelta(event: SessionEvent): number {
+  switch (event.type) {
+    case 'assistant/message':
+      return event.data.content.filter(block => block.type === 'tool-call').length
+    case 'tool/result':
+      return -1
+    default:
+      return 0
+  }
+}
+
+/** Read and validate the event named by a surface sequence. */
+function eventForSeq(events: readonly SessionEvent[], seq: number): SessionEvent {
+  const event = events[seq]
+  if (event === undefined || event.seq !== seq) {
+    throw new Error(`tool-pairing balance: surface seq ${seq} has no matching session event (corrupt surface)`)
+  }
+  return event
+}
+
+/** Fold surface sequences not yet in the cache into its balance state. */
+function extendCache(
+  session: Session,
+  cache: BalanceCache,
+  seqs: readonly number[],
+): BalanceCache {
+  const processed = cache.cutBalanced.length - 1
+  const tail = seqs.slice(processed)
+  // Validate the unseen tail before mutating the live cache, so a corrupt
+  // append cannot leave a partially advanced state behind.
+  const events = session.events
+  const pendingCuts: boolean[] = []
+  let inProgressToolCalls = cache.inProgressToolCalls
+  for (const seq of tail) {
+    inProgressToolCalls += eventDelta(eventForSeq(events, seq))
+    if (inProgressToolCalls < 0) {
+      throw new Error(`tool-pairing balance: tool/result at surface seq ${seq} has no matching tool-call (corrupt surface)`)
+    }
+    pendingCuts.push(inProgressToolCalls === 0)
+  }
+
+  tail.forEach((seq, offset) => cache.indexBySeq.set(seq, processed + offset))
+  cache.cutBalanced = cache.cutBalanced.concat(pendingCuts)
+  cache.inProgressToolCalls = inProgressToolCalls
+  return cache
+}
+
+/** Return balance state synchronized with the current session surface. */
+function balanceCache(session: Session): BalanceCache {
+  const surface = session.surface
+  const seqs = surface.nodes
+  const generation = surface.replaceGeneration
+  const cached = balanceCacheBySession.get(session)
+
+  if (cached === undefined || cached.generation !== generation || cached.cutBalanced.length - 1 > seqs.length) {
+    // A rebuild is the same fold started from the empty-surface state, whose
+    // single leading cut is trivially balanced.
+    const rebuilt = extendCache(session, {
+      generation,
+      cutBalanced: [true],
+      indexBySeq: new Map(),
+      inProgressToolCalls: 0,
+    }, seqs)
+    balanceCacheBySession.set(session, rebuilt)
+    return rebuilt
+  }
+  if (cached.cutBalanced.length - 1 < seqs.length) return extendCache(session, cached, seqs)
+  return cached
+}
+
+/** Balance of the cut at a sequence's position plus offset, rejecting seqs outside current membership. */
+function cutBalance(cache: BalanceCache, seq: number, offset: 0 | 1): boolean {
+  const index = cache.indexBySeq.get(seq)
+  const balanced = index === undefined ? undefined : cache.cutBalanced[index + offset]
+  if (balanced === undefined) {
+    throw new Error(`tool-pairing balance: surface seq ${seq} not found`)
+  }
+  return balanced
+}
+
+/**
+ * Whether the cut immediately before a current surface sequence is tool-pairing balanced.
+ * @param session - session whose surface is checked.
+ * @param seq - event sequence whose leading cut is checked.
+ * @returns true when no unanswered tool call crosses the cut.
+ * @throws when the seq is absent from the current surface, a surface sequence has no
+ * matching log event, or a tool result has no preceding open call.
+ */
+export function toolPairingBalancedBefore(session: Session, seq: number): boolean {
+  return cutBalance(balanceCache(session), seq, 0)
+}
+
+/**
+ * Whether the cut immediately after a current surface sequence is tool-pairing balanced.
+ * @param session - session whose surface is checked.
+ * @param seq - event sequence whose trailing cut is checked.
+ * @returns true when no unanswered tool call crosses the cut.
+ * @throws when the seq is absent from the current surface, a surface sequence has no
+ * matching log event, or a tool result has no preceding open call.
+ */
+export function toolPairingBalancedAfter(session: Session, seq: number): boolean {
+  return cutBalance(balanceCache(session), seq, 1)
+}
diff --git a/packages/compact/compact/src/types.ts b/packages/compact/compact/src/types.ts
index 10a5eabfcc..e831573d5a 100644
--- a/packages/compact/compact/src/types.ts
+++ b/packages/compact/compact/src/types.ts
@@ -24,6 +24,8 @@ declare module '@deepseek-ai/dsh-session' {
       shadowedRange: { start: number; end: number }
       shadowedSeqs: number[]
       shadowedTokenCount: number
+      /** The provider route that wrote the summary. */
+      provider: string
       /**
        * The model that wrote the summary — the summarize call's envelope,
        * reported by the backend that made the call, logged so the one-shot
diff --git a/packages/compact/compact/tests/compact.spec.ts b/packages/compact/compact/tests/compact.spec.ts
index c4daa8cc5a..f272b1fe85 100644
--- a/packages/compact/compact/tests/compact.spec.ts
+++ b/packages/compact/compact/tests/compact.spec.ts
@@ -41,6 +41,7 @@ class StubCompactService extends CompactService {
       shadowedRange: { start, end },
       shadowedSeqs: [],
       shadowedTokenCount: 0,
+      provider: 'mock',
       model: 'stub',
     })
     const endEvent = session.append('compact/end', { turn: 0 })
diff --git a/packages/compact/compact/tests/render.spec.ts b/packages/compact/compact/tests/render.spec.ts
index 1a22296565..3b4bb41ac2 100644
--- a/packages/compact/compact/tests/render.spec.ts
+++ b/packages/compact/compact/tests/render.spec.ts
@@ -54,7 +54,7 @@ describe('renderTranscript', () => {
       content: [{ type: 'text', text: 'fix the bug' }],
       source: { kind: 'user' },
     }, { surfaceOp: 'append' })
-    const assistant = s.append('assistant/message', {
+    const assistant = s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' },
       turn: 0, step: 0,
       content: [{ type: 'text', text: 'looking' }],
     }, { surfaceOp: 'append' })
@@ -111,7 +111,7 @@ describe('renderTranscript', () => {
       content: [{ type: 'text', text: '' }],
       source: { kind: 'user' },
     }, { surfaceOp: 'append' })
-    const emptyAssistant = s.append('assistant/message', {
+    const emptyAssistant = s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' },
       turn: 0, step: 0,
       content: [{ type: 'text', text: '' }],
     }, { surfaceOp: 'append' })
diff --git a/packages/compact/compact/tests/tool-pairing.spec.ts b/packages/compact/compact/tests/tool-pairing.spec.ts
new file mode 100644
index 0000000000..5f48e7f99f
--- /dev/null
+++ b/packages/compact/compact/tests/tool-pairing.spec.ts
@@ -0,0 +1,334 @@
+import { describe, expect, it } from 'vitest'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import { toolPairingBalancedAfter, toolPairingBalancedBefore } from '@deepseek-ai/dsh-compact'
+import { Session, SessionId } from '@deepseek-ai/dsh-session'
+import type { SessionEvent } from '@deepseek-ai/dsh-session'
+
+const SURFACE = { surfaceOp: 'append' as const }
+
+function seqOf(session: Session, type: SessionEvent['type'], nth = 0): number {
+  return session.events.filter(event => event.type === type)[nth]!.seq
+}
+
+function surfaceSeq(session: Session, seq: number): number {
+  const current = session.surface.nodes.find(candidate => candidate === seq)
+  if (current === undefined) throw new Error(`seq ${seq} is not on the surface`)
+  return current
+}
+
+function before(session: Session, type: SessionEvent['type'], nth = 0): boolean {
+  return toolPairingBalancedBefore(session, surfaceSeq(session, seqOf(session, type, nth)))
+}
+
+function after(session: Session, type: SessionEvent['type'], nth = 0): boolean {
+  return toolPairingBalancedAfter(session, surfaceSeq(session, seqOf(session, type, nth)))
+}
+
+function closedToolStep(): Session {
+  const session = new Session(SessionId('closed-tool-step'))
+  session.append('user/message', {
+    content: [{ type: 'text', text: 'go' }],
+    source: { kind: 'user' },
+  }, SURFACE)
+  session.append('assistant/message', {
+    turn: 1,
+    step: 1,
+    content: [{ type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{}' }],
+    provenance: { provider: 'mock', model: 'mock' },
+  }, SURFACE)
+  session.append('tool/result', {
+    turn: 1,
+    step: 1,
+    callId: CallId('c1'),
+    content: [{ type: 'text', text: 'done' }],
+    isError: false,
+  }, SURFACE)
+  return session
+}
+
+describe('tool-pairing boundaries', () => {
+  it('classifies closed and open single-call steps', () => {
+    const closed = closedToolStep()
+    expect(before(closed, 'user/message')).toBe(true)
+    expect(after(closed, 'user/message')).toBe(true)
+    expect(before(closed, 'assistant/message')).toBe(true)
+    expect(after(closed, 'assistant/message')).toBe(false)
+    expect(before(closed, 'tool/result')).toBe(false)
+    expect(after(closed, 'tool/result')).toBe(true)
+
+    const open = new Session(SessionId('open-tool-step'))
+    open.append('assistant/message', {
+      turn: 1,
+      step: 1,
+      content: [{ type: 'tool-call', id: CallId('open'), name: 'bash', arguments: '{}' }],
+      provenance: { provider: 'mock', model: 'mock' },
+    }, SURFACE)
+    expect(toolPairingBalancedAfter(open, open.surface.nodes[0]!)).toBe(false)
+  })
+
+  it('requires every result from a multiple-call assistant message', () => {
+    const session = new Session(SessionId('multiple-calls'))
+    session.append('assistant/message', {
+      turn: 1,
+      step: 1,
+      content: [
+        { type: 'tool-call', id: CallId('c1'), name: 'one', arguments: '{}' },
+        { type: 'tool-call', id: CallId('c2'), name: 'two', arguments: '{}' },
+      ],
+      provenance: { provider: 'mock', model: 'mock' },
+    }, SURFACE)
+    session.append('tool/result', {
+      turn: 1, step: 1, callId: CallId('c1'), content: [], isError: false,
+    }, SURFACE)
+    session.append('tool/result', {
+      turn: 1, step: 1, callId: CallId('c2'), content: [], isError: false,
+    }, SURFACE)
+
+    expect(after(session, 'tool/result', 0)).toBe(false)
+    expect(after(session, 'tool/result', 1)).toBe(true)
+  })
+
+  it('keeps neutral nodes inside an open pair unbalanced and free nodes balanced', () => {
+    const midStep = new Session(SessionId('neutral-mid-step'))
+    midStep.append('assistant/message', {
+      turn: 1,
+      step: 1,
+      content: [{ type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{}' }],
+      provenance: { provider: 'mock', model: 'mock' },
+    }, SURFACE)
+    midStep.append('context/message', {
+      content: [{ type: 'text', text: 'background update' }],
+      source: { kind: 'plugin', plugin: 'test' },
+    }, SURFACE)
+    midStep.append('tool/result', {
+      turn: 1, step: 1, callId: CallId('c1'), content: [], isError: false,
+    }, SURFACE)
+    expect(before(midStep, 'context/message')).toBe(false)
+    expect(after(midStep, 'context/message')).toBe(false)
+
+    const free = new Session(SessionId('neutral-free'))
+    free.append('context/message', {
+      content: [{ type: 'text', text: 'idle injection' }],
+      source: { kind: 'user' },
+    }, SURFACE)
+    expect(before(free, 'context/message')).toBe(true)
+    expect(after(free, 'context/message')).toBe(true)
+  })
+})
+
+describe('tool-pairing surface identity', () => {
+  it('rebuilds after replace and rejects sequences removed from current membership', () => {
+    const session = closedToolStep()
+    const staleTail = surfaceSeq(session, seqOf(session, 'tool/result'))
+    expect(toolPairingBalancedAfter(session, staleTail)).toBe(true)
+
+    const nodes = session.surface.nodes
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'checkpoint' }],
+      source: { kind: 'plugin', plugin: 'compact' },
+    }, {
+      surfaceOp: { op: 'replace', start: nodes[0]!, end: nodes.at(-1)! },
+      sourceEventSeqs: [...nodes],
+    })
+
+    const checkpoint = session.surface.nodes[0]!
+    expect(toolPairingBalancedBefore(session, checkpoint)).toBe(true)
+    expect(toolPairingBalancedAfter(session, checkpoint)).toBe(true)
+    expect(() => toolPairingBalancedBefore(session, staleTail)).toThrow(/surface seq .* not found/)
+    expect(() => toolPairingBalancedAfter(session, staleTail)).toThrow(/surface seq .* not found/)
+  })
+
+  it('answers repeated queries from cached balances', () => {
+    const session = closedToolStep()
+    const assistant = surfaceSeq(session, seqOf(session, 'assistant/message'))
+    expect(toolPairingBalancedAfter(session, assistant)).toBe(false)
+    expect(toolPairingBalancedAfter(session, assistant)).toBe(false)
+  })
+
+  it('rejects missing seqs before and after, including an empty surface', () => {
+    const session = new Session(SessionId('missing-membership'))
+    const missing = 999
+    expect(() => toolPairingBalancedBefore(session, missing)).toThrow(/surface seq 999 not found/)
+    expect(() => toolPairingBalancedAfter(session, missing)).toThrow(/surface seq 999 not found/)
+
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'first node after empty cache' }],
+      source: { kind: 'user' },
+    }, SURFACE)
+    expect(toolPairingBalancedAfter(session, session.surface.nodes[0]!)).toBe(true)
+  })
+})
+
+describe('tool-pairing cache refresh', () => {
+  it('does no event reads for unchanged or log-only growth, folds only appended nodes, and rebuilds on replace', () => {
+    const events: SessionEvent[] = [
+      {
+        type: 'user/message', seq: 0, time: 0,
+        data: { content: [{ type: 'text', text: 'user' }], source: { kind: 'user' } },
+        surfaceOp: 'append',
+      },
+      {
+        type: 'assistant/message', seq: 1, time: 1,
+        data: {
+          turn: 1,
+          step: 1,
+          content: [{ type: 'tool-call', id: CallId('c1'), name: 'one', arguments: '{}' }],
+          provenance: { provider: 'mock', model: 'mock' },
+        },
+        surfaceOp: 'append',
+      },
+      {
+        type: 'tool/result', seq: 2, time: 2,
+        data: { turn: 1, step: 1, callId: CallId('c1'), content: [], isError: false },
+        surfaceOp: 'append',
+      },
+    ]
+    const nodes: number[] = [0, 1, 2]
+    let generation = 0
+    let eventCollectionReads = 0
+    let eventIndexReads = 0
+    const trackedEvents = new Proxy(events, {
+      get(target, property, receiver) {
+        if (typeof property === 'string' && /^\d+$/.test(property)) eventIndexReads += 1
+        return Reflect.get(target, property, receiver) as unknown
+      },
+    })
+    const surface = {
+      get nodes() { return nodes },
+      get replaceGeneration() { return generation },
+    }
+    const session = {
+      surface,
+      get events() {
+        eventCollectionReads += 1
+        return trackedEvents
+      },
+    } as unknown as Session
+
+    expect(toolPairingBalancedAfter(session, nodes[2]!)).toBe(true)
+    expect(eventCollectionReads).toBe(1)
+    expect(eventIndexReads).toBe(3)
+
+    expect(toolPairingBalancedBefore(session, nodes[0]!)).toBe(true)
+    expect(toolPairingBalancedAfter(session, nodes[1]!)).toBe(false)
+    expect(eventCollectionReads).toBe(1)
+    expect(eventIndexReads).toBe(3)
+
+    events.push({
+      type: 'turn/end', seq: 3, time: 3,
+      data: { turn: 1, reason: { kind: 'completed' } },
+    })
+    expect(toolPairingBalancedAfter(session, nodes[2]!)).toBe(true)
+    expect(eventCollectionReads).toBe(1)
+    expect(eventIndexReads).toBe(3)
+
+    events.push({
+      type: 'user/message', seq: 4, time: 4,
+      data: { content: [{ type: 'text', text: 'tail' }], source: { kind: 'user' } },
+      surfaceOp: 'append',
+    })
+    nodes.push(4)
+    expect(toolPairingBalancedAfter(session, nodes[3]!)).toBe(true)
+    expect(eventCollectionReads).toBe(2)
+    expect(eventIndexReads).toBe(4)
+
+    events.push(
+      {
+        type: 'assistant/message', seq: 5, time: 5,
+        data: {
+          turn: 2,
+          step: 1,
+          content: [{ type: 'tool-call', id: CallId('c2'), name: 'two', arguments: '{}' }],
+          provenance: { provider: 'mock', model: 'mock' },
+        },
+        surfaceOp: 'append',
+      },
+      {
+        type: 'tool/result', seq: 6, time: 6,
+        data: { turn: 2, step: 1, callId: CallId('c2'), content: [], isError: false },
+        surfaceOp: 'append',
+      },
+    )
+    nodes.push(5, 6)
+    expect(toolPairingBalancedAfter(session, nodes[5]!)).toBe(true)
+    expect(eventCollectionReads).toBe(3)
+    expect(eventIndexReads).toBe(6)
+
+    events.push({
+      type: 'user/message', seq: 7, time: 7,
+      data: { content: [{ type: 'text', text: 'replacement' }], source: { kind: 'user' } },
+      surfaceOp: { op: 'replace', start: 0, end: 6 },
+    })
+    nodes.splice(0, nodes.length, 7)
+    generation += 1
+    expect(toolPairingBalancedAfter(session, nodes[0]!)).toBe(true)
+    expect(eventCollectionReads).toBe(4)
+    expect(eventIndexReads).toBe(7)
+  })
+
+  it('rebuilds defensively when a same-generation surface entry count regresses', () => {
+    const events: SessionEvent[] = [
+      {
+        type: 'user/message', seq: 0, time: 0,
+        data: { content: [], source: { kind: 'user' } }, surfaceOp: 'append',
+      },
+      {
+        type: 'user/message', seq: 1, time: 1,
+        data: { content: [], source: { kind: 'user' } }, surfaceOp: 'append',
+      },
+    ]
+    const nodes: number[] = [0, 1]
+    const session = {
+      events,
+      surface: { nodes, replaceGeneration: 0 },
+    } as unknown as Session
+    expect(toolPairingBalancedAfter(session, nodes[1]!)).toBe(true)
+    nodes.pop()
+    expect(toolPairingBalancedAfter(session, nodes[0]!)).toBe(true)
+  })
+})
+
+describe('tool-pairing corrupt surfaces', () => {
+  it('throws for an orphan result during a rebuild', () => {
+    const session = new Session(SessionId('orphan-rebuild'))
+    session.append('tool/result', {
+      turn: 1, step: 1, callId: CallId('orphan'), content: [], isError: false,
+    }, SURFACE)
+    expect(() => toolPairingBalancedAfter(session, session.surface.nodes[0]!)).toThrow(/no matching tool-call/)
+  })
+
+  it('retries an orphan result in an appended tail without committing partial cache state', () => {
+    const session = new Session(SessionId('orphan-tail'))
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'safe head' }], source: { kind: 'user' },
+    }, SURFACE)
+    expect(toolPairingBalancedAfter(session, session.surface.nodes[0]!)).toBe(true)
+    session.append('tool/result', {
+      turn: 1, step: 1, callId: CallId('orphan'), content: [], isError: false,
+    }, SURFACE)
+    expect(() => toolPairingBalancedAfter(session, session.surface.nodes[1]!)).toThrow(/no matching tool-call/)
+    expect(() => toolPairingBalancedAfter(session, session.surface.nodes[1]!)).toThrow(/no matching tool-call/)
+  })
+
+  it('throws when a current surface seq has no matching event or indexes the wrong event', () => {
+    const missingSeq = 1
+    const missing = {
+      events: [{
+        type: 'user/message', seq: 0, time: 0,
+        data: { content: [], source: { kind: 'user' } }, surfaceOp: 'append',
+      } satisfies SessionEvent],
+      surface: { nodes: [missingSeq], replaceGeneration: 0 },
+    } as unknown as Session
+    expect(() => toolPairingBalancedBefore(missing, missingSeq)).toThrow(/no matching session event/)
+
+    const mismatchedSeq = 0
+    const mismatched = {
+      events: [{
+        type: 'user/message', seq: 99, time: 0,
+        data: { content: [], source: { kind: 'user' } }, surfaceOp: 'append',
+      } satisfies SessionEvent],
+      surface: { nodes: [mismatchedSeq], replaceGeneration: 0 },
+    } as unknown as Session
+    expect(() => toolPairingBalancedBefore(mismatched, mismatchedSeq)).toThrow(/no matching session event/)
+  })
+})
diff --git a/packages/context/README.md b/packages/context/README.md
index 0045c6629c..8a947703ec 100644
--- a/packages/context/README.md
+++ b/packages/context/README.md
@@ -1,7 +1,10 @@
-# context/ — optional request context
+# context/ — request-context extensions
 
-Opt-in plugins that add bounded model-visible request context without defining a tool or service. The default `dsh-agent-spine-demo` bundle excludes them.
+Product plugins that add model-visible request context without defining a tool or service. `workspace-context` is included by the default `dsh-agent-spine-demo` bundle and can be disabled through bundle config; `time-context` is opt-in.
 
 | Package | Role | ctx key |
 |---|---|---|
-| `time-context/` | Current time and elapsed-time system-prompt context | (none) |
+| `time-context/` | Durable per-step current time and elapsed-time context | (none) |
+| `workspace-context/` | `AGENTS.md`/`CLAUDE.md` workspace context loader | (listens on `agent/session-prefix` + `tools/post-execute`) |
+
+The [`workspace-context` decision record](../../docs/rfc/implemented/feature/2026-06-24-workspace-context.md) explains its per-agent/session isolation and lifecycle split.
diff --git a/packages/context/time-context/README.md b/packages/context/time-context/README.md
index 84d487445b..fea92e726f 100644
--- a/packages/context/time-context/README.md
+++ b/packages/context/time-context/README.md
@@ -1,6 +1,6 @@
 # @deepseek-ai/dsh-time-context
 
-Opt-in dynamic system-prompt context with the current zoned time and elapsed time since the latest model-visible message before the turn. `dsh-agent-spine-demo` and shipped examples do not mount it. Decision record: [the time-context RFC](../../../docs/rfc/implemented/feature/2026-07-14-time-context-plugin.md).
+Opt-in durable context with the current zoned time and elapsed time sampled during model-request preparation. `dsh-agent-spine-demo` and shipped examples do not mount it. Decision record: [the durable time-context RFC](../../../docs/rfc/implemented/feature/2026-07-16-durable-per-step-time-context.md).
 
 ## Config
 
@@ -8,36 +8,51 @@ Opt-in dynamic system-prompt context with the current zoned time and elapsed tim
 - id: time-context
   name: '@deepseek-ai/dsh-time-context'
   config:
-    timeZone: Asia/Shanghai   # optional IANA override; omit for the process zone
-    refreshIntervalMs: 60000  # default; 0 refreshes on every step
+    timeZone: Asia/Shanghai  # optional IANA override; omit for the process zone
+    refreshIntervalMs: 60000 # optional; omit or set to 0 for every eligible attempt
 ```
 
-When `timeZone` is omitted, the plugin resolves the Node process's system zone once at plugin load. Node honors `TZ`; without that override, the host or container supplies the zone. An explicit `timeZone` must be an IANA identifier and is validated at plugin load. `refreshIntervalMs` must be a non-negative safe integer. Every turn's first request refreshes; later steps reuse the reading until its age reaches the interval. `0` refreshes every step. Refresh occurs only during request assembly and creates no timer work.
+When `timeZone` is omitted, the plugin resolves the Node process's system zone once at plugin load. Node honors `TZ`; without that override, the host or container supplies the zone. An explicit `timeZone` must be an IANA identifier and is validated at plugin load.
 
-## Message baseline
+`refreshIntervalMs` must be a non-negative safe integer. Omission or `0` appends on every pre-step attempt whose signal is not already aborted. A positive value appends only when the session has no earlier time-context injection, wall time moved backward, or at least that many milliseconds have elapsed since the latest injection.
 
-The duration starts at the latest user, assistant, tool-result, context, or steering message before the current `turn/start`. Every refresh in the turn retains that baseline, so the current prompt does not collapse the interval to approximately zero. The first turn reports that no earlier message exists. The durable clock source is session-event append time, not client send time.
+## Timing semantics
 
-The loop records the dynamic section in `request/header` / `request/header-delta`. Requests therefore remain reconstructable, carry one timing block, and retain no earlier readings in conversation history.
+The plugin prepends an `agent/pre-step` listener. When an injection is due, it appends one `context/message` through `agent.inject()` before `step/start` and ordinary automatic compaction, with source `{ kind: 'plugin', plugin: 'time-context' }`. A suppressed attempt appends nothing.
+
+Positive-interval scheduling scans the raw durable session events for the latest `context/message` with that source, including a reading shadowed by compaction. The schedule therefore applies across turns and resumed processes without process-local cache state. It reduces append frequency and history growth but never removes an existing reading, and sessions schedule independently.
+
+Step 1 measures from the latest preceding model-visible message, including the prompt that opened the turn. Later steps measure from the preceding time-context event in the same turn. Both baselines use durable session-event timestamps; backward wall-clock movement clamps elapsed time to zero. A missing first-step baseline, or a later step with no earlier same-turn reading because interval suppression skipped it, reports `unavailable`.
+
+A time reading records a request-preparation attempt, not a committed step or transmitted request. Because the listener runs first, its append may remain when a later pre-step listener cancels or fails the attempt; the log is append-only and the plugin performs no rollback.
+
+The time reading stays in derived conversation history until a later compaction shadows it. Request headers contain no time-context state. Request reconstruction uses the complete durable surface prefix at each `step/start`, so transmitted requests need not map one-to-one to readings: a failed preparation can leave an extra reading, while interval suppression can let a request reuse existing history without adding one.
 
 ## Model Experience
 
-### Temporal system prompt
+### Preparation-time temporal context
 
-**What the model sees**: Every request in an active turn includes the two lines below. `<timestamp>` is an ISO-shaped local timestamp with numeric offset and IANA zone; `<duration-or-unavailable>` is compact whole-second units or the first-turn fallback.
+**What the model sees**: On each preparation attempt that injects, one source-tagged context message containing the two lines below. `<timestamp>` is an ISO-shaped local timestamp with numeric offset and IANA zone; durations use compact whole-second units. Positive intervals can leave an attempted step without a new reading.
 
-**Token effect**: Fixed two-line cost per request. A refresh replaces the request-header section; prior readings do not accumulate.
+**Token effect**: Each injected two-line message accumulates until compaction shadows it. A positive interval reduces additions; omission or `0` adds one for every eligible preparation attempt.
 
-#### Temporal context section
+#### First step
 
 ```markdown
-Current time: <timestamp>
-Time since previous message: <duration-or-unavailable>.
+Time sampled while preparing turn <turn>, step 1: <timestamp>
+Elapsed since the preceding model-visible message: <duration-or-unavailable>.
+```
+
+#### Later steps
+
+```markdown
+Time sampled while preparing turn <turn>, step <step>: <timestamp>
+Elapsed since the preceding step context: <duration-or-unavailable>.
 ```
 
 ## Known Limitations and Deferred Work
 
-- **Request-bound refresh only** — no clock update is emitted while the agent is waiting inside a model call or tool; the next assembled step refreshes once the configured interval has elapsed.
-- **Whole-second display** — timestamps and durations omit sub-second precision even when `refreshIntervalMs` is below 1,000.
-- **Session-event baseline** — elapsed time starts from the durable append timestamp, not a client transport's original send timestamp.
+- **Whole-second display** — timestamps and durations omit sub-second precision even though durable event times retain milliseconds.
+- **Session-event baseline** — elapsed time starts from durable append timestamps, not a client transport's original send timestamp.
 - **Process-local default zone** — omission uses the Node process's `TZ`, host, or container zone captured at plugin load, not a remote user's zone; configure an explicit IANA zone when those differ.
+- **History cost between compactions** — omission or `0` retains one reading for every eligible preparation attempt, including attempts later cancelled or failed; a positive interval reduces but does not eliminate this cost.
diff --git a/packages/context/time-context/package.json b/packages/context/time-context/package.json
index f319c7a5b1..c0d69a75cb 100644
--- a/packages/context/time-context/package.json
+++ b/packages/context/time-context/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@deepseek-ai/dsh-time-context",
-  "description": "Opt-in system-prompt context with the current time and elapsed time since the previous message",
+  "description": "Opt-in durable per-step context with the current time and elapsed time",
   "version": "0.0.1",
   "private": true,
   "type": "module",
@@ -26,13 +26,14 @@
   },
   "peerDependencies": {
     "@deepseek-ai/dsh-agent": "^0.0.1",
-    "@deepseek-ai/dsh-system-prompt": "^0.0.1",
     "cordis": "^4.0.0-rc.7"
   },
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-loader-smoke": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
     "@deepseek-ai/dsh-tools": "workspace:^",
diff --git a/packages/context/time-context/src/index.ts b/packages/context/time-context/src/index.ts
index cccd433811..f8463e4bec 100644
--- a/packages/context/time-context/src/index.ts
+++ b/packages/context/time-context/src/index.ts
@@ -1,8 +1,6 @@
 /**
- * Opt-in request-time clock context. Active turns receive the current zoned
- * time and elapsed time since the preceding model-visible message. The loop
- * logs each rendered value as request-header state rather than conversation
- * history.
+ * Opt-in request-preparation clock context. Eligible pre-step attempts append
+ * durable, source-attributed time readings to conversation history.
  *
  * @module @deepseek-ai/dsh-time-context
  */
@@ -10,77 +8,30 @@
 import type { Context } from 'cordis'
 import z from 'schemastery'
 import type { Agent } from '@deepseek-ai/dsh-agent'
-import type { AssembleContext } from '@deepseek-ai/dsh-system-prompt'
+import type { Message } from '@deepseek-ai/dsh-llm'
 
 /** Cordis plugin name used by loader diagnostics. */
 export const name = 'time-context'
 
-/** The system-prompt registry that owns the dynamic request section. */
-export const inject = ['systemPrompt']
+/** The agent registry that owns the pre-step lifecycle seam. */
+export const inject = ['agents']
 
-/** Request-time clock formatting and refresh policy. Invalid values fail plugin load. */
+/** Request-preparation clock formatting and append scheduling. Invalid values fail plugin load. */
 export interface Config {
   /** IANA time zone used for the rendered timestamp. Omit to resolve the Node process's system zone at plugin load. */
   timeZone?: string
-  /** Maximum age of a reading within one turn, in milliseconds (default 60,000; `0` refreshes every step). */
+  /** Minimum milliseconds between durable injections in one session. Omit or set to 0 to inject on every eligible pre-step attempt. */
   refreshIntervalMs?: number
 }
 
-/** Schemastery validation and defaults for {@link Config}. */
+/** Schemastery validation for {@link Config}. */
 export const Config: z<Config> = z.object({
   timeZone: z.string(),
-  refreshIntervalMs: z.number().default(60_000),
+  refreshIntervalMs: z.number(),
 })
 
-interface OpenTurn {
-  turn: number
-  startSeq: number
-}
-
-/** Cached text and the fixed inter-turn baseline used by one agent's open turn. */
-interface RenderState {
-  turn: number
-  renderedAt: number
-  previousMessageTime: number | undefined
-  text: string
-}
-
 type TimestampPart = 'day' | 'hour' | 'minute' | 'month' | 'second' | 'timeZoneName' | 'year'
 
-function openTurn(agent: Agent): OpenTurn | undefined {
-  for (const event of [...agent.session.events].reverse()) {
-    switch (event.type) {
-      case 'turn/end':
-        return undefined
-      case 'turn/start':
-        return { turn: event.data.turn, startSeq: event.seq }
-      default:
-        // Merge-extensible session events: only turn boundaries matter here.
-        break
-    }
-  }
-  return undefined
-}
-
-/** Find the latest model-visible timestamp strictly before one turn boundary. */
-function previousMessageTime(agent: Agent, turnStartSeq: number): number | undefined {
-  for (const event of [...agent.session.events].reverse()) {
-    if (event.seq >= turnStartSeq) continue
-    switch (event.type) {
-      case 'user/message':
-      case 'assistant/message':
-      case 'tool/result':
-      case 'context/message':
-      case 'steering/message':
-        return event.time
-      default:
-        // Merge-extensible session events: non-surface records are not messages.
-        break
-    }
-  }
-  return undefined
-}
-
 /** Format an epoch millisecond value as an ISO-shaped timestamp with offset and IANA zone. */
 function formatTimestamp(now: number, formatter: Intl.DateTimeFormat, timeZone: string): string {
   const parts = Object.fromEntries(
@@ -107,31 +58,85 @@ function formatDuration(elapsedMs: number): string {
   return parts.join(' ')
 }
 
+/** Find the latest model-visible event, excluding this plugin's pending append. */
+function precedingMessageTime(agent: Agent): number | undefined {
+  for (const event of [...agent.session.events].reverse()) {
+    switch (event.type) {
+      case 'user/message':
+      case 'assistant/message':
+      case 'tool/result':
+      case 'context/message':
+      case 'steering/message':
+        return event.time
+      default:
+        // Merge-extensible session events: non-surface records are not messages.
+        break
+    }
+  }
+  return undefined
+}
+
+/** Find the preceding time-context event within the open turn. */
+function precedingStepContextTime(agent: Agent, turn: number): number | undefined {
+  for (const event of [...agent.session.events].reverse()) {
+    if (event.type === 'turn/start' && event.data.turn === turn) return undefined
+    if (event.type === 'context/message'
+      && event.data.source.kind === 'plugin'
+      && event.data.source.plugin === name) {
+      return event.time
+    }
+  }
+  return undefined
+}
+
+/** Find this plugin's latest durable injection, including a shadowed surface event. */
+function latestInjectionTime(agent: Agent): number | undefined {
+  for (const event of [...agent.session.events].reverse()) {
+    if (event.type === 'context/message'
+      && event.data.source.kind === 'plugin'
+      && event.data.source.plugin === name) {
+      return event.time
+    }
+  }
+  return undefined
+}
+
 function renderText(
   now: number,
+  turn: number,
+  step: number,
   previous: number | undefined,
   formatter: Intl.DateTimeFormat,
   timeZone: string,
 ): string {
-  const elapsed = previous === undefined
-    ? 'unavailable (no earlier message in this session)'
-    : formatDuration(now - previous)
-  return `Current time: ${formatTimestamp(now, formatter, timeZone)}\nTime since previous message: ${elapsed}.`
+  const elapsed = previous === undefined ? 'unavailable' : formatDuration(now - previous)
+  const baseline = step === 1 ? 'model-visible message' : 'step context'
+  return `Time sampled while preparing turn ${turn}, step ${step}: ${formatTimestamp(now, formatter, timeZone)}\n`
+    + `Elapsed since the preceding ${baseline}: ${elapsed}.`
+}
+
+/** Reject refresh intervals that cannot represent an exact elapsed-millisecond threshold. */
+function validateRefreshInterval(refreshIntervalMs: number | undefined): void {
+  if (refreshIntervalMs !== undefined && (
+    !Number.isSafeInteger(refreshIntervalMs)
+    || refreshIntervalMs < 0
+  )) {
+    throw new TypeError(
+      `time-context: refreshIntervalMs must be a non-negative safe integer, got ${String(refreshIntervalMs)}`,
+    )
+  }
 }
 
 /**
- * Register the request-time clock section for the lifetime of `ctx`.
- * @param ctx - plugin context; the section registration is disposed with it.
- * @param config - validated time zone and intra-turn refresh interval.
- * @throws when the time zone or refresh interval is invalid.
+ * Register a prepended pre-step listener for the lifetime of `ctx`.
+ * @param ctx - plugin context; the listener is disposed with it.
+ * @param config - time zone and durable refresh scheduling configuration.
+ * @throws when the refresh interval is invalid or the configured or process time zone cannot be resolved.
  */
 export function apply(ctx: Context, config: Config): void {
   const timeZone = config.timeZone
-  const refreshIntervalMs = config.refreshIntervalMs as number
-  if (!Number.isSafeInteger(refreshIntervalMs) || refreshIntervalMs < 0) {
-    throw new Error(`time-context: refreshIntervalMs must be a non-negative safe integer, got ${refreshIntervalMs}`)
-  }
-
+  const refreshIntervalMs = config.refreshIntervalMs
+  validateRefreshInterval(refreshIntervalMs)
   let formatter: Intl.DateTimeFormat
   try {
     formatter = new Intl.DateTimeFormat('en-US', {
@@ -152,32 +157,29 @@ export function apply(ctx: Context, config: Config): void {
     throw new Error(message, { cause: error })
   }
   const resolvedTimeZone = formatter.resolvedOptions().timeZone
-  const states = new WeakMap<Agent, RenderState>()
 
-  ctx.systemPrompt.section({
-    name: 'context:time',
-    order: 10,
-    text(context: AssembleContext): string {
-      const agent = context.agent
-      if (agent === undefined) return ''
-      const currentTurn = openTurn(agent)
-      if (currentTurn === undefined) return ''
-
-      const now = Date.now()
-      const prior = states.get(agent)
-      if (prior !== undefined
-        && prior.turn === currentTurn.turn
-        && now >= prior.renderedAt
-        && now - prior.renderedAt < refreshIntervalMs) {
-        return prior.text
-      }
-
-      const previous = prior?.turn === currentTurn.turn
-        ? prior.previousMessageTime
-        : previousMessageTime(agent, currentTurn.startSeq)
-      const text = renderText(now, previous, formatter, resolvedTimeZone)
-      states.set(agent, { turn: currentTurn.turn, renderedAt: now, previousMessageTime: previous, text })
-      return text
-    },
-  })
+  ctx.on('agent/pre-step', (
+    agent: Agent,
+    turn: number,
+    step: number,
+    _fullSystemPrompt: string,
+    _sessionPrefix: readonly Message[],
+    signal: AbortSignal,
+  ) => {
+    if (signal.aborted) return
+    const now = Date.now()
+    if (refreshIntervalMs !== undefined && refreshIntervalMs > 0) {
+      const lastInjection = latestInjectionTime(agent)
+      if (lastInjection !== undefined
+        && now >= lastInjection
+        && now - lastInjection < refreshIntervalMs) return
+    }
+    const previous = step === 1
+      ? precedingMessageTime(agent)
+      : precedingStepContextTime(agent, turn)
+    agent.inject(
+      [{ type: 'text', text: renderText(now, turn, step, previous, formatter, resolvedTimeZone) }],
+      { source: { kind: 'plugin', plugin: name } },
+    )
+  }, { prepend: true })
 }
diff --git a/packages/context/time-context/tests/time-context.e2e.ts b/packages/context/time-context/tests/time-context.e2e.ts
index f83b451ef7..f14981ea36 100644
--- a/packages/context/time-context/tests/time-context.e2e.ts
+++ b/packages/context/time-context/tests/time-context.e2e.ts
@@ -4,15 +4,21 @@ import { tmpdir } from 'node:os'
 import { join } from 'node:path'
 import { fileURLToPath } from 'node:url'
 import { afterEach, describe, expect, it } from 'vitest'
-import { foldRequestHeader, type SessionEvent } from '@deepseek-ai/dsh-session'
+import { type SessionEvent } from '@deepseek-ai/dsh-session'
+import { resolveExampleLaunch } from '@deepseek-ai/dsh-loader-smoke'
 
+// Keep the Loader config under examples so both modes exercise the same deployable
+// topology: local fixture source plus bare plugins owned by the examples workspace.
 const binScript = fileURLToPath(new URL('../../../examples/stdio-demo/src/bin.ts', import.meta.url))
-const configPath = fileURLToPath(new URL('./fixtures/cordis.yml', import.meta.url))
+const configPath = fileURLToPath(new URL(
+  '../../../../examples/echo-agent/tests/fixtures/context/time-context/cordis.yml',
+  import.meta.url,
+))
 const repoTsconfig = fileURLToPath(new URL('../../../../tsconfig.json', import.meta.url))
-const tsxLoader = fileURLToPath(import.meta.resolve('tsx'))
 const PROCESS_TIMEOUT_MS = 30_000
 const TEST_TIMEOUT_MS = PROCESS_TIMEOUT_MS + 15_000
-const FIRST_REPLY = 'You said: "first". Try "echo <something>" to see a tool call.'
+const FIRST_REPLY = '[main turn 1] You said: "Time sampled while preparing turn 1, step 1:'
+const SECOND_REPLY = '[main turn 2] You said: "Time sampled while preparing turn 2, step 1:'
 
 let child: ChildProcessWithoutNullStreams | undefined
 let workdir: string | undefined
@@ -38,21 +44,22 @@ async function runTwoTurns(): Promise<{ stdout: string; stderr: string }> {
   workdir = await mkdtemp(join(tmpdir(), 'time-context-e2e-'))
   const cwd = workdir
   return new Promise((resolve, reject) => {
-    const proc = spawn(
-      process.execPath,
-      ['--expose-internals', '--import', tsxLoader, binScript, configPath],
-      {
-        cwd,
-        env: {
-          ...process.env,
-          TZ: 'Asia/Shanghai',
-          TSX_TSCONFIG_PATH: repoTsconfig,
-          DSH_HOME: join(cwd, '.dsh'),
-          DSH_AGENTS_HOME: join(cwd, '.agents'),
-        },
-        stdio: ['pipe', 'pipe', 'pipe'],
+    const launch = resolveExampleLaunch({
+      srcBin: binScript,
+      configArgs: [configPath],
+      tsconfigPath: repoTsconfig,
+      exposeInternals: true,
+      env: {
+        TZ: 'Asia/Shanghai',
+        DSH_HOME: join(cwd, '.dsh'),
+        DSH_AGENTS_HOME: join(cwd, '.agents'),
       },
-    )
+    })
+    const proc = spawn(launch.command, launch.args, {
+      cwd,
+      env: { ...process.env, ...launch.env },
+      stdio: ['pipe', 'pipe', 'pipe'],
+    })
     child = proc
     let stdout = ''
     let stderr = ''
@@ -60,7 +67,7 @@ async function runTwoTurns(): Promise<{ stdout: string; stderr: string }> {
     proc.stdout.setEncoding('utf8')
     proc.stdout.on('data', (chunk: string) => {
       stdout += chunk
-      if (!sentSecond && stdout.includes(`${FIRST_REPLY}\n> `)) {
+      if (!sentSecond && stdout.includes(FIRST_REPLY) && stdout.includes('Try "echo <something>" to see a tool call.\n> ')) {
         sentSecond = true
         proc.stdin.end('second\n')
       }
@@ -84,12 +91,12 @@ async function runTwoTurns(): Promise<{ stdout: string; stderr: string }> {
 }
 
 describe('time-context through a real cordis.yml and stdio process', () => {
-  it('uses the process zone and persists both first-turn and elapsed-time request context', async () => {
+  it('uses the process zone and persists one ordered context event per request', async () => {
     const { stdout, stderr } = await runTwoTurns()
     expect(stderr).not.toContain('UNHANDLED')
     expect(stdout).toContain('time-context e2e ready.')
     expect(stdout).toContain(FIRST_REPLY)
-    expect(stdout).toContain('You said: "second".')
+    expect(stdout).toContain(SECOND_REPLY)
 
     const logs = await jsonlFiles(join(workdir as string, '.sessions'))
     expect(logs).toHaveLength(1)
@@ -97,19 +104,28 @@ describe('time-context through a real cordis.yml and stdio process', () => {
     const events = lines.slice(1).map(line => JSON.parse(line) as SessionEvent)
     expect(events.filter(event => event.type === 'turn/end')).toHaveLength(2)
 
-    const firstHeader = events.find(event => event.type === 'request/header')
-    if (firstHeader?.type !== 'request/header') throw new Error('missing initial request/header event')
-    expect(firstHeader.data.header.system).toMatch(
-      /Current time: \d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}\+08:00\[Asia\/Shanghai\]/,
+    const contexts = events.filter(event => event.type === 'context/message')
+    const starts = events.filter(event => event.type === 'step/start')
+    expect(contexts).toHaveLength(2)
+    expect(starts).toHaveLength(2)
+    for (let index = 0; index < contexts.length; index += 1) {
+      expect(contexts[index]!.seq).toBeLessThan(starts[index]!.seq)
+      expect(contexts[index]!.surfaceOp).toBe('append')
+      expect(contexts[index]!.data.source).toEqual({ kind: 'plugin', plugin: 'time-context' })
+    }
+    const contextText = contexts.map(event => event.data.content
+      .filter(block => block.type === 'text')
+      .map(block => block.text)
+      .join('\n'))
+    expect(contextText[0]).toMatch(
+      /Time sampled while preparing turn 1, step 1: \d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}\+08:00\[Asia\/Shanghai\]/,
     )
-    expect(firstHeader.data.header.system).toContain(
-      'Time since previous message: unavailable (no earlier message in this session).',
+    expect(contextText[0]).toMatch(
+      /Elapsed since the preceding model-visible message: (?:\d+d )?(?:\d+h )?(?:\d+m )?\d+s\./,
     )
+    expect(contextText[1]).toMatch(/Time sampled while preparing turn 2, step 1:/)
 
-    const finalSystem = foldRequestHeader(events)?.system
-    expect(finalSystem).toContain('[Asia/Shanghai]')
-    expect(finalSystem).toMatch(
-      /Time since previous message: (?:\d+d )?(?:\d+h )?(?:\d+m )?\d+s\./,
-    )
+    const headers = events.filter(event => event.type === 'request/header')
+    expect(JSON.stringify(headers)).not.toContain('Time sampled while preparing')
   }, TEST_TIMEOUT_MS)
 })
diff --git a/packages/context/time-context/tests/time-context.spec.ts b/packages/context/time-context/tests/time-context.spec.ts
index 562b002b68..2796e8230c 100644
--- a/packages/context/time-context/tests/time-context.spec.ts
+++ b/packages/context/time-context/tests/time-context.spec.ts
@@ -1,19 +1,20 @@
 import { afterEach, beforeEach, describe, expect, it, vi } from 'vitest'
 import { Context } from 'cordis'
 import Loader from '@cordisjs/plugin-loader'
-import LlmService, { CallId, LlmAdapter } from '@deepseek-ai/dsh-llm'
+import { CallId, LlmAdapter } from '@deepseek-ai/dsh-llm'
 import type { GenerateOptions, StreamChunk } from '@deepseek-ai/dsh-llm'
-import SessionStore, { Session, SessionId, foldRequestHeader } from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import { Session, SessionId } from '@deepseek-ai/dsh-session'
 import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
 import type { Agent } from '@deepseek-ai/dsh-agent'
-import ToolRegistry, { defineTool } from '@deepseek-ai/dsh-tools'
+import { defineTool } from '@deepseek-ai/dsh-tools'
 import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as timeContext from '@deepseek-ai/dsh-time-context'
 import type { Config } from '@deepseek-ai/dsh-time-context'
 
 const BASE = Date.parse('2026-07-14T00:00:00.000Z')
 const ORIGINAL_TIME_ZONE = process.env['TZ']
+const SIGNAL = new AbortController().signal
 
 beforeEach(() => {
   process.env['TZ'] = 'UTC'
@@ -30,18 +31,29 @@ afterEach(() => {
 
 async function mount(config: Config = {}) {
   const ctx = new Context()
-  await ctx.plugin(SystemPrompt)
+  await ctx.plugin(AgentRegistry)
   const fiber = await ctx.plugin(timeContext, config)
   return { ctx, fiber }
 }
 
 function sessionAgent(session: Session, id = 'agent'): Agent {
-  return { id: AgentId(id), session } as unknown as Agent
-}
-
-async function sectionText(ctx: Context, agent?: Agent): Promise<string | undefined> {
-  const assembly = await ctx.systemPrompt.assemble(agent === undefined ? {} : { agent })
-  return assembly.sections.find(section => section.name === 'context:time')?.text
+  return {
+    id: AgentId(id),
+    options: {},
+    session,
+    status: 'running',
+    ctx: new Context(),
+    send() {},
+    steer() {},
+    inject(content, options) {
+      session.append('context/message', {
+        content,
+        source: options?.source ?? { kind: 'user' },
+      }, { surfaceOp: 'append' })
+    },
+    cancel() {},
+    whenIdle: () => Promise.resolve(),
+  }
 }
 
 function openMessageTurn(session: Session, turn: number): void {
@@ -52,6 +64,28 @@ function openMessageTurn(session: Session, turn: number): void {
   }, { surfaceOp: 'append' })
 }
 
+function contextTexts(session: Session): string[] {
+  const texts: string[] = []
+  for (const event of session.events) {
+    if (event.type === 'context/message'
+      && event.data.source.kind === 'plugin'
+      && event.data.source.plugin === 'time-context') {
+      texts.push(event.data.content.find(block => block.type === 'text')?.text ?? '')
+    }
+  }
+  return texts
+}
+
+async function fire(
+  ctx: Context,
+  agent: Agent,
+  turn: number,
+  step: number,
+  signal: AbortSignal = SIGNAL,
+): Promise<void> {
+  await ctx.serial('agent/pre-step', agent, turn, step, '', [], signal)
+}
+
 function textResponse(text: string): StreamChunk[] {
   return [
     { type: 'block-start', index: 0, blockType: 'text' },
@@ -89,184 +123,186 @@ class ScriptedAdapter extends LlmAdapter {
 
 async function loopHarness(adapter: ScriptedAdapter, config: Config = {}): Promise<Context> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(timeContext, config)
   ctx.llm.registerAdapter(['mock'], adapter)
   return ctx
 }
 
-describe('temporal section rendering', () => {
-  it('renders the first turn in UTC with the explicit no-previous-message fallback', async () => {
-    const { ctx } = await mount()
+function requestText(request: GenerateOptions): string {
+  return request.messages
+    .flatMap(message => message.content)
+    .filter(block => block.type === 'text')
+    .map(block => block.text)
+    .join('\n')
+}
+
+describe('durable step context', () => {
+  it('records turn, step, zoned time, and the preceding model-visible message baseline', async () => {
+    const { ctx } = await mount({ timeZone: 'Asia/Shanghai' })
     const session = new Session(SessionId('first'))
     openMessageTurn(session, 1)
-
-    expect(await sectionText(ctx, sessionAgent(session))).toBe(
-      'Current time: 2026-07-14T00:00:00+00:00[UTC]\n'
-      + 'Time since previous message: unavailable (no earlier message in this session).',
-    )
-  })
-
-  it('renders a non-UTC numeric offset and all compact duration units', async () => {
-    const { ctx } = await mount({ timeZone: 'Asia/Shanghai' })
-    const session = new Session(SessionId('offset'))
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    session.append('assistant/message', {
-      turn: 1,
-      step: 1,
-      content: [{ type: 'text', text: 'previous' }],
-    }, { surfaceOp: 'append' })
-    session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
     vi.setSystemTime(BASE + 90_061_000)
-    openMessageTurn(session, 2)
 
-    expect(await sectionText(ctx, sessionAgent(session))).toBe(
-      'Current time: 2026-07-15T09:01:01+08:00[Asia/Shanghai]\n'
-      + 'Time since previous message: 1d 1h 1m 1s.',
+    await fire(ctx, sessionAgent(session), 1, 1)
+
+    expect(contextTexts(session)).toEqual([
+      'Time sampled while preparing turn 1, step 1: 2026-07-15T09:01:01+08:00[Asia/Shanghai]\n'
+      + 'Elapsed since the preceding model-visible message: 1d 1h 1m 1s.',
+    ])
+    const event = session.events.at(-1)
+    expect(event?.type).toBe('context/message')
+    if (event?.type !== 'context/message') throw new Error('missing time context')
+    expect(event.data.source).toEqual({ kind: 'plugin', plugin: 'time-context' })
+    expect(event.surfaceOp).toBe('append')
+  })
+
+  it('reports an unavailable first-step baseline when no model-visible message precedes it', async () => {
+    const { ctx } = await mount()
+    const session = new Session(SessionId('unavailable'))
+    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
+
+    await fire(ctx, sessionAgent(session), 1, 1)
+
+    expect(contextTexts(session)[0]).toContain(
+      'Elapsed since the preceding model-visible message: unavailable.',
     )
   })
 
-  it('clamps a backward wall-clock adjustment to a zero duration', async () => {
+  it.each([
+    ['omitted interval', {}],
+    ['zero interval', { refreshIntervalMs: 0 }],
+  ] as const)('uses the preceding durable step-context timestamp after step one with %s', async (_label, config) => {
+    const { ctx } = await mount(config)
+    const session = new Session(SessionId('later-step'))
+    const agent = sessionAgent(session)
+    openMessageTurn(session, 3)
+    await fire(ctx, agent, 3, 1)
+    vi.setSystemTime(BASE + 61_000)
+
+    await fire(ctx, agent, 3, 2)
+
+    expect(contextTexts(session)[1]).toBe(
+      'Time sampled while preparing turn 3, step 2: 2026-07-14T00:01:01+00:00[UTC]\n'
+      + 'Elapsed since the preceding step context: 1m 1s.',
+    )
+  })
+
+  it('reports an unavailable later-step baseline at the matching turn boundary', async () => {
     const { ctx } = await mount()
-    const session = new Session(SessionId('backward-duration'))
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    session.append('assistant/message', {
-      turn: 1,
-      step: 1,
-      content: [{ type: 'text', text: 'future by adjusted clock' }],
-    }, { surfaceOp: 'append' })
-    session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
+    const session = new Session(SessionId('later-step-boundary'))
+    openMessageTurn(session, 4)
+
+    await fire(ctx, sessionAgent(session), 4, 2)
+
+    expect(contextTexts(session)[0]).toContain(
+      'Elapsed since the preceding step context: unavailable.',
+    )
+  })
+
+  it('reports an unavailable later-step baseline when event lookup is exhausted', async () => {
+    const { ctx } = await mount()
+    const session = new Session(SessionId('later-step-exhausted'))
+
+    await fire(ctx, sessionAgent(session), 1, 2)
+
+    expect(contextTexts(session)[0]).toContain(
+      'Elapsed since the preceding step context: unavailable.',
+    )
+  })
+
+  it('injects after backward wall-clock movement and clamps elapsed time to zero', async () => {
+    const { ctx } = await mount({ refreshIntervalMs: 60_000 })
+    const session = new Session(SessionId('backward'))
+    const agent = sessionAgent(session)
+    openMessageTurn(session, 1)
+    await fire(ctx, agent, 1, 1)
     vi.setSystemTime(BASE - 5_000)
-    openMessageTurn(session, 2)
 
-    expect(await sectionText(ctx, sessionAgent(session))).toContain('Time since previous message: 0s.')
+    await fire(ctx, agent, 1, 2)
+
+    expect(contextTexts(session)).toHaveLength(2)
+    expect(contextTexts(session)[1]).toContain('Elapsed since the preceding step context: 0s.')
   })
 
-  const previousMessageCases = [
-    ['user/message', (session: Session): void => {
-      session.append('user/message', { content: [{ type: 'text', text: 'u' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    }],
-    ['assistant/message', (session: Session): void => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'a' }] }, { surfaceOp: 'append' })
-    }],
-    ['tool/result', (session: Session): void => {
-      session.append('tool/result', {
-        turn: 1,
-        step: 1,
-        callId: CallId('previous'),
-        content: [{ type: 'text', text: 'r' }],
-        isError: false,
-      }, { surfaceOp: 'append' })
-    }],
-    ['context/message', (session: Session): void => {
-      session.append('context/message', {
-        content: [{ type: 'text', text: 'c' }],
-        source: { kind: 'plugin', plugin: 'test' },
-      }, { surfaceOp: 'append' })
-    }],
-    ['steering/message', (session: Session): void => {
-      session.append('steering/message', {
-        turn: 1,
-        content: [{ type: 'text', text: 's' }],
-        source: { kind: 'user' },
-      }, { surfaceOp: 'append' })
-    }],
-  ] as const
+  it('uses a shadowed durable injection after resume and injects at the exact threshold', async () => {
+    const { ctx } = await mount({ refreshIntervalMs: 1_000 })
+    const original = new Session(SessionId('seed-source'))
+    openMessageTurn(original, 1)
+    await fire(ctx, sessionAgent(original), 1, 1)
+    const user = original.events.find(event => event.type === 'user/message')
+    const reading = original.events.find(event => event.type === 'context/message')
+    if (user === undefined || reading === undefined) throw new Error('missing source surface events')
+    original.append('context/message', {
+      content: [{ type: 'text', text: 'compacted history' }],
+      source: { kind: 'plugin', plugin: 'compact-basic' },
+    }, {
+      surfaceOp: { op: 'replace', start: user.seq, end: reading.seq },
+      sourceEventSeqs: [user.seq, reading.seq],
+    })
+    original.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
+    expect(JSON.stringify(original.deriveMessages())).not.toContain('Time sampled while preparing')
 
-  it.each(previousMessageCases)('uses a prior %s as the duration baseline', async (_name, appendPrevious) => {
-    const { ctx } = await mount()
-    const session = new Session(SessionId(`previous-${_name}`))
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    appendPrevious(session)
-    session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    vi.setSystemTime(BASE + 5_000)
-    openMessageTurn(session, 2)
+    const resumed = new Session(SessionId('resumed'), [...original.events])
+    const resumedAgent = sessionAgent(resumed)
+    vi.setSystemTime(BASE + 999)
+    openMessageTurn(resumed, 2)
+    const beforeSkip = resumed.events.length
 
-    expect(await sectionText(ctx, sessionAgent(session))).toContain('Time since previous message: 5s.')
-  })
+    await fire(ctx, resumedAgent, 2, 1)
 
-  it('contributes empty text without an active agent turn', async () => {
-    const { ctx } = await mount()
-    expect(await sectionText(ctx)).toBe('')
+    expect(resumed.events).toHaveLength(beforeSkip)
+    expect(contextTexts(resumed)).toHaveLength(1)
 
-    const empty = sessionAgent(new Session(SessionId('empty')))
-    expect(await sectionText(ctx, empty)).toBe('')
-
-    const closedSession = new Session(SessionId('closed'))
-    openMessageTurn(closedSession, 1)
-    closedSession.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    expect(await sectionText(ctx, sessionAgent(closedSession))).toBe('')
-  })
-})
-
-describe('refresh policy', () => {
-  it('reuses within the interval, refreshes at expiry, and refreshes after a backward clock jump', async () => {
-    const { ctx } = await mount({ refreshIntervalMs: 60_000 })
-    const session = new Session(SessionId('interval'))
-    const agent = sessionAgent(session)
-    openMessageTurn(session, 1)
-
-    const first = await sectionText(ctx, agent)
-    vi.setSystemTime(BASE + 30_000)
-    expect(await sectionText(ctx, agent)).toBe(first)
-    vi.setSystemTime(BASE + 60_000)
-    const expired = await sectionText(ctx, agent)
-    expect(expired).toContain('2026-07-14T00:01:00+00:00[UTC]')
-    vi.setSystemTime(BASE + 59_000)
-    expect(await sectionText(ctx, agent)).toContain('2026-07-14T00:00:59+00:00[UTC]')
-  })
-
-  it('refreshes every assembly when refreshIntervalMs is zero', async () => {
-    const { ctx } = await mount({ refreshIntervalMs: 0 })
-    const session = new Session(SessionId('every-step'))
-    const agent = sessionAgent(session)
-    openMessageTurn(session, 1)
-    const first = await sectionText(ctx, agent)
     vi.setSystemTime(BASE + 1_000)
-    expect(await sectionText(ctx, agent)).not.toBe(first)
+    await fire(ctx, resumedAgent, 2, 2)
+
+    expect(contextTexts(resumed)).toHaveLength(2)
+    expect(contextTexts(resumed)[1]).toContain(
+      'Elapsed since the preceding step context: unavailable.',
+    )
   })
 
-  it('always refreshes for a new turn and keeps the preceding message baseline', async () => {
-    const { ctx } = await mount({ refreshIntervalMs: 60_000 })
-    const session = new Session(SessionId('turn-refresh'))
+  it('applies a positive interval across turns without sharing state between sessions', async () => {
+    const { ctx } = await mount({ refreshIntervalMs: 1_000 })
+    const first = new Session(SessionId('interval-first'))
+    const firstAgent = sessionAgent(first, 'first-agent')
+    openMessageTurn(first, 1)
+    await fire(ctx, firstAgent, 1, 1)
+    first.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
+
+    vi.setSystemTime(BASE + 500)
+    openMessageTurn(first, 2)
+    const beforeSkip = first.events.length
+    await fire(ctx, firstAgent, 2, 1)
+
+    const independent = new Session(SessionId('interval-independent'))
+    openMessageTurn(independent, 1)
+    await fire(ctx, sessionAgent(independent, 'independent-agent'), 1, 1)
+
+    expect(first.events).toHaveLength(beforeSkip)
+    expect(contextTexts(first)).toHaveLength(1)
+    expect(contextTexts(independent)).toHaveLength(1)
+  })
+
+  it('runs before ordinary pre-step listeners and skips an already-aborted step', async () => {
+    const { ctx } = await mount()
+    const session = new Session(SessionId('ordering'))
     const agent = sessionAgent(session)
     openMessageTurn(session, 1)
-    const first = await sectionText(ctx, agent)
-    vi.setSystemTime(BASE + 1_000)
-    session.append('assistant/message', {
-      turn: 1,
-      step: 1,
-      content: [{ type: 'text', text: 'done' }],
-    }, { surfaceOp: 'append' })
-    session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    vi.setSystemTime(BASE + 2_000)
-    openMessageTurn(session, 2)
+    let ordinarySawContext = false
+    ctx.on('agent/pre-step', (subject) => {
+      ordinarySawContext = subject.session.events.some(event => event.type === 'context/message')
+    })
 
-    const second = await sectionText(ctx, agent)
-    expect(second).not.toBe(first)
-    expect(second).toContain('Time since previous message: 1s.')
-  })
+    await fire(ctx, agent, 1, 1)
+    const abort = new AbortController()
+    abort.abort()
+    await fire(ctx, agent, 1, 2, abort.signal)
 
-  it('keeps refresh caches independent per agent', async () => {
-    const { ctx } = await mount({ refreshIntervalMs: 60_000 })
-    const sessionA = new Session(SessionId('agent-a'))
-    const sessionB = new Session(SessionId('agent-b'))
-    const agentA = sessionAgent(sessionA, 'a')
-    const agentB = sessionAgent(sessionB, 'b')
-    openMessageTurn(sessionA, 1)
-    openMessageTurn(sessionB, 1)
-    const aFirst = await sectionText(ctx, agentA)
-    vi.setSystemTime(BASE + 30_000)
-    const bFirst = await sectionText(ctx, agentB)
-    vi.setSystemTime(BASE + 40_000)
-
-    expect(await sectionText(ctx, agentA)).toBe(aFirst)
-    expect(bFirst).toContain('2026-07-14T00:00:30+00:00[UTC]')
+    expect(ordinarySawContext).toBe(true)
+    expect(contextTexts(session)).toHaveLength(1)
   })
 })
 
@@ -278,49 +314,79 @@ describe('configuration and lifecycle', () => {
     const session = new Session(SessionId('system-zone'))
     openMessageTurn(session, 1)
 
-    expect(await sectionText(ctx, sessionAgent(session))).toContain(
-      'Current time: 2026-07-14T08:00:00+08:00[Asia/Shanghai]',
+    await fire(ctx, sessionAgent(session), 1, 1)
+
+    expect(contextTexts(session)[0]).toContain('2026-07-14T08:00:00+08:00[Asia/Shanghai]')
+  })
+
+  it('fails loud for an invalid explicit zone or an unavailable process zone', async () => {
+    const invalid = new Context()
+    await invalid.plugin(AgentRegistry)
+    await expect(invalid.plugin(timeContext, { timeZone: 'Not/A_Real_Zone' })).rejects.toThrow(
+      /invalid IANA timeZone/,
     )
-  })
 
-  it('fails loud for negative, fractional, unsafe, and invalid-zone config', async () => {
-    for (const refreshIntervalMs of [-1, 1.5, Number.MAX_SAFE_INTEGER + 1]) {
-      const ctx = new Context()
-      await ctx.plugin(SystemPrompt)
-      await expect(ctx.plugin(timeContext, { refreshIntervalMs })).rejects.toThrow(/non-negative safe integer/)
-    }
-
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-    await expect(ctx.plugin(timeContext, { timeZone: 'Not/A_Real_Zone' })).rejects.toThrow(/invalid IANA timeZone/)
-  })
-
-  it('fails loud when the process system zone cannot be resolved', async () => {
     vi.spyOn(Intl, 'DateTimeFormat').mockImplementationOnce(() => {
       throw new RangeError('system zone unavailable')
     })
-    const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
-
-    await expect(ctx.plugin(timeContext, {})).rejects.toThrow(/failed to resolve the system time zone/)
+    const unresolved = new Context()
+    await unresolved.plugin(AgentRegistry)
+    await expect(unresolved.plugin(timeContext, {})).rejects.toThrow(/failed to resolve the system time zone/)
   })
 
-  it('removes its section when the plugin fiber disposes', async () => {
+  it('rejects invalid refresh intervals at plugin load with one diagnostic', async () => {
+    const invalid = [-1, 0.5, Number.MAX_SAFE_INTEGER + 1, Number.POSITIVE_INFINITY, Number.NaN]
+    for (const refreshIntervalMs of invalid) {
+      await expect(mount({ refreshIntervalMs })).rejects.toThrow(
+        'time-context: refreshIntervalMs must be a non-negative safe integer',
+      )
+    }
+  })
+
+  it('removes its listener when the plugin fiber disposes', async () => {
     const { ctx, fiber } = await mount()
     const session = new Session(SessionId('dispose'))
     const agent = sessionAgent(session)
     openMessageTurn(session, 1)
-    expect(await sectionText(ctx, agent)).toContain('Current time:')
+    await fire(ctx, agent, 1, 1)
 
     await fiber.dispose()
-    expect(await sectionText(ctx, agent)).toBeUndefined()
+    await fire(ctx, agent, 1, 2)
+
+    expect(contextTexts(session)).toHaveLength(1)
   })
 })
 
-describe('real agent-loop request logging', () => {
-  it('refreshes a long turn in the system prompt and records the header delta without context history', async () => {
-    const adapter = new ScriptedAdapter([toolCallResponse(), textResponse('done'), textResponse('next turn')])
-    const ctx = await loopHarness(adapter, { refreshIntervalMs: 60_000 })
+describe('real agent-loop request history', () => {
+  it.each([
+    ['throws', 'error'],
+    ['cancels', 'aborted'],
+  ] as const)('retains the preparation reading when a later pre-step listener %s', async (mode, reasonKind) => {
+    const adapter = new ScriptedAdapter([textResponse('unused')])
+    const ctx = await loopHarness(adapter)
+    let laterSawReading = false
+    ctx.on('agent/pre-step', (subject) => {
+      laterSawReading = contextTexts(subject.session).length === 1
+      if (mode === 'throws') throw new Error('later pre-step failure')
+      subject.cancel('later pre-step cancellation')
+    })
+    const agent = ctx.agentLoop.create(AgentId(`late-${mode}`), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'start' }])
+    await agent.whenIdle()
+
+    expect(laterSawReading).toBe(true)
+    expect(contextTexts(agent.session)).toHaveLength(1)
+    expect(adapter.requests).toHaveLength(0)
+    expect(agent.session.events.some(event => event.type === 'step/start')).toBe(false)
+    const turnEnd = agent.session.events.findLast(event => event.type === 'turn/end')
+    expect(turnEnd?.type === 'turn/end' && turnEnd.data.reason.kind).toBe(reasonKind)
+    await ctx.fiber.dispose()
+  })
+
+  it('persists one ordered context per request, accumulates readings, and leaves system headers unchanged', async () => {
+    const adapter = new ScriptedAdapter([toolCallResponse(), textResponse('done')])
+    const ctx = await loopHarness(adapter)
     ctx.tools.register(defineTool({
       name: 'tick',
       description: 'advance fake time',
@@ -330,42 +396,57 @@ describe('real agent-loop request logging', () => {
         return [{ type: 'text' as const, text: 'advanced' }]
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('loop'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('loop'), { provider: 'mock', model: 'mock' })
 
     agent.send([{ type: 'text', text: 'start' }])
     await agent.whenIdle()
-    expect(adapter.requests).toHaveLength(2)
-    expect(adapter.requests[0]!.system).toContain('2026-07-14T00:00:00+00:00[UTC]')
-    expect(adapter.requests[1]!.system).toContain('2026-07-14T00:01:01+00:00[UTC]')
-    expect(agent.session.events.some(event => event.type === 'context/message')).toBe(false)
-    expect(agent.session.events.filter(event => event.type === 'request/header-delta')).toHaveLength(1)
-    expect(foldRequestHeader(agent.session.events)?.system).toBe(adapter.requests[1]!.system)
 
-    vi.setSystemTime(BASE + 361_000)
-    agent.send([{ type: 'text', text: 'again' }])
-    await agent.whenIdle()
-    expect(adapter.requests[2]!.system).toContain('Time since previous message: 5m 0s.')
+    expect(adapter.requests).toHaveLength(2)
+    const contexts = agent.session.events.filter(event => event.type === 'context/message')
+    const starts = agent.session.events.filter(event => event.type === 'step/start')
+    expect(contexts).toHaveLength(adapter.requests.length)
+    expect(starts).toHaveLength(adapter.requests.length)
+    for (let index = 0; index < contexts.length; index += 1) {
+      expect(contexts[index]!.seq).toBeLessThan(starts[index]!.seq)
+    }
+    expect(contexts.every(event => event.data.source.kind === 'plugin'
+      && event.data.source.plugin === 'time-context'
+      && event.surfaceOp === 'append')).toBe(true)
+
+    const firstRequestText = requestText(adapter.requests[0]!)
+    const secondRequestText = requestText(adapter.requests[1]!)
+    expect(firstRequestText).toContain('Time sampled while preparing turn 1, step 1:')
+    expect(firstRequestText).toContain('Elapsed since the preceding model-visible message: 0s.')
+    expect(firstRequestText).not.toContain('Time sampled while preparing turn 1, step 2:')
+    expect(secondRequestText).toContain('Time sampled while preparing turn 1, step 1:')
+    expect(secondRequestText).toContain('Time sampled while preparing turn 1, step 2:')
+    expect(secondRequestText).toContain('Elapsed since the preceding step context: 1m 1s.')
+
+    for (const request of adapter.requests) expect(request.system).not.toContain('Time sampled while preparing')
+    const headers = agent.session.events.filter(event => event.type === 'request/header')
+    expect(JSON.stringify(headers)).not.toContain('Time sampled while preparing')
     await ctx.fiber.dispose()
   })
 })
 
 describe('real Loader export path', () => {
-  it('keeps the namespace metadata and boots through unwrapExports', async () => {
+  it('keeps namespace metadata and boots the agent listener through unwrapExports', async () => {
     expect('default' in timeContext).toBe(false)
     const loader = Object.create(Loader.prototype) as Loader
     const unwrapped = loader.unwrapExports(timeContext) as Record<string, unknown>
     expect(unwrapped).toBe(timeContext)
     expect(unwrapped.name).toBe('time-context')
-    expect(unwrapped.inject).toEqual(['systemPrompt'])
+    expect(unwrapped.inject).toEqual(['agents'])
     expect(unwrapped.Config).toBeDefined()
     expect(typeof unwrapped.apply).toBe('function')
 
     const ctx = new Context()
-    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(AgentRegistry)
     const plugin = loader.unwrapExports(timeContext) as Parameters<Context['plugin']>[0]
     await ctx.plugin(plugin)
     const session = new Session(SessionId('loader'))
     openMessageTurn(session, 1)
-    expect(await sectionText(ctx, sessionAgent(session))).toContain('Current time:')
+    await fire(ctx, sessionAgent(session), 1, 1)
+    expect(contextTexts(session)[0]).toContain('Time sampled while preparing turn 1, step 1:')
   })
 })
diff --git a/packages/context/time-context/tsconfig.json b/packages/context/time-context/tsconfig.json
index eda3a81772..7815242b55 100644
--- a/packages/context/time-context/tsconfig.json
+++ b/packages/context/time-context/tsconfig.json
@@ -9,7 +9,10 @@
     { "path": "../../../vendor/cosmokit" },
     { "path": "../../../vendor/cordis" },
     { "path": "../../../vendor/schemastery" },
+    { "path": "../../llm/llm" },
+    { "path": "../../core/agent" },
     { "path": "../../core/system-prompt" },
-    { "path": "../../core/agent" }
+    { "path": "../../core/agent" },
+    { "path": "../../support/loader-smoke" }
   ]
 }
diff --git a/packages/context/workspace-context/README.md b/packages/context/workspace-context/README.md
new file mode 100644
index 0000000000..430ffe8a41
--- /dev/null
+++ b/packages/context/workspace-context/README.md
@@ -0,0 +1,140 @@
+# @deepseek-ai/dsh-workspace-context
+
+Per-session workspace instruction loading for `AGENTS.md`-compatible files. The plugin freezes the initial user-global and project instruction chain into the request prefix, then discovers nested files and reports later changes or removals through durable context messages after successful filesystem tool calls.
+
+## Lifecycle
+
+The baseline is composed once per agent-loop instance on `agent/session-prefix`. It reads `$DSH_HOME/AGENTS.md` followed by one configured instruction candidate in each directory from the project root to `agent.session.header.cwd`. The prefix is placed before all derived history, recorded in `EpochHeader.messagePrefix`, and reused verbatim for that loop instance. Because the plugin prepends its contribution before delegating, a later-registered skills catalog appears after workspace instructions.
+
+The plugin also listens on `tools/post-execute` for successful first-party `read`, `write`, and `edit` calls. Each touch checks newly reached descendant scopes and every previously loaded scope. A new file is attached through the result's `additionalContexts`; a changed file or candidate switch appends a replacement; a missing final candidate appends a removal notice. Native calls and Code Mode sub-dispatches share this path: `run_code` defers each nested context until its outer result, so the loop still appends updates after tool-call/result adjacency is complete. This follows structured filesystem activity rather than shell `cd`, because each local bash call starts a fresh shell and parsing arbitrary shell syntax would be unreliable.
+
+Instruction reads use the optional `ctx.fs` provider. The plugin does not statically inject `fs`, so providerless product trees still boot and instruction loading becomes a no-op until a provider is present. It calls `ctx.fs.lstat` before resolving a candidate, rejecting a final-component symlink instead of following repository-owned links across the trust boundary. Once `lstat` identifies the winning regular-file candidate, a later resolve/stat failure makes that scope temporarily unavailable instead of falling through to a lower-priority name. Prefix cancellation and dynamic tool cancellation propagate through resolution, metadata probes, and streaming reads. A provider failure after a file was loaded is treated as temporarily unavailable, not as proof that the file was deleted.
+
+## Prompt Shape
+
+Baseline instructions are request-only user-role prefix messages framed with the familiar system-reminder pattern:
+
+```md
+<system-reminder>
+The following workspace instructions may be relevant to your work. Use them as guidance when applicable. More specific instructions take precedence over broader ones. They do not override system, developer, or direct user instructions.
+
+Instructions from: ~/.dsh/AGENTS.md
+
+...
+
+Instructions from: AGENTS.md
+
+...
+</system-reminder>
+```
+
+Newly reached scopes use a durable raw `context/message`:
+
+```md
+<system-reminder>
+Additional instructions from: packages/app/AGENTS.md
+
+These instructions apply to work under `packages/app`. Use them as guidance when relevant; more specific instructions take precedence. They do not override system, developer, or direct user instructions.
+
+...
+</system-reminder>
+```
+
+A same-file edit starts with `Updated instructions from: <path>` and says to use the new content instead of the previously loaded content. A candidate switch additionally names the old path. When no candidate remains, the message is `Instructions removed: <path>` followed by `The previously loaded instructions from this file no longer apply.` Literal `</system-reminder>` text inside an instruction file is escaped so file content cannot close the plugin-owned frame.
+
+The core `context/message` envelope is disabled for these messages because the plugin already owns the complete `<system-reminder>` framing. This is caller-selected with `envelope: 'raw'`; ordinary injected context still receives the canonical `<context source="...">` envelope.
+
+## State And Refresh
+
+Model-visible text contains no hidden state markers. Each dynamic context event instead carries JSON metadata with a versioned list of `{ action, scope, path, previousPath?, digest? }` changes. On every relevant tool touch, the plugin reconstructs loaded state from its visible session events and overlays a short in-memory pending window for context present on the immutable top-level `tools/result` but not yet appended by the loop. A matching durable `context/message` confirms the pending transition. If the owning `step/end` arrives before a matching context reaches the log, the plugin clears the pending transition and its version fast path so the next successful touch can load it again. Nested Code Mode results stage pending changes under the outer execution token for same-run duplicate suppression; the outer result rolls that state back and recommits only contexts that survived outer policy.
+
+An unchanged path and SHA-1 content digest is not injected again. A per-session, per-scope metadata cache stores only `{ path, version, digest }`: when the provider's opaque `FsVersion` and the effective visible state both match, reconciliation skips the content read; a changed version triggers a bounded read and SHA-1 confirmation before any model-visible update. Resume works because SHA-1 state is persisted in the session log, while an empty in-memory version cache merely causes one confirming read. Compaction re-arms a scope after its context event leaves the visible surface even when the cached version is unchanged. A removal is a tombstone, so a later candidate reappearance is loaded again. Only model-visible changes actually rendered within the byte budget enter metadata, pending state, and the version cache; an omitted change remains eligible for a later touch, while a same-digest version refresh updates metadata only.
+
+The frozen baseline itself is not rewritten mid-instance. Its initial path/digest map is retained as comparison state; the next successful filesystem touch appends any baseline replacement or removal. A resumed loop recomposes the current baseline and also reconciles still-visible dynamic scopes during prefix composition. There is no file watcher, so an on-disk change becomes visible at the next successful `read`, `write`, or `edit` touch, or when a resumed loop composes its prefix.
+
+## Configuration
+
+```ts
+export interface Config {
+  dshHome?: string
+  projectRootMarkers?: string[]
+  maxBytes: number
+  maxSourceBytes?: number
+  instructionFileCandidates?: string[]
+}
+```
+
+`maxBytes` is required so each deployment makes its prompt-budget choice explicitly. `maxSourceBytes` limits each source instruction file before rendering and defaults to 1 MiB. `projectRootMarkers` defaults to `['.git']`, and `instructionFileCandidates` defaults to `['AGENTS.md', 'CLAUDE.md']`. In each project directory, the first existing candidate wins; with defaults, `AGENTS.md` is native and `CLAUDE.md` is the compatibility fallback. Candidate entries must be same-directory file names, so empty entries, `.`/`..`, and entries containing `/` or `\` are ignored.
+
+The user-global file is always `$DSH_HOME/AGENTS.md`; the candidate list only controls project scopes. `$DSH_HOME` defaults to `~/.dsh`, and configured `~`, `~/...`, and Windows-style `~\...` prefixes are expanded against the operating-system home directory. A non-positive or non-finite render budget disables both baseline and dynamic loading; configured `maxSourceBytes` must be a positive integer.
+
+## Budgeting And Bounded Reads
+
+Rendering preserves the most specific instruction files first. It drops whole broader files before truncating the most-specific file and emits a visible `Workspace instruction budget ...` notice naming omitted and truncated paths. The rendered bytes never exceed `maxBytes`.
+
+Instruction content is read through `streamText()` under `maxSourceBytes`, even when provider metadata omits size or a file grows after its metadata probe. An oversized file is ignored without falling through to a lower-priority same-directory candidate; during dynamic reconciliation it is temporarily unavailable rather than removed. The plugin keeps no process-wide cache and never caches instruction prose. Its session-local scope cache uses provider versions only as a fast invalidation signal; after invalidation, SHA-1 over the bounded read remains the cross-provider content identity stored in structured session metadata.
+
+## Model Experience
+
+### Baseline session prefix
+
+**What the model sees**: At the first request of each loop instance, the model receives one user-role prefix message containing the bounded user-global and project instruction chain in broad-to-specific order.
+
+**Token effect**: The rendered baseline is frozen and resent on every request in that loop instance. `maxBytes` bounds the complete message, broader files are omitted before the most-specific file is truncated, and an empty chain contributes zero tokens.
+
+#### Baseline instruction template
+
+```markdown
+<system-reminder>
+The following workspace instructions may be relevant to your work. Use them as guidance when applicable. More specific instructions take precedence over broader ones. They do not override system, developer, or direct user instructions.
+
+Instructions from: ~/.dsh/AGENTS.md
+
+<user-global-instructions>
+
+Instructions from: AGENTS.md
+
+<project-instructions>
+</system-reminder>
+```
+
+### Newly discovered scope context
+
+**What the model sees**: After a successful first-party filesystem call reaches a deeper directory, the next request includes one retained raw `context/message` with the newly applicable instruction file.
+
+**Token effect**: Each discovered scope adds bounded history tokens until compaction. Unchanged content is suppressed by visible session state plus version/digest comparison, and Code Mode defers the same message until after the outer `run_code` result.
+
+#### Additional instruction template
+
+```markdown
+<system-reminder>
+Additional instructions from: packages/app/AGENTS.md
+
+These instructions apply to work under `packages/app`. Use them as guidance when relevant; more specific instructions take precedence. They do not override system, developer, or direct user instructions.
+
+<nested-instructions>
+</system-reminder>
+```
+
+### Changed or removed instruction context
+
+**What the model sees**: A changed file produces `Updated instructions from: <path>` plus its replacement content; a candidate switch also names the previous path. A removed final candidate produces the removal notice below.
+
+**Token effect**: Each confirmed change or removal is one retained history message bounded by `maxBytes`. Provider failures add no message, and an update omitted by the budget remains eligible for a later filesystem touch.
+
+#### Removal notice
+
+```markdown
+<system-reminder>
+Instructions removed: packages/app/AGENTS.md
+
+The previously loaded instructions from this file no longer apply.
+</system-reminder>
+```
+
+## Known Limitations and Deferred Work
+
+- **Discovery follows structured fs tools, not shell navigation** — a `bash` command that changes directories does not trigger nested instruction discovery because shell syntax and per-call shell state are not a reliable filesystem seam.
+- **Refresh is touch-driven** — there is no watcher; external edits become visible on the next successful first-party `read`, `write`, or `edit`, or when a resumed loop recomposes its prefix.
+- **Candidate semantics stay intentionally small** — lowercase names, `.claude/rules/`, and `@path` imports are not interpreted; same-directory names such as `CLAUDE.local.md` require explicit `instructionFileCandidates` configuration.
+- **Instruction content is bounded, not summarized** — over-budget broad files are omitted and the most-specific file may be truncated; the plugin never asks a model to compress instruction prose.
diff --git a/packages/context/workspace-context/package.json b/packages/context/workspace-context/package.json
new file mode 100644
index 0000000000..7f704c838a
--- /dev/null
+++ b/packages/context/workspace-context/package.json
@@ -0,0 +1,51 @@
+{
+  "name": "@deepseek-ai/dsh-workspace-context",
+  "description": "Workspace context loader for AGENTS.md/CLAUDE.md instruction files",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "@deepseek-ai/dsh-agent": "^0.0.1",
+    "@deepseek-ai/dsh-fs": "^0.0.1",
+    "@deepseek-ai/dsh-llm": "^0.0.1",
+    "@deepseek-ai/dsh-paths": "^0.0.1",
+    "@deepseek-ai/dsh-session": "^0.0.1",
+    "@deepseek-ai/dsh-tools": "^0.0.1",
+    "cordis": "^4.0.0-rc.6"
+  },
+  "dependencies": {
+    "schemastery": "^3.18.0"
+  },
+  "devDependencies": {
+    "@cordisjs/plugin-loader": "workspace:^",
+    "@deepseek-ai/dsh-agent": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-fs": "workspace:^",
+    "@deepseek-ai/dsh-fs-local": "workspace:^",
+    "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-llm-deepseek": "workspace:^",
+    "@deepseek-ai/dsh-paths": "workspace:^",
+    "@deepseek-ai/dsh-session": "workspace:^",
+    "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-tool-fs": "workspace:^",
+    "@deepseek-ai/dsh-tools": "workspace:^",
+    "cordis": "^4.0.0-rc.6"
+  }
+}
diff --git a/packages/context/workspace-context/src/config.ts b/packages/context/workspace-context/src/config.ts
new file mode 100644
index 0000000000..c4bdd663c7
--- /dev/null
+++ b/packages/context/workspace-context/src/config.ts
@@ -0,0 +1,82 @@
+/**
+ * Configuration normalization for workspace instruction discovery and rendering.
+ *
+ * @module @deepseek-ai/dsh-workspace-context/config
+ */
+
+import z from 'schemastery'
+import { resolveDshHome } from '@deepseek-ai/dsh-paths'
+
+const DEFAULT_PROJECT_ROOT_MARKERS = ['.git'] as const
+const DEFAULT_INSTRUCTION_FILE_CANDIDATES = ['AGENTS.md', 'CLAUDE.md'] as const
+const DEFAULT_MAX_SOURCE_BYTES = 1_048_576
+const RESERVED_PATH_SEGMENTS = new Set(['', '.', '..'])
+
+/** User-facing workspace instruction loader configuration. */
+export interface Config {
+  /** Harness home containing the fixed user-global `AGENTS.md`; defaults to `$DSH_HOME` or `~/.dsh`. */
+  dshHome?: string
+  /** Directory entries that identify the project root while walking upward from the session cwd. */
+  projectRootMarkers?: string[]
+  /** UTF-8 byte cap for one rendered baseline or dynamic batch; non-positive or non-finite disables loading. */
+  maxBytes: number
+  /** Maximum UTF-8 bytes read from one instruction file; larger files are ignored. */
+  maxSourceBytes?: number
+  /** Ordered same-directory project candidates; the first existing regular file wins in each scope. */
+  instructionFileCandidates?: string[]
+}
+
+export const Config: z<Config> = z.object({
+  dshHome: z.string(),
+  projectRootMarkers: z.array(z.string()).default([...DEFAULT_PROJECT_ROOT_MARKERS]),
+  maxBytes: z.number().required(),
+  maxSourceBytes: z.number().step(1).min(1).default(DEFAULT_MAX_SOURCE_BYTES),
+  instructionFileCandidates: z.array(z.string()).default([...DEFAULT_INSTRUCTION_FILE_CANDIDATES]),
+})
+
+/** Normalized instruction discovery configuration. */
+export interface ResolvedDiscoveryConfig {
+  dshHome: string
+  projectRootMarkers: string[]
+  instructionFileCandidates: string[]
+}
+
+/** Normalized configuration used by discovery and reconciliation. */
+export interface ResolvedConfig extends ResolvedDiscoveryConfig {
+  maxBytes: number
+  maxSourceBytes: number
+}
+
+/**
+ * Resolve defaults, the harness home, and valid same-directory candidates.
+ * @param config - user-facing plugin configuration.
+ * @returns normalized runtime configuration.
+ */
+export function resolveConfig(config: Config): ResolvedConfig {
+  return {
+    ...resolveDiscoveryConfig(config),
+    maxBytes: config.maxBytes,
+    maxSourceBytes: config.maxSourceBytes ?? DEFAULT_MAX_SOURCE_BYTES,
+  }
+}
+
+/**
+ * Resolve the subset of configuration used before instruction content is rendered.
+ * @param config - optional discovery controls.
+ * @returns normalized home, root markers, and instruction candidates.
+ */
+export function resolveDiscoveryConfig(
+  config: Pick<Config, 'dshHome' | 'projectRootMarkers' | 'instructionFileCandidates'>,
+): ResolvedDiscoveryConfig {
+  return {
+    dshHome: resolveDshHome(config.dshHome),
+    projectRootMarkers: config.projectRootMarkers ?? [...DEFAULT_PROJECT_ROOT_MARKERS],
+    instructionFileCandidates: resolveInstructionFileCandidates(config.instructionFileCandidates),
+  }
+}
+
+function resolveInstructionFileCandidates(candidates: string[] | undefined): string[] {
+  return (candidates ?? [...DEFAULT_INSTRUCTION_FILE_CANDIDATES]).filter(candidate => (
+    !RESERVED_PATH_SEGMENTS.has(candidate) && !/[\\/]/.test(candidate)
+  ))
+}
diff --git a/packages/context/workspace-context/src/digest.ts b/packages/context/workspace-context/src/digest.ts
new file mode 100644
index 0000000000..4568371277
--- /dev/null
+++ b/packages/context/workspace-context/src/digest.ts
@@ -0,0 +1,16 @@
+/**
+ * Content identity for workspace instruction duplicate suppression.
+ *
+ * @module @deepseek-ai/dsh-workspace-context/digest
+ */
+
+import { createHash } from 'node:crypto'
+
+/**
+ * Compute the content identity used across instruction loading and session state.
+ * @param content - exact UTF-8 instruction text.
+ * @returns lowercase SHA-1 digest in hexadecimal form.
+ */
+export function instructionContentSha1(content: string): string {
+  return createHash('sha1').update(content).digest('hex')
+}
diff --git a/packages/context/workspace-context/src/files.ts b/packages/context/workspace-context/src/files.ts
new file mode 100644
index 0000000000..feb6304b4c
--- /dev/null
+++ b/packages/context/workspace-context/src/files.ts
@@ -0,0 +1,473 @@
+/**
+ * Instruction-file discovery and bounded, abort-aware provider reads.
+ *
+ * @module @deepseek-ai/dsh-workspace-context/files
+ */
+
+import { createReadStream } from 'node:fs'
+import { lstat, stat } from 'node:fs/promises'
+import { dirname, isAbsolute, join, relative, resolve } from 'node:path'
+import type { FileSystem, FsInfo, FsPathInfo, FsTarget, FsVersion } from '@deepseek-ai/dsh-fs'
+import { assertNever } from '@deepseek-ai/dsh-llm'
+import { DEFAULT_DSH_HOME_DISPLAY, defaultDshHome } from '@deepseek-ai/dsh-paths'
+import { resolveConfig, resolveDiscoveryConfig, type ResolvedConfig } from './config.ts'
+import { renderWorkspaceContext, type RenderedWorkspaceContext } from './render.ts'
+
+/** An instruction candidate identified by absolute and model-facing paths. */
+export interface InstructionFile {
+  absolutePath: string
+  displayPath: string
+}
+
+/** An instruction file whose UTF-8 content was read successfully. */
+export interface LoadedInstructionFile extends InstructionFile {
+  content: string
+  /** Provider freshness token when the file was loaded through `ctx.fs`. */
+  version?: FsVersion
+}
+
+interface DiscoveredInstructionFile extends InstructionFile {
+  target?: FsTarget
+  size?: number
+  version?: FsVersion
+}
+
+/** Provider metadata for a winning scope candidate before its content is read. */
+export interface ProbedInstructionFile extends InstructionFile {
+  target: FsTarget
+  version: FsVersion
+  size?: number
+}
+
+interface DiscoverOptions {
+  cwd: string
+  dshHome?: string
+  projectRootMarkers?: string[]
+  instructionFileCandidates?: string[]
+  signal?: AbortSignal
+}
+
+interface LoadOptions extends DiscoverOptions {
+  maxBytes: number
+  maxSourceBytes?: number
+}
+
+/** Rendered baseline plus the files that survived byte budgeting. */
+export interface RenderedInstructionSet {
+  rendered: RenderedWorkspaceContext
+  included: LoadedInstructionFile[]
+}
+
+/** Tri-state scope probe that distinguishes confirmed absence from provider failure. */
+export type ScopeInstructionProbe =
+  | { kind: 'present'; file: ProbedInstructionFile }
+  | { kind: 'absent' }
+  | { kind: 'unavailable' }
+
+interface StatFileInfo {
+  target?: FsTarget
+  size?: number
+  version?: FsVersion
+}
+
+type StatFileProbe =
+  | { kind: 'present'; info: StatFileInfo }
+  | { kind: 'absent' }
+  | { kind: 'unavailable' }
+
+function signalOptions(signal?: AbortSignal): { signal: AbortSignal } | undefined {
+  return signal === undefined ? undefined : { signal }
+}
+
+function isMissingPathError(error: unknown): boolean {
+  return error instanceof Error && 'code' in error && (error.code === 'ENOENT' || error.code === 'ENOTDIR')
+}
+
+async function nodeStatFile(path: string, signal?: AbortSignal): Promise<StatFileProbe> {
+  try {
+    signal?.throwIfAborted()
+    const info = await lstat(path)
+    signal?.throwIfAborted()
+    if (!info.isFile()) return { kind: 'absent' }
+    return { kind: 'present', info: { size: info.size } }
+  } catch (error: unknown) {
+    signal?.throwIfAborted()
+    return isMissingPathError(error) ? { kind: 'absent' } : { kind: 'unavailable' }
+  }
+}
+
+async function fsStatFile(
+  path: string,
+  fileSystem: FileSystem,
+  signal?: AbortSignal,
+): Promise<StatFileProbe> {
+  // TODO(instruction-symlink-race): replace this lstat -> resolve -> read
+  // protocol, including probeScopeInstruction below, with a provider-owned
+  // atomic no-follow read so the final component cannot change after validation.
+  let pathInfo: FsPathInfo | undefined
+  try {
+    pathInfo = await fileSystem.lstat(path, undefined, signal)
+    signal?.throwIfAborted()
+  } catch {
+    signal?.throwIfAborted()
+    return { kind: 'unavailable' }
+  }
+  if (pathInfo?.type !== 'file') return { kind: 'absent' }
+
+  try {
+    const target = await fileSystem.resolve(path, signalOptions(signal))
+    signal?.throwIfAborted()
+    const info = await fileSystem.stat(target, signal)
+    signal?.throwIfAborted()
+    if (info?.type !== 'file') return { kind: 'unavailable' }
+    return {
+      kind: 'present',
+      info: { target, version: info.version, ...info.size === undefined ? {} : { size: info.size } },
+    }
+  } catch {
+    signal?.throwIfAborted()
+    return { kind: 'unavailable' }
+  }
+}
+
+async function statFile(
+  path: string,
+  fileSystem?: FileSystem,
+  signal?: AbortSignal,
+): Promise<StatFileProbe> {
+  return fileSystem === undefined ? nodeStatFile(path, signal) : fsStatFile(path, fileSystem, signal)
+}
+
+async function existsAsMarker(path: string, fileSystem?: FileSystem, signal?: AbortSignal): Promise<boolean> {
+  if (fileSystem !== undefined) {
+    try {
+      const target = await fileSystem.resolve(path, signalOptions(signal))
+      return await fileSystem.stat(target, signal) !== undefined
+    } catch {
+      signal?.throwIfAborted()
+      // TODO(root-marker-unavailable): preserve provider failure separately from
+      // absence and stop discovery; continuing upward can cross into an ancestor project.
+      return false
+    }
+  }
+  try {
+    signal?.throwIfAborted()
+    await stat(path)
+    signal?.throwIfAborted()
+    return true
+  } catch {
+    signal?.throwIfAborted()
+    return false
+  }
+}
+
+/**
+ * Walk upward to the first directory containing a configured root marker.
+ * @param cwd - absolute session working directory where the walk begins.
+ * @param markers - child names that identify a project root.
+ * @param fileSystem - optional provider used instead of host filesystem probes.
+ * @param signal - cancellation for provider and host probes.
+ * @returns the discovered project root, or `cwd` when no marker exists.
+ */
+export async function findProjectRoot(
+  cwd: string,
+  markers: readonly string[],
+  fileSystem?: FileSystem,
+  signal?: AbortSignal,
+): Promise<string> {
+  let current = resolve(cwd)
+  for (;;) {
+    for (const marker of markers) {
+      if (await existsAsMarker(join(current, marker), fileSystem, signal)) return current
+    }
+    const parent = dirname(current)
+    if (parent === current) return resolve(cwd)
+    current = parent
+  }
+}
+
+/**
+ * Build the inclusive root-to-cwd directory chain.
+ * @param root - root directory expected to contain or equal `cwd`.
+ * @param cwd - most-specific directory in the chain.
+ * @returns directories ordered from broadest to most specific.
+ */
+export function ancestorChain(root: string, cwd: string): string[] {
+  const chain: string[] = []
+  let current = resolve(cwd)
+  const resolvedRoot = resolve(root)
+  while (current !== resolvedRoot) {
+    chain.push(current)
+    const parent = dirname(current)
+    /* v8 ignore next -- discovery always supplies cwd or an ancestor root. */
+    if (parent === current) break
+    current = parent
+  }
+  chain.push(resolvedRoot)
+  return chain.reverse()
+}
+
+/**
+ * Find descendant directories crossed between a cwd and a touched file.
+ * @param root - session cwd that bounds nested discovery.
+ * @param touchedPath - absolute path or path relative to `root`.
+ * @returns descendant directories from shallowest through the touched file's parent.
+ */
+export function descendantDirsBetween(root: string, touchedPath: string): string[] {
+  const resolvedRoot = resolve(root)
+  const targetPath = isAbsolute(touchedPath) ? resolve(touchedPath) : resolve(resolvedRoot, touchedPath)
+  const targetDir = dirname(targetPath)
+  const rel = relative(resolvedRoot, targetDir)
+  if (rel.length === 0 || rel.startsWith('..') || isAbsolute(rel)) return []
+  return ancestorChain(resolvedRoot, targetDir).slice(1)
+}
+
+/**
+ * Convert an absolute instruction path to its project-root-relative display form.
+ * @param root - project root used as the display base.
+ * @param path - absolute path to display.
+ * @returns the root-relative path.
+ */
+export function relativeDisplay(root: string, path: string): string {
+  return relative(root, path)
+}
+
+async function firstExistingInstructionFile(
+  dir: string,
+  root: string,
+  instructionFileCandidates: readonly string[],
+  fileSystem?: FileSystem,
+  signal?: AbortSignal,
+): Promise<DiscoveredInstructionFile | undefined> {
+  for (const candidate of instructionFileCandidates) {
+    const path = join(dir, candidate)
+    const probe = await statFile(path, fileSystem, signal)
+    switch (probe.kind) {
+      case 'present':
+        return {
+          absolutePath: path,
+          displayPath: relativeDisplay(root, path),
+          ...probe.info,
+        }
+      case 'absent':
+        continue
+      case 'unavailable':
+        return undefined
+      /* v8 ignore next 2 -- StatFileProbe is closed; this arm only makes adding a kind a compile error. */
+      default:
+        return assertNever(probe, 'StatFileProbe')
+    }
+  }
+  return undefined
+}
+
+async function discoverInstructionFiles(
+  options: DiscoverOptions,
+  fileSystem?: FileSystem,
+): Promise<DiscoveredInstructionFile[]> {
+  const config = resolveDiscoveryConfig(options)
+  const files: DiscoveredInstructionFile[] = []
+  const seen = new Set<string>()
+  const addFile = (file: DiscoveredInstructionFile): void => {
+    if (seen.has(file.absolutePath)) return
+    seen.add(file.absolutePath)
+    files.push(file)
+  }
+
+  const userGlobal = join(config.dshHome, 'AGENTS.md')
+  const userGlobalProbe = await statFile(userGlobal, fileSystem, options.signal)
+  switch (userGlobalProbe.kind) {
+    case 'present':
+      addFile({
+        absolutePath: userGlobal,
+        displayPath: userGlobalDisplayPath(config.dshHome),
+        ...userGlobalProbe.info,
+      })
+      break
+    case 'absent':
+    case 'unavailable':
+      break
+    /* v8 ignore next 2 -- StatFileProbe is closed; this arm only makes adding a kind a compile error. */
+    default:
+      assertNever(userGlobalProbe, 'StatFileProbe')
+  }
+
+  const cwd = resolve(options.cwd)
+  const projectRoot = await findProjectRoot(cwd, config.projectRootMarkers, fileSystem, options.signal)
+  for (const dir of ancestorChain(projectRoot, cwd)) {
+    const file = await firstExistingInstructionFile(dir, projectRoot, config.instructionFileCandidates, fileSystem, options.signal)
+    if (file !== undefined) addFile(file)
+  }
+  return files
+}
+
+/**
+ * Discover host-visible user-global and root-to-cwd instruction candidates.
+ * @param options - cwd, home, root marker, and candidate configuration.
+ * @returns de-duplicated instruction paths in model precedence order.
+ */
+export async function discoverBaselineInstructionFiles(options: DiscoverOptions): Promise<InstructionFile[]> {
+  return (await discoverInstructionFiles(options)).map(({ absolutePath, displayPath }) => ({ absolutePath, displayPath }))
+}
+
+async function* nodeTextChunks(path: string, signal?: AbortSignal): AsyncIterable<string> {
+  const stream = createReadStream(path, { encoding: 'utf8', signal })
+  for await (const chunk of stream) yield String(chunk)
+}
+
+async function readBounded(
+  file: DiscoveredInstructionFile,
+  maxSourceBytes: number,
+  fileSystem?: FileSystem,
+  signal?: AbortSignal,
+): Promise<string | undefined> {
+  // TODO(total-instruction-read-bound): enforce an aggregate source budget
+  // across a complete baseline or reconciliation batch; the render budget is
+  // applied only after every accepted file has been read under this per-file cap.
+  signal?.throwIfAborted()
+  if (file.size !== undefined && file.size > maxSourceBytes) return undefined
+  try {
+    const chunks = fileSystem === undefined || file.target === undefined
+      ? nodeTextChunks(file.absolutePath, signal)
+      : await fileSystem.streamText(file.target, signal)
+    const parts: string[] = []
+    let bytes = 0
+    for await (const chunk of chunks) {
+      signal?.throwIfAborted()
+      bytes += Buffer.byteLength(chunk, 'utf8')
+      if (bytes > maxSourceBytes) return undefined
+      parts.push(chunk)
+    }
+    signal?.throwIfAborted()
+    return parts.join('')
+  } catch {
+    signal?.throwIfAborted()
+    // A file may disappear or become unreadable after its metadata probe.
+    return undefined
+  }
+}
+
+/**
+ * Discover, read, and render the baseline instruction chain.
+ * @param options - discovery, source-size, byte-budget, and cancellation configuration.
+ * @param fileSystem - optional provider used instead of host filesystem reads.
+ * @returns rendered baseline context, or undefined when nothing can be loaded.
+ */
+export async function loadBaselineInstructions(
+  options: LoadOptions,
+  fileSystem?: FileSystem,
+): Promise<RenderedWorkspaceContext | undefined> {
+  return (await loadBaselineInstructionSet(options, fileSystem))?.rendered
+}
+
+/**
+ * Load a baseline together with the files retained after rendering.
+ * @param options - discovery, source-size, byte-budget, and cancellation configuration.
+ * @param fileSystem - optional provider used instead of host filesystem reads.
+ * @returns rendered context and retained files, or undefined when empty or disabled.
+ */
+export async function loadBaselineInstructionSet(
+  options: LoadOptions,
+  fileSystem?: FileSystem,
+): Promise<RenderedInstructionSet | undefined> {
+  const config = resolveConfig(options)
+  if (config.maxBytes <= 0 || !Number.isFinite(config.maxBytes)) return undefined
+  if (config.maxSourceBytes <= 0 || !Number.isFinite(config.maxSourceBytes)) return undefined
+  const discovered = await discoverInstructionFiles(options, fileSystem)
+  const loaded: LoadedInstructionFile[] = []
+  for (const file of discovered) {
+    const content = await readBounded(file, config.maxSourceBytes, fileSystem, options.signal)
+    if (content !== undefined) {
+      loaded.push({
+        absolutePath: file.absolutePath,
+        displayPath: file.displayPath,
+        content,
+        ...file.version === undefined ? {} : { version: file.version },
+      })
+    }
+  }
+  if (loaded.length === 0) return undefined
+  const rendered = renderWorkspaceContext(loaded, { maxBytes: config.maxBytes })
+  const omitted = new Set(rendered.omitted.map(file => file.absolutePath))
+  return { rendered, included: loaded.filter(file => !omitted.has(file.absolutePath)) }
+}
+
+/**
+ * Probe the current first-winning instruction candidate for one logical scope.
+ * @param scope - `user-global`, `.`, or a project-relative directory.
+ * @param projectRoot - project root used to resolve and display project scopes.
+ * @param resolved - normalized plugin configuration.
+ * @param fileSystem - provider used for no-follow probing.
+ * @param signal - cancellation for provider probes.
+ * @returns present metadata, confirmed absence, or temporary unavailability.
+ */
+export async function probeScopeInstruction(
+  scope: string,
+  projectRoot: string,
+  resolved: ResolvedConfig,
+  fileSystem: FileSystem,
+  signal?: AbortSignal,
+): Promise<ScopeInstructionProbe> {
+  const dir = scope === 'user-global'
+    ? resolved.dshHome
+    : scope === '.' ? projectRoot : join(projectRoot, scope)
+  const candidates = scope === 'user-global' ? ['AGENTS.md'] : resolved.instructionFileCandidates
+  for (const candidate of candidates) {
+    const absolutePath = join(dir, candidate)
+    let pathInfo: FsPathInfo | undefined
+    try {
+      pathInfo = await fileSystem.lstat(absolutePath, undefined, signal)
+    } catch {
+      signal?.throwIfAborted()
+      return { kind: 'unavailable' }
+    }
+    if (pathInfo === undefined || pathInfo.type !== 'file') continue
+    let target: FsTarget
+    let info: FsInfo | undefined
+    try {
+      target = await fileSystem.resolve(absolutePath, signalOptions(signal))
+      info = await fileSystem.stat(target, signal)
+    } catch {
+      signal?.throwIfAborted()
+      return { kind: 'unavailable' }
+    }
+    if (info?.type !== 'file') return { kind: 'unavailable' }
+    const file: ProbedInstructionFile = {
+      absolutePath,
+      displayPath: scope === 'user-global' ? userGlobalDisplayPath(resolved.dshHome) : relativeDisplay(projectRoot, absolutePath),
+      target,
+      version: info.version,
+      ...info.size === undefined ? {} : { size: info.size },
+    }
+    return { kind: 'present', file }
+  }
+  return { kind: 'absent' }
+}
+
+/**
+ * Read one already-probed scope candidate under the configured source cap.
+ * @param file - winning provider candidate and its metadata snapshot.
+ * @param maxSourceBytes - maximum UTF-8 bytes accepted from the source.
+ * @param fileSystem - provider used for the streaming read.
+ * @param signal - cancellation for provider streaming.
+ * @returns loaded content with the probed version, or undefined when unavailable.
+ */
+export async function readScopeInstruction(
+  file: ProbedInstructionFile,
+  maxSourceBytes: number,
+  fileSystem: FileSystem,
+  signal?: AbortSignal,
+): Promise<LoadedInstructionFile | undefined> {
+  const content = await readBounded(file, maxSourceBytes, fileSystem, signal)
+  if (content === undefined) return undefined
+  return {
+    absolutePath: file.absolutePath,
+    displayPath: file.displayPath,
+    content,
+    version: file.version,
+  }
+}
+
+function userGlobalDisplayPath(dshHome: string): string {
+  return dshHome === resolve(defaultDshHome()) ? `${DEFAULT_DSH_HOME_DISPLAY}/AGENTS.md` : '$DSH_HOME/AGENTS.md'
+}
diff --git a/packages/context/workspace-context/src/index.ts b/packages/context/workspace-context/src/index.ts
new file mode 100644
index 0000000000..83bbfa9d26
--- /dev/null
+++ b/packages/context/workspace-context/src/index.ts
@@ -0,0 +1,173 @@
+/**
+ * Workspace instruction loader for AGENTS.md-compatible files.
+ *
+ * Baseline instructions are frozen into `agent/session-prefix`; successful fs
+ * tool touches reconcile nested, changed, and removed instructions through
+ * `tools/post-execute` for the next model request. Plugin lifecycle reads use
+ * the optional `ctx.fs` provider, so providerless products mount it as a no-op.
+ *
+ * @module @deepseek-ai/dsh-workspace-context
+ */
+
+import type { Context } from 'cordis'
+import type { Agent } from '@deepseek-ai/dsh-agent'
+import type { Message } from '@deepseek-ai/dsh-llm'
+import type { PostToolDecision, ToolExecution, ToolExecutionResult, ToolExecutionToken } from '@deepseek-ai/dsh-tools'
+import { Config, resolveConfig, type ResolvedConfig } from './config.ts'
+import { loadBaselineInstructionSet } from './files.ts'
+import {
+  applyInstructionVersionUpdates,
+  baselineInstructionState,
+  commitPendingInstructionContexts,
+  dynamicInstructionContext,
+  name,
+  observeInstructionSessionEvent,
+  reconcileInstructionContext,
+  retainedInstructionVersionUpdates,
+  rollbackPendingInstructionChanges,
+  workspaceContextMessage,
+  type InstructionVersionCache,
+  type InstructionVersionUpdate,
+  type PendingInstructionChange,
+} from './state.ts'
+import type { WorkspaceInstructionChange } from './render.ts'
+
+export { Config, name }
+export {
+  discoverBaselineInstructionFiles,
+  loadBaselineInstructions,
+} from './files.ts'
+export type {
+  InstructionFile,
+  LoadedInstructionFile,
+} from './files.ts'
+export { renderWorkspaceContext } from './render.ts'
+export type { RenderedWorkspaceContext, TruncatedInstruction } from './render.ts'
+
+export function apply(ctx: Context, config: Config): void {
+  const resolved: ResolvedConfig = resolveConfig(config)
+  const pendingNestedChanges = new WeakMap<object, Map<string, PendingInstructionChange>>()
+  const baselineInstructionStates = new WeakMap<object, Map<string, WorkspaceInstructionChange>>()
+  const instructionVersions: InstructionVersionCache = new WeakMap()
+  const pendingVersionUpdates = new Map<ToolExecutionToken, InstructionVersionUpdate[]>()
+  const pendingByParent = new Map<ToolExecutionToken, {
+    agent: Agent
+    changes: WorkspaceInstructionChange[]
+    versionUpdates: InstructionVersionUpdate[]
+  }>()
+
+  ctx.on('session/event', (session, event) => {
+    observeInstructionSessionEvent(session, event, pendingNestedChanges, instructionVersions)
+  })
+
+  ctx.on('agent/session-prefix', async (agent: Agent, _prefix, signal, next): Promise<Message[]> => {
+    const rest = await next()
+    if (resolved.maxBytes <= 0 || !Number.isFinite(resolved.maxBytes)) return rest
+    const fileSystem = ctx.get('fs')
+    if (fileSystem === undefined) return rest
+    /* v8 ignore next -- normal agents carry an absolute session cwd. */
+    const cwd = agent.session.header.cwd ?? process.cwd()
+    const instructions = await loadBaselineInstructionSet({
+      cwd,
+      dshHome: resolved.dshHome,
+      projectRootMarkers: resolved.projectRootMarkers,
+      maxBytes: resolved.maxBytes,
+      maxSourceBytes: resolved.maxSourceBytes,
+      instructionFileCandidates: resolved.instructionFileCandidates,
+      signal,
+    }, fileSystem)
+    const baseline = baselineInstructionState(instructions?.included ?? [])
+    baselineInstructionStates.set(agent.session, baseline.changes)
+    instructionVersions.set(agent.session, baseline.versions)
+
+    const update = await reconcileInstructionContext(
+      agent,
+      resolved,
+      pendingNestedChanges,
+      baselineInstructionStates,
+      instructionVersions,
+      fileSystem,
+      { includeBaselineScopes: false, signal },
+    )
+    if (update !== undefined) {
+      agent.inject(update.context.content, {
+        source: update.context.source,
+        envelope: update.context.envelope,
+        meta: update.context.meta,
+      })
+      applyInstructionVersionUpdates(agent.session, update.versionUpdates, instructionVersions)
+    }
+    if (instructions === undefined || instructions.rendered.text.length === 0) return rest
+    return [workspaceContextMessage(instructions.rendered.text), ...rest]
+  })
+
+  ctx.on('tools/post-execute', async (
+    exec: ToolExecution,
+    result: ToolExecutionResult,
+    next,
+  ): Promise<PostToolDecision> => {
+    const downstream = await next()
+    // A downstream listener/policy blocked this call: the registry turns it
+    // into a final `isError` result, so treat it like a failed fs touch and
+    // load nothing. Reconciling here would surface workspace instructions from
+    // a call the pipeline rejected, violating the "successful fs tool touches"
+    // contract, and would advance the nested/baseline tracking state off a
+    // touch that never really happened.
+    if (downstream.kind === 'block') return downstream
+    const fileSystem = ctx.get('fs')
+    if (fileSystem === undefined) return downstream
+    const update = await dynamicInstructionContext(
+      exec.agent,
+      exec,
+      result,
+      resolved,
+      pendingNestedChanges,
+      baselineInstructionStates,
+      instructionVersions,
+      fileSystem,
+    )
+    if (update === undefined) return downstream
+    pendingVersionUpdates.set(exec.token, update.versionUpdates)
+    return {
+      kind: 'accept',
+      ...downstream.content !== undefined ? { content: downstream.content } : {},
+      additionalContexts: [update.context, ...downstream.additionalContexts ?? []],
+    }
+  })
+
+  ctx.on('tools/result', (exec: ToolExecution, result: ToolExecutionResult) => {
+    const ownVersionUpdates = pendingVersionUpdates.get(exec.token) ?? []
+    pendingVersionUpdates.delete(exec.token)
+    if (exec.parent !== undefined) {
+      if (exec.agent === undefined) return
+      // Child contexts participate in duplicate suppression within one composite
+      // run, but remain provisional until the parent reaches its final policy.
+      const changes = commitPendingInstructionContexts(exec.agent, result.additionalContexts, pendingNestedChanges)
+      if (changes.length === 0) return
+      const versionUpdates = retainedInstructionVersionUpdates(ownVersionUpdates, changes)
+      const staged = pendingByParent.get(exec.parent)
+      if (staged === undefined) pendingByParent.set(exec.parent, { agent: exec.agent, changes, versionUpdates })
+      else {
+        staged.changes.push(...changes)
+        staged.versionUpdates.push(...versionUpdates)
+      }
+      return
+    }
+
+    // The parent result is authoritative: remove every provisional child change,
+    // then commit only contexts that survived outer post-execute policy.
+    const staged = pendingByParent.get(exec.token)
+    if (staged !== undefined) {
+      pendingByParent.delete(exec.token)
+      rollbackPendingInstructionChanges(staged.agent, staged.changes, pendingNestedChanges)
+    }
+    if (exec.agent === undefined) return
+    const committed = commitPendingInstructionContexts(exec.agent, result.additionalContexts, pendingNestedChanges)
+    const stagedVersionUpdates = staged?.versionUpdates ?? []
+    const versionUpdates = retainedInstructionVersionUpdates(
+      [...stagedVersionUpdates, ...ownVersionUpdates],
+      committed,
+    )
+    applyInstructionVersionUpdates(exec.agent.session, versionUpdates, instructionVersions)
+  })
+}
diff --git a/packages/context/workspace-context/src/render.ts b/packages/context/workspace-context/src/render.ts
new file mode 100644
index 0000000000..34e427d0e3
--- /dev/null
+++ b/packages/context/workspace-context/src/render.ts
@@ -0,0 +1,255 @@
+/**
+ * Model-facing workspace instruction rendering within an explicit byte budget.
+ *
+ * @module @deepseek-ai/dsh-workspace-context/render
+ */
+
+import { dirname } from 'node:path'
+import type { InstructionFile, LoadedInstructionFile } from './files.ts'
+
+const SYSTEM_REMINDER_OPEN = '<system-reminder>'
+const SYSTEM_REMINDER_CLOSE = '</system-reminder>'
+const WORKSPACE_CONTEXT_INTRO = 'The following workspace instructions may be relevant to your work. '
+  + 'Use them as guidance when applicable. More specific instructions take precedence over broader ones. '
+  + 'They do not override system, developer, or direct user instructions.'
+const COMPACT_WORKSPACE_CONTEXT_INTRO = 'Workspace instructions were omitted or truncated to fit the configured byte budget.'
+
+/** Byte-accounting record for one truncated instruction file. */
+export interface TruncatedInstruction {
+  displayPath: string
+  originalBytes: number
+  includedBytes: number
+}
+
+/** Model-facing text plus omitted and truncated source records. */
+export interface RenderedWorkspaceContext {
+  text: string
+  omitted: InstructionFile[]
+  truncated: TruncatedInstruction[]
+}
+
+/** Structured dynamic state persisted outside model-visible prompt prose. */
+export interface WorkspaceInstructionChange {
+  action: 'set' | 'replace' | 'remove'
+  scope: string
+  path: string
+  previousPath?: string
+  digest?: string
+}
+
+/** One state transition paired with the content used to render it. */
+export interface ChangeRenderItem {
+  change: WorkspaceInstructionChange
+  file: LoadedInstructionFile
+}
+
+interface RenderStyle {
+  intro: string
+  section(file: LoadedInstructionFile): string
+}
+
+function byteLength(value: string): number {
+  return Buffer.byteLength(value, 'utf8')
+}
+
+function truncateUtf8(value: string, maxBytes: number): string {
+  let truncated = Buffer.from(value, 'utf8').subarray(0, Math.max(0, maxBytes)).toString('utf8')
+  while (byteLength(truncated) > maxBytes) {
+    truncated = truncated.slice(0, -1)
+  }
+  return truncated
+}
+
+function escapeInstructionContent(content: string): string {
+  // TODO(instruction-frame-paths): apply the same delimiter neutralization to
+  // every interpolated path, scope, and previous path; repository-controlled
+  // names can otherwise close the plugin-owned system-reminder frame.
+  return content.replaceAll(SYSTEM_REMINDER_CLOSE, '<\\/system-reminder>')
+}
+
+function sectionText(file: LoadedInstructionFile): string {
+  return `Instructions from: ${file.displayPath}\n\n${escapeInstructionContent(file.content)}`
+}
+
+/**
+ * Derive the logical instruction scope from a model-facing path.
+ * @param displayPath - project-relative or user-global instruction path.
+ * @returns `user-global`, `.`, or the containing project-relative directory.
+ */
+export function scopeForDisplayPath(displayPath: string): string {
+  if (displayPath === '~/.dsh/AGENTS.md' || displayPath === '$DSH_HOME/AGENTS.md') return 'user-global'
+  return dirname(displayPath)
+}
+
+function additionalSectionText(file: LoadedInstructionFile): string {
+  const scope = scopeForDisplayPath(file.displayPath)
+  return [
+    `Additional instructions from: ${file.displayPath}`,
+    '',
+    `These instructions apply to work under \`${scope}\`. Use them as guidance when relevant; more specific instructions take precedence. They do not override system, developer, or direct user instructions.`,
+    '',
+    escapeInstructionContent(file.content),
+  ].join('\n')
+}
+
+const BASELINE_RENDER_STYLE: RenderStyle = { intro: WORKSPACE_CONTEXT_INTRO, section: sectionText }
+
+function changedSectionText(item: ChangeRenderItem): string {
+  const { change, file } = item
+  if (change.action === 'set') return additionalSectionText(file)
+  if (change.action === 'remove') {
+    return `Instructions removed: ${change.path}\n\nThe previously loaded instructions from this file no longer apply.`
+  }
+  const description = change.previousPath === undefined
+    ? 'This file changed after it was loaded. Use the following content instead of the previously loaded instructions from this file.'
+    : `The instructions previously loaded from \`${change.previousPath}\` no longer apply. Use the following content for \`${change.scope}\` instead.`
+  return [
+    `Updated instructions from: ${change.path}`,
+    '',
+    description,
+    '',
+    escapeInstructionContent(file.content),
+  ].join('\n')
+}
+
+/**
+ * Render one reconciliation batch and retain only transitions that fit.
+ * @param items - ordered state transitions and current file contents.
+ * @param maxBytes - maximum UTF-8 bytes allowed in the rendered batch.
+ * @returns bounded prompt text and the transitions actually represented by it.
+ */
+export function renderInstructionChanges(
+  items: ChangeRenderItem[],
+  maxBytes: number,
+): { text: string; changes: WorkspaceInstructionChange[] } {
+  const byAbsolutePath = new Map(items.map(item => [item.file.absolutePath, item]))
+  const style: RenderStyle = {
+    intro: '',
+    section(file) {
+      const item = byAbsolutePath.get(file.absolutePath)
+      /* v8 ignore next -- the renderer receives exactly the files used to construct this map. */
+      return item === undefined ? '' : changedSectionText({ ...item, file })
+    },
+  }
+  const rendered = renderInstructionContext(items.map(item => item.file), maxBytes, style)
+  const omitted = new Set(rendered.omitted.map(file => file.absolutePath))
+  return {
+    text: rendered.text,
+    // TODO(rendered-change-proof): retain a transition only when its semantic
+    // notice survived rendering; a tiny compact budget can currently return
+    // unrelated notice text while still committing the full state transition.
+    changes: items.filter(item => !omitted.has(item.file.absolutePath)).map(item => item.change),
+  }
+}
+
+function markerText(maxBytes: number, omitted: InstructionFile[], truncated: TruncatedInstruction[]): string {
+  if (omitted.length === 0 && truncated.length === 0) return ''
+  const parts: string[] = []
+  if (omitted.length > 0) {
+    parts.push(`omitted ${omitted.map(file => file.displayPath).join(', ')}`)
+  }
+  if (truncated.length > 0) {
+    parts.push(`truncated ${truncated.map(item => `${item.displayPath} from ${item.originalBytes} to ${item.includedBytes} bytes`).join(', ')}`)
+  }
+  return `Workspace instruction budget ${maxBytes} bytes: ${parts.join('; ')}`
+}
+
+function buildInstructionText(
+  files: LoadedInstructionFile[],
+  maxBytes: number,
+  omitted: InstructionFile[],
+  truncated: TruncatedInstruction[],
+  style: RenderStyle,
+): string {
+  const marker = markerText(maxBytes, omitted, truncated)
+  const body = [marker, style.intro, ...files.map(file => style.section(file))].filter(block => block.length > 0)
+  return [SYSTEM_REMINDER_OPEN, body.join('\n\n'), SYSTEM_REMINDER_CLOSE].join('\n')
+}
+
+function withTruncatedContent(file: LoadedInstructionFile, includedBytes: number): LoadedInstructionFile {
+  return { ...file, content: truncateUtf8(file.content, includedBytes) }
+}
+
+function truncateToFit(
+  file: LoadedInstructionFile,
+  includedFiles: LoadedInstructionFile[],
+  maxBytes: number,
+  omitted: InstructionFile[],
+  style: RenderStyle,
+): LoadedInstructionFile {
+  const originalBytes = byteLength(file.content)
+  let low = 0
+  let high = originalBytes
+  let best = withTruncatedContent(file, 0)
+  while (low <= high) {
+    const mid = Math.floor((low + high) / 2)
+    const candidate = withTruncatedContent(file, mid)
+    const truncated = [{ displayPath: file.displayPath, originalBytes, includedBytes: byteLength(candidate.content) }]
+    const text = buildInstructionText([...includedFiles, candidate], maxBytes, omitted, truncated, style)
+    if (byteLength(text) <= maxBytes) {
+      best = candidate
+      low = mid + 1
+    } else {
+      high = mid - 1
+    }
+  }
+  return best
+}
+
+function renderInstructionContext(
+  files: LoadedInstructionFile[],
+  maxBytes: number,
+  style: RenderStyle,
+): RenderedWorkspaceContext {
+  if (maxBytes <= 0 || !Number.isFinite(maxBytes)) return { text: '', omitted: files, truncated: [] }
+
+  const fullText = buildInstructionText(files, maxBytes, [], [], style)
+  if (byteLength(fullText) <= maxBytes) return { text: fullText, omitted: [], truncated: [] }
+
+  for (let start = 1; start < files.length; start += 1) {
+    const included = files.slice(start)
+    const omitted = files.slice(0, start).map(file => ({ absolutePath: file.absolutePath, displayPath: file.displayPath }))
+    const suffixText = buildInstructionText(included, maxBytes, omitted, [], style)
+    if (byteLength(suffixText) <= maxBytes) return { text: suffixText, omitted, truncated: [] }
+  }
+
+  const mostSpecific = files.at(-1)
+  /* v8 ignore next -- callers only reach this after a non-empty fullText was built. */
+  if (mostSpecific === undefined) return { text: '', omitted: [], truncated: [] }
+  const omitted = files.slice(0, -1).map(file => ({ absolutePath: file.absolutePath, displayPath: file.displayPath }))
+
+  for (const candidateStyle of [style, { ...style, intro: COMPACT_WORKSPACE_CONTEXT_INTRO }]) {
+    const truncatedFile = truncateToFit(mostSpecific, [], maxBytes, omitted, candidateStyle)
+    const truncated = [{
+      displayPath: mostSpecific.displayPath,
+      originalBytes: byteLength(mostSpecific.content),
+      includedBytes: byteLength(truncatedFile.content),
+    }]
+    const text = buildInstructionText([truncatedFile], maxBytes, omitted, truncated, candidateStyle)
+    if (byteLength(text) <= maxBytes) return { text, omitted, truncated }
+  }
+
+  const truncated = [{
+    displayPath: mostSpecific.displayPath,
+    originalBytes: byteLength(mostSpecific.content),
+    includedBytes: 0,
+  }]
+  const compactNotice = markerText(maxBytes, omitted, truncated)
+  const compactWithHeading = [compactNotice, style.section(withTruncatedContent(mostSpecific, 0))].join('\n\n')
+  if (byteLength(compactWithHeading) <= maxBytes) return { text: compactWithHeading, omitted, truncated }
+  const text = byteLength(compactNotice) <= maxBytes ? compactNotice : truncateUtf8(compactNotice, maxBytes)
+  return { text, omitted, truncated }
+}
+
+/**
+ * Render the baseline instruction chain with deterministic precedence budgeting.
+ * @param files - loaded files ordered from broadest to most specific.
+ * @param options - required rendering byte budget.
+ * @returns bounded baseline prompt text and budget diagnostics.
+ */
+export function renderWorkspaceContext(
+  files: LoadedInstructionFile[],
+  options: { maxBytes: number },
+): RenderedWorkspaceContext {
+  return renderInstructionContext(files, options.maxBytes, BASELINE_RENDER_STYLE)
+}
diff --git a/packages/context/workspace-context/src/state.ts b/packages/context/workspace-context/src/state.ts
new file mode 100644
index 0000000000..4e8ee7de56
--- /dev/null
+++ b/packages/context/workspace-context/src/state.ts
@@ -0,0 +1,507 @@
+/**
+ * Session-visible workspace instruction state and dynamic reconciliation.
+ *
+ * @module @deepseek-ai/dsh-workspace-context/state
+ */
+
+import type { Agent, HookContext } from '@deepseek-ai/dsh-agent'
+import type { Message } from '@deepseek-ai/dsh-llm'
+import type { JsonValue, Session, SessionEvent } from '@deepseek-ai/dsh-session'
+import type { FileSystem, FsVersion } from '@deepseek-ai/dsh-fs'
+import type { ToolExecution, ToolExecutionResult } from '@deepseek-ai/dsh-tools'
+import type { ResolvedConfig } from './config.ts'
+import { instructionContentSha1 } from './digest.ts'
+import {
+  ancestorChain,
+  descendantDirsBetween,
+  findProjectRoot,
+  probeScopeInstruction,
+  readScopeInstruction,
+  relativeDisplay,
+  type LoadedInstructionFile,
+} from './files.ts'
+import {
+  renderInstructionChanges,
+  scopeForDisplayPath,
+  type ChangeRenderItem,
+  type WorkspaceInstructionChange,
+} from './render.ts'
+
+export const name = 'workspace-context'
+
+const PLUGIN_SOURCE = { kind: 'plugin', plugin: name } as const
+const FILE_TOUCH_TOOL_NAMES = new Set(['read', 'write', 'edit'])
+
+/** Dynamic state waiting for the loop to append its returned context event. */
+export interface PendingInstructionChange {
+  change: WorkspaceInstructionChange
+  afterSeq: number
+  step?: { turn: number; step: number }
+}
+
+/** Per-scope metadata cache; instruction prose is deliberately not retained. */
+export interface InstructionVersionState {
+  path: string
+  version: FsVersion
+  digest: string
+}
+
+/** Session-isolated fast-path state keyed by logical instruction scope. */
+export type InstructionVersionCache = WeakMap<Session, Map<string, InstructionVersionState>>
+
+/** A cache transition coupled to the model-visible change that authorizes it. */
+export interface InstructionVersionUpdate {
+  change: WorkspaceInstructionChange
+  state?: InstructionVersionState
+}
+
+/** Rendered reconciliation plus cache transitions awaiting final policy. */
+export interface ReconciledInstructionContext {
+  context: WorkspaceHookContext
+  versionUpdates: InstructionVersionUpdate[]
+}
+
+/** Plugin-owned raw context with required replay metadata. */
+export interface WorkspaceHookContext extends HookContext {
+  envelope: 'raw'
+  meta: JsonValue
+}
+
+function workspaceContextHook(text: string, changes: WorkspaceInstructionChange[]): WorkspaceHookContext {
+  const serializedChanges: JsonValue[] = changes.map(change => ({
+    action: change.action,
+    scope: change.scope,
+    path: change.path,
+    ...change.previousPath !== undefined ? { previousPath: change.previousPath } : {},
+    ...change.digest !== undefined ? { digest: change.digest } : {},
+  }))
+  const meta: JsonValue = { kind: 'workspace-instructions', version: 1, changes: serializedChanges }
+  return { content: [{ type: 'text', text }], source: PLUGIN_SOURCE, envelope: 'raw', meta }
+}
+
+/**
+ * Build the request-prefix message for a rendered baseline.
+ * @param text - complete plugin-owned system-reminder text.
+ * @returns a user-role prefix message.
+ */
+export function workspaceContextMessage(text: string): Message {
+  return { role: 'user', content: [{ type: 'text', text }] }
+}
+
+function filePathFromExecution(exec: ToolExecution): string | undefined {
+  if (!FILE_TOUCH_TOOL_NAMES.has(exec.name)) return undefined
+  if (typeof exec.arguments !== 'object' || exec.arguments === null) return undefined
+  if (!('file_path' in exec.arguments) || typeof exec.arguments.file_path !== 'string') return undefined
+  const filePath = exec.arguments.file_path.trim()
+  return filePath.length > 0 ? filePath : undefined
+}
+
+function isWorkspaceContextSource(source: unknown): source is typeof PLUGIN_SOURCE {
+  return typeof source === 'object' && source !== null
+    && 'kind' in source && source.kind === 'plugin'
+    && 'plugin' in source && source.plugin === name
+}
+
+function isRecord(value: JsonValue | undefined): value is { [key: string]: JsonValue } {
+  return typeof value === 'object' && value !== null && !Array.isArray(value)
+}
+
+function workspaceInstructionChanges(meta: JsonValue | undefined): WorkspaceInstructionChange[] {
+  if (!isRecord(meta) || meta.kind !== 'workspace-instructions' || meta.version !== 1 || !Array.isArray(meta.changes)) return []
+  const changes: WorkspaceInstructionChange[] = []
+  for (const value of meta.changes) {
+    if (!isRecord(value)) continue
+    if (value.action !== 'set' && value.action !== 'replace' && value.action !== 'remove') continue
+    if (typeof value.scope !== 'string' || typeof value.path !== 'string') continue
+    if (value.previousPath !== undefined && typeof value.previousPath !== 'string') continue
+    if (value.digest !== undefined && typeof value.digest !== 'string') continue
+    changes.push({
+      action: value.action,
+      scope: value.scope,
+      path: value.path,
+      ...value.previousPath !== undefined ? { previousPath: value.previousPath } : {},
+      ...value.digest !== undefined ? { digest: value.digest } : {},
+    })
+  }
+  return changes
+}
+
+function sameInstructionChange(a: WorkspaceInstructionChange, b: WorkspaceInstructionChange): boolean {
+  return a.action === b.action
+    && a.scope === b.scope
+    && a.path === b.path
+    && a.previousPath === b.previousPath
+    && a.digest === b.digest
+}
+
+function visibleInstructionChanges(
+  agent: Agent,
+  pending: Map<string, PendingInstructionChange>,
+): Map<string, WorkspaceInstructionChange> {
+  const visibleSeqs = new Set(agent.session.surface.nodes)
+  const visible = new Map<string, WorkspaceInstructionChange>()
+  for (const [seq, event] of agent.session.events.entries()) {
+    if (event.type !== 'context/message' || !isWorkspaceContextSource(event.data.source)) continue
+    const changes = workspaceInstructionChanges(event.data.meta)
+    for (const change of changes) {
+      const waiting = pending.get(change.scope)
+      if (waiting !== undefined && seq >= waiting.afterSeq && sameInstructionChange(waiting.change, change)) {
+        pending.delete(change.scope)
+      }
+      if (visibleSeqs.has(seq)) visible.set(change.scope, change)
+    }
+  }
+  for (const { change } of pending.values()) visible.set(change.scope, change)
+  return visible
+}
+
+/**
+ * Convert retained baseline files into comparison and metadata-cache state.
+ * @param files - baseline files that survived rendering.
+ * @returns latest baseline changes and provider versions keyed by logical scope.
+ */
+export function baselineInstructionState(files: LoadedInstructionFile[]): {
+  changes: Map<string, WorkspaceInstructionChange>
+  versions: Map<string, InstructionVersionState>
+} {
+  const changes = new Map<string, WorkspaceInstructionChange>()
+  const versions = new Map<string, InstructionVersionState>()
+  for (const file of files) {
+    const digest = instructionContentSha1(file.content)
+    const change: WorkspaceInstructionChange = {
+      action: 'set',
+      scope: scopeForDisplayPath(file.displayPath),
+      path: file.displayPath,
+      digest,
+    }
+    changes.set(change.scope, change)
+    if (file.version !== undefined) {
+      versions.set(change.scope, { path: file.displayPath, version: file.version, digest })
+    }
+  }
+  return { changes, versions }
+}
+
+function versionStatesFor(session: Session, cache: InstructionVersionCache): Map<string, InstructionVersionState> {
+  let states = cache.get(session)
+  if (states === undefined) {
+    states = new Map()
+    cache.set(session, states)
+  }
+  return states
+}
+
+/**
+ * Keep only cache updates whose model-visible changes survived final policy.
+ * @param updates - proposed updates from one or more reconciliations.
+ * @param committedChanges - transitions retained on the authoritative result.
+ * @returns updates authorized by an exact retained transition.
+ */
+export function retainedInstructionVersionUpdates(
+  updates: readonly InstructionVersionUpdate[],
+  committedChanges: readonly WorkspaceInstructionChange[],
+): InstructionVersionUpdate[] {
+  return updates.filter(update => committedChanges.some(change => sameInstructionChange(update.change, change)))
+}
+
+/**
+ * Apply authorized metadata-cache transitions without retaining instruction prose.
+ * @param session - owning session.
+ * @param updates - ordered set/delete transitions.
+ * @param cache - session-isolated metadata cache.
+ */
+export function applyInstructionVersionUpdates(
+  session: Session,
+  updates: readonly InstructionVersionUpdate[],
+  cache: InstructionVersionCache,
+): void {
+  if (updates.length === 0) return
+  const states = versionStatesFor(session, cache)
+  for (const update of updates) {
+    if (update.state === undefined) states.delete(update.change.scope)
+    else states.set(update.change.scope, update.state)
+  }
+  if (states.size === 0) cache.delete(session)
+}
+
+function pendingChangesFor(
+  session: object,
+  pendingBySession: WeakMap<object, Map<string, PendingInstructionChange>>,
+): Map<string, PendingInstructionChange> {
+  let pending = pendingBySession.get(session)
+  if (pending === undefined) {
+    pending = new Map()
+    pendingBySession.set(session, pending)
+  }
+  return pending
+}
+
+function openStep(session: Session): { turn: number; step: number } | undefined {
+  const boundary = session.events.findLast(event => event.type === 'step/start' || event.type === 'step/end')
+  return boundary?.type === 'step/start' ? boundary.data : undefined
+}
+
+function invalidateInstructionVersions(
+  session: Session,
+  scopes: readonly string[],
+  cache: InstructionVersionCache,
+): void {
+  const states = cache.get(session)
+  if (states === undefined) return
+  for (const scope of scopes) states.delete(scope)
+  if (states.size === 0) cache.delete(session)
+}
+
+/**
+ * Settle provisional tool-result state against durable session events.
+ * A matching context event confirms the transition. If its owning step closes
+ * first, both duplicate suppression and the metadata fast path are re-armed for
+ * the next successful touch.
+ * @param session - session whose append-only log emitted `event`.
+ * @param event - newly committed session event.
+ * @param pendingBySession - provisional transitions awaiting log confirmation.
+ * @param versionCache - metadata fast path coupled to those transitions.
+ */
+export function observeInstructionSessionEvent(
+  session: Session,
+  event: SessionEvent,
+  pendingBySession: WeakMap<object, Map<string, PendingInstructionChange>>,
+  versionCache: InstructionVersionCache,
+): void {
+  const pending = pendingBySession.get(session)
+  if (pending === undefined) return
+
+  switch (event.type) {
+    case 'context/message': {
+      if (!isWorkspaceContextSource(event.data.source)) return
+      for (const change of workspaceInstructionChanges(event.data.meta)) {
+        const waiting = pending.get(change.scope)
+        if (waiting !== undefined && event.seq >= waiting.afterSeq && sameInstructionChange(waiting.change, change)) {
+          pending.delete(change.scope)
+        }
+      }
+      if (pending.size === 0) pendingBySession.delete(session)
+      return
+    }
+    case 'step/end': {
+      const discardedScopes: string[] = []
+      for (const [scope, waiting] of pending) {
+        const step = waiting.step
+        if (step === undefined || step.turn !== event.data.turn || step.step !== event.data.step) continue
+        pending.delete(scope)
+        discardedScopes.push(scope)
+      }
+      if (pending.size === 0) pendingBySession.delete(session)
+      invalidateInstructionVersions(session, discardedScopes, versionCache)
+      return
+    }
+    default:
+      // SessionEventMap is merge-extensible; unrelated events do not settle workspace state.
+      return
+  }
+}
+
+/**
+ * Commit only workspace contexts that survived the complete tool pipeline.
+ * The observe-only `tools/result` notification calls this before the loop can
+ * append the returned contexts, closing that short pending window without
+ * trusting an intermediate post-execute decision.
+ * @param agent - session that will receive the final result contexts.
+ * @param contexts - immutable contexts on the authoritative top-level result.
+ * @param pendingBySession - per-session pending transition maps.
+ * @returns transitions committed into the short pending window.
+ */
+export function commitPendingInstructionContexts(
+  agent: Agent,
+  contexts: readonly HookContext[] | undefined,
+  pendingBySession: WeakMap<object, Map<string, PendingInstructionChange>>,
+): WorkspaceInstructionChange[] {
+  const committed: WorkspaceInstructionChange[] = []
+  const step = openStep(agent.session)
+  for (const context of contexts ?? []) {
+    if (!isWorkspaceContextSource(context.source)) continue
+    const changes = workspaceInstructionChanges(context.meta)
+    if (changes.length === 0) continue
+    const pending = pendingChangesFor(agent.session, pendingBySession)
+    for (const change of changes) {
+      pending.set(change.scope, {
+        change,
+        afterSeq: agent.session.seq,
+        ...step === undefined ? {} : { step },
+      })
+      committed.push(change)
+    }
+  }
+  return committed
+}
+
+/**
+ * Roll back parent-token state when an enclosing tool result discards deferred
+ * contexts. A newer transition for the same scope is left intact.
+ * @param agent - session whose pending state was staged.
+ * @param changes - exact staged transitions to remove when still current.
+ * @param pendingBySession - per-session pending transition maps.
+ */
+export function rollbackPendingInstructionChanges(
+  agent: Agent,
+  changes: readonly WorkspaceInstructionChange[],
+  pendingBySession: WeakMap<object, Map<string, PendingInstructionChange>>,
+): void {
+  const pending = pendingBySession.get(agent.session)
+  if (pending === undefined) return
+  for (const change of changes) {
+    const current = pending.get(change.scope)
+    if (current !== undefined && sameInstructionChange(current.change, change)) pending.delete(change.scope)
+  }
+  if (pending.size === 0) pendingBySession.delete(agent.session)
+}
+
+function relativeScope(projectRoot: string, dir: string): string {
+  const scope = relativeDisplay(projectRoot, dir)
+  return scope.length === 0 ? '.' : scope
+}
+
+/**
+ * Compare visible/pending state with provider-visible files and render transitions.
+ * @param agent - session owner whose visible surface supplies durable state.
+ * @param resolved - normalized plugin configuration.
+ * @param pendingBySession - short pending window before returned context is logged.
+ * @param baselineBySession - frozen baseline comparison state per session.
+ * @param versionCache - per-session scope metadata used to skip unchanged reads.
+ * @param fileSystem - provider used for current file probes.
+ * @param options - touched path and whether baseline scopes should be checked.
+ * @returns rendered context plus deferred cache updates, or undefined when unchanged/unavailable.
+ */
+export async function reconcileInstructionContext(
+  agent: Agent,
+  resolved: ResolvedConfig,
+  pendingBySession: WeakMap<object, Map<string, PendingInstructionChange>>,
+  baselineBySession: WeakMap<object, Map<string, WorkspaceInstructionChange>>,
+  versionCache: InstructionVersionCache,
+  fileSystem: FileSystem,
+  options: { touchedPath?: string; includeBaselineScopes: boolean; signal?: AbortSignal },
+): Promise<ReconciledInstructionContext | undefined> {
+  const session = agent.session
+  const pending = pendingChangesFor(session, pendingBySession)
+  const visible = visibleInstructionChanges(agent, pending)
+  const effective = new Map(baselineBySession.get(session) ?? [])
+  for (const [scope, change] of visible) effective.set(scope, change)
+  /* v8 ignore next -- normal agents carry an absolute session cwd. */
+  const cwd = session.header.cwd ?? process.cwd()
+  // TODO(frozen-project-root): retain the baseline root for the loop instance;
+  // recomputing it after marker edits reinterprets the existing relative scope keys.
+  const projectRoot = await findProjectRoot(cwd, resolved.projectRootMarkers, fileSystem, options.signal)
+  const scopes = new Set<string>()
+  if (options.includeBaselineScopes) {
+    scopes.add('user-global')
+    for (const dir of ancestorChain(projectRoot, cwd)) scopes.add(relativeScope(projectRoot, dir))
+  }
+  for (const scope of effective.keys()) scopes.add(scope)
+  if (options.touchedPath !== undefined) {
+    for (const dir of descendantDirsBetween(cwd, options.touchedPath)) scopes.add(relativeScope(projectRoot, dir))
+  }
+
+  const versions = versionStatesFor(session, versionCache)
+  const seenAbsolutePaths = new Set<string>()
+  const items: ChangeRenderItem[] = []
+  const versionUpdates: InstructionVersionUpdate[] = []
+  for (const scope of scopes) {
+    const previous = effective.get(scope)
+    const probe = await probeScopeInstruction(scope, projectRoot, resolved, fileSystem, options.signal)
+    if (probe.kind === 'unavailable') continue
+    if (probe.kind === 'absent') {
+      if (previous === undefined || previous.action === 'remove') {
+        versions.delete(scope)
+        continue
+      }
+      const change: WorkspaceInstructionChange = { action: 'remove', scope, path: previous.path }
+      items.push({
+        change,
+        file: { absolutePath: `removed:${scope}`, displayPath: previous.path, content: '' },
+      })
+      versionUpdates.push({ change })
+      continue
+    }
+    const { file: probedFile } = probe
+    if (seenAbsolutePaths.has(probedFile.absolutePath)) continue
+    seenAbsolutePaths.add(probedFile.absolutePath)
+    const cached = versions.get(scope)
+    if (
+      cached !== undefined
+      && cached.path === probedFile.displayPath
+      && cached.version === probedFile.version
+      && previous !== undefined
+      && previous.action !== 'remove'
+      && previous.path === cached.path
+      && previous.digest === cached.digest
+    ) continue
+
+    const file = await readScopeInstruction(probedFile, resolved.maxSourceBytes, fileSystem, options.signal)
+    if (file === undefined) continue
+    const currentDigest = instructionContentSha1(file.content)
+    const nextVersion: InstructionVersionState = {
+      path: file.displayPath,
+      version: probedFile.version,
+      digest: currentDigest,
+    }
+    if (previous !== undefined && previous.action !== 'remove' && previous.path === file.displayPath && previous.digest === currentDigest) {
+      versions.set(scope, nextVersion)
+      continue
+    }
+    const action = previous === undefined || previous.action === 'remove' ? 'set' : 'replace'
+    const previousPath = action === 'replace' && previous !== undefined && previous.path !== file.displayPath
+      ? previous.path
+      : undefined
+    const change: WorkspaceInstructionChange = {
+      action,
+      scope,
+      path: file.displayPath,
+      ...previousPath === undefined ? {} : { previousPath },
+      digest: currentDigest,
+    }
+    items.push({ change, file })
+    versionUpdates.push({ change, state: nextVersion })
+  }
+  if (items.length === 0) return undefined
+  const rendered = renderInstructionChanges(items, resolved.maxBytes)
+  if (rendered.text.length === 0 || rendered.changes.length === 0) return undefined
+  return {
+    context: workspaceContextHook(rendered.text, rendered.changes),
+    versionUpdates: retainedInstructionVersionUpdates(versionUpdates, rendered.changes),
+  }
+}
+
+/**
+ * Validate a successful structured file touch and reconcile its applicable scopes.
+ * @param agent - optional agent attached to the tool execution.
+ * @param exec - completed tool execution descriptor.
+ * @param result - original tool result before post-execute decisions.
+ * @param resolved - normalized plugin configuration.
+ * @param pendingNestedChanges - per-session pending transition maps.
+ * @param baselineInstructionStates - retained baseline comparison state.
+ * @param versionCache - per-session scope metadata used to skip unchanged reads.
+ * @param fileSystem - provider used for current file probes.
+ * @returns rendered context plus deferred cache updates, or undefined for irrelevant/failed/unchanged calls.
+ */
+export async function dynamicInstructionContext(
+  agent: Agent | undefined,
+  exec: ToolExecution,
+  result: ToolExecutionResult,
+  resolved: ResolvedConfig,
+  pendingNestedChanges: WeakMap<object, Map<string, PendingInstructionChange>>,
+  baselineInstructionStates: WeakMap<object, Map<string, WorkspaceInstructionChange>>,
+  versionCache: InstructionVersionCache,
+  fileSystem: FileSystem,
+): Promise<ReconciledInstructionContext | undefined> {
+  if (agent === undefined || result.isError) return undefined
+  const touchedPath = filePathFromExecution(exec)
+  if (touchedPath === undefined) return undefined
+  return reconcileInstructionContext(
+    agent, resolved, pendingNestedChanges, baselineInstructionStates, versionCache, fileSystem,
+    {
+      touchedPath,
+      includeBaselineScopes: baselineInstructionStates.has(agent.session),
+      ...exec.signal === undefined ? {} : { signal: exec.signal },
+    },
+  )
+}
diff --git a/packages/context/workspace-context/tests/workspace-context.e2e.ts b/packages/context/workspace-context/tests/workspace-context.e2e.ts
new file mode 100644
index 0000000000..eab9603fa5
--- /dev/null
+++ b/packages/context/workspace-context/tests/workspace-context.e2e.ts
@@ -0,0 +1,124 @@
+import { mkdtemp, mkdir, rm, writeFile } from 'node:fs/promises'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { afterEach, describe, expect, it } from 'vitest'
+import { Context } from 'cordis'
+import LlmService from '@deepseek-ai/dsh-llm'
+import SessionStore, { SessionId } from '@deepseek-ai/dsh-session'
+import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import ToolRegistry from '@deepseek-ai/dsh-tools'
+import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import type { Agent } from '@deepseek-ai/dsh-agent'
+import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import * as LlmDeepSeek from '@deepseek-ai/dsh-llm-deepseek'
+import * as WorkspaceContext from '@deepseek-ai/dsh-workspace-context'
+import LocalFileSystem from '@deepseek-ai/dsh-fs-local'
+import * as ToolFs from '@deepseek-ai/dsh-tool-fs'
+import type { SessionEvent } from '@deepseek-ai/dsh-session'
+
+const PROBE = 'banana-271828'
+const NESTED_PROBE = 'papaya-314159'
+const UPDATED_PROBE = 'guava-161803'
+
+let ctx: Context | undefined
+let workdir: string | undefined
+
+afterEach(async () => {
+  await ctx?.fiber.dispose()
+  ctx = undefined
+  if (workdir !== undefined) await rm(workdir, { recursive: true, force: true })
+  workdir = undefined
+})
+
+async function harness(): Promise<{ ctx: Context; agent: Agent }> {
+  workdir = await mkdtemp(join(tmpdir(), 'dsh-workspace-context-e2e-'))
+  await mkdir(join(workdir, '.git'), { recursive: true })
+  await writeFile(join(workdir, 'AGENTS.md'), `If the user asks for the workspace context handshake, reply with exactly this string and nothing else: ${PROBE}.\n`)
+  ctx = new Context()
+  await ctx.plugin(LlmService)
+  await ctx.plugin(SessionStore)
+  await ctx.plugin(SystemPrompt, { persona: 'Answer the user exactly and concisely.' })
+  await ctx.plugin(ToolRegistry)
+  await ctx.plugin(AgentRegistry)
+  await ctx.plugin(LocalFileSystem, { cwd: '/' })
+  await ctx.plugin(ToolFs)
+  await ctx.plugin(WorkspaceContext, { maxBytes: 65536 })
+  await ctx.plugin(AgentLoop, { agents: [] })
+  await ctx.plugin(LlmDeepSeek, { models: [{ id: 'deepseek-v4-flash' }] })
+  const handle = await ctx.agents.create({
+    agentId: AgentId('workspace-context-e2e'),
+    sessionId: SessionId('workspace-context-e2e-session'),
+    meta: { cwd: workdir },
+    agentOptions: { provider: 'deepseek', model: 'deepseek-v4-flash' },
+  })
+  return { ctx, agent: handle.agent }
+}
+
+function waitForIdle(ctx: Context, agent: Agent): Promise<void> {
+  return new Promise((resolve) => {
+    const dispose = ctx.on('agent/status', (subject, status) => {
+      if (subject === agent && status === 'idle') {
+        dispose()
+        resolve()
+      }
+    })
+  })
+}
+
+function finalText(events: SessionEvent[]): string {
+  const message = events.findLast(event => event.type === 'assistant/message')
+  if (message?.type !== 'assistant/message') return ''
+  return message.data.content
+    .filter(block => block.type === 'text')
+    .map(block => block.text)
+    .join('')
+}
+
+describe.skipIf(!process.env.DEEPSEEK_API_KEY)('workspace context e2e: real model sees AGENTS.md baseline', () => {
+  it('obeys a probe instruction loaded from the workspace', async () => {
+    const live = await harness()
+
+    live.agent.send([{ type: 'text', text: 'Workspace context handshake?' }])
+    await waitForIdle(live.ctx, live.agent)
+
+    expect(finalText([...live.agent.session.events])).toContain(PROBE)
+  }, 120_000)
+
+  it('loads a nested AGENTS.md after the real read tool touches a descendant file', async () => {
+    const live = await harness()
+    await mkdir(join(workdir!, 'pkg/deep'), { recursive: true })
+    await writeFile(join(workdir!, 'pkg/AGENTS.md'), `If the user asks for the nested instruction handshake, reply with exactly this string and nothing else: ${NESTED_PROBE}.\n`)
+    await writeFile(join(workdir!, 'pkg/deep/file.txt'), 'This file exists only to trigger nested workspace instructions.\n')
+
+    live.agent.send([{ type: 'text', text: 'Use the read tool to inspect pkg/deep/file.txt. After reading it, answer: nested instruction handshake?' }])
+    await waitForIdle(live.ctx, live.agent)
+
+    expect(finalText([...live.agent.session.events])).toContain(NESTED_PROBE)
+  }, 120_000)
+
+  it('appends changed baseline instructions after a real file-tool touch without rewriting the frozen prefix', async () => {
+    const live = await harness()
+    await writeFile(join(workdir!, 'trigger.txt'), 'This file triggers workspace instruction reconciliation.\n')
+    live.agent.send([{ type: 'text', text: 'Workspace context handshake?' }])
+    await waitForIdle(live.ctx, live.agent)
+    await writeFile(join(workdir!, 'AGENTS.md'), `The old workspace handshake no longer applies. If the user asks for the updated workspace context handshake, reply with exactly this string and nothing else: ${UPDATED_PROBE}.\n`)
+
+    live.agent.send([{ type: 'text', text: 'You must use the read tool to inspect trigger.txt. After reading it, answer: updated workspace context handshake?' }])
+    await waitForIdle(live.ctx, live.agent)
+
+    const events = [...live.agent.session.events]
+    const update = events.find(event => event.type === 'context/message'
+      && typeof event.data.meta === 'object'
+      && event.data.meta !== null
+      && !Array.isArray(event.data.meta)
+      && event.data.meta.kind === 'workspace-instructions')
+    expect(update?.type === 'context/message' && update.data.meta).toMatchObject({
+      changes: [{ action: 'replace', scope: '.', path: 'AGENTS.md' }],
+    })
+    const updateText = update?.type === 'context/message'
+      ? update.data.content.filter(block => block.type === 'text').map(block => block.text).join('')
+      : ''
+    expect(updateText).toContain('Updated instructions from: AGENTS.md')
+    expect(finalText(events)).toContain(UPDATED_PROBE)
+  }, 120_000)
+})
diff --git a/packages/context/workspace-context/tests/workspace-context.spec.ts b/packages/context/workspace-context/tests/workspace-context.spec.ts
new file mode 100644
index 0000000000..2894937c9c
--- /dev/null
+++ b/packages/context/workspace-context/tests/workspace-context.spec.ts
@@ -0,0 +1,2890 @@
+import { chmod, mkdtemp, mkdir, rm, stat, symlink, utimes, writeFile } from 'node:fs/promises'
+import { dirname, join } from 'node:path'
+import { tmpdir } from 'node:os'
+import { describe, expect, it, vi } from 'vitest'
+import { Context } from 'cordis'
+import Loader from '@cordisjs/plugin-loader'
+import * as workspaceContext from '@deepseek-ai/dsh-workspace-context'
+import LlmService, { CallId, type Message, type StreamChunk } from '@deepseek-ai/dsh-llm'
+import SessionStore, { Session, SessionId, SESSION_FORMAT_VERSION, type SessionEvent } from '@deepseek-ai/dsh-session'
+import AgentRegistry, { AgentId, type Agent, type HookContext } from '@deepseek-ai/dsh-agent'
+import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { FileSystem, FsTargetKey, FsVersion } from '@deepseek-ai/dsh-fs'
+import type {
+  FsDirEntry,
+  FsEditOutcome,
+  FsEditRequest,
+  FsInfo,
+  FsPathInfo,
+  FsTarget,
+  FsWriteIntent,
+  FsWriteOutcome,
+} from '@deepseek-ai/dsh-fs'
+import LocalFileSystem from '@deepseek-ai/dsh-fs-local'
+import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import ToolRegistry, { defineTool } from '@deepseek-ai/dsh-tools'
+import type { ToolExecution, ToolExecutionToken } from '@deepseek-ai/dsh-tools'
+import * as ToolFs from '@deepseek-ai/dsh-tool-fs'
+import {
+  discoverBaselineInstructionFiles,
+  loadBaselineInstructions,
+  renderWorkspaceContext,
+} from '@deepseek-ai/dsh-workspace-context'
+import {
+  baselineInstructionState,
+  commitPendingInstructionContexts,
+  observeInstructionSessionEvent,
+  rollbackPendingInstructionChanges,
+  type InstructionVersionCache,
+  type PendingInstructionChange,
+} from '../src/state.ts'
+import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
+
+async function tempRepo(): Promise<string> {
+  return mkdtemp(join(tmpdir(), 'dsh-workspace-context-'))
+}
+
+async function write(path: string, content: string): Promise<void> {
+  await mkdir(join(path, '..'), { recursive: true })
+  await writeFile(path, content)
+}
+
+class RecordingFileSystem extends FileSystem {
+  entries = new Map<string, { type: FsInfo['type']; content?: string; version?: FsVersion }>()
+  lstatTypes = new Map<string, FsPathInfo['type']>()
+  throwOnStat = new Set<string>()
+  omitSizes = new Set<string>()
+  readTargets: string[] = []
+  readTextTargets: string[] = []
+  signals: AbortSignal[] = []
+
+  override async resolve(path: string, opts?: { cwd?: string; signal?: AbortSignal }): Promise<FsTarget> {
+    if (opts?.signal !== undefined) this.signals.push(opts.signal)
+    opts?.signal?.throwIfAborted()
+    const absolute = join(opts?.cwd ?? '/', path)
+    return { targetKey: FsTargetKey(absolute), displayPath: absolute }
+  }
+
+  override async stat(target: FsTarget, signal?: AbortSignal): Promise<FsInfo | undefined> {
+    if (signal !== undefined) this.signals.push(signal)
+    signal?.throwIfAborted()
+    if (this.throwOnStat.has(target.targetKey)) throw new Error(`stat failed: ${target.displayPath}`)
+    const entry = this.entries.get(target.targetKey)
+    if (entry === undefined) return undefined
+    const info: FsInfo = {
+      version: entry.version ?? FsVersion(`v:${target.targetKey}:${entry.type}:${entry.content ?? ''}`),
+      type: entry.type,
+    }
+    if (entry.content !== undefined && !this.omitSizes.has(target.targetKey)) info.size = Buffer.byteLength(entry.content, 'utf8')
+    return info
+  }
+
+  override async lstat(path: string, opts?: { cwd?: string }, signal?: AbortSignal): Promise<FsPathInfo | undefined> {
+    if (signal !== undefined) this.signals.push(signal)
+    signal?.throwIfAborted()
+    const target = await this.resolve(path, { ...opts, ...signal === undefined ? {} : { signal } })
+    const lstatType = this.lstatTypes.get(target.targetKey)
+    if (lstatType !== undefined) return { version: FsVersion(`lstat:${target.targetKey}`), type: lstatType }
+    const info = await this.stat(target, signal)
+    if (info === undefined) return undefined
+    return {
+      version: info.version,
+      type: info.type,
+      ...(info.size !== undefined ? { size: info.size } : {}),
+    }
+  }
+
+  override async readText(target: FsTarget, signal?: AbortSignal): Promise<string> {
+    if (signal !== undefined) this.signals.push(signal)
+    signal?.throwIfAborted()
+    this.readTextTargets.push(target.targetKey)
+    return this.entries.get(target.targetKey)?.content ?? ''
+  }
+
+  override async streamText(target: FsTarget, signal?: AbortSignal): Promise<AsyncIterable<string>> {
+    if (signal !== undefined) this.signals.push(signal)
+    signal?.throwIfAborted()
+    this.readTargets.push(target.targetKey)
+    const content = this.entries.get(target.targetKey)?.content ?? ''
+    return (async function* () {
+      const midpoint = Math.ceil(content.length / 2)
+      yield content.slice(0, midpoint)
+      signal?.throwIfAborted()
+      yield content.slice(midpoint)
+    })()
+  }
+
+  override async listDir(_target: FsTarget): Promise<FsDirEntry[]> {
+    return []
+  }
+
+  override async writeText(_target: FsTarget, _content: string, _expected?: FsWriteIntent): Promise<FsWriteOutcome> {
+    return { operation: 'update', version: FsVersion('unused'), before: '', after: _content }
+  }
+
+  override async editText(_target: FsTarget, _edit: FsEditRequest): Promise<FsEditOutcome> {
+    return { version: FsVersion('unused'), before: '', after: '' }
+  }
+}
+
+class BlockingReadFileSystem extends RecordingFileSystem {
+  readonly started = Promise.withResolvers<undefined>()
+
+  override async streamText(target: FsTarget, signal?: AbortSignal): Promise<AsyncIterable<string>> {
+    if (signal !== undefined) this.signals.push(signal)
+    this.readTargets.push(target.targetKey)
+    this.started.resolve(undefined)
+    return (async function* () {
+      await new Promise<void>((_resolve, reject) => {
+        const abortReason = (): Error => signal?.reason instanceof Error ? signal.reason : new Error('aborted')
+        if (signal?.aborted) { reject(abortReason()); return }
+        signal?.addEventListener('abort', () => { reject(abortReason()) }, { once: true })
+      })
+      yield 'unreachable'
+    })()
+  }
+}
+
+async function mountWorkspaceContext(ctx: Context, config: workspaceContext.Config): Promise<Awaited<ReturnType<Context['plugin']>>> {
+  await ctx.plugin(LocalFileSystem, { cwd: '/' })
+  return ctx.plugin(workspaceContext, config)
+}
+
+async function mountFileToolsAndWorkspaceContext(ctx: Context, config: workspaceContext.Config): Promise<Awaited<ReturnType<Context['plugin']>>> {
+  await ctx.plugin(SystemPrompt)
+  await ctx.plugin(ToolRegistry)
+  await ctx.plugin(LocalFileSystem, { cwd: '/' })
+  await ctx.plugin(ToolFs)
+  return ctx.plugin(workspaceContext, config)
+}
+
+function stubAgent(cwd?: string, seed: SessionEvent[] = []): Agent {
+  const id = SessionId('s1')
+  const session = new Session(id, seed, cwd === undefined ? undefined : { version: SESSION_FORMAT_VERSION, id, createdAt: 0, cwd })
+  return {
+    ctx: new Context(),
+    id: AgentId('a1'),
+    options: {},
+    session,
+    status: 'idle',
+    send() {},
+    steer() {},
+    inject(content, options) {
+      session.append('context/message', {
+        content,
+        source: options?.source ?? { kind: 'user' },
+        ...options?.envelope !== undefined ? { envelope: options.envelope } : {},
+        ...options?.meta !== undefined ? { meta: options.meta } : {},
+      }, { surfaceOp: 'append' })
+    },
+    cancel() {},
+    whenIdle: () => Promise.resolve(),
+  }
+}
+
+function stubToolExecution(input: Omit<ToolExecution, 'token'>): ToolExecution {
+  return {
+    token: Symbol('workspace-context-test-execution') as ToolExecutionToken,
+    ...input,
+  }
+}
+
+function blocksText(blocks: { type: string; text?: string }[] | undefined): string {
+  return blocks?.map(block => block.type === 'text' ? block.text ?? '' : '').join('\n') ?? ''
+}
+
+function workspaceContextOf(result: { additionalContexts?: HookContext[] }): HookContext | undefined {
+  return result.additionalContexts?.find(context =>
+    context.source.kind === 'plugin' && context.source.plugin === 'workspace-context')
+}
+
+function workspaceChangeContext(scope: string, digest: string): HookContext {
+  return {
+    content: [{ type: 'text', text: `instructions for ${scope}` }],
+    source: { kind: 'plugin', plugin: 'workspace-context' },
+    envelope: 'raw',
+    meta: {
+      kind: 'workspace-instructions',
+      version: 1,
+      changes: [{ action: 'set', scope, path: `${scope}/AGENTS.md`, digest }],
+    },
+  }
+}
+
+function appendAdditionalContexts(agent: Agent, result: { additionalContexts?: HookContext[] }): number | undefined {
+  let lastSeq: number | undefined
+  for (const context of result.additionalContexts ?? []) {
+    lastSeq = agent.session.append('context/message', {
+      content: context.content,
+      source: context.source,
+      ...context.envelope !== undefined ? { envelope: context.envelope } : {},
+      ...context.meta !== undefined ? { meta: context.meta } : {},
+    }, { surfaceOp: 'append' }).seq
+  }
+  return lastSeq
+}
+
+const composedPrefixes = new WeakMap<object, Message[]>()
+
+async function composeBaselinePrefix(ctx: Context, agent: Agent): Promise<Message[]> {
+  const empty: Message[] = []
+  const prefix = await ctx.waterfall(
+    'agent/session-prefix', agent, empty, AbortSignal.timeout(1000),
+    () => Promise.resolve(empty),
+  )
+  composedPrefixes.set(agent, prefix)
+  return prefix
+}
+
+function derivedText(agent: Agent): string {
+  return blocksText(composedPrefixes.get(agent)?.[0]?.content)
+}
+
+function expectNoDerivedMessages(agent: Agent): void {
+  expect(agent.session.deriveMessages()).toEqual([])
+  expect(composedPrefixes.get(agent) ?? []).toEqual([])
+}
+
+describe('workspace context instruction discovery', () => {
+  it('treats ENOTDIR while probing a host candidate as confirmed absence', async () => {
+    const root = await tempRepo()
+    const homeFile = join(root, 'not-a-directory')
+    try {
+      await writeFile(homeFile, 'file')
+
+      const files = await discoverBaselineInstructionFiles({ cwd: root, dshHome: homeFile })
+
+      expect(files).toEqual([])
+    } finally {
+      await rm(root, { recursive: true, force: true })
+    }
+  })
+
+  it('loads user-global first, then root-to-cwd workspace instructions using the default candidate order', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      const cwd = join(root, 'packages/app')
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(home, 'AGENTS.md'), 'global rules')
+      await write(join(root, 'AGENTS.md'), 'root agents')
+      await write(join(root, 'CLAUDE.md'), 'root claude ignored')
+      await write(join(root, 'packages/CLAUDE.md'), 'package claude')
+      await write(join(cwd, 'AGENTS.md'), 'app agents')
+
+      const files = await discoverBaselineInstructionFiles({ cwd, dshHome: home })
+
+      expect(files.map(file => file.displayPath)).toEqual([
+        '$DSH_HOME/AGENTS.md',
+        'AGENTS.md',
+        'packages/CLAUDE.md',
+        'packages/app/AGENTS.md',
+      ])
+      expect(files.map(file => file.absolutePath)).not.toContain(join(root, 'CLAUDE.md'))
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('treats a .git file as a project root marker and does not search above it', async () => {
+    const outer = await tempRepo()
+    const home = await tempRepo()
+    try {
+      const root = join(outer, 'worktree')
+      const cwd = join(root, 'src')
+      await write(join(outer, 'AGENTS.md'), 'outer must not load')
+      await write(join(root, '.git'), 'gitdir: ../.git/worktrees/worktree')
+      await write(join(root, 'AGENTS.md'), 'root')
+      await mkdir(cwd, { recursive: true })
+
+      const files = await discoverBaselineInstructionFiles({ cwd, dshHome: home })
+
+      expect(files.map(file => file.displayPath)).toEqual(['AGENTS.md'])
+    } finally {
+      await rm(outer, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('re-reads content after a same-version, same-size rewrite', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      const cwd = join(root, 'pkg')
+      await mkdir(join(root, '.git'), { recursive: true })
+      await mkdir(cwd, { recursive: true })
+
+      expect(await loadBaselineInstructions({ cwd, dshHome: home, maxBytes: 65536 })).toBeUndefined()
+
+      const leaf = join(cwd, 'AGENTS.md')
+      await write(leaf, 'first')
+      const first = await loadBaselineInstructions({ cwd, dshHome: home, maxBytes: 65536 })
+      expect(first?.text).toContain('first')
+      const again = await loadBaselineInstructions({ cwd, dshHome: home, maxBytes: 65536 })
+      expect(again?.text).toContain('first')
+
+      const before = await stat(leaf)
+      await writeFile(leaf, 'other')
+      await utimes(leaf, before.atime, before.mtime)
+      const second = await loadBaselineInstructions({ cwd, dshHome: home, maxBytes: 65536 })
+      expect(second?.text).toContain('other')
+      expect(second?.text).not.toContain('first')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('skips a file that becomes unreadable after discovery without failing the request', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      const cwd = join(root, 'pkg')
+      await mkdir(join(root, '.git'), { recursive: true })
+      await mkdir(cwd, { recursive: true })
+      const leaf = join(cwd, 'AGENTS.md')
+      await write(leaf, 'secret-ish rule')
+      await chmod(leaf, 0)
+
+      const loaded = await loadBaselineInstructions({ cwd, dshHome: home, maxBytes: 65536 })
+
+      expect(loaded).toBeUndefined()
+      await chmod(leaf, 0o600)
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('rejects symlinked instruction files instead of following repository-controlled links', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    const outside = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(outside, 'secret.txt'), 'outside secret')
+      await symlink(join(outside, 'secret.txt'), join(root, 'AGENTS.md'))
+
+      const files = await discoverBaselineInstructionFiles({ cwd: root, dshHome: home })
+      const loaded = await loadBaselineInstructions({ cwd: root, dshHome: home, maxBytes: 65536 })
+
+      expect(files).toEqual([])
+      expect(loaded).toBeUndefined()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+      await rm(outside, { recursive: true, force: true })
+    }
+  })
+
+  it('rejects symlinked instruction files through ctx.fs instead of following repository-controlled links', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    const outside = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(outside, 'secret.txt'), 'outside secret')
+      await symlink(join(outside, 'secret.txt'), join(root, 'AGENTS.md'))
+      const ctx = new Context()
+      await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expectNoDerivedMessages(agent)
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+      await rm(outside, { recursive: true, force: true })
+    }
+  })
+
+  it('disables baseline loading when the byte budget is zero', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'repo rule')
+
+      await expect(loadBaselineInstructions({ cwd: root, dshHome: home, maxBytes: 0 })).resolves.toBeUndefined()
+      await expect(loadBaselineInstructions({ cwd: root, dshHome: home, maxBytes: 65536, maxSourceBytes: 0 })).resolves.toBeUndefined()
+      await expect(loadBaselineInstructions({
+        cwd: root, dshHome: home, maxBytes: 65536, maxSourceBytes: Infinity,
+      })).resolves.toBeUndefined()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('honors configured instruction candidates that exclude CLAUDE.md', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'CLAUDE.md'), 'claude only')
+
+      const files = await discoverBaselineInstructionFiles({
+        cwd: root,
+        dshHome: home,
+        instructionFileCandidates: ['AGENTS.md'],
+      })
+
+      expect(files).toEqual([])
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('uses the configured instruction candidate order without hard-coding AGENTS.md priority', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'native rule')
+      await write(join(root, 'CLAUDE.local.md'), 'local claude rule')
+      await write(join(root, 'CLAUDE.md'), 'claude rule')
+
+      const files = await discoverBaselineInstructionFiles({
+        cwd: root,
+        dshHome: home,
+        instructionFileCandidates: ['CLAUDE.local.md', 'AGENTS.md', 'CLAUDE.md'],
+      })
+
+      expect(files.map(file => file.displayPath)).toEqual(['CLAUDE.local.md'])
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('ignores configured instruction candidates that are not same-directory file names', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'native rule')
+      await write(join(root, '.claude/CLAUDE.md'), 'nested claude rule')
+
+      const files = await discoverBaselineInstructionFiles({
+        cwd: root,
+        dshHome: home,
+        instructionFileCandidates: ['', '.', '..', '.claude/CLAUDE.md', 'nested\\CLAUDE.md', 'AGENTS.md'],
+      })
+
+      expect(files.map(file => file.displayPath)).toEqual(['AGENTS.md'])
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('defaults dshHome and uses cwd itself as root when no project marker exists', async () => {
+    const root = await tempRepo()
+    try {
+      const cwd = join(root, 'child')
+      await mkdir(cwd, { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'parent without marker')
+      await write(join(cwd, 'AGENTS.md'), 'cwd without marker')
+
+      const files = await discoverBaselineInstructionFiles({ cwd })
+
+      expect(files.map(file => file.displayPath)).toEqual(['AGENTS.md'])
+      expect(files.map(file => file.absolutePath)).toEqual([join(cwd, 'AGENTS.md')])
+    } finally {
+      await rm(root, { recursive: true, force: true })
+    }
+  })
+
+  it('honors DSH_HOME when dshHome is not configured explicitly', async () => {
+    const root = await tempRepo()
+    const envHome = await tempRepo()
+    try {
+      await write(join(envHome, 'AGENTS.md'), 'env global rule')
+      vi.stubEnv('DSH_HOME', envHome)
+
+      const files = await discoverBaselineInstructionFiles({ cwd: root })
+
+      expect(files).toEqual([{ absolutePath: join(envHome, 'AGENTS.md'), displayPath: '$DSH_HOME/AGENTS.md' }])
+    } finally {
+      vi.unstubAllEnvs()
+      await rm(root, { recursive: true, force: true })
+      await rm(envHome, { recursive: true, force: true })
+    }
+  })
+
+  it('labels the default DSH home as ~/.dsh when HOME points at the configured default', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await write(join(home, '.dsh/AGENTS.md'), 'global default rule')
+
+      vi.resetModules()
+      vi.doMock('node:os', () => ({ homedir: () => home }))
+      const isolated = await import('@deepseek-ai/dsh-workspace-context')
+      const files = await isolated.discoverBaselineInstructionFiles({ cwd: root })
+
+      expect(files.map(file => file.displayPath)).toEqual(['~/.dsh/AGENTS.md'])
+    } finally {
+      vi.doUnmock('node:os')
+      vi.resetModules()
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('expands a configured ~/.dsh home to the operating-system home directory', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await write(join(home, '.dsh/AGENTS.md'), 'global tilde rule')
+
+      vi.resetModules()
+      vi.doMock('node:os', () => ({ homedir: () => home }))
+      const isolated = await import('@deepseek-ai/dsh-workspace-context')
+      const files = await isolated.discoverBaselineInstructionFiles({ cwd: root, dshHome: '~/.dsh' })
+
+      expect(files).toEqual([{ absolutePath: join(home, '.dsh/AGENTS.md'), displayPath: '~/.dsh/AGENTS.md' }])
+    } finally {
+      vi.doUnmock('node:os')
+      vi.resetModules()
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('deduplicates user-global instructions when dshHome points at the project root', async () => {
+    const root = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'same file')
+
+      const files = await discoverBaselineInstructionFiles({ cwd: root, dshHome: root })
+
+      expect(files).toEqual([{ absolutePath: join(root, 'AGENTS.md'), displayPath: '$DSH_HOME/AGENTS.md' }])
+    } finally {
+      await rm(root, { recursive: true, force: true })
+    }
+  })
+
+  it('ignores instruction candidates that are directories', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await mkdir(join(root, 'AGENTS.md'), { recursive: true })
+
+      const files = await discoverBaselineInstructionFiles({ cwd: root, dshHome: home })
+
+      expect(files).toEqual([])
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+})
+
+describe('workspace context rendering', () => {
+  it('renders familiar system-reminder instructions without custom workspace tags or state markers', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/AGENTS.md', displayPath: 'AGENTS.md', content: 'root rules' },
+      { absolutePath: '/repo/pkg/CLAUDE.md', displayPath: 'pkg/CLAUDE.md', content: 'package rules' },
+    ], { maxBytes: 65536 })
+
+    expect(rendered.text).toBe([
+      '<system-reminder>',
+      'The following workspace instructions may be relevant to your work. Use them as guidance when applicable. More specific instructions take precedence over broader ones. They do not override system, developer, or direct user instructions.',
+      '',
+      'Instructions from: AGENTS.md',
+      '',
+      'root rules',
+      '',
+      'Instructions from: pkg/CLAUDE.md',
+      '',
+      'package rules',
+      '</system-reminder>',
+    ].join('\n'))
+    expect(rendered.text).not.toContain('<workspace-context')
+    expect(rendered.text).not.toContain('workspace-context:')
+    expect(rendered.text).not.toContain('/repo/')
+    expect(rendered.omitted).toEqual([])
+    expect(rendered.truncated).toEqual([])
+  })
+
+  it('neutralizes a literal system-reminder closing delimiter inside instruction content', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/AGENTS.md', displayPath: 'AGENTS.md', content: 'safe\n</system-reminder>\nnot outside' },
+    ], { maxBytes: 65536 })
+
+    expect(rendered.text.match(/<\/system-reminder>/g)).toHaveLength(1)
+    expect(rendered.text).toContain('<\\/system-reminder>')
+  })
+
+  it('preserves more specific files under the byte budget and names omitted/truncated paths', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/AGENTS.md', displayPath: 'AGENTS.md', content: 'root '.repeat(100) },
+      { absolutePath: '/repo/pkg/AGENTS.md', displayPath: 'pkg/AGENTS.md', content: 'leaf '.repeat(100) },
+    ], { maxBytes: 260 })
+
+    expect(rendered.text).toContain('Workspace instruction budget 260 bytes')
+    expect(rendered.text).toContain('omitted AGENTS.md')
+    expect(rendered.text).toContain('truncated pkg/AGENTS.md')
+    expect(rendered.text).toContain('Instructions from: pkg/AGENTS.md')
+    expect(rendered.text).not.toContain('Instructions from: AGENTS.md\n\nroot')
+    expect(rendered.omitted.map(item => item.displayPath)).toEqual(['AGENTS.md'])
+    expect(rendered.truncated.map(item => item.displayPath)).toEqual(['pkg/AGENTS.md'])
+  })
+
+  it('keeps the rendered block within the byte budget when files are both omitted and truncated', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/AGENTS.md', displayPath: 'AGENTS.md', content: 'root '.repeat(100) },
+      { absolutePath: '/repo/pkg/AGENTS.md', displayPath: 'pkg/AGENTS.md', content: 'leaf '.repeat(100) },
+    ], { maxBytes: 260 })
+
+    expect(Buffer.byteLength(rendered.text, 'utf8')).toBeLessThanOrEqual(260)
+    expect(rendered.text).not.toContain(':;')
+    expect(rendered.omitted.map(item => item.displayPath)).toEqual(['AGENTS.md'])
+    expect(rendered.truncated.map(item => item.displayPath)).toEqual(['pkg/AGENTS.md'])
+  })
+
+  it('drops a parent file while keeping a specific child file intact when the child fits', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/AGENTS.md', displayPath: 'AGENTS.md', content: 'root '.repeat(200) },
+      { absolutePath: '/repo/pkg/AGENTS.md', displayPath: 'pkg/AGENTS.md', content: 'leaf rule' },
+    ], { maxBytes: 700 })
+
+    expect(rendered.text).toContain('omitted AGENTS.md')
+    expect(rendered.text).toContain('Instructions from: pkg/AGENTS.md\n\nleaf rule')
+    expect(rendered.text).not.toContain('root root')
+    expect(rendered.omitted.map(item => item.displayPath)).toEqual(['AGENTS.md'])
+    expect(rendered.truncated).toEqual([])
+  })
+
+  it('keeps the longest most-specific suffix that fits under the byte budget', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/AGENTS.md', displayPath: 'AGENTS.md', content: 'root '.repeat(200) },
+      { absolutePath: '/repo/pkg/AGENTS.md', displayPath: 'pkg/AGENTS.md', content: 'package rule' },
+      { absolutePath: '/repo/pkg/app/AGENTS.md', displayPath: 'pkg/app/AGENTS.md', content: 'app rule' },
+    ], { maxBytes: 760 })
+
+    expect(rendered.text).toContain('omitted AGENTS.md')
+    expect(rendered.text).toContain('Instructions from: pkg/AGENTS.md\n\npackage rule')
+    expect(rendered.text).toContain('Instructions from: pkg/app/AGENTS.md\n\napp rule')
+    expect(rendered.text).not.toContain('root root')
+    expect(rendered.omitted.map(item => item.displayPath)).toEqual(['AGENTS.md'])
+    expect(rendered.truncated).toEqual([])
+  })
+
+  it('truncates a single oversized file to the largest content slice that fits', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/AGENTS.md', displayPath: 'AGENTS.md', content: 'x'.repeat(1000) },
+    ], { maxBytes: 700 })
+
+    expect(rendered.text).toContain('truncated AGENTS.md')
+    expect(rendered.text).toContain('Instructions from: AGENTS.md')
+    expect(rendered.truncated).toHaveLength(1)
+    expect(rendered.truncated[0]?.originalBytes).toBe(1000)
+    expect(rendered.truncated[0]!.includedBytes).toBeGreaterThan(0)
+    expect(Buffer.byteLength(rendered.text, 'utf8')).toBeLessThanOrEqual(700)
+  })
+
+  it('omits all text when the render budget is disabled', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/AGENTS.md', displayPath: 'AGENTS.md', content: 'root rules' },
+    ], { maxBytes: 0 })
+
+    expect(rendered).toEqual({
+      text: '',
+      omitted: [{ absolutePath: '/repo/AGENTS.md', displayPath: 'AGENTS.md', content: 'root rules' }],
+      truncated: [],
+    })
+  })
+
+  it('falls back to a compact truncation notice when even the empty heading cannot fit', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/pkg/AGENTS.md', displayPath: 'pkg/AGENTS.md', content: 'x'.repeat(1000) },
+    ], { maxBytes: 100 })
+
+    expect(rendered.text).toBe('Workspace instruction budget 100 bytes: truncated pkg/AGENTS.md from 1000 to 0 bytes')
+    expect(rendered.truncated).toEqual([{ displayPath: 'pkg/AGENTS.md', originalBytes: 1000, includedBytes: 0 }])
+    expect(Buffer.byteLength(rendered.text, 'utf8')).toBeLessThanOrEqual(100)
+  })
+
+  it('keeps the empty instruction heading when it fits beside the compact notice', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/pkg/AGENTS.md', displayPath: 'pkg/AGENTS.md', content: 'x'.repeat(1000) },
+    ], { maxBytes: 120 })
+
+    expect(rendered.text).toBe([
+      'Workspace instruction budget 120 bytes: truncated pkg/AGENTS.md from 1000 to 0 bytes',
+      '',
+      'Instructions from: pkg/AGENTS.md',
+      '',
+      '',
+    ].join('\n'))
+    expect(Buffer.byteLength(rendered.text, 'utf8')).toBe(120)
+  })
+
+  it('truncates the compact notice itself when the render budget is smaller than the notice', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/pkg/AGENTS.md', displayPath: 'pkg/AGENTS.md', content: 'x'.repeat(1000) },
+    ], { maxBytes: 20 })
+
+    expect(rendered.text).toBe('Workspace instructio')
+    expect(rendered.truncated).toEqual([{ displayPath: 'pkg/AGENTS.md', originalBytes: 1000, includedBytes: 0 }])
+    expect(Buffer.byteLength(rendered.text, 'utf8')).toBe(20)
+  })
+
+  it('keeps compact truncation notices within budget when a multibyte display path is cut', () => {
+    const rendered = renderWorkspaceContext([
+      { absolutePath: '/repo/路径/AGENTS.md', displayPath: '路径/AGENTS.md', content: 'x'.repeat(1000) },
+    ], { maxBytes: 51 })
+
+    expect(Buffer.byteLength(rendered.text, 'utf8')).toBeLessThanOrEqual(51)
+  })
+})
+
+describe('workspace context request injection', () => {
+  it('requires an explicit maxBytes configuration', async () => {
+    const ctx = new Context()
+
+    await expect(ctx.plugin(workspaceContext, {} as workspaceContext.Config)).rejects.toThrow(/maxBytes/)
+  })
+
+  it('mounts without requiring a filesystem provider', async () => {
+    const ctx = new Context()
+    try {
+      const outcome = await Promise.race([
+        ctx.plugin(workspaceContext, { maxBytes: 65536 }).then(() => {
+          return 'settled' as const
+        }),
+        new Promise<'pending'>((resolve) => {
+          setTimeout(() => {
+            resolve('pending')
+          }, 50)
+        }),
+      ])
+
+      expect(outcome).toBe('settled')
+    } finally {
+      await ctx.fiber.dispose()
+    }
+  })
+
+  it('does not declare fs as a static inject dependency', () => {
+    expect('inject' in workspaceContext).toBe(false)
+  })
+
+  it('does not inject baseline context when no filesystem provider is present', async () => {
+    const ctx = new Context()
+    try {
+      await ctx.plugin(workspaceContext, { maxBytes: 65536 })
+      const agent = stubAgent('/virtual/repo')
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expectNoDerivedMessages(agent)
+    } finally {
+      await ctx.fiber.dispose()
+    }
+  })
+
+  it('leaves post-execute decisions unchanged when no filesystem provider is present', async () => {
+    const ctx = new Context()
+    try {
+      await ctx.plugin(workspaceContext, { maxBytes: 65536 })
+
+      const decision = await ctx.waterfall('tools/post-execute', stubToolExecution({
+        callId: CallId('no-fs-post-execute'),
+        name: 'read',
+        arguments: { file_path: 'pkg/file.txt' },
+        agent: stubAgent('/virtual/repo'),
+      }), {
+        isError: false,
+        content: [{ type: 'text', text: 'file content' }],
+      }, async () => ({
+        kind: 'accept',
+        content: [{ type: 'text', text: 'downstream content' }],
+      }))
+
+      expect(decision).toEqual({ kind: 'accept', content: [{ type: 'text', text: 'downstream content' }] })
+    } finally {
+      await ctx.fiber.dispose()
+    }
+  })
+
+  it('does not load workspace instructions when a downstream listener blocks the tool call', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    const ctx = new Context()
+    try {
+      await ctx.plugin(SystemPrompt)
+      await ctx.plugin(ToolRegistry)
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(root, 'pkg/AGENTS.md'), { type: 'file', content: 'nested package rule' })
+      fs.entries.set(join(root, 'pkg/file.txt'), { type: 'file', content: 'hello' })
+      await ctx.plugin(ToolFs)
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const exec = stubToolExecution({
+        callId: CallId('read-blocked-post-execute'),
+        name: 'read',
+        arguments: { file_path: 'pkg/file.txt' },
+        agent,
+      })
+      const result = {
+        isError: false,
+        content: [{ type: 'text' as const, text: 'hello' }],
+      }
+
+      // A later PostToolUse-style policy blocks this otherwise-successful read.
+      const blocked = await ctx.waterfall('tools/post-execute', exec, result, async () => ({
+        kind: 'block' as const,
+        feedback: [{ type: 'text' as const, text: 'blocked by policy' }],
+      }))
+
+      expect(blocked).toEqual({
+        kind: 'block',
+        feedback: [{ type: 'text', text: 'blocked by policy' }],
+      })
+      expect(blocked.additionalContexts).toBeUndefined()
+
+      // The same read, when the downstream accepts, DOES surface the nested
+      // instructions — proving the block branch above is what suppressed them,
+      // and that the block did not consume the pending nested change.
+      const accepted = await ctx.waterfall('tools/post-execute', exec, result, async () => ({
+        kind: 'accept' as const,
+      }))
+      expect(accepted.kind).toBe('accept')
+      expect(workspaceContextOf(accepted)?.source).toEqual({ kind: 'plugin', plugin: 'workspace-context' })
+      expect(blocksText(workspaceContextOf(accepted)?.content)).toContain('nested package rule')
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('contributes baseline instructions through the frozen session prefix instead of durable history', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'repo rule')
+      const ctx = new Context()
+      await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expect(agent.session.deriveMessages()).toEqual([])
+      expect(composedPrefixes.get(agent)).toHaveLength(1)
+      expect(derivedText(agent)).toContain('<system-reminder>')
+      expect(derivedText(agent)).toContain('Instructions from: AGENTS.md')
+      expect(derivedText(agent)).toContain('repo rule')
+      expect(derivedText(agent)).not.toContain('<context source=')
+      expect(derivedText(agent)).not.toContain('<workspace-context')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('returns one baseline contribution per session-prefix composition without appending context events', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'repo rule')
+      const ctx = new Context()
+      await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const first = await composeBaselinePrefix(ctx, agent)
+      const second = await composeBaselinePrefix(ctx, agent)
+
+      expect(second).toEqual(first)
+      expect(agent.session.events.filter(event => event.type === 'context/message')).toHaveLength(0)
+      expect(derivedText(agent)).toContain('repo rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('tracks only baseline files that were actually included under the byte budget', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      const cwd = join(root, 'pkg')
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'root '.repeat(200))
+      await write(join(cwd, 'AGENTS.md'), 'package rule')
+      const ctx = new Context()
+      await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: 700 })
+      const agent = stubAgent(cwd)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expect(derivedText(agent)).toContain('omitted AGENTS.md')
+      expect(derivedText(agent)).toContain('Instructions from: pkg/AGENTS.md\n\npackage rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('places workspace instructions before later session-prefix contributors such as a skills catalog', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'repo rule')
+      const ctx = new Context()
+      await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      ctx.on('agent/session-prefix', async (_agent, _prefix, _signal, next) => {
+        const rest = await next()
+        return [{ role: 'user', content: [{ type: 'text', text: '<system-reminder>Available skills</system-reminder>' }] }, ...rest]
+      })
+
+      const prefix = await composeBaselinePrefix(ctx, stubAgent(root))
+
+      expect(prefix).toHaveLength(2)
+      expect(blocksText(prefix[0]?.content)).toContain('Instructions from: AGENTS.md')
+      expect(blocksText(prefix[1]?.content)).toBe('<system-reminder>Available skills</system-reminder>')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('appends a replacement when a frozen baseline file changes before a later fs tool call', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'old root rule')
+      await write(join(root, 'file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+      await write(join(root, 'AGENTS.md'), 'new root rule with more detail')
+      const result = await ctx.tools.execute({
+        callId: CallId('read-after-baseline-change'), name: 'read', arguments: { file_path: 'file.txt' }, agent,
+      })
+
+      expect(workspaceContextOf(result)?.meta).toMatchObject({
+        changes: [{ action: 'replace', scope: '.', path: 'AGENTS.md' }],
+      })
+      expect(blocksText(workspaceContextOf(result)?.content)).toContain('Updated instructions from: AGENTS.md')
+      expect(blocksText(workspaceContextOf(result)?.content)).toContain('new root rule with more detail')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('appends a removal when a frozen baseline file is deleted before a later fs tool call', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'root rule')
+      await write(join(root, 'file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+      await rm(join(root, 'AGENTS.md'))
+      const result = await ctx.tools.execute({
+        callId: CallId('read-after-baseline-remove'), name: 'read', arguments: { file_path: 'file.txt' }, agent,
+      })
+
+      expect(workspaceContextOf(result)?.meta).toMatchObject({
+        changes: [{ action: 'remove', scope: '.', path: 'AGENTS.md' }],
+      })
+      expect(blocksText(workspaceContextOf(result)?.content)).toContain('Instructions removed: AGENTS.md')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('deduplicates one AGENTS.md that is both user-global and the project-root candidate', async () => {
+    const root = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'shared root and global rule')
+      await write(join(root, 'file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: root, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+      const result = await ctx.tools.execute({
+        callId: CallId('read-with-shared-global-root'), name: 'read', arguments: { file_path: 'file.txt' }, agent,
+      })
+
+      expect(derivedText(agent).match(/shared root and global rule/g)).toHaveLength(1)
+      expect(result.additionalContexts).toBeUndefined()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+    }
+  })
+
+  it('does not expose state markers when a tiny budget reduces the baseline contribution', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'repo rule')
+      const ctx = new Context()
+      await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: 10 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expect(agent.session.events.filter(event => event.type === 'context/message')).toHaveLength(0)
+      expect(derivedText(agent)).not.toContain('workspace-context:')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('loads instruction file content through ctx.fs instead of direct node reads', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'node fs rule')
+      const ctx = new Context()
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(root, 'AGENTS.md'), { type: 'file', content: 'ctx.fs rule' })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expect(derivedText(agent)).toContain('ctx.fs rule')
+      expect(derivedText(agent)).not.toContain('node fs rule')
+      expect(fs.readTargets).toEqual([join(root, 'AGENTS.md')])
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('loads provider-visible instruction files that do not exist on the host filesystem', async () => {
+    const root = join(await tempRepo(), 'virtual-repo')
+    const home = join(await tempRepo(), 'virtual-home')
+    const ctx = new Context()
+    try {
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(root, 'AGENTS.md'), { type: 'file', content: 'provider-only rule' })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expect(derivedText(agent)).toContain('provider-only rule')
+      expect(fs.readTargets).toEqual([join(root, 'AGENTS.md')])
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(dirname(root), { recursive: true, force: true })
+      await rm(dirname(home), { recursive: true, force: true })
+    }
+  })
+
+  it('rejects a provider-sized instruction file before reading content', async () => {
+    const root = join(await tempRepo(), 'virtual-repo')
+    const home = join(await tempRepo(), 'virtual-home')
+    const ctx = new Context()
+    try {
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(root, 'AGENTS.md'), { type: 'file', content: 'far too large' })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536, maxSourceBytes: 4 })
+
+      const prefix = await composeBaselinePrefix(ctx, stubAgent(root))
+
+      expect(prefix).toEqual([])
+      expect(fs.readTargets).toEqual([])
+      expect(fs.readTextTargets).toEqual([])
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(dirname(root), { recursive: true, force: true })
+      await rm(dirname(home), { recursive: true, force: true })
+    }
+  })
+
+  it('bounds streamed instruction content when provider size is unavailable', async () => {
+    const root = join(await tempRepo(), 'virtual-repo')
+    const home = join(await tempRepo(), 'virtual-home')
+    const ctx = new Context()
+    try {
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      const instructionPath = join(root, 'AGENTS.md')
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(instructionPath, { type: 'file', content: 'far too large' })
+      fs.omitSizes.add(instructionPath)
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536, maxSourceBytes: 4 })
+
+      const prefix = await composeBaselinePrefix(ctx, stubAgent(root))
+
+      expect(prefix).toEqual([])
+      expect(fs.readTargets).toEqual([instructionPath])
+      expect(fs.readTextTargets).toEqual([])
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(dirname(root), { recursive: true, force: true })
+      await rm(dirname(home), { recursive: true, force: true })
+    }
+  })
+
+  it('aborts an in-flight baseline stream with the session-prefix signal', async () => {
+    const root = join(await tempRepo(), 'virtual-repo')
+    const home = join(await tempRepo(), 'virtual-home')
+    const ctx = new Context()
+    try {
+      await ctx.plugin(BlockingReadFileSystem)
+      const fs = ctx.fs as BlockingReadFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(root, 'AGENTS.md'), { type: 'file', content: 'blocked' })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const controller = new AbortController()
+      const reason = new Error('cancel prefix')
+      const empty: Message[] = []
+      const pending = ctx.waterfall(
+        'agent/session-prefix', stubAgent(root), empty, controller.signal,
+        () => Promise.resolve(empty),
+      )
+
+      await fs.started.promise
+      controller.abort(reason)
+
+      await expect(pending).rejects.toBe(reason)
+      expect(fs.signals).toContain(controller.signal)
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(dirname(root), { recursive: true, force: true })
+      await rm(dirname(home), { recursive: true, force: true })
+    }
+  })
+
+  it('loads user-global and CLAUDE fallback content through ctx.fs', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(home, 'AGENTS.md'), 'node global rule')
+      await write(join(root, 'CLAUDE.md'), 'node claude rule')
+      const ctx = new Context()
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(home, 'AGENTS.md'), { type: 'file', content: 'ctx global rule' })
+      fs.entries.set(join(root, 'CLAUDE.md'), { type: 'file', content: 'ctx claude rule' })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expect(derivedText(agent)).toContain('ctx global rule')
+      expect(derivedText(agent)).toContain('ctx claude rule')
+      expect(derivedText(agent)).not.toContain('node global rule')
+      expect(derivedText(agent)).not.toContain('node claude rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('skips provider-visible instruction candidates when ctx.fs reports a non-file target', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'node fs rule')
+      const ctx = new Context()
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(root, 'AGENTS.md'), { type: 'directory' })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expectNoDerivedMessages(agent)
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('skips provider-visible instruction candidates when ctx.fs stat disagrees after no-follow preflight', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'node fs rule')
+      const ctx = new Context()
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(root, 'AGENTS.md'), { type: 'directory' })
+      fs.lstatTypes.set(join(root, 'AGENTS.md'), 'file')
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expectNoDerivedMessages(agent)
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('loads instruction files when ctx.fs omits the metadata size', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'node fs rule')
+      const ctx = new Context()
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(root, 'AGENTS.md'), { type: 'file' })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expect(derivedText(agent)).toContain('Instructions from: AGENTS.md')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('skips provider-visible instruction candidates when ctx.fs cannot stat them', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'node fs rule')
+      const ctx = new Context()
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.throwOnStat.add(join(root, 'AGENTS.md'))
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expectNoDerivedMessages(agent)
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not fall through to a lower-priority candidate when the winning provider file becomes unavailable', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      const ctx = new Context()
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.lstatTypes.set(join(root, 'AGENTS.md'), 'file')
+      fs.throwOnStat.add(join(root, 'AGENTS.md'))
+      fs.entries.set(join(root, 'CLAUDE.md'), { type: 'file', content: 'must not bypass AGENTS failure' })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expectNoDerivedMessages(agent)
+      expect(fs.readTargets).not.toContain(join(root, 'CLAUDE.md'))
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('treats ctx.fs marker lookup failures as absent root markers', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'repo rule')
+      const ctx = new Context()
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.throwOnStat.add(join(root, '.git'))
+      fs.entries.set(join(root, 'AGENTS.md'), { type: 'file', content: 'repo rule' })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expect(derivedText(agent)).toContain('repo rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('keeps different session cwd instruction files isolated in one context', async () => {
+    const repoA = await tempRepo()
+    const repoB = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(repoA, '.git'), { recursive: true })
+      await mkdir(join(repoB, '.git'), { recursive: true })
+      await write(join(repoA, 'AGENTS.md'), 'repo A only')
+      await write(join(repoB, 'AGENTS.md'), 'repo B only')
+      const ctx = new Context()
+      await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agentA = stubAgent(repoA)
+      const agentB = stubAgent(repoB)
+
+      await composeBaselinePrefix(ctx, agentA)
+      await composeBaselinePrefix(ctx, agentB)
+
+      expect(derivedText(agentA)).toContain('repo A only')
+      expect(derivedText(agentA)).not.toContain('repo B only')
+      expect(derivedText(agentB)).toContain('repo B only')
+      expect(derivedText(agentB)).not.toContain('repo A only')
+    } finally {
+      await rm(repoA, { recursive: true, force: true })
+      await rm(repoB, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('uses schema defaults on the plugin path so ancestor discovery still finds .git roots', async () => {
+    const root = await tempRepo()
+    try {
+      const cwd = join(root, 'child')
+      await mkdir(join(root, '.git'), { recursive: true })
+      await mkdir(cwd, { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'root schema default rule')
+      await write(join(cwd, 'AGENTS.md'), 'child schema default rule')
+      const ctx = new Context()
+      await ctx.plugin(LocalFileSystem, { cwd: '/' })
+      await ctx.plugin(workspaceContext, { maxBytes: 65536 })
+      const agent = stubAgent(cwd)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expect(derivedText(agent)).toContain('Instructions from: AGENTS.md\n\nroot schema default rule')
+      expect(derivedText(agent)).toContain('Instructions from: child/AGENTS.md\n\nchild schema default rule')
+      await ctx.fiber.dispose()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+    }
+  })
+
+  it('cleans up its agent/session-prefix listener when the plugin fiber is disposed', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'repo rule')
+      const ctx = new Context()
+      const fiber = await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      await fiber.dispose()
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expectNoDerivedMessages(agent)
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not inject anything when maxBytes is zero', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'repo rule')
+      const ctx = new Context()
+      await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: 0 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expectNoDerivedMessages(agent)
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not inject an empty workspace-context message when maxBytes is negative', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'repo rule')
+      const ctx = new Context()
+      await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: -1 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expectNoDerivedMessages(agent)
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('leaves the request unchanged when no instruction files are present', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      const ctx = new Context()
+      await mountWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      await composeBaselinePrefix(ctx, agent)
+
+      expectNoDerivedMessages(agent)
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('labels a custom dshHome as DSH_HOME instead of pretending it is ~/.dsh', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await write(join(home, 'AGENTS.md'), 'global custom rule')
+      const files = await discoverBaselineInstructionFiles({ cwd: root, dshHome: home })
+
+      expect(files.map(file => file.displayPath)).toEqual(['$DSH_HOME/AGENTS.md'])
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not repeat a candidate metadata probe during one discovery and read pass', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'repo rule')
+
+      const observedStats = new Map<string, number>()
+      vi.resetModules()
+      vi.doMock('node:fs/promises', async (importOriginal) => {
+        const actual = await importOriginal<typeof import('node:fs/promises')>()
+        return {
+          ...actual,
+          lstat: async (path: string) => {
+            observedStats.set(path, (observedStats.get(path) ?? 0) + 1)
+            return actual.lstat(path)
+          },
+        }
+      })
+      const isolated = await import('@deepseek-ai/dsh-workspace-context')
+      await isolated.loadBaselineInstructions({ cwd: root, dshHome: home, maxBytes: 65536 })
+      observedStats.clear()
+      await isolated.loadBaselineInstructions({ cwd: root, dshHome: home, maxBytes: 65536 })
+
+      expect(observedStats.get(join(root, 'AGENTS.md'))).toBe(1)
+    } finally {
+      vi.doUnmock('node:fs/promises')
+      vi.resetModules()
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not bypass an unavailable host AGENTS.md with a lower-priority candidate', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'CLAUDE.md'), 'must not bypass unavailable AGENTS')
+      vi.resetModules()
+      vi.doMock('node:fs/promises', async (importOriginal) => {
+        const actual = await importOriginal<typeof import('node:fs/promises')>()
+        return {
+          ...actual,
+          lstat: async (path: string) => {
+            if (path === join(root, 'AGENTS.md')) {
+              throw Object.assign(new Error('permission denied'), { code: 'EACCES' })
+            }
+            return actual.lstat(path)
+          },
+        }
+      })
+      const isolated = await import('@deepseek-ai/dsh-workspace-context')
+
+      const rendered = await isolated.loadBaselineInstructions({ cwd: root, dshHome: home, maxBytes: 65536 })
+
+      expect(rendered).toBeUndefined()
+    } finally {
+      vi.doUnmock('node:fs/promises')
+      vi.resetModules()
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+})
+
+describe('dynamic nested workspace context injection', () => {
+  it('commits a buffered instruction change before a later tool abort closes the step', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    const ctx = new Context()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested rule survives an aborted tool batch')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const adapter = new MockAdapter([
+        [
+          { type: 'block-start', index: 0, blockType: 'tool-call' },
+          { type: 'block-end', index: 0, block: { type: 'tool-call', id: CallId('read-before-abort'), name: 'read', arguments: '{"file_path":"pkg/deep/file.txt"}' } },
+          { type: 'block-start', index: 1, blockType: 'tool-call' },
+          { type: 'block-end', index: 1, block: { type: 'tool-call', id: CallId('abort-after-read'), name: 'abort_step', arguments: '{}' } },
+          { type: 'finish', reason: { kind: 'tool-calls' } },
+        ] satisfies StreamChunk[],
+        toolCallResponse('read-after-abort', 'read', { file_path: 'pkg/deep/file.txt' }),
+        textResponse('done'),
+      ])
+      await ctx.plugin(LlmService)
+      await ctx.plugin(SessionStore)
+      await ctx.plugin(SystemPrompt)
+      await ctx.plugin(ToolRegistry)
+      await ctx.plugin(AgentRegistry)
+      await ctx.plugin(LocalFileSystem, { cwd: '/' })
+      await ctx.plugin(ToolFs)
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      await ctx.plugin(AgentLoop, { agents: [] })
+      ctx.llm.registerAdapter(['mock'], adapter)
+      const agent = ctx.agentLoop.create(AgentId('workspace-context-abort'), { provider: 'mock', model: 'mock' }, { cwd: root })
+      ctx.tools.register(defineTool({
+        name: 'abort_step',
+        description: 'Abort the current test step.',
+        parameters: {},
+        async execute() {
+          ;(agent as unknown as { currentAbort?: AbortController }).currentAbort?.abort('test abort')
+          return [{ type: 'text', text: 'aborted' }]
+        },
+      }))
+
+      agent.send([{ type: 'text', text: 'read and abort' }])
+      await agent.whenIdle()
+      expect(agent.session.events.filter(event => event.type === 'context/message')).toHaveLength(1)
+
+      agent.send([{ type: 'text', text: 'retry the read' }])
+      await agent.whenIdle()
+
+      const contexts = agent.session.events.filter(event => event.type === 'context/message')
+      // The aborted batch drained its accepted context before step close, so the
+      // retry sees durable history without producing a duplicate instruction.
+      expect(contexts).toHaveLength(1)
+      expect(adapter.requests).toHaveLength(3)
+      expect(adapter.requests[2]?.messages.map(blocks => blocksText(blocks.content)).join('\n'))
+        .toContain('nested rule survives an aborted tool batch')
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('builds persisted digest state without inventing a provider version', () => {
+    const state = baselineInstructionState([{
+      absolutePath: '/repo/AGENTS.md',
+      displayPath: 'AGENTS.md',
+      content: 'root rule',
+    }])
+
+    const change = state.changes.get('.')
+    expect(change).toMatchObject({
+      action: 'set',
+      path: 'AGENTS.md',
+    })
+    expect(change?.digest).toMatch(/^[a-f0-9]{40}$/)
+    expect(state.versions).toEqual(new Map())
+  })
+
+  it('propagates the tool execution signal into dynamic filesystem reconciliation', async () => {
+    const root = join(await tempRepo(), 'virtual-repo')
+    const home = join(await tempRepo(), 'virtual-home')
+    const ctx = new Context()
+    try {
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(root, 'pkg/AGENTS.md'), { type: 'file', content: 'nested' })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const controller = new AbortController()
+      const reason = new Error('cancel dynamic reconciliation')
+      controller.abort(reason)
+      const exec = stubToolExecution({
+        callId: CallId('cancelled-dynamic-read'),
+        name: 'read',
+        arguments: { file_path: 'pkg/file.txt' },
+        agent: stubAgent(root),
+        signal: controller.signal,
+      })
+
+      const pending = ctx.waterfall('tools/post-execute', exec, {
+        content: [{ type: 'text', text: 'ok' }],
+        isError: false,
+      }, () => Promise.resolve({ kind: 'accept' as const }))
+
+      await expect(pending).rejects.toBe(reason)
+      expect(fs.signals).toContain(controller.signal)
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(dirname(root), { recursive: true, force: true })
+      await rm(dirname(home), { recursive: true, force: true })
+    }
+  })
+
+  it('attaches newly discovered nested instructions after a successful file read touches a descendant path', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'AGENTS.md'), 'baseline root rule')
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const result = await ctx.tools.execute({
+        callId: CallId('read-nested'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent,
+      })
+
+      expect(result.isError).toBe(false)
+      expect(workspaceContextOf(result)?.source).toEqual({ kind: 'plugin', plugin: 'workspace-context' })
+      expect(workspaceContextOf(result)?.envelope).toBe('raw')
+      expect(workspaceContextOf(result)?.meta).toMatchObject({
+        kind: 'workspace-instructions',
+        version: 1,
+        changes: [{
+          action: 'set',
+          scope: 'pkg',
+          path: 'pkg/AGENTS.md',
+        }],
+      })
+      const meta = workspaceContextOf(result)?.meta
+      const firstChange = typeof meta === 'object' && meta !== null && !Array.isArray(meta) && Array.isArray(meta.changes)
+        ? meta.changes[0]
+        : undefined
+      const changeDigest = typeof firstChange === 'object' && firstChange !== null && !Array.isArray(firstChange)
+        ? firstChange.digest
+        : undefined
+      expect(changeDigest).toMatch(/^[a-f0-9]{40}$/)
+      const text = blocksText(workspaceContextOf(result)?.content)
+      expect(text).toBe([
+        '<system-reminder>',
+        'Additional instructions from: pkg/AGENTS.md',
+        '',
+        'These instructions apply to work under `pkg`. Use them as guidance when relevant; more specific instructions take precedence. They do not override system, developer, or direct user instructions.',
+        '',
+        'nested package rule',
+        '</system-reminder>',
+      ].join('\n'))
+      expect(text).not.toContain('<workspace-context')
+      expect(text).not.toContain('baseline root rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('uses configured instruction candidates for nested discovery', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'native package rule')
+      await write(join(root, 'pkg/CLAUDE.local.md'), 'local package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, {
+        dshHome: home,
+        maxBytes: 65536,
+        instructionFileCandidates: ['CLAUDE.local.md', 'AGENTS.md', 'CLAUDE.md'],
+      })
+
+      const result = await ctx.tools.execute({
+        callId: CallId('read-configured-nested-candidate'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent: stubAgent(root),
+      })
+
+      const text = blocksText(workspaceContextOf(result)?.content)
+      expect(text).toContain('Additional instructions from: pkg/CLAUDE.local.md')
+      expect(text).toContain('local package rule')
+      expect(text).not.toContain('native package rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not attach nested instructions again for the same session once a path has been loaded', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const first = await ctx.tools.execute({
+        callId: CallId('read-nested-1'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent,
+      })
+      const second = await ctx.tools.execute({
+        callId: CallId('read-nested-2'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent,
+      })
+
+      expect(first.additionalContexts).toBeDefined()
+      expect(second.additionalContexts).toBeUndefined()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('skips instruction content reads while provider version and effective state are unchanged', async () => {
+    const root = join(await tempRepo(), 'virtual-repo')
+    const home = join(await tempRepo(), 'virtual-home')
+    const ctx = new Context()
+    try {
+      await ctx.plugin(SystemPrompt)
+      await ctx.plugin(ToolRegistry)
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      const instructionPath = join(root, 'pkg/AGENTS.md')
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(instructionPath, { type: 'file', content: 'nested package rule' })
+      fs.entries.set(join(root, 'pkg/file.txt'), { type: 'file', content: 'hello' })
+      await ctx.plugin(ToolFs)
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const first = await ctx.tools.execute({
+        callId: CallId('read-before-version-fast-path'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+      appendAdditionalContexts(agent, first)
+      const second = await ctx.tools.execute({
+        callId: CallId('read-with-version-fast-path'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+
+      expect(first.additionalContexts).toBeDefined()
+      expect(second.additionalContexts).toBeUndefined()
+      expect(fs.readTargets.filter(path => path === instructionPath)).toHaveLength(1)
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(dirname(root), { recursive: true, force: true })
+      await rm(dirname(home), { recursive: true, force: true })
+    }
+  })
+
+  it('re-reads a changed provider version, then refreshes metadata when SHA-1 is unchanged', async () => {
+    const root = join(await tempRepo(), 'virtual-repo')
+    const home = join(await tempRepo(), 'virtual-home')
+    const ctx = new Context()
+    try {
+      await ctx.plugin(SystemPrompt)
+      await ctx.plugin(ToolRegistry)
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      const instructionPath = join(root, 'pkg/AGENTS.md')
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(instructionPath, { type: 'file', content: 'same package rule', version: FsVersion('revision-1') })
+      fs.entries.set(join(root, 'pkg/file.txt'), { type: 'file', content: 'hello' })
+      await ctx.plugin(ToolFs)
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const first = await ctx.tools.execute({
+        callId: CallId('read-before-same-digest-version-change'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+      appendAdditionalContexts(agent, first)
+      fs.entries.set(instructionPath, { type: 'file', content: 'same package rule', version: FsVersion('revision-2') })
+      const afterVersionChange = await ctx.tools.execute({
+        callId: CallId('read-after-same-digest-version-change'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+      const afterRefresh = await ctx.tools.execute({
+        callId: CallId('read-after-version-cache-refresh'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+
+      expect(afterVersionChange.additionalContexts).toBeUndefined()
+      expect(afterRefresh.additionalContexts).toBeUndefined()
+      expect(fs.readTargets.filter(path => path === instructionPath)).toHaveLength(2)
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(dirname(root), { recursive: true, force: true })
+      await rm(dirname(home), { recursive: true, force: true })
+    }
+  })
+
+  it('isolates instruction version caches between sessions that touch the same scope', async () => {
+    const root = join(await tempRepo(), 'virtual-repo')
+    const home = join(await tempRepo(), 'virtual-home')
+    const ctx = new Context()
+    try {
+      await ctx.plugin(SystemPrompt)
+      await ctx.plugin(ToolRegistry)
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      const instructionPath = join(root, 'pkg/AGENTS.md')
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(instructionPath, { type: 'file', content: 'shared path, separate sessions' })
+      fs.entries.set(join(root, 'pkg/file.txt'), { type: 'file', content: 'hello' })
+      await ctx.plugin(ToolFs)
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+
+      const first = await ctx.tools.execute({
+        callId: CallId('read-from-first-session'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent: stubAgent(root),
+      })
+      const second = await ctx.tools.execute({
+        callId: CallId('read-from-second-session'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent: stubAgent(root),
+      })
+
+      expect(first.additionalContexts).toBeDefined()
+      expect(second.additionalContexts).toBeDefined()
+      expect(fs.readTargets.filter(path => path === instructionPath)).toHaveLength(2)
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(dirname(root), { recursive: true, force: true })
+      await rm(dirname(home), { recursive: true, force: true })
+    }
+  })
+
+  it('replaces previously loaded instructions when the same file content changes', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'old package rule')
+      await write(join(root, 'pkg/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const first = await ctx.tools.execute({
+        callId: CallId('read-before-change'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+      appendAdditionalContexts(agent, first)
+      await write(join(root, 'pkg/AGENTS.md'), 'new package rule with more detail')
+      const changed = await ctx.tools.execute({
+        callId: CallId('read-after-change'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+
+      expect(workspaceContextOf(changed)?.meta).toMatchObject({
+        kind: 'workspace-instructions',
+        changes: [{ action: 'replace', scope: 'pkg', path: 'pkg/AGENTS.md' }],
+      })
+      expect(blocksText(workspaceContextOf(changed)?.content)).toBe([
+        '<system-reminder>',
+        'Updated instructions from: pkg/AGENTS.md',
+        '',
+        'This file changed after it was loaded. Use the following content instead of the previously loaded instructions from this file.',
+        '',
+        'new package rule with more detail',
+        '</system-reminder>',
+      ].join('\n'))
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('replaces an AGENTS candidate with the configured fallback in the same scope', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'native package rule')
+      await write(join(root, 'pkg/CLAUDE.md'), 'fallback package rule')
+      await write(join(root, 'pkg/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const first = await ctx.tools.execute({
+        callId: CallId('read-before-fallback'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+      appendAdditionalContexts(agent, first)
+      await rm(join(root, 'pkg/AGENTS.md'))
+      const changed = await ctx.tools.execute({
+        callId: CallId('read-after-fallback'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+      appendAdditionalContexts(agent, changed)
+      const unchanged = await ctx.tools.execute({
+        callId: CallId('read-after-logged-fallback'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+
+      expect(workspaceContextOf(changed)?.meta).toMatchObject({
+        changes: [{
+          action: 'replace', scope: 'pkg', path: 'pkg/CLAUDE.md', previousPath: 'pkg/AGENTS.md',
+        }],
+      })
+      expect(blocksText(workspaceContextOf(changed)?.content)).toContain('Updated instructions from: pkg/CLAUDE.md')
+      expect(blocksText(workspaceContextOf(changed)?.content)).toContain('The instructions previously loaded from `pkg/AGENTS.md` no longer apply. Use the following content for `pkg` instead.')
+      expect(blocksText(workspaceContextOf(changed)?.content)).toContain('fallback package rule')
+      expect(unchanged.additionalContexts).toBeUndefined()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('removes previously loaded instructions when no candidate remains in the scope', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'package rule')
+      await write(join(root, 'pkg/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const first = await ctx.tools.execute({
+        callId: CallId('read-before-remove'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+      appendAdditionalContexts(agent, first)
+      await rm(join(root, 'pkg/AGENTS.md'))
+      const removed = await ctx.tools.execute({
+        callId: CallId('read-after-remove'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+
+      expect(workspaceContextOf(removed)?.meta).toEqual({
+        kind: 'workspace-instructions',
+        version: 1,
+        changes: [{ action: 'remove', scope: 'pkg', path: 'pkg/AGENTS.md' }],
+      })
+      expect(blocksText(workspaceContextOf(removed)?.content)).toBe([
+        '<system-reminder>',
+        'Instructions removed: pkg/AGENTS.md',
+        '',
+        'The previously loaded instructions from this file no longer apply.',
+        '</system-reminder>',
+      ].join('\n'))
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('loads a candidate again after a logged removal tombstone', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'first package rule')
+      await write(join(root, 'pkg/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const first = await ctx.tools.execute({
+        callId: CallId('read-before-tombstone'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+      appendAdditionalContexts(agent, first)
+      await rm(join(root, 'pkg/AGENTS.md'))
+      const removed = await ctx.tools.execute({
+        callId: CallId('read-to-create-tombstone'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+      appendAdditionalContexts(agent, removed)
+      await write(join(root, 'pkg/AGENTS.md'), 'restored package rule')
+
+      const restored = await ctx.tools.execute({
+        callId: CallId('read-after-tombstone'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+
+      expect(workspaceContextOf(restored)?.meta).toMatchObject({
+        changes: [{ action: 'set', scope: 'pkg', path: 'pkg/AGENTS.md' }],
+      })
+      expect(blocksText(workspaceContextOf(restored)?.content)).toContain('Additional instructions from: pkg/AGENTS.md')
+      expect(blocksText(workspaceContextOf(restored)?.content)).toContain('restored package rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not report removal when a previously loaded scope is temporarily unavailable', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    const ctx = new Context()
+    try {
+      await ctx.plugin(SystemPrompt)
+      await ctx.plugin(ToolRegistry)
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.entries.set(join(root, 'pkg/AGENTS.md'), { type: 'file', content: 'provider package rule' })
+      fs.entries.set(join(root, 'pkg/file.txt'), { type: 'file', content: 'hello' })
+      await ctx.plugin(ToolFs)
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const first = await ctx.tools.execute({
+        callId: CallId('read-before-provider-failure'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+      appendAdditionalContexts(agent, first)
+      fs.throwOnStat.add(join(root, 'pkg/AGENTS.md'))
+      const duringFailure = await ctx.tools.execute({
+        callId: CallId('read-during-provider-failure'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      })
+
+      expect(first.additionalContexts).toBeDefined()
+      expect(duringFailure.additionalContexts).toBeUndefined()
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('derives loaded nested instructions from resumed session history instead of duplicating them', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+      const first = await ctx.tools.execute({
+        callId: CallId('read-before-resume'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent,
+      })
+      appendAdditionalContexts(agent, first)
+      const resumed = {
+        ...agent,
+        session: new Session(agent.session.id, [...agent.session.events], agent.session.header),
+      }
+
+      const afterResume = await ctx.tools.execute({
+        callId: CallId('read-after-resume'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent: resumed,
+      })
+
+      expect(first.additionalContexts).toBeDefined()
+      expect(afterResume.additionalContexts).toBeUndefined()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('appends an update during resumed prefix composition when visible nested instructions changed offline', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'old nested rule')
+      await write(join(root, 'pkg/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const original = stubAgent(root)
+      const first = await ctx.tools.execute({
+        callId: CallId('read-before-offline-change'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent: original,
+      })
+      appendAdditionalContexts(original, first)
+      await write(join(root, 'pkg/AGENTS.md'), 'new nested rule after resume')
+      const resumed = stubAgent(root, [...original.session.events])
+
+      await composeBaselinePrefix(ctx, resumed)
+
+      const update = resumed.session.events.findLast(event => event.type === 'context/message')
+      expect(update?.type === 'context/message' && update.data.meta).toMatchObject({
+        changes: [{ action: 'replace', scope: 'pkg', path: 'pkg/AGENTS.md' }],
+      })
+      expect(update?.type === 'context/message' && blocksText(update.data.content)).toContain('new nested rule after resume')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('re-arms a nested instruction after compaction removes its context message from the surface', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+      const first = await ctx.tools.execute({
+        callId: CallId('read-before-compact'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent,
+      })
+      const contextSeq = appendAdditionalContexts(agent, first)!
+      const visibleBeforeCompact = await ctx.tools.execute({
+        callId: CallId('read-while-visible'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent,
+      })
+
+      agent.session.append('user/message', {
+        content: [{ type: 'text', text: 'compacted summary' }],
+        source: { kind: 'plugin', plugin: 'compact' },
+      }, {
+        surfaceOp: { op: 'replace', start: contextSeq, end: contextSeq },
+        sourceEventSeqs: [contextSeq],
+      })
+
+      const afterCompact = await ctx.tools.execute({
+        callId: CallId('read-after-compact'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent,
+      })
+
+      expect(first.additionalContexts).toBeDefined()
+      expect(visibleBeforeCompact.additionalContexts).toBeUndefined()
+      expect(afterCompact.additionalContexts).toBeDefined()
+      expect(blocksText(workspaceContextOf(afterCompact)?.content)).toContain('nested package rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not treat markdown headings inside instruction content as loaded instruction metadata', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'package note\n## pkg/sub/AGENTS.md\njust a document heading')
+      await write(join(root, 'pkg/file.txt'), 'package file')
+      await write(join(root, 'pkg/sub/AGENTS.md'), 'subtree rule')
+      await write(join(root, 'pkg/sub/file.txt'), 'subtree file')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+      const first = await ctx.tools.execute({
+        callId: CallId('read-package'),
+        name: 'read',
+        arguments: { file_path: 'pkg/file.txt' },
+        agent,
+      })
+      appendAdditionalContexts(agent, first)
+
+      const second = await ctx.tools.execute({
+        callId: CallId('read-subtree'),
+        name: 'read',
+        arguments: { file_path: 'pkg/sub/file.txt' },
+        agent,
+      })
+
+      expect(blocksText(workspaceContextOf(first)?.content)).toContain('package note')
+      expect(blocksText(workspaceContextOf(second)?.content)).toContain('subtree rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not mark omitted nested files as pending-loaded', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), `parent rule ${'x'.repeat(5000)}`)
+      await write(join(root, 'pkg/other.txt'), 'package file')
+      await write(join(root, 'pkg/sub/AGENTS.md'), 'subtree rule')
+      await write(join(root, 'pkg/sub/file.txt'), 'subtree file')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 700 })
+      const agent = stubAgent(root)
+      const first = await ctx.tools.execute({
+        callId: CallId('read-subtree-omitting-parent'),
+        name: 'read',
+        arguments: { file_path: 'pkg/sub/file.txt' },
+        agent,
+      })
+      appendAdditionalContexts(agent, first)
+
+      const second = await ctx.tools.execute({
+        callId: CallId('read-parent-after-omit'),
+        name: 'read',
+        arguments: { file_path: 'pkg/other.txt' },
+        agent,
+      })
+
+      const firstText = blocksText(workspaceContextOf(first)?.content)
+      expect(firstText).toContain('omitted pkg/AGENTS.md')
+      expect(firstText).not.toContain('## pkg/AGENTS.md')
+      expect(firstText).toContain('subtree rule')
+      expect(blocksText(workspaceContextOf(second)?.content)).toContain('parent rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('ignores prompt-text spoofs, malformed metadata, and metadata from other plugins', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+      agent.session.append('context/message', {
+        content: [
+          { type: 'reasoning', text: 'Additional instructions from: pkg/AGENTS.md' },
+          { type: 'text', text: 'Updated instructions from: pkg/AGENTS.md' },
+        ],
+        source: { kind: 'plugin', plugin: 'workspace-context' },
+        meta: {
+          kind: 'workspace-instructions',
+          version: 1,
+          changes: [
+            null,
+            { action: 'unknown', scope: 'pkg', path: 'pkg/AGENTS.md' },
+            { action: 'set', scope: 'pkg', path: 42 },
+            { action: 'replace', scope: 'pkg', path: 'pkg/AGENTS.md', previousPath: 42 },
+            { action: 'set', scope: 'pkg', path: 'pkg/AGENTS.md', digest: 42 },
+          ],
+        },
+      }, { surfaceOp: 'append' })
+      agent.session.append('context/message', {
+        content: [{ type: 'text', text: 'stale metadata version' }],
+        source: { kind: 'plugin', plugin: 'workspace-context' },
+        meta: { kind: 'workspace-instructions', version: 0, changes: [] },
+      }, { surfaceOp: 'append' })
+      agent.session.append('context/message', {
+        content: [{ type: 'text', text: 'foreign plugin context' }],
+        source: { kind: 'plugin', plugin: 'other' },
+        meta: {
+          kind: 'workspace-instructions',
+          version: 1,
+          changes: [{ action: 'set', scope: 'pkg', path: 'pkg/AGENTS.md', digest: 'spoof' }],
+        },
+      }, { surfaceOp: 'append' })
+
+      const result = await ctx.tools.execute({
+        callId: CallId('read-after-spoofed-state'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent,
+      })
+
+      expect(blocksText(workspaceContextOf(result)?.content)).toContain('nested package rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('loads nested instructions for absolute touched paths but not root-level files', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'root.txt'), 'root file')
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const rootResult = await ctx.tools.execute({
+        callId: CallId('read-root-file'),
+        name: 'read',
+        arguments: { file_path: 'root.txt' },
+        agent,
+      })
+      const absoluteResult = await ctx.tools.execute({
+        callId: CallId('read-absolute-nested-file'),
+        name: 'read',
+        arguments: { file_path: join(root, 'pkg/deep/file.txt') },
+        agent,
+      })
+
+      expect(rootResult.additionalContexts).toBeUndefined()
+      expect(blocksText(workspaceContextOf(absoluteResult)?.content)).toContain('nested package rule')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('treats provider failures and type disagreement after lstat as unavailable, not removed', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    const ctx = new Context()
+    try {
+      await ctx.plugin(RecordingFileSystem)
+      const fs = ctx.fs as RecordingFileSystem
+      fs.entries.set(join(root, '.git'), { type: 'directory' })
+      fs.lstatTypes.set(join(root, 'pkg/AGENTS.md'), 'file')
+      fs.throwOnStat.add(join(root, 'pkg/AGENTS.md'))
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+      const result = {
+        callId: CallId('provider-probe-result'),
+        content: [{ type: 'text' as const, text: 'ok' }],
+        isError: false,
+      }
+
+      const failedStat = await ctx.waterfall('tools/post-execute', stubToolExecution({
+        callId: CallId('provider-stat-failure'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      }), result, async () => ({ kind: 'accept' as const }))
+      fs.throwOnStat.clear()
+      fs.entries.set(join(root, 'pkg/AGENTS.md'), { type: 'directory' })
+      const mismatchedStat = await ctx.waterfall('tools/post-execute', stubToolExecution({
+        callId: CallId('provider-stat-mismatch'), name: 'read', arguments: { file_path: 'pkg/file.txt' }, agent,
+      }), result, async () => ({ kind: 'accept' as const }))
+
+      expect(failedStat).toEqual({ kind: 'accept' })
+      expect(mismatchedStat).toEqual({ kind: 'accept' })
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('skips unreadable nested instruction files without attaching empty context', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      const nested = join(root, 'pkg/AGENTS.md')
+      await write(nested, 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      await chmod(nested, 0)
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+
+      const result = await ctx.tools.execute({
+        callId: CallId('read-with-unreadable-nested-instruction'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent: stubAgent(root),
+      })
+
+      expect(result.isError).toBe(false)
+      expect(result.additionalContexts).toBeUndefined()
+      await chmod(nested, 0o600)
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('preserves nested and downstream post-execute contexts as separate entries', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      ctx.on('tools/post-execute', async () => ({
+        kind: 'accept' as const,
+        content: [{ type: 'text' as const, text: 'downstream replacement' }],
+        additionalContexts: [{
+          content: [{ type: 'text' as const, text: 'downstream context' }],
+          source: { kind: 'plugin' as const, plugin: 'downstream' },
+        }],
+      }))
+
+      const result = await ctx.tools.execute({
+        callId: CallId('read-with-downstream'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent: stubAgent(root),
+      })
+
+      expect(blocksText(result.content)).toBe('downstream replacement')
+      expect(result.additionalContexts).toHaveLength(2)
+      expect(workspaceContextOf(result)?.source).toEqual({ kind: 'plugin', plugin: 'workspace-context' })
+      expect(workspaceContextOf(result)?.envelope).toBe('raw')
+      expect(workspaceContextOf(result)?.meta).toMatchObject({
+        kind: 'workspace-instructions',
+        changes: [{ action: 'set', scope: 'pkg', path: 'pkg/AGENTS.md' }],
+      })
+      expect(blocksText(workspaceContextOf(result)?.content)).toContain('nested package rule')
+      expect(blocksText(workspaceContextOf(result)?.content)).not.toContain('downstream context')
+      expect(result.additionalContexts?.[1]).toEqual({
+        content: [{ type: 'text', text: 'downstream context' }],
+        source: { kind: 'plugin', plugin: 'downstream' },
+      })
+      const agent = stubAgent(root)
+      appendAdditionalContexts(agent, result)
+      expect(blocksText(agent.session.deriveMessages()[1]?.content)).toContain('<context source="plugin">\ndownstream context\n</context>')
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not attach discovered instructions when a downstream listener blocks the tool call', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      ctx.on('tools/post-execute', async () => ({
+        kind: 'block' as const,
+        feedback: [{ type: 'text' as const, text: 'blocked downstream' }],
+      }))
+
+      const result = await ctx.tools.execute({
+        callId: CallId('read-blocked-downstream'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent: stubAgent(root),
+      })
+
+      // The pipeline rejected this touch, so no workspace instructions from it
+      // should reach the model, and the block feedback must survive unchanged.
+      expect(result.isError).toBe(true)
+      expect(blocksText(result.content)).toBe('blocked downstream')
+      expect(result.additionalContexts).toBeUndefined()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not commit pending state when an outer post-execute listener blocks the final result', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    const ctx = new Context()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      await ctx.plugin(SystemPrompt)
+      await ctx.plugin(ToolRegistry)
+      await ctx.plugin(LocalFileSystem, { cwd: '/' })
+      await ctx.plugin(ToolFs)
+      let shouldBlock = true
+      ctx.on('tools/post-execute', async (_exec, _result, next) => {
+        const downstream = await next()
+        return shouldBlock
+          ? { kind: 'block' as const, feedback: [{ type: 'text' as const, text: 'outer policy block' }] }
+          : downstream
+      })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const blocked = await ctx.tools.execute({
+        callId: CallId('outer-block-first'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent,
+      })
+      shouldBlock = false
+      const accepted = await ctx.tools.execute({
+        callId: CallId('outer-block-retry'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent,
+      })
+
+      expect(blocked.isError).toBe(true)
+      expect(blocked.additionalContexts).toBeUndefined()
+      expect(accepted.isError).toBe(false)
+      expect(blocksText(workspaceContextOf(accepted)?.content)).toContain('nested package rule')
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('rolls back parent-token pending state when a composite result is blocked', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    const ctx = new Context()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      await ctx.plugin(SystemPrompt)
+      await ctx.plugin(ToolRegistry)
+      await ctx.plugin(LocalFileSystem, { cwd: '/' })
+      await ctx.plugin(ToolFs)
+      ctx.tools.register(defineTool({
+        name: 'composite-read',
+        description: 'read through a nested dispatch',
+        parameters: {},
+        async execute(_args, exec) {
+          const nested = await ctx.tools.execute({
+            callId: CallId(`${exec.callId}:nested`),
+            name: 'read',
+            arguments: { file_path: 'pkg/deep/file.txt' },
+            ...exec.agent === undefined ? {} : { agent: exec.agent },
+            parent: exec.token,
+            ...exec.signal === undefined ? {} : { signal: exec.signal },
+          })
+          for (const context of nested.additionalContexts ?? []) exec.deferContext(context)
+          return nested.content
+        },
+      }))
+      let shouldBlock = true
+      ctx.on('tools/post-execute', async (exec, _result, next) => {
+        const downstream = await next()
+        return exec.name === 'composite-read' && shouldBlock
+          ? { kind: 'block' as const, feedback: [{ type: 'text' as const, text: 'outer composite block' }] }
+          : downstream
+      })
+      await ctx.plugin(workspaceContext, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+
+      const blocked = await ctx.tools.execute({
+        callId: CallId('composite-first'), name: 'composite-read', arguments: {}, agent,
+      })
+      shouldBlock = false
+      const accepted = await ctx.tools.execute({
+        callId: CallId('composite-retry'), name: 'composite-read', arguments: {}, agent,
+      })
+
+      expect(blocked.isError).toBe(true)
+      expect(blocked.additionalContexts).toBeUndefined()
+      expect(accepted.isError).toBe(false)
+      expect(blocksText(workspaceContextOf(accepted)?.content)).toContain('nested package rule')
+    } finally {
+      await ctx.fiber.dispose()
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('handles defensive tools/result observer branches without retaining staged state', async () => {
+    const ctx = new Context()
+    try {
+      await ctx.plugin(workspaceContext, { maxBytes: 65536 })
+      const agent = stubAgent('/')
+      const parent = Symbol('parent') as ToolExecutionToken
+      const plainResult = { callId: CallId('plain'), content: [], isError: false }
+
+      ctx.emit('tools/result', stubToolExecution({
+        callId: CallId('agentless-child'), name: 'read', arguments: {}, parent,
+      }), plainResult)
+      ctx.emit('tools/result', stubToolExecution({
+        callId: CallId('contextless-child'), name: 'read', arguments: {}, agent, parent,
+      }), { ...plainResult, additionalContexts: [{ content: [], source: { kind: 'plugin', plugin: 'workspace-context' } }] })
+      ctx.emit('tools/result', stubToolExecution({
+        callId: CallId('first-child'), name: 'read', arguments: {}, agent, parent,
+      }), { ...plainResult, additionalContexts: [workspaceChangeContext('first', 'one')] })
+      ctx.emit('tools/result', stubToolExecution({
+        callId: CallId('second-child'), name: 'read', arguments: {}, agent, parent,
+      }), { ...plainResult, additionalContexts: [workspaceChangeContext('second', 'two')] })
+      ctx.emit('tools/result', {
+        ...stubToolExecution({ callId: CallId('agentless-parent'), name: 'composite', arguments: {} }),
+        token: parent,
+      }, plainResult)
+
+      expect(agent.session.deriveMessages()).toEqual([])
+    } finally {
+      await ctx.fiber.dispose()
+    }
+  })
+
+  it('ignores post-execute events that are not successful structured file touches', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      const agent = stubAgent(root)
+      const result = {
+        callId: CallId('manual'),
+        content: [{ type: 'text' as const, text: 'manual result' }],
+        isError: false,
+      }
+      const cases = [
+        { name: 'read', arguments: { file_path: 'pkg/deep/file.txt' }, agent: undefined },
+        { name: 'bash', arguments: { file_path: 'pkg/deep/file.txt' }, agent },
+        { name: 'read', arguments: null, agent },
+        { name: 'read', arguments: {}, agent },
+        { name: 'read', arguments: { file_path: 1 }, agent },
+        { name: 'read', arguments: { file_path: '   ' }, agent },
+      ]
+
+      for (const item of cases) {
+        const decision = await ctx.waterfall('tools/post-execute', stubToolExecution({
+          callId: CallId(`manual-${item.name}-${cases.indexOf(item)}`),
+          name: item.name,
+          arguments: item.arguments,
+          ...item.agent === undefined ? {} : { agent: item.agent },
+        }), result, async () => ({ kind: 'accept' as const }))
+        expect(decision).toEqual({ kind: 'accept' })
+      }
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not attach nested instructions when the byte budget is disabled', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 0 })
+
+      const result = await ctx.tools.execute({
+        callId: CallId('read-with-disabled-budget'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent: stubAgent(root),
+      })
+
+      expect(result.isError).toBe(false)
+      expect(result.additionalContexts).toBeUndefined()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('does not attach nested instructions after a failed file read', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      const ctx = new Context()
+      await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+
+      const result = await ctx.tools.execute({
+        callId: CallId('read-missing'),
+        name: 'read',
+        arguments: { file_path: 'pkg/missing.txt' },
+        agent: stubAgent(root),
+      })
+
+      expect(result.isError).toBe(true)
+      expect(result.additionalContexts).toBeUndefined()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+
+  it('cleans up its tools/post-execute listener when the plugin fiber is disposed', async () => {
+    const root = await tempRepo()
+    const home = await tempRepo()
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await write(join(root, 'pkg/AGENTS.md'), 'nested package rule')
+      await write(join(root, 'pkg/deep/file.txt'), 'hello')
+      const ctx = new Context()
+      const fiber = await mountFileToolsAndWorkspaceContext(ctx, { dshHome: home, maxBytes: 65536 })
+      await fiber.dispose()
+
+      const result = await ctx.tools.execute({
+        callId: CallId('read-after-dispose'),
+        name: 'read',
+        arguments: { file_path: 'pkg/deep/file.txt' },
+        agent: stubAgent(root),
+      })
+
+      expect(result.isError).toBe(false)
+      expect(result.additionalContexts).toBeUndefined()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+      await rm(home, { recursive: true, force: true })
+    }
+  })
+})
+
+describe('workspace context pending state', () => {
+  it('leaves pending transitions from other or untracked steps untouched', () => {
+    const agent = stubAgent('/')
+    const change = (scope: string) => ({
+      action: 'set' as const, scope, path: `${scope}/AGENTS.md`, digest: scope,
+    })
+    const pending = new WeakMap<object, Map<string, PendingInstructionChange>>([[
+      agent.session,
+      new Map([
+        ['untracked', { change: change('untracked'), afterSeq: 0 }],
+        ['other-turn', { change: change('other-turn'), afterSeq: 0, step: { turn: 2, step: 1 } }],
+        ['other-step', { change: change('other-step'), afterSeq: 0, step: { turn: 1, step: 2 } }],
+        ['current', { change: change('current'), afterSeq: 0, step: { turn: 1, step: 1 } }],
+      ]),
+    ]])
+    const versions: InstructionVersionCache = new WeakMap()
+    const ended = agent.session.append('step/end', { turn: 1, step: 1 })
+
+    observeInstructionSessionEvent(agent.session, ended, pending, versions)
+
+    expect([...pending.get(agent.session)?.keys() ?? []]).toEqual(['untracked', 'other-turn', 'other-step'])
+  })
+
+  it('confirms a pending transition only when its matching workspace context reaches the log', () => {
+    const agent = stubAgent('/')
+    const pending = new WeakMap<object, Map<string, PendingInstructionChange>>()
+    const versions: InstructionVersionCache = new WeakMap()
+    const [change] = commitPendingInstructionContexts(agent, [workspaceChangeContext('pkg', 'one')], pending)
+    expect(change).toBeDefined()
+    versions.set(agent.session, new Map([['pkg', {
+      path: 'pkg/AGENTS.md', version: FsVersion('v1'), digest: 'one',
+    }]]))
+
+    const unrelated = agent.session.append('context/message', {
+      content: [], source: { kind: 'plugin', plugin: 'other' },
+    }, { surfaceOp: 'append' })
+    observeInstructionSessionEvent(agent.session, unrelated, pending, versions)
+    expect(pending.get(agent.session)?.has('pkg')).toBe(true)
+
+    const otherContext = workspaceChangeContext('other', 'other')
+    const otherWorkspaceEvent = agent.session.append('context/message', {
+      content: otherContext.content,
+      source: otherContext.source,
+      ...otherContext.envelope !== undefined ? { envelope: otherContext.envelope } : {},
+      ...otherContext.meta !== undefined ? { meta: otherContext.meta } : {},
+    }, { surfaceOp: 'append' })
+    observeInstructionSessionEvent(agent.session, otherWorkspaceEvent, pending, versions)
+    expect(pending.get(agent.session)?.has('pkg')).toBe(true)
+
+    const context = workspaceChangeContext('pkg', 'one')
+    const confirmed = agent.session.append('context/message', {
+      content: context.content,
+      source: context.source,
+      ...context.envelope !== undefined ? { envelope: context.envelope } : {},
+      ...context.meta !== undefined ? { meta: context.meta } : {},
+    }, { surfaceOp: 'append' })
+    observeInstructionSessionEvent(agent.session, confirmed, pending, versions)
+
+    expect(pending.has(agent.session)).toBe(false)
+    expect(versions.get(agent.session)?.has('pkg')).toBe(true)
+  })
+
+  it('discards pending state and its version fast path when the owning step closes first', () => {
+    const agent = stubAgent('/')
+    const pending = new WeakMap<object, Map<string, PendingInstructionChange>>()
+    const versions: InstructionVersionCache = new WeakMap()
+    agent.session.append('step/start', { turn: 1, step: 1 })
+    commitPendingInstructionContexts(agent, [workspaceChangeContext('pkg', 'one')], pending)
+    versions.set(agent.session, new Map([['pkg', {
+      path: 'pkg/AGENTS.md', version: FsVersion('v1'), digest: 'one',
+    }]]))
+
+    const ended = agent.session.append('step/end', { turn: 1, step: 1 })
+    observeInstructionSessionEvent(agent.session, ended, pending, versions)
+
+    expect(pending.has(agent.session)).toBe(false)
+    expect(versions.has(agent.session)).toBe(false)
+  })
+
+  it('rolls back only the exact current transition and releases empty session state', () => {
+    const agent = stubAgent('/')
+    const pending = new WeakMap<object, Map<string, PendingInstructionChange>>()
+
+    rollbackPendingInstructionChanges(agent, [{
+      action: 'set', scope: 'missing', path: 'missing/AGENTS.md', digest: 'none',
+    }], pending)
+    expect(commitPendingInstructionContexts(agent, [{
+      content: [], source: { kind: 'plugin', plugin: 'workspace-context' },
+    }], pending)).toEqual([])
+
+    const committed = commitPendingInstructionContexts(agent, [
+      workspaceChangeContext('first', 'one'),
+      workspaceChangeContext('second', 'two'),
+    ], pending)
+    const [first, second] = committed
+    expect(first).toBeDefined()
+    expect(second).toBeDefined()
+
+    const [newer] = commitPendingInstructionContexts(agent, [workspaceChangeContext('first', 'newer')], pending)
+    rollbackPendingInstructionChanges(agent, [first!], pending)
+    rollbackPendingInstructionChanges(agent, [{
+      action: 'set', scope: 'unknown', path: 'unknown/AGENTS.md', digest: 'unknown',
+    }], pending)
+    rollbackPendingInstructionChanges(agent, [second!], pending)
+    expect(pending.get(agent.session)?.get('first')?.change).toEqual(newer)
+
+    rollbackPendingInstructionChanges(agent, [newer!], pending)
+    expect(pending.has(agent.session)).toBe(false)
+  })
+})
+
+describe('workspace context plugin export shape', () => {
+  it('has the namespace-plugin export shape (no stray default) so the Loader keeps name/Config/apply', () => {
+    expect('default' in workspaceContext).toBe(false)
+    expect(typeof workspaceContext.apply).toBe('function')
+
+    const loader = Object.create(Loader.prototype) as Loader
+    const unwrapped = loader.unwrapExports(workspaceContext) as Record<string, unknown>
+    expect(unwrapped).toBe(workspaceContext)
+    expect(unwrapped.name).toBe('workspace-context')
+    expect(unwrapped.Config).toBeDefined()
+    expect(typeof unwrapped.apply).toBe('function')
+  })
+})
diff --git a/packages/context/workspace-context/tsconfig.json b/packages/context/workspace-context/tsconfig.json
new file mode 100644
index 0000000000..b4807ded65
--- /dev/null
+++ b/packages/context/workspace-context/tsconfig.json
@@ -0,0 +1,36 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": [
+    "src"
+  ],
+  "references": [
+    {
+      "path": "../../../vendor/cordis"
+    },
+    {
+      "path": "../../../vendor/schemastery"
+    },
+    {
+      "path": "../../llm/llm"
+    },
+    {
+      "path": "../../core/agent"
+    },
+    {
+      "path": "../../core/session"
+    },
+    {
+      "path": "../../core/tools"
+    },
+    {
+      "path": "../../fs/fs"
+    },
+    {
+      "path": "../../util/paths"
+    }
+  ]
+}
diff --git a/packages/cordis/tool-cordis/package.json b/packages/cordis/tool-cordis/package.json
index fd9c35e48e..e13de3e58b 100644
--- a/packages/cordis/tool-cordis/package.json
+++ b/packages/cordis/tool-cordis/package.json
@@ -33,6 +33,7 @@
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-scope": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
diff --git a/packages/cordis/tool-cordis/src/api-catalog.ts b/packages/cordis/tool-cordis/src/api-catalog.ts
index 5d16b5d496..07b6228d62 100644
--- a/packages/cordis/tool-cordis/src/api-catalog.ts
+++ b/packages/cordis/tool-cordis/src/api-catalog.ts
@@ -84,17 +84,20 @@ export const SERVICE_API: readonly ServiceApiEntry[] = [
   },
   {
     key: 'bash',
-    summary: 'Registers one `ctx.bash` implementation.',
+    summary: 'Abstract bash execution service.',
     methods: [
       'abstract resolve(request: BashExecRequest): BashExecSpec',
       'abstract run(spec: BashExecSpec): Promise<BashRunResult>',
-      'abstract start(spec: BashExecSpec): BashTask',
-      'abstract get(id: BashTaskId): BashTask | undefined',
-      'abstract ownerOf(id: BashTaskId): OwnerToken | undefined',
-      'abstract list(): BashTask[]',
-      'abstract readOutput(id: BashTaskId): BashTaskRead',
-      'abstract kill(id: BashTaskId): boolean',
-      'onTaskDone(listener: BashTaskListener): () => void',
+      'abstract start(spec: BashExecSpec): BashProcess',
+    ],
+  },
+  {
+    key: 'bashEnv',
+    summary: 'Registry (`ctx.bashEnv`) for trusted, per-execution `DSH_*` variables.',
+    methods: [
+      'register(contributor: BashEnvContributor): () => void',
+      'collect(execution: ToolExecution): DshEnvironment',
+      'list(): BashEnvVariableInfo[]',
     ],
   },
   {
@@ -116,8 +119,9 @@ export const SERVICE_API: readonly ServiceApiEntry[] = [
     key: 'fs',
     summary: 'Abstract filesystem provider.',
     methods: [
-      'abstract resolve(path: string, opts?: { cwd?: string }): Promise<FsTarget>',
+      'abstract resolve(path: string, opts?: { cwd?: string; signal?: AbortSignal }): Promise<FsTarget>',
       'abstract stat(target: FsTarget, signal?: AbortSignal): Promise<FsInfo | undefined>',
+      'abstract lstat(path: string, opts?: { cwd?: string }, signal?: AbortSignal): Promise<FsPathInfo | undefined>',
       'abstract readText(target: FsTarget, signal?: AbortSignal): Promise<string>',
       'abstract streamText(target: FsTarget, signal?: AbortSignal): Promise<AsyncIterable<string>>',
       'abstract listDir(target: FsTarget, signal?: AbortSignal): Promise<FsDirEntry[]>',
@@ -129,8 +133,9 @@ export const SERVICE_API: readonly ServiceApiEntry[] = [
     key: 'llm',
     summary: 'The abstract `llm` service: an adapter registry plus a streaming model-call surface, interceptable via the `llm/stream` waterfall.',
     methods: [
-      'registerAdapter(models: string[], adapter: LlmAdapter): () => void',
-      'models(): string[]',
+      'registerAdapter(providers: string[], adapter: LlmAdapter): () => void',
+      'listProviders(): LlmProviderInfo[]',
+      'async listModels(provider: string): Promise<LlmModelInfo[]>',
       'stream(options: GenerateOptions): AsyncIterable<StreamChunk>',
     ],
   },
@@ -155,6 +160,7 @@ export const SERVICE_API: readonly ServiceApiEntry[] = [
     key: 'sessionPersistence',
     summary: 'Durable append-only session storage.',
     methods: [
+      'abstract locate(meta: SessionHeader): SessionLocation | undefined',
       'abstract create(meta: SessionHeader): Promise<void>',
       'abstract append(id: SessionId, events: readonly SessionEvent[]): Promise<void>',
       'abstract load(id: SessionId): Promise<{ meta: SessionHeader; events: SessionEvent[] }>',
@@ -163,10 +169,12 @@ export const SERVICE_API: readonly ServiceApiEntry[] = [
   },
   {
     key: 'sessionQuery',
-    summary: 'Live-preferred logical-corpus and exact-event read service.',
+    summary: 'Live-preferred logical-corpus exact-read and relationship-tracing service.',
     methods: [
       'listSessions(): Promise<SessionRecord[]>',
       'async listEvents(sessionId: SessionId): Promise<SessionEventRecord[]>',
+      'async traceSession(sessionId: SessionId): Promise<SessionLineageTrace>',
+      'async traceEvent(request: SessionEventTraceRequest): Promise<SessionEventTrace>',
       'async readEvent(request: SessionEventReadRequest): Promise<SessionEventWindow>',
     ],
   },
@@ -194,6 +202,13 @@ export const SERVICE_API: readonly ServiceApiEntry[] = [
       'async get(name: string, options: SkillLookupOptions = {}): Promise<SkillDefinition | undefined>',
     ],
   },
+  {
+    key: 'spillStore',
+    summary: 'Abstract spill storage service.',
+    methods: [
+      'abstract saveText(input: SaveTextSpill): Promise<SpillRef>',
+    ],
+  },
   {
     key: 'subagents',
     summary: 'Named provider registry and capability-checked start surface.',
@@ -214,6 +229,28 @@ export const SERVICE_API: readonly ServiceApiEntry[] = [
       'async assemble(context: AssembleContext = {}): Promise<PromptAssembly>',
     ],
   },
+  {
+    key: 'tasks',
+    summary: 'The `tasks` service: the runtime-global background task registry.',
+    methods: [
+      'start(spec: TaskStart): TaskId',
+      'list(caller?: Agent): TaskSnapshot[]',
+      'get(id: TaskId, caller?: Agent): TaskSnapshot',
+      'read(id: TaskId, caller?: Agent): TaskRead',
+      'kill(id: TaskId, caller?: Agent, reason?: string): \'requested\' | \'already-finished\'',
+      'async wait(id: TaskId, timeoutMs: number, caller?: Agent, signal?: AbortSignal): Promise<TaskSnapshot>',
+      'onTaskDone(listener: TaskDoneListener): () => void',
+      'attachSurface(name: string): () => void',
+    ],
+  },
+  {
+    key: 'tokenMeter',
+    summary: 'Replay owner for one service-wide estimator and isolated per-session folds.',
+    methods: [
+      'measure(session: Session, requestHeader?: EpochHeader): TokenMeasurement',
+      'estimateMessage(message: Message): number',
+    ],
+  },
   {
     key: 'tools',
     summary: 'Tool registry and execution pipeline.',
@@ -223,6 +260,7 @@ export const SERVICE_API: readonly ServiceApiEntry[] = [
       'guard(guard: ToolGuard): () => void',
       'get(name: string, scope?: ScopeKey): ToolDefinition | undefined',
       'schemas(scope?: ScopeKey): ToolSchema[]',
+      'executionMode(exec: ToolExecutionInput): ToolExecutionMode',
       'async execute(exec: ToolExecutionInput): Promise<ToolExecutionResult>',
     ],
   },
@@ -495,7 +533,7 @@ export const EVENT_API: readonly EventApiEntry[] = [
 export const TYPE_API: readonly TypeApiEntry[] = [
   {
     name: 'Agent',
-    declaration: 'export interface Agent {\n    readonly id: AgentId;\n    readonly options: AgentOptions;\n    readonly session: Session;\n    readonly status: AgentStatus;\n    readonly ctx: Context;\n    send(content: ContentBlock[], options?: SendOptions): void;\n    steer(content: ContentBlock[], options?: SendOptions): void;\n    inject(content: ContentBlock[], options?: SendOptions): void;\n    cancel(reason?: string): void;\n    whenIdle(): Promise<void>;\n}',
+    declaration: 'export interface Agent {\n    readonly id: AgentId;\n    readonly options: AgentOptions;\n    readonly session: Session;\n    readonly status: AgentStatus;\n    readonly ctx: Context;\n    send(content: ContentBlock[], options?: SendOptions): void;\n    steer(content: ContentBlock[], options?: SendOptions): void;\n    inject(content: ContentBlock[], options?: InjectOptions): void;\n    cancel(reason?: string): void;\n    whenIdle(): Promise<void>;\n}',
   },
   {
     name: 'AgentFactory',
@@ -511,7 +549,7 @@ export const TYPE_API: readonly TypeApiEntry[] = [
   },
   {
     name: 'AgentOptions',
-    declaration: 'export interface AgentOptions {\n    model?: string;\n}',
+    declaration: 'export interface AgentOptions {\n    provider?: string;\n    model?: string;\n}',
   },
   {
     name: 'AgentStatus',
@@ -557,13 +595,41 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'AssembledSection',
     declaration: 'export interface AssembledSection {\n    name: string;\n    text: string;\n}',
   },
+  {
+    name: 'AssistantProvenance',
+    declaration: 'export interface AssistantProvenance {\n    provider: string;\n    model: string;\n    replayState?: unknown;\n}',
+  },
+  {
+    name: 'BashEnvContributor',
+    declaration: 'export interface BashEnvContributor {\n    name: string;\n    variables: Readonly<Record<DshEnvironmentKey, BashEnvVariable>>;\n    resolve(execution: ToolExecution): Readonly<Partial<Record<DshEnvironmentKey, string>>>;\n}',
+  },
+  {
+    name: 'BashEnvVariable',
+    declaration: 'export interface BashEnvVariable {\n    description: string;\n}',
+  },
+  {
+    name: 'BashEnvVariableInfo',
+    declaration: 'export interface BashEnvVariableInfo extends BashEnvVariable {\n    contributor: string;\n    key: DshEnvironmentKey;\n}',
+  },
   {
     name: 'BashExecRequest',
-    declaration: 'export interface BashExecRequest {\n    command: string;\n    workdir?: string | undefined;\n    timeoutMs?: number | undefined;\n    signal?: AbortSignal | undefined;\n    stdin?: string | undefined;\n    env?: Record<string, string> | undefined;\n    owner?: OwnerToken | undefined;\n    sandboxMode?: SandboxMode | undefined;\n}',
+    declaration: 'export interface BashExecRequest {\n    command: string;\n    workdir?: string | undefined;\n    timeoutMs?: number | undefined;\n    stdoutMaxBytes?: number | undefined;\n    signal?: AbortSignal | undefined;\n    stdin?: string | undefined;\n    env?: Record<string, string> | undefined;\n    dshEnv?: DshEnvironment | undefined;\n    sandboxMode?: SandboxMode | undefined;\n}',
   },
   {
     name: 'BashExecSpec',
-    declaration: 'export interface BashExecSpec {\n    command: string;\n    workdir: string;\n    timeoutMs: number;\n    signal?: AbortSignal | undefined;\n    stdin?: string | undefined;\n    env?: Record<string, string> | undefined;\n    owner: OwnerToken | undefined;\n    sandboxMode: SandboxMode | undefined;\n}',
+    declaration: 'export interface BashExecSpec {\n    command: string;\n    workdir: string;\n    timeoutMs: number;\n    stdoutMaxBytes: number;\n    signal?: AbortSignal | undefined;\n    stdin?: string | undefined;\n    env?: Record<string, string> | undefined;\n    dshEnv?: DshEnvironment | undefined;\n    sandboxMode: SandboxMode | undefined;\n}',
+  },
+  {
+    name: 'BashProcess',
+    declaration: 'export interface BashProcess {\n    status: BashProcessStatus;\n    exitCode: number | null;\n    signal: NodeJS.Signals | null;\n    readonly done: Promise<void>;\n    sandbox?: BashSandboxInfo;\n    readOutput(): BashProcessRead;\n    kill(): boolean;\n}',
+  },
+  {
+    name: 'BashProcessRead',
+    declaration: 'export interface BashProcessRead {\n    delta: string;\n    lossy: boolean;\n    stdoutSpillPath?: string;\n    stderrSpillPath?: string;\n}',
+  },
+  {
+    name: 'BashProcessStatus',
+    declaration: 'export type BashProcessStatus = \'running\' | \'completed\' | \'killed\';',
   },
   {
     name: 'BashRunResult',
@@ -573,26 +639,6 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'BashSandboxInfo',
     declaration: 'export interface BashSandboxInfo {\n    mode: SandboxMode;\n    denied: boolean;\n    enforcement?: SandboxEnforcement;\n    runnerFailed?: boolean;\n}',
   },
-  {
-    name: 'BashTask',
-    declaration: 'export interface BashTask {\n    readonly id: BashTaskId;\n    status: BashTaskStatus;\n    exitCode: number | null;\n    signal: NodeJS.Signals | null;\n    readonly done: Promise<void>;\n    sandbox?: BashSandboxInfo;\n}',
-  },
-  {
-    name: 'BashTaskId',
-    declaration: 'export type BashTaskId = Branded<\'BashTaskId\'>;',
-  },
-  {
-    name: 'BashTaskListener',
-    declaration: 'export type BashTaskListener = (task: BashTask) => void;',
-  },
-  {
-    name: 'BashTaskRead',
-    declaration: 'export interface BashTaskRead {\n    task: BashTask;\n    delta: string;\n    lossy: boolean;\n    stdoutSpillPath?: string;\n    stderrSpillPath?: string;\n}',
-  },
-  {
-    name: 'BashTaskStatus',
-    declaration: 'export type BashTaskStatus = \'running\' | \'completed\' | \'killed\';',
-  },
   {
     name: 'Branded',
     declaration: 'export type Branded<B extends string> = string & {\n    readonly [BRAND]: B;\n};',
@@ -649,6 +695,10 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'ContentBlockType',
     declaration: 'export type ContentBlockType = keyof ContentBlockMap;',
   },
+  {
+    name: 'ContextEnvelope',
+    declaration: 'export type ContextEnvelope = \'context\' | \'raw\';',
+  },
   {
     name: 'CreateAgentOptions',
     declaration: 'export interface CreateAgentOptions {\n    readonly agentId: AgentId;\n    readonly sessionId: SessionId;\n    readonly meta?: {\n        readonly cwd?: string;\n        readonly parentSession?: SessionId;\n        readonly seedLength?: number;\n    };\n    readonly seed?: readonly SessionEvent[];\n    readonly agentOptions?: AgentOptions;\n    readonly signal?: AbortSignal;\n    readonly setup?: (agentCtx: Context) => Promise<void> | void;\n}',
@@ -665,6 +715,18 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'DiffResultView',
     declaration: 'export interface DiffResultView {\n    card: \'diff\';\n    title?: string;\n    diffs: FileDiff[];\n}',
   },
+  {
+    name: 'DshEnvironment',
+    declaration: 'export type DshEnvironment = Readonly<Record<DshEnvironmentKey, string>>;',
+  },
+  {
+    name: 'DshEnvironmentKey',
+    declaration: 'export type DshEnvironmentKey = `${typeof DSH_ENV_PREFIX}${string}`;',
+  },
+  {
+    name: 'EpochHeader',
+    declaration: 'export interface EpochHeader {\n    config: LlmCallConfig;\n    system?: string;\n    tools?: ToolSchema[];\n    messagePrefix?: Message[];\n}',
+  },
   {
     name: 'FileDiff',
     declaration: 'export interface FileDiff {\n    path: string;\n    oldText: string | null;\n    newText: string;\n}',
@@ -697,6 +759,10 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'FsInfo',
     declaration: 'export interface FsInfo {\n    version: FsVersion;\n    type: \'file\' | \'directory\' | \'other\';\n    size?: number;\n}',
   },
+  {
+    name: 'FsPathInfo',
+    declaration: 'export interface FsPathInfo {\n    version: FsVersion;\n    type: \'file\' | \'directory\' | \'symlink\' | \'other\';\n    size?: number;\n}',
+  },
   {
     name: 'FsTarget',
     declaration: 'export interface FsTarget {\n    targetKey: FsTargetKey;\n    displayPath: string;\n}',
@@ -719,7 +785,7 @@ export const TYPE_API: readonly TypeApiEntry[] = [
   },
   {
     name: 'GenerateOptions',
-    declaration: 'export interface GenerateOptions {\n    model: string;\n    messages: Message[];\n    system?: string;\n    tools?: ToolSchema[];\n    temperature?: number;\n    maxTokens?: number;\n    stop?: string[];\n    signal?: AbortSignal;\n    sessionId?: Branded<\'SessionId\'>;\n}',
+    declaration: 'export interface GenerateOptions {\n    provider: string;\n    model: string;\n    messages: Message[];\n    system?: string;\n    tools?: ToolSchema[];\n    temperature?: number;\n    maxTokens?: number;\n    stop?: string[];\n    signal?: AbortSignal;\n    sessionId?: Branded<\'SessionId\'>;\n}',
   },
   {
     name: 'GenericCallView',
@@ -731,11 +797,31 @@ export const TYPE_API: readonly TypeApiEntry[] = [
   },
   {
     name: 'HookContext',
-    declaration: 'export interface HookContext {\n    content: ContentBlock[];\n    source: MessageSource;\n}',
+    declaration: 'export interface HookContext {\n    content: ContentBlock[];\n    source: MessageSource;\n    envelope?: ContextEnvelope;\n    meta?: JsonValue;\n}',
+  },
+  {
+    name: 'InjectOptions',
+    declaration: 'export interface InjectOptions extends SendOptions {\n    envelope?: ContextEnvelope;\n    meta?: JsonValue;\n}',
+  },
+  {
+    name: 'JsonValue',
+    declaration: 'export type JsonValue = null | boolean | number | string | JsonValue[] | {\n    [key: string]: JsonValue;\n};',
+  },
+  {
+    name: 'LlmCallConfig',
+    declaration: 'export interface LlmCallConfig {\n    provider: string;\n    model: string;\n    temperature?: number;\n    maxTokens?: number;\n    stop?: string[];\n}',
+  },
+  {
+    name: 'LlmModelInfo',
+    declaration: 'export interface LlmModelInfo {\n    provider: string;\n    id: string;\n    name: string;\n    description?: string;\n}',
+  },
+  {
+    name: 'LlmProviderInfo',
+    declaration: 'export interface LlmProviderInfo {\n    id: string;\n    name: string;\n}',
   },
   {
     name: 'Message',
-    declaration: 'export interface Message {\n    role: \'system\' | \'user\' | \'assistant\';\n    content: ContentBlock[];\n}',
+    declaration: 'export interface Message {\n    role: \'system\' | \'user\' | \'assistant\';\n    content: ContentBlock[];\n    provenance?: AssistantProvenance;\n}',
   },
   {
     name: 'MessageSource',
@@ -745,10 +831,6 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'MessageSourceMap',
     declaration: 'export interface MessageSourceMap {\n    user: {\n        kind: \'user\';\n    };\n    plugin: {\n        kind: \'plugin\';\n        plugin: string;\n    };\n}',
   },
-  {
-    name: 'OwnerToken',
-    declaration: 'export type OwnerToken = Branded<\'OwnerToken\'>;',
-  },
   {
     name: 'PresetOption',
     declaration: 'export interface PresetOption {\n    value: string;\n    name: string;\n    description?: string;\n}',
@@ -785,6 +867,10 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'SandboxPolicy',
     declaration: 'export interface SandboxPolicy {\n    mode: ConfinedSandboxMode;\n    workspaceRoot: string;\n}',
   },
+  {
+    name: 'SaveTextSpill',
+    declaration: 'export interface SaveTextSpill {\n    owner: SpillOwner;\n    source: SpillSource;\n    suggestedName: string;\n    content: string;\n}',
+  },
   {
     name: 'ScopeKey',
     declaration: 'export type ScopeKey = object;',
@@ -799,7 +885,7 @@ export const TYPE_API: readonly TypeApiEntry[] = [
   },
   {
     name: 'SessionEventMap',
-    declaration: 'export interface SessionEventMap {\n    \'turn/start\': {\n        turn: number;\n        trigger: TurnTrigger;\n    };\n    \'turn/end\': {\n        turn: number;\n        reason: TurnEndReason;\n    };\n    \'step/start\': {\n        turn: number;\n        step: number;\n    };\n    \'step/end\': {\n        turn: number;\n        step: number;\n    };\n    \'user/message\': {\n        content: ContentBlock[];\n        source: MessageSource;\n    };\n    \'prompt/blocked\': {\n        content: ContentBlock[];\n        source: MessageSource;\n        reason: string;\n    };\n    \'context/message\': {\n        content: ContentBlock[];\n        source: MessageSource;\n    };\n    \'assistant/chunk\': {\n        turn: number;\n        step: number;\n        chunk: StreamChunk;\n    };\n    \'assistant/message\': {\n        turn: number;\n        step: number;\n        content: ContentBlock[];\n        usage?: TokenUsage;\n    };\n    \'tool/call\': {\n        turn: number;\n        step: number;\n        callId: CallId;\n        name: string;\n        arguments: string;\n    };\n    \'tool/result\': {\n        turn: number;\n        step: number;\n        callId: CallId;\n        content: ContentBlock[];\n        isError: boolean;\n        error?: {\n            name: string;\n            code: string;\n        };\n        meta?: unknown;\n    };\n    \'steering/message\': {\n        turn: number;\n        content: ContentBlock[];\n        source: MessageSource;\n    };\n    \'todo/write\': {\n        todos: TodoItem[];\n    };\n    \'request/header\': {\n        header: E /* …truncated — full shape in source */',
+    declaration: 'export interface SessionEventMap {\n    \'turn/start\': {\n        turn: number;\n        trigger: TurnTrigger;\n    };\n    \'turn/end\': {\n        turn: number;\n        reason: TurnEndReason;\n    };\n    \'step/start\': {\n        turn: number;\n        step: number;\n    };\n    \'step/end\': {\n        turn: number;\n        step: number;\n    };\n    \'user/message\': {\n        content: ContentBlock[];\n        source: MessageSource;\n    };\n    \'prompt/blocked\': {\n        content: ContentBlock[];\n        source: MessageSource;\n        reason: string;\n    };\n    \'context/message\': {\n        content: ContentBlock[];\n        source: MessageSource;\n        envelope?: ContextEnvelope;\n        meta?: JsonValue;\n    };\n    \'assistant/chunk\': {\n        turn: number;\n        step: number;\n        chunk: StreamChunk;\n    };\n    \'assistant/message\': {\n        turn: number;\n        step: number;\n        content: ContentBlock[];\n        provenance: AssistantProvenance;\n        usage?: TokenUsage;\n    };\n    \'tool/call\': {\n        turn: number;\n        step: number;\n        callId: CallId;\n        name: string;\n        arguments: string;\n    };\n    \'tool/result\': {\n        turn: number;\n        step: number;\n        callId: CallId;\n        content: ContentBlock[];\n        isError: boolean;\n        error?: {\n            name: string;\n            code: string;\n        };\n        meta?: unknown;\n    };\n    \'steering/message\': {\n        turn: number;\n        content: ContentBlock[];\n        source: MessageSource; /* …truncated — full shape in source */',
   },
   {
     name: 'SessionEventReadRequest',
@@ -813,6 +899,14 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'SessionEventSurface',
     declaration: 'export type SessionEventSurface = \'current\' | \'shadowed\' | \'log-only\';',
   },
+  {
+    name: 'SessionEventTrace',
+    declaration: 'export interface SessionEventTrace {\n    target: SessionEventRecord;\n    replacedBy?: number;\n    replacementChain: number[];\n    replacedEventSeqs: number[];\n    sourceEventSeqs: number[];\n    derivedEventSeqs: number[];\n}',
+  },
+  {
+    name: 'SessionEventTraceRequest',
+    declaration: 'export interface SessionEventTraceRequest {\n    sessionId: SessionId;\n    seq: number;\n}',
+  },
   {
     name: 'SessionEventType',
     declaration: 'export type SessionEventType = keyof SessionEventMap;',
@@ -833,6 +927,18 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'SessionId',
     declaration: 'export type SessionId = Branded<\'SessionId\'>;',
   },
+  {
+    name: 'SessionLineageNode',
+    declaration: 'export interface SessionLineageNode {\n    session: SessionRecord;\n    descendants: SessionLineageNode[];\n}',
+  },
+  {
+    name: 'SessionLineageTrace',
+    declaration: 'export type SessionLineageTrace = {\n    target: SessionRecord;\n    ancestors: SessionRecord[];\n    descendants: SessionLineageNode[];\n} & ({\n    complete: true;\n    root: SessionRecord;\n} | {\n    complete: false;\n    unresolvedParentId: SessionId;\n});',
+  },
+  {
+    name: 'SessionLocation',
+    declaration: 'export interface SessionLocation {\n    readonly kind: string;\n    readonly path: string;\n}',
+  },
   {
     name: 'SessionRecord',
     declaration: 'export interface SessionRecord {\n    header: SessionHeader;\n    live: boolean;\n    persisted: boolean;\n}',
@@ -869,9 +975,25 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'SkillSummary',
     declaration: 'export interface SkillSummary {\n    readonly name: string;\n    readonly description: string;\n    readonly whenToUse?: string;\n    readonly disableModelInvocation?: boolean;\n    readonly source: SkillSource;\n    readonly provider: string;\n    readonly resourceBase?: SkillResourceBase;\n}',
   },
+  {
+    name: 'SpillLocator',
+    declaration: 'export type SpillLocator = Branded<\'SpillLocator\'>;',
+  },
+  {
+    name: 'SpillOwner',
+    declaration: 'export interface SpillOwner {\n    sessionId: SessionId;\n}',
+  },
+  {
+    name: 'SpillRef',
+    declaration: 'export interface SpillRef {\n    locator: SpillLocator;\n    bytes: number;\n    retrievalHint: string;\n}',
+  },
+  {
+    name: 'SpillSource',
+    declaration: 'export interface SpillSource {\n    toolName: string;\n    callId: CallId;\n    label: string;\n}',
+  },
   {
     name: 'StreamChunk',
-    declaration: 'export type StreamChunk = {\n    type: \'block-start\';\n    index: number;\n    blockType: ContentBlockType;\n} | {\n    type: \'text-delta\';\n    index: number;\n    text: string;\n} | {\n    type: \'reasoning-delta\';\n    index: number;\n    text: string;\n} | {\n    type: \'tool-call-delta\';\n    index: number;\n    id: CallId;\n    name?: string;\n    argumentsDelta: string;\n} | {\n    type: \'block-end\';\n    index: number;\n    block: ContentBlock;\n} | {\n    type: \'usage\';\n    usage: TokenUsage;\n} | {\n    type: \'finish\';\n    reason: FinishReason;\n};',
+    declaration: 'export type StreamChunk = {\n    type: \'block-start\';\n    index: number;\n    blockType: ContentBlockType;\n} | {\n    type: \'text-delta\';\n    index: number;\n    text: string;\n} | {\n    type: \'reasoning-delta\';\n    index: number;\n    text: string;\n} | {\n    type: \'tool-call-delta\';\n    index: number;\n    id: CallId;\n    name?: string;\n    argumentsDelta: string;\n} | {\n    type: \'block-end\';\n    index: number;\n    block: ContentBlock;\n} | {\n    type: \'usage\';\n    usage: TokenUsage;\n} | {\n    type: \'finish\';\n    reason: FinishReason;\n    replayState?: unknown;\n};',
   },
   {
     name: 'StructuredOutputSchema',
@@ -925,6 +1047,46 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'SurfaceOp',
     declaration: 'export type SurfaceOp = \'append\' | {\n    op: \'replace\';\n    start: number;\n    end: number;\n};',
   },
+  {
+    name: 'TaskDoneListener',
+    declaration: 'export type TaskDoneListener = (snapshot: TaskSnapshot, owner: Agent | undefined) => void | PromiseLike<void>;',
+  },
+  {
+    name: 'TaskHooks',
+    declaration: 'export interface TaskHooks {\n    cancel(reason?: string): void;\n    done: Promise<TaskOutcome>;\n    readOutput?(): string;\n}',
+  },
+  {
+    name: 'TaskId',
+    declaration: 'export type TaskId = Branded<\'TaskId\'>;',
+  },
+  {
+    name: 'TaskKind',
+    declaration: 'export type TaskKind = TaskKindMap[keyof TaskKindMap];',
+  },
+  {
+    name: 'TaskKindMap',
+    declaration: 'export interface TaskKindMap {\n    bash: \'bash\';\n    subagent: \'subagent\';\n}',
+  },
+  {
+    name: 'TaskOutcome',
+    declaration: 'export interface TaskOutcome {\n    status: \'completed\' | \'killed\' | \'failed\';\n    detail?: string;\n    output?: string;\n}',
+  },
+  {
+    name: 'TaskRead',
+    declaration: 'export interface TaskRead {\n    text: string;\n    snapshot: TaskSnapshot;\n}',
+  },
+  {
+    name: 'TaskSnapshot',
+    declaration: 'export interface TaskSnapshot {\n    id: TaskId;\n    kind: TaskKind;\n    label: string;\n    ownerSession?: SessionId;\n    status: TaskStatus;\n    detail?: string;\n    startedAt: number;\n    finishedAt?: number;\n    reported: boolean;\n}',
+  },
+  {
+    name: 'TaskStart',
+    declaration: 'export interface TaskStart {\n    kind: TaskKind;\n    label: string;\n    owner?: Agent;\n    run(): TaskHooks;\n}',
+  },
+  {
+    name: 'TaskStatus',
+    declaration: 'export type TaskStatus = \'running\' | \'stopping\' | \'completed\' | \'killed\' | \'failed\';',
+  },
   {
     name: 'TerminalCallView',
     declaration: 'export interface TerminalCallView {\n    card: \'terminal\';\n    title: string;\n    description?: string;\n    cwd?: string;\n}',
@@ -934,8 +1096,16 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     declaration: 'export interface TerminalResultView {\n    card: \'terminal\';\n    title?: string;\n    output?: string;\n    exitCode?: number;\n    signal?: string;\n}',
   },
   {
-    name: 'TodoItem',
-    declaration: 'export interface TodoItem {\n    content: string;\n    status: \'pending\' | \'in_progress\' | \'completed\';\n}',
+    name: 'TokenMeasurement',
+    declaration: 'export interface TokenMeasurement {\n    readonly logRevision: number;\n    readonly baseline: TokenMeasurementBaseline;\n    readonly surfaceDeltaTokens: number;\n    readonly totalTokens: number;\n    readonly surfaceTokens: number;\n    readonly nodes: readonly TokenSurfaceNode[];\n}',
+  },
+  {
+    name: 'TokenMeasurementBaseline',
+    declaration: 'export type TokenMeasurementBaseline = {\n    readonly kind: \'none\';\n    readonly tokens: 0;\n} | {\n    readonly kind: \'estimated\';\n    readonly tokens: number;\n} | {\n    readonly kind: \'usage\';\n    readonly tokens: number;\n    readonly usage: Readonly<TokenUsage>;\n};',
+  },
+  {
+    name: 'TokenSurfaceNode',
+    declaration: 'export interface TokenSurfaceNode {\n    readonly seq: number;\n    readonly tokens: number;\n}',
   },
   {
     name: 'TokenUsage',
@@ -955,7 +1125,7 @@ export const TYPE_API: readonly TypeApiEntry[] = [
   },
   {
     name: 'ToolDefinition',
-    declaration: 'export interface ToolDefinition extends ToolSchema {\n    execute(args: unknown, exec: ToolExecution): Promise<ToolExecuteReturn>;\n    timeoutMs?: number;\n    presentCall?(args: unknown): ToolCallView | undefined;\n    presentResult?(args: unknown, result: ToolResult): ToolResultView | undefined;\n}',
+    declaration: 'export interface ToolDefinition extends ToolSchema {\n    execute(args: unknown, exec: ToolRunContext): Promise<ToolExecuteReturn>;\n    timeoutMs?: number;\n    isConcurrencySafe?(args: unknown): boolean;\n    presentCall?(args: unknown): ToolCallView | undefined;\n    presentResult?(args: unknown, result: ToolResult): ToolResultView | undefined;\n}',
   },
   {
     name: 'ToolErrorInfo',
@@ -973,9 +1143,13 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'ToolExecutionInput',
     declaration: 'export interface ToolExecutionInput {\n    readonly callId: CallId;\n    readonly name: string;\n    readonly arguments: unknown;\n    readonly agent?: Agent;\n    readonly parent?: ToolExecutionToken;\n    signal?: AbortSignal;\n}',
   },
+  {
+    name: 'ToolExecutionMode',
+    declaration: 'export type ToolExecutionMode = {\n    kind: \'parallel\';\n} | {\n    kind: \'exclusive\';\n};',
+  },
   {
     name: 'ToolExecutionResult',
-    declaration: 'export interface ToolExecutionResult {\n    content: ContentBlock[];\n    isError: boolean;\n    error?: ToolErrorInfo;\n    additionalContext?: HookContext;\n    meta?: unknown;\n}',
+    declaration: 'export interface ToolExecutionResult {\n    content: ContentBlock[];\n    isError: boolean;\n    error?: ToolErrorInfo;\n    additionalContexts?: HookContext[];\n    meta?: unknown;\n}',
   },
   {
     name: 'ToolExecutionToken',
@@ -1005,6 +1179,10 @@ export const TYPE_API: readonly TypeApiEntry[] = [
     name: 'ToolResultView',
     declaration: 'export type ToolResultView = GenericResultView | TerminalResultView | DiffResultView;',
   },
+  {
+    name: 'ToolRunContext',
+    declaration: 'export interface ToolRunContext extends ToolExecution {\n    deferContext(context: HookContext): void;\n}',
+  },
   {
     name: 'ToolSchema',
     declaration: 'export interface ToolSchema {\n    name: string;\n    description: string;\n    parameters: Record<string, unknown>;\n}',
@@ -1067,7 +1245,7 @@ export const TYPE_API: readonly TypeApiEntry[] = [
   },
   {
     name: 'WorkflowPhase',
-    declaration: 'export interface WorkflowPhase {\n    title: string;\n    detail?: string;\n    model?: string;\n}',
+    declaration: 'export interface WorkflowPhase {\n    title: string;\n    detail?: string;\n    provider?: string;\n    model?: string;\n}',
   },
   {
     name: 'WorkflowResult',
diff --git a/packages/cordis/tool-cordis/tests/integration.spec.ts b/packages/cordis/tool-cordis/tests/integration.spec.ts
index 94331df7c0..9823e119b5 100644
--- a/packages/cordis/tool-cordis/tests/integration.spec.ts
+++ b/packages/cordis/tool-cordis/tests/integration.spec.ts
@@ -1,11 +1,8 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import { AgentId } from '@deepseek-ai/dsh-agent'
 import AgentLoop, { ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as ToolCordis from '../src/index.ts'
 import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
 import { REVERSE_TOOL_CODE } from './helpers.ts'
@@ -20,11 +17,7 @@ import { REVERSE_TOOL_CODE } from './helpers.ts'
 
 async function harness(adapter: MockAdapter): Promise<Context> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(ToolCordis)
   ctx.llm.registerAdapter(['mock'], adapter)
@@ -51,7 +44,7 @@ describe('cordis tools through the agent loop', () => {
       textResponse('Done.'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('it-cordis'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('it-cordis'), { provider: 'mock', model: 'mock' })
 
     agent.send([{ type: 'text', text: 'give yourself reverse_text, use it, clean up' }])
     await waitForIdle(ctx, agent)
diff --git a/packages/core/README.md b/packages/core/README.md
index 921591d85e..d2d0c60fda 100644
--- a/packages/core/README.md
+++ b/packages/core/README.md
@@ -15,4 +15,4 @@ The session log, system-prompt assembly, tool registry, agent vocabulary, and co
 
 `agent-loop` is the one concrete implementation of the `agent` seam and lives here because it is the harness's default product loop; everything else in `core/` is interface/vocabulary. Plugins depend on the `agent` vocabulary, never on `agent-loop` directly, so the loop stays swappable.
 
-The default composition that wires this spine into a runnable agent lives in [`examples/agent-spine-demo`](../examples/agent-spine-demo/README.md): one bundle plugin that loads the control spine plus selected default capabilities (`timer` + `llm` + sessions + system-prompt + tools + agents + invariants + the local [skill family](../skill/README.md) + `tool-bash` + `agent-loop`) and forwards `agent-loop`'s `agents` list as its own config. It sits in `examples/` — ready-to-run demo/reference bundles — not in `core/`: `core/` ships the swappable spine pieces, while a demo bundle picks one concrete composition of them and adds a front door.
+The default composition that wires this spine into a runnable agent lives in [`examples/agent-spine-demo`](../examples/agent-spine-demo/README.md): one bundle plugin that loads the control spine plus selected default capabilities (`timer` + `llm` + sessions + system-prompt + tools + agents + invariants + the local [skill family](../skill/README.md) + `tool-bash` + workspace-context + `agent-loop`) and forwards `agent-loop`'s `agents` list as its own config. It sits in `examples/` — ready-to-run demo/reference bundles — not in `core/`: `core/` ships the swappable spine pieces, while a demo bundle picks one concrete composition of them and adds a front door.
diff --git a/packages/core/agent-loop/README.md b/packages/core/agent-loop/README.md
index 91ed9d5049..cdbf2e49a2 100644
--- a/packages/core/agent-loop/README.md
+++ b/packages/core/agent-loop/README.md
@@ -29,8 +29,10 @@ The config-driven `ctx.agentLoop.create()` path keeps its agent owned by the loo
 
 ```ts
 interface Config {
+  maxParallelToolCalls?: number // default 10; 1 is serial
   agents: Array<{
     id: string                 // required
+    provider?: string
     model?: string
     resumeSessionId?: string   // load this persisted session instead of creating one
     cwd?: string               // optional workspace cwd for the fresh session
@@ -38,27 +40,31 @@ interface Config {
 }
 ```
 
-Configured agents start automatically. `cwd` applies only to fresh sessions; `resumeSessionId` retains persisted metadata. They use the deployment persona. Programmatic setup can shadow it per agent. This plugin supplies the per-agent `model` and `cwd` prompt variables; harness identity and deployment persona belong to `dsh-system-prompt`.
+Configured agents start automatically. A model call requires both `provider` and `model`; `agent/request` may supply a missing pair before dispatch. `maxParallelToolCalls` bounds every agent's rolling pool for parallel-safe calls and defaults to `10`. `cwd` applies only to fresh sessions, while `resumeSessionId` retains persisted metadata. Configured agents use the deployment persona, and programmatic setup can shadow it per agent. This plugin supplies the per-agent `provider`, `model`, and `cwd` prompt variables; harness identity and deployment persona belong to `dsh-system-prompt`.
 
 ### Exported concrete class
 
 - `ReactLoopAgent` — the concrete `Agent` implementation. Its inbox is a JavaScript native-private field, and one prepared session can be claimed by only one concrete driver. Everything observable happens through session events and the `agent/*` event taxonomy.
 
-`Inbox`, `runLoop`, and the instance-bound publication/start controls are package-internal. The package root does not export them, and the package exports map exposes no `./src/*` escape hatch; lifecycle owners create agents through `ctx.agents` rather than constructing or starting the driver internals. `ReactLoopAgent.send()` and running `steer()` materialize content plus resolved source once as detached, deeply frozen lossless JSON, then share that accepted record between `agent/queued` and the inbox; malformed data throws before either boundary.
+`Inbox`, `runLoop`, and the instance-bound publication/start controls are package-internal. The package root does not export them, and the package exports map exposes no `./src/*` escape hatch; lifecycle owners create agents through `ctx.agents` rather than constructing or starting the driver internals. `ReactLoopAgent.send()`, running `steer()`, and open-turn `inject()` materialize content plus resolved source once as detached, deeply frozen lossless JSON; malformed data throws before enqueue or append. An injection that arrives while the current step executes assistant tool calls stays in a FIFO until the batch settles; successful batches place it after the complete result batch, and interrupted batches drain it before the turn closes.
 
 ### Loop lifecycle (`loop.ts`)
 
 The driver owns one agent for its lifetime. It records turn, step, request, stream, and tool boundaries in the session log; live extension events coordinate policy around those durable facts. The [architecture turn flow](../../../docs/architecture.md#turn-flow) and generated [event catalog](../../../docs/cordis-catalog/events.md) are the authoritative sequence and signatures.
 
+Every provider call that reaches a successful finish appends exactly one `assistant/message` completion anchor, including content-less calls and `max-tokens` finishes. A successful `agent/step-result` stores its transformed content; a rejected result records empty content before the original failure continues. The anchor retains exact chunk provenance (`[]` for a stream with no chunks) and usage when available, while empty content stays out of derived message history.
+
 Plugin failure ends the current turn, not the loop. Cancellation clears pending work and aborts the current step without leaking to the next prompt. Terminal continuation stops remain authoritative through turn close and durability flush.
 
+Within a step, exclusive calls form barriers; parallel-safe calls use a bounded rolling pool and are reclassified before start. Only dispatch/body overlaps. Policy, durable results, and result context remain model-ordered. Abort stops new calls, drains started results, then drains accepted batch context before the turn closes through the normal abort path.
+
 ### What belongs to plugins
 
 Everything that goes beyond "call the model, run the tools, repeat" belongs to plugins listening on the event taxonomy:
 - Hooks and policy: the relevant `agent/*` checkpoints plus the guarded `tools/pre-execute` → `tools/execute` → `tools/post-execute` → `tools/result` pipeline; exact signatures and modes live in the [generated event catalog](../../../docs/cordis-catalog/events.md)
 - Compaction: `agent/pre-step`
 - Sandbox, permission, plan mode: `tools/pre-execute` for extensible deny/ask, `tools.guard()` for monotonic owner policy, `tools/post-execute` for result decisions, and `tools/result` for final observation
-- Sub-agents: implemented outside the loop as `ctx.subagents` providers; in-process providers use `ctx.agents.create()` and owned `AgentHandle` teardown, while child streaming/progress and background/poll collection remain deferred.
+- Sub-agents: implemented outside the loop as `ctx.subagents` providers; in-process providers use `ctx.agents.create()` and owned `AgentHandle` teardown, while generic [`ctx.tasks`](../../tasks/tasks/) plus [`dsh-tool-subagent`](../../subagent/tool-subagent/) own background collection.
 - Persistence: `session/event` + `session/flush`
 - UI: `session/event` (assistant token stream, boundaries, tool activity) + `agent/*` control events (`agent/status`, `agent/created`/`agent/disposed`)
 
@@ -78,7 +84,7 @@ Everything that goes beyond "call the model, run the tools, repeat" belongs to p
 
 ## Known Limitations and Deferred Work
 
-- **Tool calls within a step execute sequentially** — parallel execution waits on concurrency-safety metadata in the tool contract (see `dsh-tools`).
+- **Classification is unary** — calls whose safety depends on comparing siblings or resources must remain exclusive ([rationale](../../../docs/rfc/implemented/feature/2026-07-10-parallel-tool-call-execution.md)).
 - **No resume-or-create policy on the config path** — config-driven `create()` starts a fresh `${id}-session-<uuid>` every run (`TODO(demo)`), and a config `resumeSessionId` whose resume fails logs a warning and creates no agent.
 - **Config agents have no per-agent persona field or setup hook** — they use the deployment persona; scoped persona/tool composition is available only through the programmatic `ctx.agents.create()` / `resume()` factory options.
 - **No built-in turn budget** — the default continuation is `continue` whenever a step had tool calls or steering; bounding a runaway turn requires an `agent/turn-continuation` force-stop plugin.
diff --git a/packages/core/agent-loop/src/agent.ts b/packages/core/agent-loop/src/agent.ts
index a2288c65b7..c8784453b9 100644
--- a/packages/core/agent-loop/src/agent.ts
+++ b/packages/core/agent-loop/src/agent.ts
@@ -8,7 +8,7 @@
 
 import type { Context } from 'cordis'
 import { agentEvents } from '@deepseek-ai/dsh-agent'
-import type { AgentId, AgentOptions, AgentStatus, SendOptions } from '@deepseek-ai/dsh-agent'
+import type { AgentId, AgentOptions, AgentStatus, HookContext, InjectOptions, SendOptions } from '@deepseek-ai/dsh-agent'
 import type { Agent } from '@deepseek-ai/dsh-agent'
 import { deepFreeze } from '@deepseek-ai/dsh-llm'
 import type { ContentBlock, MessageSource } from '@deepseek-ai/dsh-llm'
@@ -54,15 +54,16 @@ export interface PreparedReactLoopAgent {
  * @param id - the concrete agent identity.
  * @param options - loop options for the agent.
  * @param session - the prepared session the agent will own.
+ * @param maxParallelToolCalls - resolved in-flight cap for this agent.
  * @returns the agent and closures bound only to that exact instance.
  */
 export function prepareReactLoopAgent(
-  ctx: Context, id: AgentId, options: AgentOptions, session: Session,
+  ctx: Context, id: AgentId, options: AgentOptions, session: Session, maxParallelToolCalls: number,
 ): PreparedReactLoopAgent {
   if (claimedDriverSessions.has(session)) {
     throw new Error(`session "${session.id}" already has a concrete agent driver`)
   }
-  const agent = new ReactLoopAgent(ctx, id, options, session)
+  const agent = new ReactLoopAgent(ctx, id, options, session, maxParallelToolCalls)
   claimedDriverSessions.add(session)
   const dispose = () => agent[stopDriver]()
   return {
@@ -143,19 +144,27 @@ export class ReactLoopAgent implements Agent {
    * the `disposed` transition fires and leave the promise hanging.
    */
   private idleWaiters: (() => void)[] = []
+  /** Maximum parallel-safe calls allowed in one step. */
+  private readonly maxParallelToolCalls: number
   /**
    * Durability checkpoints started by idle {@link inject} calls. `inject()` is
    * synchronous, so it cannot await them itself; the driver disposer drains
    * this set before the lifecycle unregisters the agent or detaches its session.
    */
   private pendingIdleFlushes = new Set<Promise<void>>()
+  /** Whether the current step is executing an assistant tool-call batch. */
+  private toolBatchActive = false
+  /** Open-turn injections waiting for the active assistant tool-call batch to close. */
+  private deferredInjections: HookContext[] = []
 
   constructor(
     private loopCtx: Context,
     public readonly id: AgentId,
     public readonly options: AgentOptions,
     public readonly session: Session,
+    maxParallelToolCalls: number,
   ) {
+    this.maxParallelToolCalls = maxParallelToolCalls
     const { promise, resolve } = Promise.withResolvers<void>()
     this.disposed = promise
     this.resolveDisposed = resolve
@@ -189,12 +198,11 @@ export class ReactLoopAgent implements Agent {
   }
 
   /**
-   * Accept one public send/steer payload as the exact detached record shared by
-   * the live notification and inbox. Lossless-JSON materialization reads every
-   * nested field once; deep freeze prevents an observer from rewriting queued
-   * work before the loop drains it.
+   * Accept one public message payload as a detached record. Lossless-JSON
+   * materialization reads every nested field once; deep freeze prevents later
+   * caller mutation before an inbox or deferred-injection queue drains it.
    */
-  private acceptInboxMessage(content: ContentBlock[], options?: SendOptions): InboxMessage {
+  private acceptMessage(content: ContentBlock[], options?: SendOptions): InboxMessage {
     const source = this.resolveSource(options)
     const accepted = snapshotJsonValue({ content, source })
     if (accepted === undefined) {
@@ -203,6 +211,15 @@ export class ReactLoopAgent implements Agent {
     return deepFreeze(accepted)
   }
 
+  /** Detach one context before it can outlive its caller in the active-batch FIFO. */
+  private acceptContext(context: HookContext): HookContext {
+    const accepted = snapshotJsonValue(context)
+    if (accepted === undefined) {
+      throw new TypeError('agent context must be losslessly JSON-serializable')
+    }
+    return deepFreeze(accepted)
+  }
+
   /** Reject a driving operation once teardown has synchronously closed the agent. */
   private assertNotDisposed(): void {
     if (this._status === 'disposed') throw new Error(`agent "${this.id}" is disposed`)
@@ -210,7 +227,7 @@ export class ReactLoopAgent implements Agent {
 
   send(content: ContentBlock[], options?: SendOptions): void {
     this.assertNotDisposed()
-    const accepted = this.acceptInboxMessage(content, options)
+    const accepted = this.acceptMessage(content, options)
     this.#inbox.enqueue(accepted)
     const info = { source: accepted.source, steering: false } as const
     agentEvents(this.loopCtx, this).emit('agent/queued', accepted.content, info)
@@ -219,20 +236,31 @@ export class ReactLoopAgent implements Agent {
   steer(content: ContentBlock[], options?: SendOptions): void {
     this.assertNotDisposed()
     if (this._status !== 'running') { this.send(content, options); return }
-    const accepted = this.acceptInboxMessage(content, options)
+    const accepted = this.acceptMessage(content, options)
     this.#inbox.steer(accepted)
     const info = { source: accepted.source, steering: true } as const
     agentEvents(this.loopCtx, this).emit('agent/queued', accepted.content, info)
   }
 
-  inject(content: ContentBlock[], options?: SendOptions): void {
+  inject(content: ContentBlock[], options?: InjectOptions): void {
     this.assertNotDisposed()
     const source = this.resolveSource(options)
+    const context = {
+      content,
+      source,
+      ...options?.envelope !== undefined ? { envelope: options.envelope } : {},
+      ...options?.meta !== undefined ? { meta: options.meta } : {},
+    }
     if (isTurnOpen(this.session)) {
-      // A turn is open in the LOG (decided from the log, not agent status —
-      // status can be `running` with no turn open): the context/message is
-      // turn-enclosed by that turn, so append it directly.
-      this.session.append('context/message', { content, source }, { surfaceOp: 'append' })
+      const accepted = this.acceptContext(context)
+      // Provider protocols require every assistant tool-call batch to be
+      // followed only by its tool results. Historical interrupted batches do
+      // not own new context; only the currently executing batch may defer it.
+      if (this.toolBatchActive) {
+        this.deferredInjections.push(accepted)
+        return
+      }
+      this.session.append('context/message', accepted, { surfaceOp: 'append' })
       return
     }
     // No turn open: wrap the injection in a one-shot turn so every event stays
@@ -244,7 +272,7 @@ export class ReactLoopAgent implements Agent {
     // are contained by Session and cannot create a false append failure.
     try {
       this.session.append('turn/start', { turn, trigger: { kind: 'injection', source } })
-      this.session.append('context/message', { content, source }, { surfaceOp: 'append' })
+      this.session.append('context/message', context, { surfaceOp: 'append' })
     } finally {
       // Close the turn if turn/start made it into the log. A pre-commit veto
       // must escape rather than being mistaken for a committed turn/end.
@@ -272,6 +300,34 @@ export class ReactLoopAgent implements Agent {
     }
   }
 
+  /** Append deferred open-turn injections after the loop closes a tool-result batch. */
+  private drainDeferredInjections(): void {
+    const pending = this.deferredInjections.splice(0)
+    for (const accepted of pending) {
+      this.session.append('context/message', accepted, { surfaceOp: 'append' })
+    }
+  }
+
+  /**
+   * Run one tool-call batch and drain its deferred context before settlement.
+   * The loop-owned acceptor remains valid after public disposal begins because
+   * the interrupted turn stays open until this batch settles.
+   */
+  private async withToolBatch<T>(
+    run: (acceptContext: (context: HookContext) => void) => Promise<T>,
+  ): Promise<T> {
+    this.toolBatchActive = true
+    const acceptContext = (context: HookContext): void => {
+      this.deferredInjections.push(this.acceptContext(context))
+    }
+    try {
+      return await run(acceptContext)
+    } finally {
+      this.toolBatchActive = false
+      this.drainDeferredInjections()
+    }
+  }
+
   cancel(reason?: string): void {
     // Arm only for current work; an idle marker would cancel the next prompt.
     if (this._status === 'running' || this.currentAbort !== undefined || this.#inbox.hasQueued || this.#inbox.hasSteering) {
@@ -329,6 +385,7 @@ export class ReactLoopAgent implements Agent {
     this.driverStarted = true
     this.done = runLoop(this.loopCtx, this, {
       inbox: this.#inbox,
+      maxParallelToolCalls: this.maxParallelToolCalls,
       setStatus: (status) => { this.setStatus(status) },
       setAbort: controller => void (this.currentAbort = controller),
       disposed: this.disposed,
@@ -336,6 +393,7 @@ export class ReactLoopAgent implements Agent {
       isCancelled: () => this.cancelRequested,
       cancelReason: () => this.cancelReason,
       clearCancel: () => { this.cancelRequested = false },
+      withToolBatch: run => this.withToolBatch(run),
       // Pre-step cancellation re-parks without emitting a status transition.
       settleIdle: () => { this.settleIdleWaiters() },
     })
diff --git a/packages/core/agent-loop/src/constants.ts b/packages/core/agent-loop/src/constants.ts
new file mode 100644
index 0000000000..3f5510967a
--- /dev/null
+++ b/packages/core/agent-loop/src/constants.ts
@@ -0,0 +1,6 @@
+/** Shared agent-loop scheduler defaults.
+ * @module dsh-agent-loop/constants
+ */
+
+/** Default maximum in-flight parallel-safe calls per agent step. */
+export const DEFAULT_MAX_PARALLEL_TOOL_CALLS = 10
diff --git a/packages/core/agent-loop/src/index.ts b/packages/core/agent-loop/src/index.ts
index 6cd66f622c..5f5b9a9eb6 100644
--- a/packages/core/agent-loop/src/index.ts
+++ b/packages/core/agent-loop/src/index.ts
@@ -32,6 +32,7 @@ import {
   ReactLoopAgent,
 } from './agent.ts'
 import type { PreparedReactLoopAgent } from './agent.ts'
+import { DEFAULT_MAX_PARALLEL_TOOL_CALLS } from './constants.ts'
 
 export { ReactLoopAgent } from './agent.ts'
 
@@ -73,6 +74,15 @@ function signalAbortError(id: AgentId, signal: AbortSignal): Error {
   return new Error(`agent "${id}" creation aborted`, { cause: signal.reason })
 }
 
+/** Resolve the deployment-wide scheduler cap at the owning config boundary. */
+function resolveMaxParallelToolCalls(value: number | undefined): number {
+  const maxParallelToolCalls = value ?? DEFAULT_MAX_PARALLEL_TOOL_CALLS
+  if (!Number.isInteger(maxParallelToolCalls) || maxParallelToolCalls < 1) {
+    throw new Error('maxParallelToolCalls must be a positive integer')
+  }
+  return maxParallelToolCalls
+}
+
 /**
  * Caller-owned create/resume transaction through rollback-covered publication
  * and quiescent teardown. Resources remain private until the final registry
@@ -163,13 +173,13 @@ class AgentCreationTransaction {
   }
 
   /** Construct the driver and scope, then install their complete ordered lifecycle. */
-  prepare(options: AgentOptions, session: Session): ReactLoopAgent {
+  prepare(options: AgentOptions, session: Session, maxParallelToolCalls: number): ReactLoopAgent {
     this.assertActive()
     const gate = Promise.withResolvers<void>()
     this.preparing = gate.promise
     try {
       this.session = session
-      const driver = prepareReactLoopAgent(this.loopCtx, this.id, options, session)
+      const driver = prepareReactLoopAgent(this.loopCtx, this.id, options, session, maxParallelToolCalls)
       this.driver = driver
       const agent = driver.agent
       const scope = createScope(this.loopCtx, agent)
@@ -318,8 +328,15 @@ declare module 'cordis' {
   }
 }
 
-/** Plugin configuration for declarative startup agents. */
+export { DEFAULT_MAX_PARALLEL_TOOL_CALLS }
+
+/** Agent-loop plugin configuration. */
 export interface Config {
+  /**
+   * Maximum parallel-safe calls in flight per agent step. `1` is serial;
+   * omission defaults to {@link DEFAULT_MAX_PARALLEL_TOOL_CALLS}.
+   */
+  maxParallelToolCalls?: number
   /** Agents created or resumed at plugin startup. */
   agents: (AgentOptions & {
     /** Registry identity for the live agent. */
@@ -337,8 +354,10 @@ export class AgentLoop extends Service implements AgentFactory {
 
   /** Runtime schema for declarative agents. */
   static Config = z.object({
+    maxParallelToolCalls: z.number().step(1).min(1).default(DEFAULT_MAX_PARALLEL_TOOL_CALLS),
     agents: z.array(z.object({
       id: z.string().required(),
+      provider: z.string(),
       model: z.string(),
       cwd: z.string(),
       resumeSessionId: z.string(),
@@ -346,15 +365,19 @@ export class AgentLoop extends Service implements AgentFactory {
   }) as unknown as z<Config>
 
   private readonly ownership: FactoryOwnership
+  /** Resolved concurrency cap for every driver created by this factory. */
+  private readonly maxParallelToolCalls: number
   /** Plain holder prevents Cordis from re-tracing the factory's dependency context through a caller shadow. */
   private readonly runtime: { ctx: Context }
 
   constructor(ctx: Context, public config: Config) {
     super(ctx, 'agentLoop')
+    this.maxParallelToolCalls = resolveMaxParallelToolCalls(config.maxParallelToolCalls)
     this.ownership = new FactoryOwnership(ctx.fiber)
     this.runtime = { ctx }
     ctx.effect(() => () => this.ownership.dispose(), 'agentLoop.transactions()')
     ctx.effect(() => ctx.agents.setFactory(this), 'agentLoop.setFactory()')
+    ctx.systemPrompt.variable('provider', context => context.agent?.options.provider)
     ctx.systemPrompt.variable('model', context => context.agent?.options.model)
     ctx.systemPrompt.variable('cwd', context => context.agent?.session.header.cwd)
 
@@ -392,7 +415,7 @@ export class AgentLoop extends Service implements AgentFactory {
     try {
       const sessionId = SessionId(`${id}-session-${randomUUID()}`)
       const session = loopCtx.sessions.prepare(sessionId, { meta })
-      const agent = transaction.prepare(options, session)
+      const agent = transaction.prepare(options, session, this.maxParallelToolCalls)
       transaction.publish('startup')
       return agent
     } catch (error: unknown) {
@@ -410,6 +433,7 @@ export class AgentLoop extends Service implements AgentFactory {
    * @returns the published handle.
    */
   async createAgent(ownerCtx: Context, options: CreateAgentOptions): Promise<AgentHandle> {
+    const agentOptions = options.agentOptions ?? {}
     const transaction = new AgentCreationTransaction(
       this.runtime.ctx,
       ownerCtx,
@@ -422,7 +446,7 @@ export class AgentLoop extends Service implements AgentFactory {
         ...options.seed === undefined ? {} : { seed: options.seed },
         ...options.meta === undefined ? {} : { meta: options.meta },
       })
-      const agent = transaction.prepare(options.agentOptions ?? {}, session)
+      const agent = transaction.prepare(agentOptions, session, this.maxParallelToolCalls)
       await transaction.waitFor(options.setup?.(agent.ctx))
       transaction.assertActive()
       return transaction.publish('startup')
@@ -454,6 +478,7 @@ export class AgentLoop extends Service implements AgentFactory {
     persistence: SessionPersistence,
     options: ResumeAgentOptions,
   ): Promise<AgentHandle> {
+    const agentOptions = options.agentOptions ?? {}
     const transaction = new AgentCreationTransaction(
       this.runtime.ctx,
       ownerCtx,
@@ -473,7 +498,7 @@ export class AgentLoop extends Service implements AgentFactory {
           ...loaded.meta.seedLength === undefined ? {} : { seedLength: loaded.meta.seedLength },
         },
       })
-      const agent = transaction.prepare(options.agentOptions ?? {}, session)
+      const agent = transaction.prepare(agentOptions, session, this.maxParallelToolCalls)
       await transaction.waitFor(options.setup?.(agent.ctx))
       transaction.assertActive()
       return transaction.publish('resume')
diff --git a/packages/core/agent-loop/src/loop.ts b/packages/core/agent-loop/src/loop.ts
index 15cc0aa410..a153eba7e4 100644
--- a/packages/core/agent-loop/src/loop.ts
+++ b/packages/core/agent-loop/src/loop.ts
@@ -6,7 +6,8 @@
  */
 
 import type { Context } from 'cordis'
-import type { FinishReason, GenerateOptions, LlmCallConfig, Message } from '@deepseek-ai/dsh-llm'
+import type { ContentBlock, FinishReason, GenerateOptions, LlmCallConfig, Message } from '@deepseek-ai/dsh-llm'
+import { isDeepStrictEqual } from 'node:util'
 import { BlockAssembler, HarnessError, deepFreeze } from '@deepseek-ai/dsh-llm'
 import { agentEvents, assembleContextFor } from '@deepseek-ai/dsh-agent'
 import type { AgentEventDispatch, ContinuationDecision, HookContext, PromptDecision } from '@deepseek-ai/dsh-agent'
@@ -17,6 +18,7 @@ import type { TransmissionLog } from './request-log.ts'
 import { renderPrompt } from '@deepseek-ai/dsh-system-prompt'
 import type { PromptAssembly } from '@deepseek-ai/dsh-system-prompt'
 import type {} from '@deepseek-ai/dsh-tools'
+import { executeToolCalls } from './tool-calls.ts'
 import type { ReactLoopAgent } from './agent.ts'
 import type { Inbox } from './inbox.ts'
 
@@ -72,6 +74,8 @@ function stepFinishReason(finish: FinishReason): TurnEndReason | undefined {
 export interface LoopHandle {
   /** Native-private agent inbox handed to the driver only at internal startup. */
   readonly inbox: Inbox
+  /** Maximum parallel-safe calls allowed in one step. */
+  readonly maxParallelToolCalls: number
   setStatus(status: 'idle' | 'running'): void
   setAbort(controller: AbortController | undefined): void
   /** Resolves when the agent is disposed — unblocks the idle wait. */
@@ -85,6 +89,8 @@ export interface LoopHandle {
   clearCancel(): void
   /** Settle idle waiters when pre-running cancellation skips a turn, without emitting `agent/status`. */
   settleIdle(): void
+  /** Run an active tool-call batch, accepting post-tool context into the FIFO drained before settlement. */
+  readonly withToolBatch: <T>(run: (acceptContext: (context: HookContext) => void) => Promise<T>) => Promise<T>
 }
 
 /**
@@ -234,10 +240,15 @@ async function runTurn(
       // `allow.content` REPLACES the prompt bytes (a rewrite); absent keeps them.
       const content = decision.content ?? message.content
       session.append('user/message', { content, source: message.source }, { surfaceOp: 'append' })
-      // `allow.additionalContext` is a SEPARATE context/message the next request
-      // also sees. The turn is open, so inject() appends it into THIS turn.
-      if (decision.additionalContext) {
-        agent.inject(decision.additionalContext.content, { source: decision.additionalContext.source })
+      // Every `allow.additionalContexts` entry is a separate context/message the
+      // next request also sees. The turn is open, so inject() appends each one
+      // into THIS turn without flattening provenance, framing, or metadata.
+      for (const context of decision.additionalContexts ?? []) {
+        agent.inject(context.content, {
+          source: context.source,
+          ...context.envelope !== undefined ? { envelope: context.envelope } : {},
+          ...context.meta !== undefined ? { meta: context.meta } : {},
+        })
       }
     }
 
@@ -325,7 +336,7 @@ async function runTurn(
       let stepOutcome: { hadToolCalls: boolean; finish: FinishReason } | { error: Error }
       try {
         stepOutcome = await runStep(
-          ctx, events, agent, turn, step, assembly, fullSystemPrompt, boundaryMessages, transmission, abort.signal)
+          ctx, events, agent, handle, turn, step, assembly, fullSystemPrompt, boundaryMessages, transmission, abort.signal)
       } catch (error: unknown) {
         stepOutcome = { error: toError(error) }
       } finally {
@@ -462,6 +473,7 @@ async function runStep(
   ctx: Context,
   events: AgentEventDispatch,
   agent: ReactLoopAgent,
+  handle: LoopHandle,
   turn: number,
   step: number,
   assembly: PromptAssembly,
@@ -477,12 +489,12 @@ async function runStep(
   const seedConfig: LlmCallConfig = deepFreeze(structuredClone(transmission.loggedHeader
     // eslint-disable-next-line @typescript-eslint/no-non-null-assertion -- loggedHeader ⟹ a snapshot is in the log
     ? session.requestHeader()!.config
-    : { model: options.model ?? '' }))
+    : { provider: options.provider ?? '', model: options.model ?? '' }))
 
   // Listener replacements are recorded in the request header before dispatch.
   const config = await events.waterfall('agent/request', turn, step, seedConfig, () => Promise.resolve(seedConfig))
-  if (!config.model) {
-    throw new Error(`agent "${agent.id}" has no model: set AgentOptions.model or supply one via the agent/request waterfall`)
+  if (!config.provider || !config.model) {
+    throw new Error(`agent "${agent.id}" has no provider/model: set AgentOptions.provider and AgentOptions.model or supply both via the agent/request waterfall`)
   }
 
   // eslint-disable-next-line @typescript-eslint/no-non-null-assertion -- runTurn composes the prefix before every runStep call
@@ -499,6 +511,7 @@ async function runStep(
 
   // Freeze the logged header plus boundary snapshot; the prefix precedes derived history.
   const request: GenerateOptions = deepFreeze({
+    provider: header.config.provider,
     model: header.config.model,
     messages: [...header.messagePrefix ?? [], ...boundaryMessages],
     ...header.system !== undefined ? { system: header.system } : {},
@@ -526,81 +539,108 @@ async function runStep(
   if (stepError) throw stepError
 
   if (assembler.finish.kind === 'max-tokens') {
-    let message: Message = withoutToolCalls(assembler.message())
-    message = withoutToolCalls(await events.waterfall('agent/step-result', turn, step, message, () => Promise.resolve(message)))
+    const assembled = assembler.message()
+    const assembledContent = structuredClone(assembled.content)
+    let message: Message = withoutToolCalls(assembled)
+    message = withoutToolCalls(await processStepResult(
+      events, session, turn, step, header.config, assembledContent, message, assembler, chunkSeqs,
+    ))
     // Preserve usage even when max-token truncation produced no content.
-    if (message.content.length > 0 || assembler.usage) {
-      // The finish chunk guarantees non-empty provenance here.
-      session.append(
-        'assistant/message',
-        { turn, step, content: message.content, ...(assembler.usage ? { usage: assembler.usage } : {}) },
-        { surfaceOp: 'append', sourceEventSeqs: chunkSeqs },
-      )
-    }
+    recordAssistantMessage(session, turn, step, header.config, assembledContent, message, assembler, chunkSeqs)
     return { hadToolCalls: false, finish: assembler.finish }
   }
 
   // Record the post-waterfall message that tool dispatch uses.
-  let message: Message = assembler.message()
-  message = await events.waterfall('agent/step-result', turn, step, message, () => Promise.resolve(message))
+  const assembled = assembler.message()
+  const assembledContent = structuredClone(assembled.content)
+  let message: Message = assembled
+  message = await processStepResult(
+    events, session, turn, step, header.config, assembledContent, message, assembler, chunkSeqs,
+  )
 
-  // Empty messages exist only to carry usage; omit empty provenance.
-  if (message.content.length > 0 || assembler.usage) {
-    session.append(
-      'assistant/message',
-      { turn, step, content: message.content, ...(assembler.usage ? { usage: assembler.usage } : {}) },
-      { surfaceOp: 'append', ...(chunkSeqs.length > 0 ? { sourceEventSeqs: chunkSeqs } : {}) },
-    )
-  }
+  // Every successful call records its completion anchor, including explicit
+  // empty chunk provenance for a contentless, usage-less provider response.
+  recordAssistantMessage(session, turn, step, header.config, assembledContent, message, assembler, chunkSeqs)
 
-  // Tool execution stays sequential; recheck abort around each normalized result.
+  // Dispatch may overlap; policy, durable results, and result context stay model-ordered.
   const toolCalls = message.content.filter(block => block.type === 'tool-call')
-  // Buffer context until all results are appended to preserve call/result adjacency.
-  const pendingContext: HookContext[] = []
-  for (const call of toolCalls) {
-    /* v8 ignore next -- signal.reason always set: cancel()/disposal provide a default */
-    if (signal.aborted) throw new Error(String(signal.reason ?? 'aborted'))
-    const callEvent = session.append('tool/call', { turn, step, callId: call.id, name: call.name, arguments: call.arguments })
-    let parsedArguments: unknown
-    try {
-      parsedArguments = call.arguments ? JSON.parse(call.arguments) : {}
-    } catch {
-      parsedArguments = call.arguments
-    }
-    // TODO(pre-tool-input-rewrite): Keep logged history and live presentation aligned;
-    // see docs/rfc/proposed/feature/2026-06-30-pre-tool-input-rewrite.md.
-    const result = await ctx.tools.execute({
-      callId: call.id,
-      name: call.name,
-      arguments: parsedArguments,
-      agent,
-      signal,
-    })
-    session.append('tool/result', {
-      turn, step,
-      // Correlation comes from the immutable execution input; the result does
-      // not duplicate this authoritative transcript identity.
-      callId: call.id,
-      content: result.content,
-      isError: result.isError,
-      ...result.error ? { error: result.error } : {},
-      // Persist tool-owned presentation data for replay.
-      ...result.meta !== undefined ? { meta: result.meta } : {},
-    }, { surfaceOp: 'append', sourceEventSeqs: [callEvent.seq] })
-    if (result.additionalContext) pendingContext.push(result.additionalContext)
-    // The signal may flip while the tool is awaited.
-    /* v8 ignore start -- signal.reason default unreachable: cancel()/disposal always set it */
-    // eslint-disable-next-line @typescript-eslint/no-unnecessary-condition
-    if (signal.aborted) throw new Error(String(signal.reason ?? 'aborted'))
-    /* v8 ignore stop */
-  }
+  if (toolCalls.length === 0) return { hadToolCalls: false, finish: assembler.finish }
+  return handle.withToolBatch(async (acceptContext) => {
+    await executeToolCalls(
+      ctx, agent, turn, step, toolCalls, signal, handle.maxParallelToolCalls, acceptContext,
+    )
+    return { hadToolCalls: true, finish: assembler.finish }
+  })
+}
 
-  // Append buffered context after the complete result batch.
-  for (const context of pendingContext) {
-    agent.inject(context.content, { source: context.source })
+/** Preserve successful-call accounting without retaining output that result processing rejected. */
+async function processStepResult(
+  events: AgentEventDispatch,
+  session: Session,
+  turn: number,
+  step: number,
+  config: LlmCallConfig,
+  assembledContent: ContentBlock[],
+  message: Message,
+  assembler: BlockAssembler,
+  chunkSeqs: number[],
+): Promise<Message> {
+  try {
+    return await events.waterfall(
+      'agent/step-result', turn, step, message, () => Promise.resolve(message),
+    )
+  } catch (error: unknown) {
+    recordAssistantMessage(
+      session,
+      turn,
+      step,
+      config,
+      assembledContent,
+      { ...message, content: [] },
+      assembler,
+      chunkSeqs,
+      false,
+    )
+    throw error
   }
+}
 
-  return { hadToolCalls: toolCalls.length > 0, finish: assembler.finish }
+/** Record one content-or-usage assistant message with replay-safe provenance. */
+function recordAssistantMessage(
+  session: Session,
+  turn: number,
+  step: number,
+  config: LlmCallConfig,
+  assembledContent: ContentBlock[],
+  message: Message,
+  assembler: BlockAssembler,
+  chunkSeqs: number[],
+  preserveReplayState = true,
+): void {
+  session.append(
+    'assistant/message',
+    {
+      turn,
+      step,
+      content: message.content,
+      provenance: assistantProvenance(
+        config,
+        assembler.replayState,
+        preserveReplayState && isDeepStrictEqual(message.content, assembledContent),
+      ),
+      ...assembler.usage === undefined ? {} : { usage: assembler.usage },
+    },
+    { surfaceOp: 'append', sourceEventSeqs: chunkSeqs },
+  )
+}
+
+/** Build durable assistant provenance, dropping replay state after any content rewrite. */
+function assistantProvenance(config: LlmCallConfig, replayState: unknown, contentUnchanged: boolean): NonNullable<Message['provenance']> {
+  return {
+    provider: config.provider,
+    model: config.model,
+    ...contentUnchanged && replayState !== undefined ? { replayState } : {},
+  }
 }
 
 function withoutToolCalls(message: Message): Message {
diff --git a/packages/core/agent-loop/src/request-log.ts b/packages/core/agent-loop/src/request-log.ts
index c7314c2409..ea6141fea9 100644
--- a/packages/core/agent-loop/src/request-log.ts
+++ b/packages/core/agent-loop/src/request-log.ts
@@ -1,11 +1,12 @@
 /**
  * Per-loop-instance request-header bookkeeping for reconstructability. The
- * comparison baseline is the header folded from the session log, so a fresh
- * loop instance needs no special resume or fork state.
+ * comparison baseline is folded from the session log; a fresh instance anchors
+ * it with an initial/resume snapshot and later logs full changed snapshots.
+ *
  * @module dsh-agent-loop/request-log
  */
 
-import { diffHeader, headerEquals, applyHeaderDelta } from '@deepseek-ai/dsh-session'
+import { headerEquals } from '@deepseek-ai/dsh-session'
 import type { EpochHeader, Session } from '@deepseek-ai/dsh-session'
 import type { Message } from '@deepseek-ai/dsh-llm'
 
@@ -32,10 +33,8 @@ export function createTransmissionLog(): TransmissionLog {
 }
 
 /**
- * Append whatever header event makes the log reproduce this request's header.
- * The first request from an instance always records a full `initial` or `resume`
- * snapshot. Later requests record nothing when unchanged, a round-tripping
- * delta when expressible, or a full `fallback` snapshot otherwise.
+ * Append the full header snapshot owed by this request: initial/resume for the
+ * instance's first request, nothing when unchanged, or change otherwise.
  *
  * @param session - the session whose log explains the request.
  * @param state - this loop instance's bookkeeping (mutated on first log).
@@ -52,12 +51,5 @@ export function recordRequestHeader(session: Session, state: TransmissionLog, he
   // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
   const baseline = session.requestHeader()!
   if (headerEquals(baseline, header)) return
-  const delta = diffHeader(baseline, header)
-  /* v8 ignore next -- headerEquals false ⟹ diffHeader defined: both compare the same four parts */
-  if (delta === undefined) return
-  if (headerEquals(applyHeaderDelta(baseline, delta), header)) {
-    session.append('request/header-delta', delta)
-  } else {
-    session.append('request/header', { header, reason: 'fallback' })
-  }
+  session.append('request/header', { header, reason: 'change' })
 }
diff --git a/packages/core/agent-loop/src/tool-calls.ts b/packages/core/agent-loop/src/tool-calls.ts
new file mode 100644
index 0000000000..3f3581c70a
--- /dev/null
+++ b/packages/core/agent-loop/src/tool-calls.ts
@@ -0,0 +1,229 @@
+/**
+ * Schedules one assistant step's tool calls. Exclusive calls form barriers;
+ * parallel calls use a bounded rolling pool and are reclassified before start.
+ * Dispatch may overlap, while policy, results, and result context remain
+ * model-ordered. Abort stops replenishment and drains started calls.
+ *
+ * Each started call records `tool/call`; `tool/result` commits in model order,
+ * preserving derived history when audit events interleave with earlier results.
+ * @module dsh-agent-loop/tool-calls
+ */
+
+import type { Context } from 'cordis'
+import { assertNever, type ToolCallBlock } from '@deepseek-ai/dsh-llm'
+import type { HookContext } from '@deepseek-ai/dsh-agent'
+import type { Session } from '@deepseek-ai/dsh-session'
+import { TOOL_REGISTRY_SCHEDULER, type ToolExecutionInput, type ToolExecutionMode, type ToolExecutionResult, type ToolRunContext } from '@deepseek-ai/dsh-tools'
+import type { ReactLoopAgent } from './agent.ts'
+
+/** One tool call after argument parsing, ready to schedule. */
+interface PlannedCall {
+  block: ToolCallBlock
+  exec: ToolExecutionInput
+}
+
+/** Settled dispatch awaiting model-order finalization. */
+interface Slot {
+  exec: ToolRunContext
+  result: ToolExecutionResult
+  needsPost: boolean
+}
+
+/**
+ * Schedule one assistant step's tool calls by their live concurrency mode.
+ * Started calls receive ordered results. Abort drains them and rethrows after
+ * accepting their context into the batch FIFO owned by the caller.
+ *
+ * @param ctx - loop context that owns the tool registry.
+ * @param agent - agent and session receiving the call lifecycle.
+ * @param turn - current turn number.
+ * @param step - current step number.
+ * @param toolCalls - assistant calls in model order.
+ * @param signal - abort signal shared by the step.
+ * @param maxParallel - validated in-flight cap.
+ * @param acceptContext - accepts committed result context into the active batch.
+ */
+export async function executeToolCalls(
+  ctx: Context,
+  agent: ReactLoopAgent,
+  turn: number,
+  step: number,
+  toolCalls: ToolCallBlock[],
+  signal: AbortSignal,
+  maxParallel: number,
+  acceptContext: (context: HookContext) => void,
+): Promise<void> {
+  const { session } = agent
+
+  // Inputs are distinct because tools/execute wrappers may replace `exec.signal`.
+  const planned: PlannedCall[] = toolCalls.map(block => ({
+    block,
+    exec: {
+      callId: block.id,
+      name: block.name,
+      arguments: parseArguments(block.arguments),
+      agent,
+      signal,
+    },
+  }))
+
+  let next = 0
+  while (next < planned.length) {
+    // Commit before classifying again so registry changes affect unstarted calls.
+    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion -- bounded by the loop condition
+    const first = planned[next]!
+    const mode = ctx.tools.executionMode(first.exec).kind
+    const group = mode === 'parallel' ? planned.slice(next) : [first]
+    next += await runGroup(ctx, session, turn, step, group, mode, signal, maxParallel, acceptContext)
+  }
+}
+
+/** Parse model arguments, preserving invalid JSON as text and mapping empty input to `{}`. */
+function parseArguments(raw: string): unknown {
+  try {
+    return raw ? JSON.parse(raw) : {}
+  } catch {
+    return raw
+  }
+}
+
+/**
+ * Run one exclusive barrier or parallel pool. Later calls are reclassified
+ * before start; an exclusive reclassification waits for the current pool to
+ * drain and remains for the caller's next barrier. Results and contexts commit
+ * in model order. Abort stops starts, drains and commits started calls, accepts
+ * their contexts into the owning batch, and throws.
+ */
+async function runGroup(
+  ctx: Context,
+  session: Session,
+  turn: number,
+  step: number,
+  group: PlannedCall[],
+  mode: ToolExecutionMode['kind'],
+  signal: AbortSignal,
+  maxParallel: number,
+  acceptContext: (context: HookContext) => void,
+): Promise<number> {
+  /* v8 ignore next -- signal.reason always set: cancel()/disposal provide a default */
+  if (signal.aborted) throw new Error(String(signal.reason ?? 'aborted'))
+  const slots: (Slot | undefined)[] = group.map(() => undefined)
+  // Started slots retain their tool/call seq for result provenance.
+  const callSeqs: number[] = group.map(() => -1)
+  let nextToStart = 0
+  let committed = 0
+  let started = 0
+  let aborted: boolean = signal.aborted
+
+  // `committed` advances only across contiguous model-order slots.
+  const commitReady = async (): Promise<void> => {
+    while (committed < group.length) {
+      const slot = slots[committed]
+      if (slot === undefined) break
+      const call = group[committed]
+      const result = slot.needsPost
+        ? await ctx.tools[TOOL_REGISTRY_SCHEDULER].finalize(slot.exec, slot.result)
+        : ctx.tools[TOOL_REGISTRY_SCHEDULER].finish(slot.exec, slot.result)
+      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion -- bounded index
+      appendToolResult(session, turn, step, call!.block, result, callSeqs[committed]!)
+      for (const context of result.additionalContexts ?? []) acceptContext(context)
+      committed++
+    }
+  }
+
+  const inFlight = new Map<number, Promise<number>>()
+
+  const startCall = async (index: number): Promise<void> => {
+    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion -- bounded index
+    const call = group[index]!
+    callSeqs[index] = appendToolCall(session, turn, step, call.block)
+    started++
+    const prepared = await ctx.tools[TOOL_REGISTRY_SCHEDULER].prepare(call.exec)
+    switch (prepared.kind) {
+      case 'dispatch': {
+        const promise = ctx.tools[TOOL_REGISTRY_SCHEDULER].dispatch(prepared.exec).then((outcome) => {
+          slots[index] = { exec: prepared.exec, result: outcome.result, needsPost: outcome.kind === 'post-result' }
+          return index
+        })
+        inFlight.set(index, promise)
+        break
+      }
+      case 'post-result':
+        slots[index] = { exec: prepared.exec, result: prepared.result, needsPost: true }
+        break
+      case 'final-result':
+        slots[index] = { exec: prepared.exec, result: prepared.result, needsPost: false }
+        break
+      /* v8 ignore next -- closed-union exhaustiveness guard */
+      default:
+        assertNever(prepared, 'tool-call scheduler prepare result')
+    }
+  }
+
+  const fillPool = async (): Promise<void> => {
+    while (!aborted && nextToStart < group.length && inFlight.size < maxParallel) {
+      // Re-read later modes after ordered commits so registry changes can create a barrier.
+      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion -- bounded by the loop condition
+      const nextCall = group[nextToStart]!
+      if (nextToStart > 0 && mode === 'parallel'
+        && ctx.tools.executionMode(nextCall.exec).kind !== 'parallel') break
+      await startCall(nextToStart)
+      nextToStart++
+      await commitReady()
+      // Abort may arrive while pre-execute awaits.
+      if (signal.aborted) aborted = true
+    }
+  }
+
+  // Ordered pre-execute may await; only dispatch/body overlaps.
+  // TODO: Drain every started call before rethrowing a scheduler error; tool
+  // bodies must not outlive the failed turn.
+  await fillPool()
+  while (inFlight.size > 0) {
+    const settledIndex = await Promise.race(inFlight.values())
+    inFlight.delete(settledIndex)
+    await commitReady()
+    // Abort may arrive while a tool or ordered commit awaits.
+    // eslint-disable-next-line @typescript-eslint/no-unnecessary-condition
+    if (signal.aborted) aborted = true
+    await fillPool()
+  }
+
+  if (aborted) {
+    // Started calls and accepted context settle before the turn records the abort.
+    /* v8 ignore next -- signal.reason always set: cancel()/disposal provide a default */
+    throw new Error(String(signal.reason ?? 'aborted'))
+  }
+  /* v8 ignore next -- unreachable: a non-aborted group commits every started call */
+  if (committed !== started) throw new Error('tool-call scheduler: uncommitted settled calls')
+  return started
+}
+
+/** Append a started call and return its provenance sequence. */
+function appendToolCall(session: Session, turn: number, step: number, block: ToolCallBlock): number {
+  const event = session.append('tool/call', { turn, step, callId: block.id, name: block.name, arguments: block.arguments })
+  return event.seq
+}
+
+/** Append a model-ordered result linked to its call event. */
+function appendToolResult(
+  session: Session,
+  turn: number,
+  step: number,
+  block: ToolCallBlock,
+  result: ToolExecutionResult,
+  callSeq: number,
+): void {
+  session.append('tool/result', {
+    turn, step,
+    // Correlation stays with the loop's authoritative model-transcript call id;
+    // registry results deliberately do not duplicate it.
+    callId: block.id,
+    content: result.content,
+    isError: result.isError,
+    ...result.error ? { error: result.error } : {},
+    // The tool's private presentation payload (e.g. a result-time diff),
+    // persisted so a UI bridge reproduces the card on replay.
+    ...result.meta !== undefined ? { meta: result.meta } : {},
+  }, { surfaceOp: 'append', sourceEventSeqs: [callSeq] })
+}
diff --git a/packages/core/agent-loop/tests/agent.spec.ts b/packages/core/agent-loop/tests/agent.spec.ts
index 581b6fa207..8c2a550b12 100644
--- a/packages/core/agent-loop/tests/agent.spec.ts
+++ b/packages/core/agent-loop/tests/agent.spec.ts
@@ -6,7 +6,7 @@ import SessionStore, { SessionId } from '@deepseek-ai/dsh-session'
 import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
 import ToolRegistry from '@deepseek-ai/dsh-tools'
 import AgentRegistry from '@deepseek-ai/dsh-agent'
-import AgentLoop, { ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import AgentLoop, { DEFAULT_MAX_PARALLEL_TOOL_CALLS, ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
 import { bindReactLoopAgentContext, prepareReactLoopAgent } from '../src/agent.ts'
 import { MockAdapter, textResponse } from './mock-adapter.ts'
 
@@ -53,10 +53,14 @@ describe('ReactLoopAgent', () => {
     const ctx = new Context()
     await ctx.plugin(SessionStore)
     const session = ctx.sessions.create(SessionId('exclusive-driver'))
-    const prepared = prepareReactLoopAgent(ctx, AgentId('first-driver'), { model: 'mock' }, session)
+    const prepared = prepareReactLoopAgent(
+      ctx, AgentId('first-driver'), { provider: 'mock', model: 'mock' }, session, DEFAULT_MAX_PARALLEL_TOOL_CALLS,
+    )
 
     expect(() => prepared.agent.ctx).toThrow('context is not bound')
-    expect(() => prepareReactLoopAgent(ctx, AgentId('second-driver'), { model: 'mock' }, session))
+    expect(() => prepareReactLoopAgent(
+      ctx, AgentId('second-driver'), { provider: 'mock', model: 'mock' }, session, DEFAULT_MAX_PARALLEL_TOOL_CALLS,
+    ))
       .toThrow('already has a concrete agent driver')
 
     await prepared.dispose()
@@ -65,7 +69,7 @@ describe('ReactLoopAgent', () => {
 
   it('borrows caller options and binds its scoped context exactly once', async () => {
     const ctx = await harness(new MockAdapter([textResponse('unused')]))
-    const options = { model: 'mock' }
+    const options = { provider: 'mock', model: 'mock' }
     const agent = ctx.agentLoop.create(AgentId('owned-bindings'), options)
 
     expect(agent.options).toBe(options)
@@ -81,7 +85,7 @@ describe('ReactLoopAgent', () => {
     const ctx = await harness(adapter)
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('scoped'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('scoped'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
     send(agent, 'go')
     await new Promise(r => setTimeout(r, 30))
@@ -96,7 +100,7 @@ describe('ReactLoopAgent', () => {
     const ctx = await harness(adapter)
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('scoped'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('scoped'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
     send(agent, 'go')
     await new Promise(r => setTimeout(r, 30))
@@ -111,7 +115,7 @@ describe('ReactLoopAgent', () => {
     const ctx = await harness(adapter)
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('scoped'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('scoped'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
     send(agent, 'go')
     await new Promise(r => setTimeout(r, 30))
@@ -124,7 +128,7 @@ describe('ReactLoopAgent', () => {
   it('inject() decides enclosure from the LOG (open turn), not agent status', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // Simulate an OPEN turn in the log while the agent is idle (status is not a
     // reliable open-turn signal). inject must append into that open turn, NOT
@@ -150,7 +154,7 @@ describe('ReactLoopAgent', () => {
     // A persistence-like listener whose flush rejects.
     ctx.on('session/flush', () => { throw new Error('disk gone') })
     const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => undefined)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // inject() is synchronous and fires a fire-and-forget flush; a rejecting
     // flush must be contained (logged), never thrown into the caller.
@@ -163,7 +167,7 @@ describe('ReactLoopAgent', () => {
   it('idle inject() closes its one-shot turn AND still checkpoints even if the append throws', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     let flushes = 0
     ctx.on('session/flush', () => { flushes += 1 })
 
@@ -181,7 +185,7 @@ describe('ReactLoopAgent', () => {
   it('idle inject() still checkpoints when a listener throws on the synthetic turn/end', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     let flushes = 0
     ctx.on('session/flush', () => { flushes += 1 })
     // Session contains a throwing post-commit turn/end observer. The accepted
@@ -204,7 +208,7 @@ describe('ReactLoopAgent', () => {
     // A non-Error rejection exercises the String() normalization branch.
     ctx.on('session/flush', () => { throw 'disk gone' })
     const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => undefined)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     const errors: { turn: number; step: number; message: string }[] = []
     ctx.on('agent/error', (_a, turn, step, error) => void errors.push({ turn, step, message: error.message }))
 
@@ -223,7 +227,7 @@ describe('ReactLoopAgent', () => {
   it('idle inject() with a non-serializable source opens no turn (nothing to close)', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // A non-serializable source makes the turn/start append throw BEFORE the
     // event is pushed (Session.append validates before push), so NO turn opens.
@@ -238,7 +242,7 @@ describe('ReactLoopAgent', () => {
   it('steer() when idle falls through to send() and starts a turn', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // steer while idle delegates to send
     agent.steer([{ type: 'text', text: 'steer idle' }], { source: { kind: 'plugin', plugin: 'test' } })
@@ -254,7 +258,9 @@ describe('ReactLoopAgent', () => {
     const ctx = new Context()
     await ctx.plugin(SessionStore)
     const session = ctx.sessions.create(SessionId('test'))
-    const prepared = prepareReactLoopAgent(ctx, AgentId('bare'), { model: 'mock' }, session)
+    const prepared = prepareReactLoopAgent(
+      ctx, AgentId('bare'), { provider: 'mock', model: 'mock' }, session, DEFAULT_MAX_PARALLEL_TOOL_CALLS,
+    )
     const { agent } = prepared
 
     prepared.markPublished()
@@ -272,7 +278,9 @@ describe('ReactLoopAgent', () => {
     const ctx = new Context()
     await ctx.plugin(SessionStore)
     const session = ctx.sessions.create(SessionId('pre-start-dispose'))
-    const prepared = prepareReactLoopAgent(ctx, AgentId('pre-start-dispose'), { model: 'mock' }, session)
+    const prepared = prepareReactLoopAgent(
+      ctx, AgentId('pre-start-dispose'), { provider: 'mock', model: 'mock' }, session, DEFAULT_MAX_PARALLEL_TOOL_CALLS,
+    )
 
     await prepared.dispose()
     expect(prepared.agent.status).toBe('disposed')
@@ -286,7 +294,7 @@ describe('ReactLoopAgent', () => {
   it('setting the same status does not emit agent/status again', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const statuses: string[] = []
     ctx.on('agent/status', (subject, status) => {
@@ -305,7 +313,7 @@ describe('ReactLoopAgent', () => {
   it('whenIdle() resolves immediately when the agent is not running', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // Fresh agent is idle — whenIdle() takes the not-running fast path and
     // resolves without subscribing. await must not hang.
@@ -316,7 +324,7 @@ describe('ReactLoopAgent', () => {
   it('whenIdle() waits for queued work that has not flipped status yet', async () => {
     const adapter = new MockAdapter(['hang'])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'queued')
     let settled = false
@@ -334,8 +342,8 @@ describe('ReactLoopAgent', () => {
   it('whenIdle() awaits the running→idle transition, ignoring other subjects/running events', async () => {
     const adapter = new MockAdapter([textResponse('ok'), textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    const other = ctx.agentLoop.create(AgentId('a2'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+    const other = ctx.agentLoop.create(AgentId('a2'), { provider: 'mock', model: 'mock' })
 
     // Drive `agent` into `running`, then await whenIdle() — it subscribes to
     // agent/status and resolves on the first transition out of running.
@@ -369,7 +377,9 @@ describe('ReactLoopAgent', () => {
     const adapter = new MockAdapter(['hang'])
     ctx.llm.registerAdapter(['mock'], adapter)
     const session = ctx.sessions.create(SessionId('bare'))
-    const prepared = prepareReactLoopAgent(ctx, AgentId('bare'), { model: 'mock' }, session)
+    const prepared = prepareReactLoopAgent(
+      ctx, AgentId('bare'), { provider: 'mock', model: 'mock' }, session, DEFAULT_MAX_PARALLEL_TOOL_CALLS,
+    )
     const { agent } = prepared
     prepared.markPublished()
     const dispose = prepared.startDriver()
@@ -391,7 +401,7 @@ describe('ReactLoopAgent', () => {
     const ctx = await harness(adapter)
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('scoped'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('scoped'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
     send(agent, 'go')
     await new Promise(r => setTimeout(r, 30))
@@ -410,7 +420,7 @@ describe('ReactLoopAgent', () => {
     const ctx = await harness(adapter)
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('scoped'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('scoped'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
     send(agent, 'go')
     await new Promise(r => setTimeout(r, 30))
@@ -431,7 +441,7 @@ describe('ReactLoopAgent', () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
     const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => undefined)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     ctx.on('agent/status', (_subject, status) => {
       if (status === 'running') throw new Error('bad running listener')
     })
@@ -449,7 +459,7 @@ describe('ReactLoopAgent', () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
     const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => undefined)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     ctx.on('agent/status', (_subject, status) => {
       if (status === 'idle') throw new Error('bad idle listener')
     })
diff --git a/packages/core/agent-loop/tests/cancel.spec.ts b/packages/core/agent-loop/tests/cancel.spec.ts
index cad830e827..fc6055a00c 100644
--- a/packages/core/agent-loop/tests/cancel.spec.ts
+++ b/packages/core/agent-loop/tests/cancel.spec.ts
@@ -54,7 +54,7 @@ describe('Agent.cancel()', () => {
   it('cancel() on an idle agent with nothing queued is a no-op; the next prompt runs (F2 leak guard)', async () => {
     const adapter = new MockAdapter([textResponse('reply')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // The loop is parked at the idle wait with nothing queued. A cancel here must
     // NOT arm the marker — otherwise the next legitimate prompt would be dropped.
@@ -71,7 +71,7 @@ describe('Agent.cancel()', () => {
   it('pre-step cancel drops the about-to-start turn (no turn is opened)', async () => {
     const adapter = new MockAdapter([textResponse('should not run')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // send() queues synchronously (status still idle, loop microtask not yet
     // resumed). Cancel in that pre-step window: the queued turn must not run.
@@ -90,7 +90,7 @@ describe('Agent.cancel()', () => {
   it('a whenIdle() waiter registered BEFORE a pre-step cancel resolves (F1 hang guard)', async () => {
     const adapter = new MockAdapter([textResponse('x')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // This waiter cannot rely on a running→idle transition because cancellation
     // drops the turn before it runs; the skip path must settle it directly.
@@ -109,7 +109,7 @@ describe('Agent.cancel()', () => {
   it('cancel() mid-step aborts the in-flight model call; the turn ends aborted', async () => {
     const adapter = new MockAdapter(['hang'])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -126,7 +126,7 @@ describe('Agent.cancel()', () => {
   it('cancel() with no reason defaults to "cancelled" when aborting an in-flight step', async () => {
     const adapter = new MockAdapter(['hang'])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -142,7 +142,7 @@ describe('Agent.cancel()', () => {
   it('a prompt sent AFTER a cancelled turn settles runs normally (marker reset)', async () => {
     const adapter = new MockAdapter(['hang', textResponse('second reply')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // First turn hangs; cancel it mid-step.
     send(agent, 'first')
@@ -164,7 +164,7 @@ describe('Agent.cancel()', () => {
   it('cancel from inside the agent/session-prefix waterfall drops the step (prefix-composition window)', async () => {
     const adapter = new MockAdapter([textResponse('should not stream')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // Prefix composition runs before the pre-step seam on the instance's first
     // step; a cancel landing inside it must drop the about-to-start step
@@ -200,7 +200,7 @@ describe('Agent.cancel()', () => {
     const handle = await ctx.agents.create({
       agentId: AgentId('a-dispose-prefix'),
       sessionId: SessionId('dispose-prefix-session'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     const agent = handle.agent as ReactLoopAgent
 
@@ -227,7 +227,7 @@ describe('Agent.cancel()', () => {
   it('a cancel-interrupted prefix composition is discarded: the next send recomposes and ships the fresh prefix (stale-cache guard)', async () => {
     const adapter = new MockAdapter([textResponse('reply')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // The interrupted first composition must not cache its degraded empty value;
     // the next prompt recomposes and logs/sends the fresh prefix.
@@ -257,7 +257,7 @@ describe('Agent.cancel()', () => {
   it('cancel from a synchronous turn/start session-event listener drops the step (step-start window)', async () => {
     const adapter = new MockAdapter([textResponse('should not stream')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // A turn/start listener fires before a step controller exists, so the
     // turn-scoped marker—not step abort—must drop the pending step.
@@ -284,7 +284,7 @@ describe('Agent.cancel()', () => {
   it('cancel from a synchronous step/start session-event listener drops the step (post-step-start window)', async () => {
     const adapter = new MockAdapter([textResponse('should not stream')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // A step/start session-event listener fires AFTER step/start is appended
     // (and after the pre-step seam), so cancelling there lands in the SECOND
@@ -325,7 +325,7 @@ describe('Agent.cancel()', () => {
     const handle = await ctx.agents.create({
       agentId: AgentId('a-dispose-step-start'),
       sessionId: SessionId('dispose-step-start-session'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     const agent = handle.agent as ReactLoopAgent
 
@@ -355,7 +355,7 @@ describe('Agent.cancel()', () => {
     // `aborted` and run NO second step.
     const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let steps = 0
     const reasons: TurnEndReason[] = []
@@ -387,7 +387,7 @@ describe('Agent.cancel()', () => {
   it('cancel from a synchronous agent/status(running) listener drops the turn (window 2)', async () => {
     const adapter = new MockAdapter([textResponse('should not run')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // `agent/status` is synchronous, so cancellation can land after the first
     // pre-step check; the second check must drop the now-empty turn.
@@ -411,7 +411,7 @@ describe('Agent.cancel()', () => {
     // Cancellation must not settle idle while replacement work remains queued.
     const adapter = new MockAdapter([textResponse('A reply'), textResponse('B reply')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let replaced = false
     const dispose = ctx.on('agent/status', (subject, status) => {
@@ -438,7 +438,7 @@ describe('Agent.cancel()', () => {
     // prompt B is queued before the loop resumes from the idle wait.
     const adapter = new MockAdapter([textResponse('A reply'), textResponse('B reply')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'A')           // queues A (status still idle, loop microtask pending)
     const idle = agent.whenIdle() // registers a waiter (idle + hasQueued → no fast path)
@@ -457,7 +457,7 @@ describe('Agent.cancel()', () => {
   it("cancel clears the turn's steering — it is not re-enqueued as a fresh turn", async () => {
     const adapter = new MockAdapter(['hang'])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'go')
     await new Promise(r => setTimeout(r, 30))
diff --git a/packages/core/agent-loop/tests/config-session-id.spec.ts b/packages/core/agent-loop/tests/config-session-id.spec.ts
index 8a5d9ce264..86db33a8ca 100644
--- a/packages/core/agent-loop/tests/config-session-id.spec.ts
+++ b/packages/core/agent-loop/tests/config-session-id.spec.ts
@@ -32,7 +32,7 @@ describe('config-driven session id', () => {
     await ctx.plugin(ToolRegistry)
     await ctx.plugin(AgentRegistry)
     const loopFiber = await ctx.plugin(AgentLoop, {
-      agents: [{ id: AgentId('main'), model: 'mock', resumeSessionId: SessionId('deferred') }],
+      agents: [{ id: AgentId('main'), provider: 'mock', model: 'mock', resumeSessionId: SessionId('deferred') }],
     })
 
     const resumeEffect = loopFiber.getEffects().find(effect => effect.label === 'agentLoop.resume(main)')
@@ -53,7 +53,7 @@ describe('config-driven session id', () => {
     await ctx1.plugin(SystemPrompt)
     await ctx1.plugin(ToolRegistry)
     await ctx1.plugin(AgentRegistry)
-    await ctx1.plugin(AgentLoop, { agents: [{ id: AgentId('cfg'), model: 'mock' }] })
+    await ctx1.plugin(AgentLoop, { agents: [{ id: AgentId('cfg'), provider: 'mock', model: 'mock' }] })
     await ctx1.plugin(SessionPersistenceJsonl, { root })
     ctx1.llm.registerAdapter(['mock'], new MockAdapter([textResponse('cfg')]))
     const a1 = ctx1.agents.get(AgentId('cfg')) as ReactLoopAgent
@@ -70,7 +70,7 @@ describe('config-driven session id', () => {
     await ctx2.plugin(SystemPrompt)
     await ctx2.plugin(ToolRegistry)
     await ctx2.plugin(AgentRegistry)
-    await ctx2.plugin(AgentLoop, { agents: [{ id: AgentId('cfg'), model: 'mock' }] })
+    await ctx2.plugin(AgentLoop, { agents: [{ id: AgentId('cfg'), provider: 'mock', model: 'mock' }] })
     await ctx2.plugin(SessionPersistenceJsonl, { root })
     ctx2.llm.registerAdapter(['mock'], new MockAdapter([textResponse('cfg2')]))
     const a2 = ctx2.agents.get(AgentId('cfg')) as ReactLoopAgent
@@ -109,7 +109,7 @@ describe('config-driven session id', () => {
     await ctx2.plugin(SystemPrompt)
     await ctx2.plugin(ToolRegistry)
     await ctx2.plugin(AgentRegistry)
-    await ctx2.plugin(AgentLoop, { agents: [{ id: AgentId('main'), model: 'mock', resumeSessionId: SessionId('sticky-1') }] })
+    await ctx2.plugin(AgentLoop, { agents: [{ id: AgentId('main'), provider: 'mock', model: 'mock', resumeSessionId: SessionId('sticky-1') }] })
     await ctx2.plugin(SessionPersistenceJsonl, { root })
     ctx2.llm.registerAdapter(['mock'], new MockAdapter([textResponse('second')]))
 
@@ -137,7 +137,7 @@ describe('config-driven session id', () => {
     await ctx.plugin(SystemPrompt)
     await ctx.plugin(ToolRegistry)
     await ctx.plugin(AgentRegistry)
-    await ctx.plugin(AgentLoop, { agents: [{ id: AgentId('main'), model: 'mock', resumeSessionId: SessionId('does-not-exist') }] })
+    await ctx.plugin(AgentLoop, { agents: [{ id: AgentId('main'), provider: 'mock', model: 'mock', resumeSessionId: SessionId('does-not-exist') }] })
     const warn = vi.spyOn((ctx.agentLoop as unknown as { ctx: { logger: { warn: (...a: unknown[]) => void } } }).ctx.logger, 'warn')
       .mockImplementation(() => undefined)
     await ctx.plugin(SessionPersistenceJsonl, { root })
diff --git a/packages/core/agent-loop/tests/contract-regressions.spec.ts b/packages/core/agent-loop/tests/contract-regressions.spec.ts
index ee50261e9d..12460039b1 100644
--- a/packages/core/agent-loop/tests/contract-regressions.spec.ts
+++ b/packages/core/agent-loop/tests/contract-regressions.spec.ts
@@ -3,12 +3,12 @@ import { Context } from 'cordis'
 import LlmService, { CallId, ContentBlock, MessageSource, StreamChunk } from '@deepseek-ai/dsh-llm'
 import SessionStore, { Session, SessionEvent, SessionId, TurnEndReason } from '@deepseek-ai/dsh-session'
 import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry, { defineTool } from '@deepseek-ai/dsh-tools'
+import ToolRegistry, { defineTool, type PostToolDecision } from '@deepseek-ai/dsh-tools'
 import AgentRegistry, { AgentId, type ContinuationDecision } from '@deepseek-ai/dsh-agent'
-import AgentLoop, { ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import AgentLoop, { DEFAULT_MAX_PARALLEL_TOOL_CALLS, ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
 import { prepareReactLoopAgent } from '../src/agent.ts'
 import * as Invariants from '@deepseek-ai/dsh-invariants'
-import { MockAdapter, textResponse, toolCallResponse } from './mock-adapter.ts'
+import { maxTokensResponse, MockAdapter, textResponse, toolCallResponse } from './mock-adapter.ts'
 
 /** Regression tests for agent-loop boundary, identity, and lifecycle contracts. */
 
@@ -41,7 +41,9 @@ function send(agent: ReactLoopAgent, text: string) {
 
 describe('session log records what agent/step-result actually produced', () => {
   it('a step-result rewrite is what the log, derived history, and tool dispatch all see', async () => {
-    const adapter = new MockAdapter([textResponse('original'), textResponse('done')])
+    const original = textResponse('original')
+    original[original.length - 1] = { type: 'finish', reason: { kind: 'stop' }, replayState: { private: 'original-state' } }
+    const adapter = new MockAdapter([original, textResponse('done')])
     const ctx = await harness(adapter)
     const executed: string[] = []
     ctx.tools.register(defineTool({
@@ -53,7 +55,7 @@ describe('session log records what agent/step-result actually produced', () => {
         return [{ type: 'text', text: 'ran' }]
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // Plugin rewrites the message: replaces the text AND adds a tool call.
     let rewritten = false
@@ -78,6 +80,7 @@ describe('session log records what agent/step-result actually produced', () => {
     const recorded = agent.session.events.find(e => e.type === 'assistant/message')!
     expect(JSON.stringify(recorded.data)).toContain('rewritten')
     expect(JSON.stringify(recorded.data)).not.toContain('original')
+    expect(recorded.type === 'assistant/message' && recorded.data.provenance.replayState).toBeUndefined()
     // tool/call + tool/result correlate with the injected call id
     const callEvent = agent.session.events.find(e => e.type === 'tool/call')!
     if (callEvent.type !== 'tool/call') throw new Error('wrong event type')
@@ -87,6 +90,113 @@ describe('session log records what agent/step-result actually produced', () => {
     expect(JSON.stringify(derived)).toContain('rewritten')
     expect(JSON.stringify(derived)).not.toContain('original')
   })
+
+  it('records adapter replay state when step-result preserves the assembled content', async () => {
+    const response = textResponse('unchanged')
+    const replayState = { private: 'state' }
+    response[response.length - 1] = { type: 'finish', reason: { kind: 'stop' }, replayState }
+    const adapter = new MockAdapter([response])
+    const ctx = await harness(adapter)
+    const agent = ctx.agentLoop.create(AgentId('replay-state'), { provider: 'mock', model: 'next-model' })
+
+    send(agent, 'go')
+    await waitForIdle(ctx, agent)
+
+    const recorded = agent.session.events.find(e => e.type === 'assistant/message')
+    expect(recorded?.type === 'assistant/message' && recorded.data.provenance).toEqual({
+      provider: 'mock', model: 'next-model', replayState,
+    })
+    expect(agent.session.deriveMessages().at(-1)?.provenance).toEqual({
+      provider: 'mock', model: 'next-model', replayState,
+    })
+  })
+
+  it('drops adapter replay state when step-result mutates assembled content in place', async () => {
+    const response = textResponse('original')
+    response[response.length - 1] = { type: 'finish', reason: { kind: 'stop' }, replayState: { private: 'state' } }
+    const adapter = new MockAdapter([response])
+    const ctx = await harness(adapter)
+    ctx.on('agent/step-result', async (_agent, _turn, _step, message) => {
+      const block = message.content[0]
+      if (block?.type === 'text') block.text = 'mutated'
+      return message
+    })
+    const agent = ctx.agentLoop.create(AgentId('mutated-replay-state'), { provider: 'mock', model: 'next-model' })
+
+    send(agent, 'go')
+    await waitForIdle(ctx, agent)
+
+    const recorded = agent.session.events.find(event => event.type === 'assistant/message')
+    expect(recorded?.type === 'assistant/message' && recorded.data.content).toEqual([{ type: 'text', text: 'mutated' }])
+    expect(recorded?.type === 'assistant/message' && recorded.data.provenance.replayState).toBeUndefined()
+  })
+})
+
+describe('successful provider completion survives agent/step-result failure', () => {
+  async function expectContentlessCompletionAnchor(
+    response: StreamChunk[],
+    id: string,
+    providerText: string,
+  ): Promise<void> {
+    const adapter = new MockAdapter([response])
+    const ctx = await harness(adapter)
+    await ctx.plugin(Invariants)
+    const agent = ctx.agentLoop.create(AgentId(id), { provider: 'mock', model: 'mock' })
+    const failure = new Error(`${id} result processing failed`)
+    const reported: Error[] = []
+
+    ctx.on('agent/step-result', async () => {
+      throw failure
+    })
+    ctx.on('agent/error', (subject, _turn, _step, error) => {
+      if (subject === agent) reported.push(error)
+    })
+
+    send(agent, 'go')
+    await waitForIdle(ctx, agent)
+
+    const events = [...agent.session.events]
+    const chunks = events.filter(event => event.type === 'assistant/chunk')
+    const completions = events.filter(event => event.type === 'assistant/message')
+    expect(completions).toHaveLength(1)
+    expect(completions[0]?.type === 'assistant/message' && completions[0].data).toEqual({
+      turn: 1,
+      step: 1,
+      content: [],
+      provenance: { provider: 'mock', model: 'mock' },
+      usage: { inputTokens: 10, outputTokens: providerText.length },
+    })
+    expect(completions[0]?.sourceEventSeqs).toEqual(chunks.map(event => event.seq))
+    expect(agent.session.deriveMessages()).toEqual([
+      { role: 'user', content: [{ type: 'text', text: 'go' }] },
+    ])
+    expect(reported).toHaveLength(1)
+    expect(reported[0]).toBe(failure)
+    const turnEnd = events.findLast(event => event.type === 'turn/end')
+    expect(turnEnd?.type === 'turn/end' && turnEnd.data.reason).toEqual({
+      kind: 'error',
+      step: 1,
+      message: failure.message,
+    })
+  }
+
+  it('records one content-less anchor when ordinary stop result processing rejects', async () => {
+    const providerText = 'ordinary provider output'
+    await expectContentlessCompletionAnchor(
+      textResponse(providerText),
+      'a-step-result-stop-failure',
+      providerText,
+    )
+  })
+
+  it('records one content-less anchor when max-token result processing rejects', async () => {
+    const providerText = 'truncated provider output'
+    await expectContentlessCompletionAnchor(
+      maxTokensResponse(providerText),
+      'a-step-result-max-token-failure',
+      providerText,
+    )
+  })
 })
 
 describe('abort during tool execution ends the turn', () => {
@@ -104,7 +214,7 @@ describe('abort during tool execution ends the turn', () => {
     ])
     const ctx = await harness(adapter)
     const executed: string[] = []
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     ctx.tools.register(defineTool({
       name: 'aborter',
       description: '',
@@ -139,6 +249,190 @@ describe('abort during tool execution ends the turn', () => {
     expect(adapter.requests).toHaveLength(1)        // no follow-up model call
     expect(reasons).toEqual([{ kind: 'aborted', reason: 'user interrupt' }])
   })
+
+  it('records context accepted before a tool-step abort in the same turn', async () => {
+    const adapter = new MockAdapter([toolCallResponse('c1', 'aborter', {})])
+    const ctx = await harness(adapter)
+    const agent = ctx.agentLoop.create(AgentId('a-abort-injection'), { provider: 'mock', model: 'mock' })
+    ctx.tools.register(defineTool({
+      name: 'aborter',
+      description: '',
+      parameters: {},
+      async execute() {
+        agent.inject([{ type: 'text', text: 'accepted before abort' }], { source: { kind: 'plugin', plugin: 'test' } })
+        ;(agent as unknown as { currentAbort?: AbortController }).currentAbort?.abort('user interrupt')
+        return [{ type: 'text', text: 'done' }]
+      },
+    }))
+    ctx.on('tools/post-execute', async (): Promise<PostToolDecision> => ({
+      kind: 'accept',
+      additionalContexts: [{
+        content: [{ type: 'text', text: 'accepted result context after abort' }],
+        source: { kind: 'plugin', plugin: 'test' },
+      }],
+    }))
+
+    send(agent, 'go')
+    await waitForIdle(ctx, agent)
+
+    const events = [...agent.session.events]
+    expect(events
+      .filter(event => event.type === 'tool/result' || event.type === 'context/message'
+        || event.type === 'step/end' || event.type === 'turn/end')
+      .map(event => event.type))
+      .toEqual(['tool/result', 'context/message', 'context/message', 'step/end', 'turn/end'])
+    expect(events
+      .filter(event => event.type === 'context/message')
+      .map(event => event.data.content))
+      .toEqual([
+        [{ type: 'text', text: 'accepted before abort' }],
+        [{ type: 'text', text: 'accepted result context after abort' }],
+      ])
+  })
+
+  it('records post-tool context when a later call aborts the batch', async () => {
+    const adapter = new MockAdapter([[
+      { type: 'block-start', index: 0, blockType: 'tool-call' },
+      { type: 'block-end', index: 0, block: { type: 'tool-call', id: CallId('c1'), name: 'first', arguments: '{}' } },
+      { type: 'block-start', index: 1, blockType: 'tool-call' },
+      { type: 'block-end', index: 1, block: { type: 'tool-call', id: CallId('c2'), name: 'aborter', arguments: '{}' } },
+      { type: 'finish', reason: { kind: 'tool-calls' } },
+    ] satisfies StreamChunk[]])
+    const ctx = await harness(adapter)
+    const agent = ctx.agentLoop.create(AgentId('a-later-abort-context'), { provider: 'mock', model: 'mock' })
+    ctx.tools.register(defineTool({
+      name: 'first',
+      description: '',
+      parameters: {},
+      async execute() {
+        return [{ type: 'text', text: 'first done' }]
+      },
+    }))
+    ctx.tools.register(defineTool({
+      name: 'aborter',
+      description: '',
+      parameters: {},
+      async execute() {
+        ;(agent as unknown as { currentAbort?: AbortController }).currentAbort?.abort('user interrupt')
+        return [{ type: 'text', text: 'aborted' }]
+      },
+    }))
+    ctx.on('tools/post-execute', async (exec, _result, next): Promise<PostToolDecision> => {
+      if (exec.callId !== CallId('c1')) return next()
+      return {
+        kind: 'accept',
+        additionalContexts: [{
+          content: [{ type: 'text', text: 'accepted after first result' }],
+          source: { kind: 'plugin', plugin: 'test' },
+        }],
+      }
+    })
+
+    send(agent, 'go')
+    await waitForIdle(ctx, agent)
+
+    const events = [...agent.session.events]
+    expect(events
+      .filter(event => event.type === 'tool/result' || event.type === 'context/message'
+        || event.type === 'step/end' || event.type === 'turn/end')
+      .map(event => event.type))
+      .toEqual(['tool/result', 'tool/result', 'context/message', 'step/end', 'turn/end'])
+    expect(events.find(event => event.type === 'context/message')?.data.content)
+      .toEqual([{ type: 'text', text: 'accepted after first result' }])
+  })
+
+  it('drains deferred context before disposal reaches quiescence', async () => {
+    const adapter = new MockAdapter([toolCallResponse('c1', 'waiter', {})])
+    const ctx = await harness(adapter)
+    const started = Promise.withResolvers<undefined>()
+    let agent!: ReactLoopAgent
+    const fiber = await ctx.plugin(Object.assign((inner: Context) => {
+      agent = inner.agentLoop.create(AgentId('a-dispose-injection'), { provider: 'mock', model: 'mock' })
+    }, { inject: ['agentLoop'] }))
+    ctx.tools.register(defineTool({
+      name: 'waiter',
+      description: '',
+      parameters: {},
+      async execute(_args, exec) {
+        agent.inject([{ type: 'text', text: 'accepted before disposal' }], { source: { kind: 'plugin', plugin: 'test' } })
+        started.resolve(undefined)
+        const signal = exec.signal
+        if (!signal) throw new Error('tool execution signal is missing')
+        await new Promise<void>((resolve) => {
+          if (signal.aborted) resolve()
+          else signal.addEventListener('abort', () => { resolve() }, { once: true })
+        })
+        return [{ type: 'text', text: 'done' }]
+      },
+    }))
+    ctx.on('tools/post-execute', async (): Promise<PostToolDecision> => ({
+      kind: 'accept',
+      additionalContexts: [{
+        content: [{ type: 'text', text: 'accepted result context during disposal' }],
+        source: { kind: 'plugin', plugin: 'test' },
+      }],
+    }))
+
+    send(agent, 'go')
+    await started.promise
+    await fiber.dispose()
+
+    expect(agent.session.events
+      .filter(event => event.type === 'context/message')
+      .map(event => event.data.content))
+      .toEqual([
+        [{ type: 'text', text: 'accepted before disposal' }],
+        [{ type: 'text', text: 'accepted result context during disposal' }],
+      ])
+    expect(agent.session.events.find(event => event.type === 'turn/end')?.data.reason)
+      .toEqual({ kind: 'disposed' })
+  })
+
+  it('limits injection deferral to the current tool batch', async () => {
+    const adapter = new MockAdapter([
+      [
+        { type: 'block-start', index: 0, blockType: 'tool-call' },
+        { type: 'block-end', index: 0, block: { type: 'tool-call', id: CallId('c1'), name: 'aborter', arguments: '{}' } },
+        { type: 'block-start', index: 1, blockType: 'tool-call' },
+        { type: 'block-end', index: 1, block: { type: 'tool-call', id: CallId('c2'), name: 'second', arguments: '{}' } },
+        { type: 'finish', reason: { kind: 'tool-calls' } },
+      ] satisfies StreamChunk[],
+      textResponse('later turn'),
+    ])
+    const ctx = await harness(adapter)
+    const agent = ctx.agentLoop.create(AgentId('a-historical-tool-pair'), { provider: 'mock', model: 'mock' })
+    ctx.tools.register(defineTool({
+      name: 'aborter',
+      description: '',
+      parameters: {},
+      async execute() {
+        ;(agent as unknown as { currentAbort?: AbortController }).currentAbort?.abort('user interrupt')
+        return [{ type: 'text', text: 'done' }]
+      },
+    }))
+    ctx.tools.register(defineTool({
+      name: 'second',
+      description: '',
+      parameters: {},
+      async execute() {
+        return [{ type: 'text', text: 'must not run' }]
+      },
+    }))
+
+    send(agent, 'leave an unmatched historical call')
+    await waitForIdle(ctx, agent)
+    ctx.on('agent/pre-step', (subject, turn) => {
+      if (subject === agent && turn === 2) {
+        agent.inject([{ type: 'text', text: 'new turn context' }], { source: { kind: 'plugin', plugin: 'test' } })
+      }
+    })
+    send(agent, 'start a text-only turn')
+    await waitForIdle(ctx, agent)
+
+    expect(agent.session.events.find(event => event.type === 'context/message')?.data.content)
+      .toEqual([{ type: 'text', text: 'new turn context' }])
+    expect(JSON.stringify(adapter.requests[1]?.messages)).toContain('new turn context')
+  })
 })
 
 describe('steering from late extension points is never stranded', () => {
@@ -148,7 +442,7 @@ describe('steering from late extension points is never stranded', () => {
       textResponse('continued because of steering'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let steeredOnce = false
     ctx.on('agent/turn-continuation', async (_agent, _turn, _decision, next) => {
@@ -174,7 +468,7 @@ describe('steering from late extension points is never stranded', () => {
       textResponse('after goal reminder'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let steeredOnce = false
     ctx.on('session/event', (subject, event) => {
@@ -202,7 +496,7 @@ describe('steering from late extension points is never stranded', () => {
   it('steer() from a turn/end session-event listener becomes a queued message for the next turn', async () => {
     const adapter = new MockAdapter([textResponse('turn 1'), textResponse('turn 2')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const turns: number[] = []
     let steeredOnce = false
@@ -228,7 +522,7 @@ describe('steering from late extension points is never stranded', () => {
   it('steering queued during an aborted step is re-delivered, not silently consumed', async () => {
     const adapter = new MockAdapter(['hang', textResponse('recovered')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'go')
     await new Promise(r => setTimeout(r, 30))
@@ -251,7 +545,7 @@ describe('plugin exceptions are contained', () => {
   it('a throwing agent/turn-continuation listener ends the turn with an error, loop survives', async () => {
     const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let threwOnce = false
     ctx.on('agent/turn-continuation', async (): Promise<ContinuationDecision> => {
@@ -279,7 +573,7 @@ describe('plugin exceptions are contained', () => {
   it('a rejecting session/flush listener is reported but does not kill the agent', async () => {
     const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let rejectedOnce = false
     ctx.on('session/flush', async () => {
@@ -309,7 +603,7 @@ describe('disposed status is part of the agent/status contract', () => {
 
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('scoped'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('scoped'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     const statuses: string[] = []
@@ -332,7 +626,7 @@ describe('disposed status is part of the agent/status contract', () => {
 
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('scoped'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('scoped'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     ctx.on('agent/status', (_agent, status) => {
@@ -358,7 +652,7 @@ describe('adapter registration, routing, and accepted-input ownership', () => {
     expect(() => ctx.llm.registerAdapter(['m1'], new MockAdapter([])))
       .toThrow('already registered')
     // the original registration survives the failed attempt
-    expect(ctx.llm.models()).toEqual(['m1'])
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'm1', name: 'm1' }])
   })
 
   it('an agent without a model fails the step with a clear error (not NO_ADAPTER for "default")', async () => {
@@ -372,7 +666,7 @@ describe('adapter registration, routing, and accepted-input ownership', () => {
     send(agent, 'go')
     await waitForIdle(ctx, agent)
     expect(errors).toHaveLength(1)
-    expect(errors[0]!.message).toContain('has no model')
+    expect(errors[0]!.message).toContain('has no provider/model')
     expect(errors[0]!.message).toContain('agent/request')
   })
 
@@ -382,7 +676,7 @@ describe('adapter registration, routing, and accepted-input ownership', () => {
     const agent = ctx.agentLoop.create(AgentId('a1'), {}) // no model — router plugin decides
 
     ctx.on('agent/request', async (_agent, _turn, _step, config, _next) => {
-      return { ...config, model: 'mock' }
+      return { ...config, provider: 'mock', model: 'mock' }
     })
 
     send(agent, 'go')
@@ -394,7 +688,7 @@ describe('adapter registration, routing, and accepted-input ownership', () => {
   it('agent/queued carries the resolved source; steering/message records its source', async () => {
     const adapter = new MockAdapter([toolCallResponse('c1', 'noop', {}), textResponse('done')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     ctx.tools.register(defineTool({
       name: 'noop',
       description: '',
@@ -422,7 +716,7 @@ describe('adapter registration, routing, and accepted-input ownership', () => {
   it('send() owns content and source before notification and delivery', async () => {
     const adapter = new MockAdapter([textResponse('done')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('owned-send'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('owned-send'), { provider: 'mock', model: 'mock' })
     const content = [{ type: 'text' as const, text: 'accepted-send' }]
     const source = { kind: 'plugin' as const, plugin: 'accepted-source' }
     let notifiedContent: ContentBlock[] | undefined
@@ -458,7 +752,7 @@ describe('adapter registration, routing, and accepted-input ownership', () => {
   it('running steer() owns content and source before notification and delivery', async () => {
     const adapter = new MockAdapter([toolCallResponse('c1', 'gate', {}), textResponse('done')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('owned-steer'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('owned-steer'), { provider: 'mock', model: 'mock' })
     const entered = Promise.withResolvers<undefined>()
     const release = Promise.withResolvers<undefined>()
     ctx.tools.register(defineTool({
@@ -512,7 +806,7 @@ describe('turn numbering continues across seeded sessions', () => {
   it('a forked agent continues turn numbers after the seed log', async () => {
     const first = new MockAdapter([textResponse('turn one')])
     const ctx = await harness(first)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     send(agent, 'first')
     await waitForIdle(ctx, agent)
 
@@ -528,7 +822,9 @@ describe('turn numbering continues across seeded sessions', () => {
     ctx2.llm.registerAdapter(['mock'], second)
 
     const seeded = ctx2.sessions.create(SessionId('forked'), { seed: [...agent.session.events] })
-    const prepared = prepareReactLoopAgent(ctx2, AgentId('forked-agent'), { model: 'mock' }, seeded)
+    const prepared = prepareReactLoopAgent(
+      ctx2, AgentId('forked-agent'), { provider: 'mock', model: 'mock' }, seeded, DEFAULT_MAX_PARALLEL_TOOL_CALLS,
+    )
     const forked = prepared.agent
     prepared.markPublished()
     ctx2.effect(() => prepared.startDriver())
@@ -572,7 +868,7 @@ describe('a finish-error stream chunk ends the turn as error, not completed', ()
     ]
     const adapter = new MockAdapter([errorStream])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-finish-error'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-finish-error'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -597,7 +893,7 @@ describe('a finish-error stream chunk ends the turn as error, not completed', ()
     ]
     const adapter = new MockAdapter([abortedStream])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-finish-aborted'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-finish-aborted'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -615,7 +911,7 @@ describe('a finish-error stream chunk ends the turn as error, not completed', ()
     ]
     const adapter = new MockAdapter([errorStream])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-finish-error-nocode'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-finish-error-nocode'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -631,7 +927,7 @@ describe('step boundary publication order', () => {
   it('the step/start event is in session.events when its session/event listener fires', async () => {
     const adapter = new MockAdapter([textResponse('done')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-step-order'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-step-order'), { provider: 'mock', model: 'mock' })
 
     // Append commits before observers run.
     const observed: { turn: number; step: number; lastEventType: string | undefined; sawStepStart: boolean }[] = []
@@ -686,7 +982,7 @@ describe('turn and step boundary recovery', () => {
   it('a throwing step/start observer cannot change a successful turn', async () => {
     const adapter = new MockAdapter([textResponse('request completed')])
     const ctx = await balancedHarness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-stepstart'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-stepstart'), { provider: 'mock', model: 'mock' })
 
     // Session owns post-commit containment. The loop sees a successful append,
     // runs the request, and balances the ordinary step and turn boundaries.
@@ -715,7 +1011,7 @@ describe('turn and step boundary recovery', () => {
   it('a pre-commit step/start validation failure does not invent a step boundary', async () => {
     const adapter = new MockAdapter([textResponse('never reached')])
     const ctx = await balancedHarness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-stepstart-veto'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-stepstart-veto'), { provider: 'mock', model: 'mock' })
     let rejected = false
     ctx.on('internal/dispatch', (_mode, name, args) => {
       if (name !== 'session/event') return
@@ -746,7 +1042,7 @@ describe('turn and step boundary recovery', () => {
     const errorStream: StreamChunk[] = [{ type: 'finish', reason: { kind: 'error', message: 'provider failed' } }]
     const adapter = new MockAdapter([errorStream])
     const ctx = await balancedHarness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-turnend-veto'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-turnend-veto'), { provider: 'mock', model: 'mock' })
     let rejected = false
     ctx.on('internal/dispatch', (_mode, name, args) => {
       if (name !== 'session/event') return
@@ -780,7 +1076,7 @@ describe('turn and step boundary recovery', () => {
   it('a one-shot step/end validation failure keeps the step open until retry succeeds', async () => {
     const adapter = new MockAdapter([textResponse('completed before close validation')])
     const ctx = await balancedHarness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-stepend-veto'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-stepend-veto'), { provider: 'mock', model: 'mock' })
     let rejected = false
     ctx.on('internal/dispatch', (_mode, name, args) => {
       if (name !== 'session/event') return
@@ -812,7 +1108,7 @@ describe('turn and step boundary recovery', () => {
     const errorStream: StreamChunk[] = [{ type: 'finish', reason: { kind: 'error', message: 'provider 500' } }]
     const adapter = new MockAdapter([errorStream, textResponse('turn 2 ok')])
     const ctx = await balancedHarness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-errorlistener'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-errorlistener'), { provider: 'mock', model: 'mock' })
 
     let threw = false
     ctx.on('agent/error', () => { if (!threw) { threw = true; throw new Error('boom error-listener') } })
@@ -845,7 +1141,7 @@ describe('turn and step boundary recovery', () => {
     const ctx = await balancedHarness(adapter)
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('a-dispose'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('a-dispose'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     const reasons: TurnEndReason[] = []
@@ -872,7 +1168,7 @@ describe('turn and step boundary recovery', () => {
     const ctx = await balancedHarness(adapter)
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('a-prestep-dispose-throw'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('a-prestep-dispose-throw'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     let threw = false
@@ -903,7 +1199,7 @@ describe('turn and step boundary recovery', () => {
   it('a throwing turn/start observer cannot starve the loop or later turns', async () => {
     const adapter = new MockAdapter([textResponse('turn 1'), textResponse('turn 2')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-preturn'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-preturn'), { provider: 'mock', model: 'mock' })
 
     let threw = false
     ctx.on('session/event', (_session, event) => {
@@ -934,7 +1230,7 @@ describe('turn and step boundary recovery', () => {
   it('a throwing step/end observer cannot rewrite the turn outcome', async () => {
     const adapter = new MockAdapter([textResponse('all good'), textResponse('turn 2 ok')])
     const ctx = await balancedHarness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-stepend-throw'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-stepend-throw'), { provider: 'mock', model: 'mock' })
 
     let threw = false
     ctx.on('session/event', (_s, event) => {
@@ -973,7 +1269,7 @@ describe('turn and step boundary recovery', () => {
     const errorStream: StreamChunk[] = [{ type: 'finish', reason: { kind: 'error', message: 'provider 500' } }]
     const adapter = new MockAdapter([errorStream, textResponse('turn 2 ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-stependthrow'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-stependthrow'), { provider: 'mock', model: 'mock' })
 
     let threw = false
     ctx.on('session/event', (_s, event) => {
@@ -1003,7 +1299,7 @@ describe('turn and step boundary recovery', () => {
     // boundary stays authoritative and the loop continues normally.
     const adapter = new MockAdapter([textResponse('turn 1'), textResponse('turn 2')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a-turnendappend'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-turnendappend'), { provider: 'mock', model: 'mock' })
 
     let threw = false
     ctx.on('session/event', (_s, event) => {
@@ -1049,7 +1345,7 @@ describe('tool result call identity', () => {
       return Promise.resolve({ kind: 'accept', content: [{ type: 'text', text: 'ok' }] })
     }, { prepend: true })
 
-    const agent = ctx.agentLoop.create(AgentId('a-callid'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-callid'), { provider: 'mock', model: 'mock' })
     send(agent, 'use tool')
     await waitForIdle(ctx, agent)
 
@@ -1073,13 +1369,14 @@ describe('tool result call identity', () => {
   })
 })
 
-describe('surface: assistant/message omits sourceEventSeqs when no chunks streamed', () => {
-  it('a step-result listener injecting content over an empty stream appends with surfaceOp but no sourceEventSeqs', async () => {
-    // Injected result content with no chunks must omit empty sourceEventSeqs.
+describe('surface: assistant/message records exact empty provenance when no chunks streamed', () => {
+  it('a step-result listener injecting content over an empty stream records sourceEventSeqs []', async () => {
+    // The explicit empty source set distinguishes a known empty provider
+    // stream from legacy events whose provenance was not recorded.
     const adapter = new MockAdapter([[]])
     const ctx = await harness(adapter)
     await ctx.plugin(Invariants)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     ctx.on('agent/step-result', async (_agent, _turn, _step, _message, _next) => ({
       role: 'assistant' as const,
@@ -1092,7 +1389,7 @@ describe('surface: assistant/message omits sourceEventSeqs when no chunks stream
     const recorded = agent.session.events.find(e => e.type === 'assistant/message')!
     expect(recorded.type).toBe('assistant/message')
     expect(recorded.surfaceOp).toBe('append')
-    expect(recorded.sourceEventSeqs).toBeUndefined()
+    expect(recorded.sourceEventSeqs).toEqual([])
     // The injected content reaches derived history.
     expect(JSON.stringify(agent.session.deriveMessages())).toContain('injected')
   })
@@ -1126,7 +1423,7 @@ describe('disposal and cancellation during pre-step assembly', () => {
 
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('a-dispose-assemble'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('a-dispose-assemble'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     const reasons: TurnEndReason[] = []
@@ -1176,7 +1473,7 @@ describe('disposal and cancellation during pre-step assembly', () => {
 
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('a-cancel-assemble'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('a-cancel-assemble'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     const reasons: TurnEndReason[] = []
@@ -1230,7 +1527,7 @@ describe('disposal and cancellation during pre-step assembly', () => {
 
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('a-dispose-prestep'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('a-dispose-prestep'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     const reasons: TurnEndReason[] = []
@@ -1281,7 +1578,7 @@ describe('disposal and cancellation during pre-step assembly', () => {
 
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('a-cancel-prestep'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('a-cancel-prestep'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     const reasons: TurnEndReason[] = []
@@ -1331,7 +1628,7 @@ describe('disposal and cancellation during pre-step assembly', () => {
 
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('a-dispose-no-leak'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('a-dispose-no-leak'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     send(agent, 'go')
diff --git a/packages/core/agent-loop/tests/coverage-edges.spec.ts b/packages/core/agent-loop/tests/coverage-edges.spec.ts
index 5deee8e159..ddbac2716d 100644
--- a/packages/core/agent-loop/tests/coverage-edges.spec.ts
+++ b/packages/core/agent-loop/tests/coverage-edges.spec.ts
@@ -40,7 +40,7 @@ describe('inbox acceptance', () => {
   it('rejects non-serializable content or source synchronously before notification or enqueue', async () => {
     const adapter = new MockAdapter([textResponse('turn 1')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     let queued = 0
     ctx.on('agent/queued', () => { queued += 1 })
 
@@ -80,7 +80,7 @@ describe('tool JSON parse', () => {
         return [{ type: 'text', text: typeof args === 'string' ? `raw: ${args}` : JSON.stringify(args) }]
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'use tool')
     await waitForIdle(ctx, agent)
@@ -113,7 +113,7 @@ describe('tool JSON parse', () => {
         return [{ type: 'text', text: 'ran with empty args' }]
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'use tool')
     await waitForIdle(ctx, agent)
@@ -126,7 +126,7 @@ describe('toError normalization', () => {
   it('normalizes non-Error throws from pre-commit dispatch validation via the runLoop backstop', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let threwOnce = false
     ctx.on('internal/dispatch', (_mode, name, args) => {
@@ -152,7 +152,7 @@ describe('toError normalization', () => {
   it('normalizes non-Error throws from agent/request waterfall via inline toError in runStep catch', async () => {
     const adapter = new MockAdapter([textResponse('irrelevant')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let threwOnce = false
     ctx.on('agent/request', async (_agent, _turn, _step, _options, _next) => {
@@ -180,7 +180,7 @@ describe('coded error data emission', () => {
   it('errorData includes code when a coded error (LlmError) is thrown from a plugin', async () => {
     const adapter = new MockAdapter([textResponse('turn 1')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let threwOnce = false
     ctx.on('agent/request', async (_agent, _turn, _step, _options, next) => {
@@ -214,7 +214,7 @@ describe('disposed vs aborted branching', () => {
     const ctx = await harness(adapter)
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('scoped'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('scoped'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     const reasons: TurnEndReason[] = []
@@ -240,7 +240,7 @@ describe('structured tool error propagation (the runtime-validation RFC, part 2)
       textResponse('done'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     ctx.tools.register(defineTool({
       name: 'boom',
       description: 'always fails',
diff --git a/packages/core/agent-loop/tests/interception.spec.ts b/packages/core/agent-loop/tests/interception.spec.ts
index 4b37210993..9932d9ca25 100644
--- a/packages/core/agent-loop/tests/interception.spec.ts
+++ b/packages/core/agent-loop/tests/interception.spec.ts
@@ -17,7 +17,7 @@ import { MockAdapter, textResponse, toolCallResponse } from './mock-adapter.ts'
  * The interception seams introduced by the hooks taxonomy: `agent/prompt-submit`,
  * `agent/session-start`, the reshaped `agent/turn-continuation`
  * ({@link ContinuationDecision}), and the `tools/pre-execute` / `tools/post-execute`
- * split with `additionalContext` buffering. These verify the canonical event
+ * split with `additionalContexts` buffering. These verify the canonical event
  * surface a hook bridge (or a native plugin) programs against, WITHOUT any
  * external protocol — a native plugin uses the typed decisions directly.
  */
@@ -57,7 +57,7 @@ describe('agent/prompt-submit', () => {
   it('allow (default via next) records the user/message unchanged', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const seen: string[] = []
     ctx.on('agent/prompt-submit', async (_agent, content, _source, next) => {
@@ -76,7 +76,7 @@ describe('agent/prompt-submit', () => {
   it('allow with content REWRITES the prompt before it is recorded', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     ctx.on('agent/prompt-submit', async (): Promise<PromptDecision> =>
       ({ kind: 'allow', content: [{ type: 'text', text: 'REWRITTEN' }] }))
@@ -91,15 +91,21 @@ describe('agent/prompt-submit', () => {
     expect(JSON.stringify(adapter.requests[0]!.messages)).not.toContain('original')
   })
 
-  it('allow with additionalContext injects a separate context/message into the turn', async () => {
+  it('allow with additionalContexts injects separate context/message events into the turn', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
+    const meta = { kind: 'prompt-context', version: 1 }
     ctx.on('agent/prompt-submit', async (): Promise<PromptDecision> =>
       ({
         kind: 'allow',
-        additionalContext: { content: [{ type: 'text', text: 'extra ctx' }], source: { kind: 'plugin', plugin: 'test' } },
+        additionalContexts: [{
+          content: [{ type: 'text', text: '<system-reminder>extra ctx</system-reminder>' }],
+          source: { kind: 'plugin', plugin: 'test' },
+          envelope: 'raw',
+          meta,
+        }],
       }))
 
     send(agent, 'go')
@@ -109,25 +115,27 @@ describe('agent/prompt-submit', () => {
     const userMsg = log.find(e => e.type === 'user/message')
     const ctxMsg = log.find(e => e.type === 'context/message')
     expect(userMsg).toBeDefined()
-    expect(ctxMsg?.type === 'context/message' && ctxMsg.data.content).toEqual([{ type: 'text', text: 'extra ctx' }])
+    expect(ctxMsg?.type === 'context/message' && ctxMsg.data.content).toEqual([{ type: 'text', text: '<system-reminder>extra ctx</system-reminder>' }])
     expect(ctxMsg?.type === 'context/message' && ctxMsg.data.source).toEqual({ kind: 'plugin', plugin: 'test' })
+    expect(ctxMsg?.type === 'context/message' && ctxMsg.data.envelope).toBe('raw')
+    expect(ctxMsg?.type === 'context/message' && ctxMsg.data.meta).toEqual(meta)
     // both the prompt and the injected context reach the model
     const sent = JSON.stringify(adapter.requests[0]!.messages)
     expect(sent).toContain('extra ctx')
   })
 
-  it('a prompt-submit rewrite + additionalContext is VISIBLE to the agent/pre-step seam (merged ordering)', async () => {
+  it('a prompt-submit rewrite + additionalContexts is VISIBLE to the agent/pre-step seam (merged ordering)', async () => {
     // Prompt rewrites and injected context land before `agent/pre-step`, so a
     // compaction listener measures the current surface before the single derive.
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     ctx.on('agent/prompt-submit', async (): Promise<PromptDecision> =>
       ({
         kind: 'allow',
         content: [{ type: 'text', text: 'REWRITTEN prompt' }],
-        additionalContext: { content: [{ type: 'text', text: 'injected ctx' }], source: { kind: 'plugin', plugin: 'test' } },
+        additionalContexts: [{ content: [{ type: 'text', text: 'injected ctx' }], source: { kind: 'plugin', plugin: 'test' } }],
       }))
 
     // The pre-step seam (where compaction lives) derives the surface it would act
@@ -151,7 +159,7 @@ describe('agent/prompt-submit', () => {
   it('block drops the (only) prompt → zero-step turn ends rejected, model never called', async () => {
     const adapter = new MockAdapter([textResponse('should not run')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     ctx.on('agent/prompt-submit', async (): Promise<PromptDecision> =>
       ({ kind: 'block', reason: 'blocked by policy' }))
@@ -187,7 +195,7 @@ describe('agent/prompt-submit', () => {
     // the allowed prompt keeps the turn from ending rejected.
     const adapter = new MockAdapter([textResponse('ran once')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     ctx.on('agent/prompt-submit', async (_agent, content, _source, next): Promise<PromptDecision> => {
       const text = content.map(b => (b.type === 'text' ? b.text : '')).join('')
@@ -223,7 +231,7 @@ describe('agent/prompt-submit', () => {
   it('a throwing prompt-submit listener ends the turn balanced (error), loop survives', async () => {
     const adapter = new MockAdapter([textResponse('after')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let threw = false
     ctx.on('agent/prompt-submit', async () => {
@@ -256,7 +264,7 @@ describe('agent/session-start', () => {
     const sources: SessionStartSource[] = []
     ctx.on('agent/session-start', (_agent, source) => void sources.push(source))
 
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     // fires synchronously at create, before any turn
     expect(sources).toEqual(['startup'])
     expect(events(agent).some(e => e.type === 'turn/start')).toBe(false)
@@ -275,7 +283,7 @@ describe('agent/session-start', () => {
       agent.inject([{ type: 'text', text: 'session preamble' }], { source: { kind: 'plugin', plugin: 'test' } })
     })
 
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     send(agent, 'go')
     await waitForIdle(ctx, agent)
 
@@ -293,7 +301,7 @@ describe('agent/session-start', () => {
     ctx.on('agent/session-start', () => { throw new Error('session-start hook broke') })
 
     // create must not throw — the listener error is contained/logged
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     expect(agent.id).toBe(AgentId('a1'))
 
     // and the agent still runs
@@ -307,8 +315,8 @@ describe('agent/session-prefix', () => {
   it('dispatches to global and matching agent-scope listeners only', async () => {
     const adapter = new MockAdapter([textResponse('a done'), textResponse('b done')])
     const ctx = await harness(adapter)
-    const agentA = ctx.agentLoop.create(AgentId('prefix-a'), { model: 'mock' })
-    const agentB = ctx.agentLoop.create(AgentId('prefix-b'), { model: 'mock' })
+    const agentA = ctx.agentLoop.create(AgentId('prefix-a'), { provider: 'mock', model: 'mock' })
+    const agentB = ctx.agentLoop.create(AgentId('prefix-b'), { provider: 'mock', model: 'mock' })
     const seen: string[] = []
     ctx.on('agent/session-prefix', async (agent, _prefix, _signal, next) => {
       seen.push(`global:${agent.id}`)
@@ -345,7 +353,7 @@ describe('agent/session-prefix', () => {
       name: 'echo', description: 'echo', parameters: { text: { type: 'string' } },
       async execute(args) { return [{ type: 'text', text: String(args.text) }] },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const reminder: Message = { role: 'user', content: [{ type: 'text', text: '<system-reminder>catalog</system-reminder>' }] }
     let composed = 0
@@ -367,8 +375,8 @@ describe('agent/session-prefix', () => {
       expect(request.messages[0]).toEqual(reminder)
     }
     // The anchoring snapshot is the prefix's durable record — and the ONLY
-    // header event: reuse means no request/header-delta ever.
-    const headerEvents = events(agent).filter(e => e.type === 'request/header' || e.type === 'request/header-delta')
+    // header event: reuse means no changed snapshot ever.
+    const headerEvents = events(agent).filter(e => e.type === 'request/header')
     expect(headerEvents).toHaveLength(1)
     expect(headerEvents[0]?.type === 'request/header' && headerEvents[0].data.header.messagePrefix).toEqual([reminder])
     // Never session history: the derivation starts at the real user prompt.
@@ -378,7 +386,7 @@ describe('agent/session-prefix', () => {
   it('composes before the first pre-step and hands the prefix to the seam (pressure gates see the real value)', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const reminder: Message = { role: 'user', content: [{ type: 'text', text: 'opener' }] }
     const order: string[] = []
@@ -405,7 +413,7 @@ describe('agent/session-prefix', () => {
   it('the canonical prepend pattern composes contributions in registration order', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // Both listeners use the canonical `[mine, ...await next()]` prepend: the
     // waterfall unwinds innermost-first (the second listener's array is built
@@ -427,7 +435,7 @@ describe('agent/session-prefix', () => {
   it('with no contributions the header omits messagePrefix and the request is the bare derivation', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // A listener that delegates without contributing — the canonical no-op.
     ctx.on('agent/session-prefix', async (_agent, _prefix, _signal, next) => next())
@@ -443,7 +451,7 @@ describe('agent/session-prefix', () => {
   it('the frozen seed rejects in-place mutation — a contribution is a returned extension', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let mutationError: unknown
     ctx.on('agent/session-prefix', async (_agent, prefix, _signal, next): Promise<Message[]> => {
@@ -472,7 +480,7 @@ describe('agent/session-prefix', () => {
       name: 'echo', description: 'echo', parameters: { text: { type: 'string' } },
       async execute(args) { return [{ type: 'text', text: String(args.text) }] },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const held: Message = { role: 'user', content: [{ type: 'text', text: 'v1' }] }
     ctx.on('agent/session-prefix', async (_agent, _prefix, _signal, next): Promise<Message[]> => [...await next(), held])
@@ -484,7 +492,7 @@ describe('agent/session-prefix', () => {
     // cached prefix is a deep-frozen clone, so step 2's request is unchanged.
     held.content = [{ type: 'text', text: 'v2' }]
     expect(adapter.requests[1]!.messages[0]).toEqual({ role: 'user', content: [{ type: 'text', text: 'v1' }] })
-    expect(events(agent).filter(e => e.type === 'request/header-delta')).toHaveLength(0)
+    expect(events(agent).filter(e => e.type === 'request/header')).toHaveLength(1)
   })
 })
 
@@ -493,7 +501,7 @@ describe('agent/turn-continuation (ContinuationDecision)', () => {
   it('a continue decision with a reason records next-step steering in the same turn', async () => {
     const adapter = new MockAdapter([textResponse('step 1 no tools'), textResponse('step 2')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let forced = false
     ctx.on('agent/turn-continuation', async (_agent, _turn, _default, next): Promise<ContinuationDecision> => {
@@ -525,7 +533,7 @@ describe('agent/turn-continuation (ContinuationDecision)', () => {
       name: 'echo', description: 'echo', parameters: { text: { type: 'string' } },
       async execute(args) { return [{ type: 'text', text: String(args.text) }] },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     ctx.on('agent/turn-continuation', async (): Promise<ContinuationDecision> => ({ action: 'stop' }))
 
@@ -538,8 +546,8 @@ describe('agent/turn-continuation (ContinuationDecision)', () => {
   })
 })
 
-describe('tools/post-execute additionalContext buffering across a multi-call step', () => {
-  it('appends each call\'s additionalContext only AFTER all tool/results, preserving adjacency', async () => {
+describe('tool additionalContexts buffering across a step', () => {
+  it('appends each call\'s contexts only AFTER all tool/results, preserving adjacency', async () => {
     // One assistant step with TWO tool calls; the second model response stops.
     const twoCalls = [
       { type: 'block-start' as const, index: 0, blockType: 'tool-call' as const },
@@ -555,11 +563,19 @@ describe('tools/post-execute additionalContext buffering across a multi-call ste
       name: 'echo', description: 'echo', parameters: { text: { type: 'string' } },
       async execute(args) { return [{ type: 'text', text: String(args.text) }] },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
-    // Each call attaches additionalContext naming itself.
+    // Each call attaches one context naming itself.
     ctx.on('tools/post-execute', async (exec, _result): Promise<PostToolDecision> =>
-      ({ kind: 'accept', additionalContext: { content: [{ type: 'text', text: `ctx-${exec.callId}` }], source: { kind: 'plugin', plugin: 'p' } } }))
+      ({
+        kind: 'accept',
+        additionalContexts: [{
+          content: [{ type: 'text', text: `ctx-${exec.callId}` }],
+          source: { kind: 'plugin', plugin: 'p' },
+          envelope: 'raw',
+          meta: { callId: exec.callId },
+        }],
+      }))
 
     send(agent, 'go')
     await waitForIdle(ctx, agent)
@@ -579,6 +595,37 @@ describe('tools/post-execute additionalContext buffering across a multi-call ste
       .flatMap(e => (e.type === 'context/message' ? e.data.content : []))
       .map(b => (b.type === 'text' ? b.text : ''))
     expect(ctxTexts).toEqual(['ctx-c1', 'ctx-c2'])
+    const contextEvents = events(agent).filter(e => e.type === 'context/message')
+    expect(contextEvents.map(e => e.type === 'context/message' && e.data.envelope)).toEqual(['raw', 'raw'])
+    expect(contextEvents.map(e => e.type === 'context/message' && e.data.meta)).toEqual([{ callId: 'c1' }, { callId: 'c2' }])
+  })
+
+  it('appends multiple contexts deferred by one composite tool after its outer result', async () => {
+    const adapter = new MockAdapter([toolCallResponse('c1', 'composite', {}), textResponse('done')])
+    const ctx = await harness(adapter)
+    ctx.tools.register(defineTool({
+      name: 'composite', description: 'composite', parameters: {},
+      async execute(_args, exec) {
+        exec.deferContext({ content: [{ type: 'text', text: 'nested-a' }], source: { kind: 'plugin', plugin: 'a' }, meta: { order: 1 } })
+        exec.deferContext({ content: [{ type: 'text', text: 'nested-b' }], source: { kind: 'plugin', plugin: 'b' }, envelope: 'raw', meta: { order: 2 } })
+        return [{ type: 'text', text: 'outer result' }]
+      },
+    }))
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    send(agent, 'go')
+    await waitForIdle(ctx, agent)
+
+    const log = events(agent)
+    const resultIndex = log.findIndex(event => event.type === 'tool/result')
+    const contextEvents = log.filter(event => event.type === 'context/message')
+    expect(resultIndex).toBeGreaterThanOrEqual(0)
+    expect(log.findIndex(event => event === contextEvents[0])).toBeGreaterThan(resultIndex)
+    expect(contextEvents.map(event => event.type === 'context/message' && event.data.source)).toEqual([
+      { kind: 'plugin', plugin: 'a' },
+      { kind: 'plugin', plugin: 'b' },
+    ])
+    expect(contextEvents.map(event => event.type === 'context/message' && event.data.meta)).toEqual([{ order: 1 }, { order: 2 }])
   })
 })
 
@@ -591,7 +638,7 @@ describe('tools/pre-execute gate (native-plugin permission pattern, end-to-end t
       name: 'danger', description: 'danger', parameters: {},
       async execute() { ran = true; return [{ type: 'text', text: 'should not run' }] },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     ctx.on('tools/pre-execute', async (exec, next): Promise<PreToolDecision> => {
       if (exec.name === 'danger') return { kind: 'deny', reason: 'blocked dangerous tool' }
@@ -638,7 +685,7 @@ describe('worked example: a native hook plugin is just a cordis plugin on the se
       ctx.on('tools/post-execute', async (_exec, _result, next): Promise<PostToolDecision> => {
         const decision = await next()
         if (decision.kind === 'accept') {
-          return { kind: 'accept', additionalContext: { content: [{ type: 'text', text: 'audited' }], source: { kind: 'plugin', plugin: 'native-guard' } } }
+          return { kind: 'accept', additionalContexts: [{ content: [{ type: 'text', text: 'audited' }], source: { kind: 'plugin', plugin: 'native-guard' } }] }
         }
         return decision
       })
@@ -653,7 +700,7 @@ describe('worked example: a native hook plugin is just a cordis plugin on the se
       name: 'echo', description: 'echo', parameters: { text: { type: 'string' } },
       async execute(args) { return [{ type: 'text', text: String(args.text) }] },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'please echo hi')
     await waitForIdle(ctx, agent)
@@ -676,7 +723,7 @@ describe('worked example: a native hook plugin is just a cordis plugin on the se
     const adapter = new MockAdapter([textResponse('should not run')])
     const ctx = await harness(adapter)
     await ctx.plugin(NativeGuard)
-    const agent = ctx.agentLoop.create(AgentId('a2'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a2'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event: SessionEvent) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -695,7 +742,7 @@ describe('worked example: a native hook plugin is just a cordis plugin on the se
     await fiber.dispose()
 
     // After disposal, a destructive prompt is NOT blocked (the listener is gone).
-    const agent = ctx.agentLoop.create(AgentId('a3'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a3'), { provider: 'mock', model: 'mock' })
     send(agent, 'run rm -rf /')
     await waitForIdle(ctx, agent)
     // the prompt ran (not rejected) — proving the prompt-submit listener was disposed
diff --git a/packages/core/agent-loop/tests/loop.spec.ts b/packages/core/agent-loop/tests/loop.spec.ts
index fb686928b1..6f4f15ba0c 100644
--- a/packages/core/agent-loop/tests/loop.spec.ts
+++ b/packages/core/agent-loop/tests/loop.spec.ts
@@ -44,7 +44,7 @@ describe('agent loop', () => {
   it('runs a simple turn: queued message → model → idle, with ordered events', async () => {
     const adapter = new MockAdapter([textResponse('hello there')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     // All boundaries — turn and step — are durable session events on the
     // session/event feed (no agent/* mirror). Record them in fire order to
@@ -92,7 +92,7 @@ describe('agent loop', () => {
         return [{ type: 'text', text: `echo: ${args.text}` }]
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'use the tool')
     await waitForIdle(ctx, agent)
@@ -131,7 +131,7 @@ describe('agent loop', () => {
         return { content: [{ type: 'text', text: 'ok' }], meta: { diffs: [{ path: 'a.txt', oldText: null, newText: 'x' }] } }
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'use the tool')
     await waitForIdle(ctx, agent)
@@ -155,7 +155,7 @@ describe('agent loop', () => {
         return []
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'hi')
     await waitForIdle(ctx, agent)
@@ -172,7 +172,7 @@ describe('agent loop', () => {
       agentId: AgentId('a-cwd'),
       sessionId: SessionId('s-cwd'),
       meta: { cwd: '/work/space' },
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
 
     const agent = handle.agent as ReactLoopAgent
@@ -188,7 +188,7 @@ describe('agent loop', () => {
     const ctx = await harness(adapter, 'In {{cwd}}.')
     const errors: Error[] = []
     ctx.on('agent/error', (_agent, _turn, _step, error) => void errors.push(error))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'hi')
     await waitForIdle(ctx, agent)
@@ -223,11 +223,12 @@ describe('agent loop', () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter, 'You run on {{model}}.')
     ctx.on('system-prompt/assemble', async (assembly, _context, next) => {
+      assembly.variables['provider'] = 'mock'
       assembly.variables['model'] = 'mock'
       return next()
     })
     ctx.on('agent/request', async (_agent, _turn, _step, config, _next) => {
-      return { ...config, model: 'mock' }
+      return { ...config, provider: 'mock', model: 'mock' }
     })
     const agent = ctx.agentLoop.create(AgentId('a-late-model'), {})
 
@@ -255,7 +256,7 @@ describe('agent loop', () => {
       parameters: {},
       execute: () => Promise.resolve({ content: [{ type: 'text' as const, text: 'apparent success' }], meta }),
     }))
-    const agent = ctx.agentLoop.create(AgentId('bad-meta-agent'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('bad-meta-agent'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'use the tool')
     await waitForIdle(ctx, agent)
@@ -284,7 +285,7 @@ describe('agent loop', () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
     ctx.on('system-prompt/assemble', async () => ({ sections: [], tools: [], variables: {} }))
-    const agent = ctx.agentLoop.create(AgentId('a-no-system'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a-no-system'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'hi')
     await waitForIdle(ctx, agent)
@@ -296,7 +297,7 @@ describe('agent loop', () => {
   it('records raw chunks for replay as assistant/chunk session events', async () => {
     const adapter = new MockAdapter([textResponse('abc')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'hi')
     await waitForIdle(ctx, agent)
@@ -320,7 +321,7 @@ describe('agent loop', () => {
     ])
     const ctx = await harness(adapter)
 
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     ctx.tools.register(defineTool({
       name: 'slow',
       description: '',
@@ -352,7 +353,7 @@ describe('agent loop', () => {
   it('steering while idle behaves like send (starts a turn)', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     agent.steer([{ type: 'text', text: 'hello' }])
     await waitForIdle(ctx, agent)
@@ -362,7 +363,7 @@ describe('agent loop', () => {
   it('inject() while idle wraps context in a one-shot turn, visible to the next request', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     agent.inject([{ type: 'text', text: 'file changed: a.ts' }], { source: { kind: 'plugin', plugin: 'watcher' } })
     // The idle inject records a self-contained turn (turn/start → context/message
@@ -383,22 +384,56 @@ describe('agent loop', () => {
     expect(flat).toContain('<context source=\\"plugin\\">')
   })
 
-  it('inject() while running appends into the open turn (no extra synthetic turn)', async () => {
+  it('inject() can persist raw structured context without the generic context envelope', async () => {
+    const adapter = new MockAdapter([textResponse('ok')])
+    const ctx = await harness(adapter)
+    const agent = ctx.agentLoop.create(AgentId('raw-context'), { provider: 'mock', model: 'mock' })
+    const text = '<system-reminder>Additional instructions from: pkg/AGENTS.md</system-reminder>'
+    const meta = {
+      kind: 'workspace-instructions',
+      version: 1,
+      changes: [{ action: 'set', scope: 'pkg', path: 'pkg/AGENTS.md', digest: 'abc123' }],
+    }
+
+    agent.inject([{ type: 'text', text }], {
+      source: { kind: 'plugin', plugin: 'workspace-context' },
+      envelope: 'raw',
+      meta,
+    })
+    send(agent, 'go')
+    await waitForIdle(ctx, agent)
+
+    const contextEvent = agent.session.events.find(event => event.type === 'context/message')
+    expect(contextEvent?.type === 'context/message' && contextEvent.data).toMatchObject({ envelope: 'raw', meta })
+    const requestText = JSON.stringify(adapter.requests[0]!.messages)
+    expect(requestText).toContain('Additional instructions from: pkg/AGENTS.md')
+    expect(requestText).not.toContain('<context source=')
+  })
+
+  it('defers inject() during tool execution until after the tool result', async () => {
     const adapter = new MockAdapter([
       toolCallResponse('c1', 'noticer', {}, 'calling'),
       textResponse('done'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    // A tool that injects mid-execution: at this point the agent is running, so
-    // inject must append the context/message into the ALREADY-open turn rather
-    // than wrap it in its own one-shot turn.
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+    let visibleDuringTool = false
+    const meta = { kind: 'deferred-test', version: 1 }
     ctx.tools.register(defineTool({
       name: 'noticer',
       description: 'injects a notice',
       parameters: {},
       async execute() {
-        agent.inject([{ type: 'text', text: 'mid-turn notice' }], { source: { kind: 'plugin', plugin: 'x' } })
+        await Promise.resolve()
+        const first = { type: 'text' as const, text: 'mid-turn notice' }
+        agent.inject([first], {
+          source: { kind: 'plugin', plugin: 'x' },
+          envelope: 'raw',
+          meta,
+        })
+        first.text = 'mutated after inject'
+        agent.inject([{ type: 'text', text: 'second notice' }], { source: { kind: 'plugin', plugin: 'x' } })
+        visibleDuringTool = agent.session.events.some(e => e.type === 'context/message')
         return [{ type: 'text', text: 'ok' }]
       },
     }))
@@ -406,13 +441,67 @@ describe('agent loop', () => {
     send(agent, 'go')
     await waitForIdle(ctx, agent)
 
-    // Exactly ONE turn ran (no synthetic injection turn), and the mid-turn
-    // context/message sits inside it.
+    expect(visibleDuringTool).toBe(false)
+
+    // The injection stays in the open turn, but its user-role context cannot
+    // split the assistant tool call from the provider's tool-result message.
     const turnStarts = agent.session.events.filter(e => e.type === 'turn/start')
     expect(turnStarts).toHaveLength(1)
     const ts0 = turnStarts[0]!
     expect(ts0.type === 'turn/start' && ts0.data.trigger.kind).toBe('message')
-    expect(agent.session.events.some(e => e.type === 'context/message')).toBe(true)
+    const result = agent.session.events.find(e => e.type === 'tool/result')!
+    const contexts = agent.session.events.filter(e => e.type === 'context/message')
+    expect(contexts).toHaveLength(2)
+    expect(result.seq).toBeLessThan(contexts[0]!.seq)
+    expect(contexts[0]?.type === 'context/message' && contexts[0].data).toMatchObject({
+      envelope: 'raw',
+      meta,
+    })
+    expect(contexts.flatMap(event => event.type === 'context/message' ? event.data.content : []))
+      .toEqual([
+        { type: 'text', text: 'mid-turn notice' },
+        { type: 'text', text: 'second notice' },
+      ])
+
+    const secondRequest = adapter.requests[1]!.messages
+    const resultIndex = secondRequest.findIndex(message =>
+      message.content.some(block => block.type === 'tool-result'))
+    const contextIndexes = secondRequest.flatMap((message, index) =>
+      message.content.some(block => block.type === 'text'
+        && (block.text.includes('mid-turn notice') || block.text.includes('second notice')))
+        ? [index]
+        : [])
+    expect(resultIndex).toBeGreaterThanOrEqual(0)
+    expect(contextIndexes).toHaveLength(2)
+    expect(contextIndexes.every(index => index > resultIndex)).toBe(true)
+  })
+
+  it('rejects non-JSON context before it enters the active tool-batch FIFO', async () => {
+    const adapter = new MockAdapter([
+      toolCallResponse('c1', 'invalid-injector', {}, 'calling'),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter)
+    const agent = ctx.agentLoop.create(AgentId('invalid-context'), { provider: 'mock', model: 'mock' })
+    ctx.tools.register(defineTool({
+      name: 'invalid-injector',
+      description: 'attempts an invalid context injection',
+      parameters: {},
+      async execute() {
+        expect(() => {
+          agent.inject([{ type: 'text', text: 'invalid' }], {
+            source: { kind: 'plugin', plugin: 'test' },
+            meta: { bigint: 1n } as never,
+          })
+        }).toThrow('agent context must be losslessly JSON-serializable')
+        return [{ type: 'text', text: 'rejected invalid context' }]
+      },
+    }))
+
+    send(agent, 'go')
+    await waitForIdle(ctx, agent)
+
+    expect(agent.session.events.some(event => event.type === 'context/message')).toBe(false)
   })
 
   it('agent/turn-continuation can force-continue (/loop pattern) and force-stop', async () => {
@@ -423,7 +512,7 @@ describe('agent loop', () => {
       textResponse('step 3'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let steps = 0
     ctx.on('session/event', (_session, event) => { if (event.type === 'step/end') steps++ })
@@ -449,7 +538,7 @@ describe('agent loop', () => {
         return [{ type: 'text', text: String(args.text) }]
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     ctx.on('agent/turn-continuation', async () => ({ action: 'stop' }) as const)
 
@@ -464,8 +553,7 @@ describe('agent loop', () => {
   it('agent/request waterfall switches models by returning a replacement config; the switch is logged', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    ctx.llm.registerAdapter(['other-model'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     ctx.on('agent/request', async (_agent, _turn, _step, config, _next) => {
       // The seed is frozen — config is not a mutable per-call knob; a switch
@@ -498,7 +586,7 @@ describe('agent loop', () => {
       name: 'echo', description: 'echo', parameters: {},
       async execute() { return [{ type: 'text', text: 'echoed' }] },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const fires: { turn: number; step: number; fullSystemPrompt: string }[] = []
     ctx.on('agent/pre-step', (subject, turn, step, fullSystemPrompt) => {
@@ -522,7 +610,7 @@ describe('agent loop', () => {
     // same step's request must include it.
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let injected = false
     ctx.on('agent/pre-step', (subject) => {
@@ -556,7 +644,7 @@ describe('agent loop', () => {
     // closing, the turn records error, and the loop remains available.
     const adapter = new MockAdapter([textResponse('second turn ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let throwOnce = true
     ctx.on('agent/pre-step', () => {
@@ -590,7 +678,7 @@ describe('agent loop', () => {
   it('cancel() mid-stream ends the turn with reason aborted', async () => {
     const adapter = new MockAdapter(['hang'])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -610,7 +698,7 @@ describe('agent loop', () => {
     // turn stops by default and ends max-tokens, not completed.
     const adapter = new MockAdapter([maxTokensResponse('truncat')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -633,7 +721,7 @@ describe('agent loop', () => {
       textResponse('second half'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let steps = 0
     ctx.on('session/event', (_session, event) => { if (event.type === 'step/end') steps++ })
@@ -654,7 +742,7 @@ describe('agent loop', () => {
     expect(adapter.requests).toHaveLength(2)
     expect(adapter.requests[1]!.messages).toEqual([
       { role: 'user', content: [{ type: 'text', text: 'go' }] },
-      { role: 'assistant', content: [{ type: 'text', text: 'first half' }] },
+      { role: 'assistant', content: [{ type: 'text', text: 'first half' }], provenance: { provider: 'mock', model: 'mock' } },
     ])
     expect(reasons).toEqual([{ kind: 'max-tokens' }])
   })
@@ -664,7 +752,7 @@ describe('agent loop', () => {
     // stop. The per-turn reason must be independent — turn 2 ends completed.
     const adapter = new MockAdapter([maxTokensResponse('cut'), textResponse('clean')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -697,7 +785,7 @@ describe('agent loop', () => {
         return [{ type: 'text', text: 'should not run' }]
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -713,14 +801,13 @@ describe('agent loop', () => {
     // skips that host so it does not create a spurious assistant turn.
     const assistantMessage = agent.session.events.find(e => e.type === 'assistant/message')
     expect(assistantMessage?.type === 'assistant/message' && assistantMessage.data).toEqual({
-      turn: 1, step: 1, content: [], usage: { inputTokens: 10, outputTokens: 5 },
+      turn: 1, step: 1, content: [], provenance: { provider: 'mock', model: 'mock' }, usage: { inputTokens: 10, outputTokens: 5 },
     })
   })
 
-  it('appends no assistant/message for a max-tokens step with empty content and no usage', async () => {
-    // A max-tokens step truncated to a dropped tool call AND with no usage chunk has nothing to
-    // record: empty content and no accounting → no assistant/message (the empty-content host
-    // exists only to carry usage).
+  it('appends an empty completion anchor for a max-tokens step with no usage', async () => {
+    // The truncated tool call is dropped from durable content, while the
+    // successful provider call still needs an exact replay anchor.
     const callId = CallId('c1')
     const adapter = new MockAdapter([[
       { type: 'block-start', index: 0, blockType: 'tool-call' },
@@ -735,7 +822,7 @@ describe('agent loop', () => {
       parameters: { text: { type: 'string' } },
       async execute() { return [{ type: 'text', text: 'should not run' }] },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -744,17 +831,23 @@ describe('agent loop', () => {
     await waitForIdle(ctx, agent)
 
     expect(reasons).toEqual([{ kind: 'max-tokens' }])
-    expect(agent.session.events.some(e => e.type === 'assistant/message')).toBe(false)
+    const assistant = agent.session.events.find(e => e.type === 'assistant/message')!
+    expect(assistant.type === 'assistant/message' && assistant.data).toEqual({
+      turn: 1,
+      step: 1,
+      content: [],
+      provenance: { provider: 'mock', model: 'mock' },
+    })
+    expect(assistant.sourceEventSeqs?.length).toBeGreaterThan(0)
     expect(agent.session.deriveMessages()).toEqual([{ role: 'user', content: [{ type: 'text', text: 'go' }] }])
   })
 
-  it('appends no assistant/message for a normal stop finish with empty content and no usage', async () => {
-    // A clean `stop` finish that streamed nothing assembled (no blocks) and
-    // carried no usage chunk has nothing to record: the content-or-usage guard
-    // on the normal step path suppresses a pure trace-only empty assistant/message.
+  it('appends an empty completion anchor for a normal stop with no usage', async () => {
+    // A clean content-less call stays absent from derived messages but remains
+    // a durable successful-call boundary for replay consumers.
     const adapter = new MockAdapter([[{ type: 'finish', reason: { kind: 'stop' } }]])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const reasons: TurnEndReason[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/end') reasons.push(event.data.reason) })
@@ -763,7 +856,14 @@ describe('agent loop', () => {
     await waitForIdle(ctx, agent)
 
     expect(reasons).toEqual([{ kind: 'completed' }])
-    expect(agent.session.events.some(e => e.type === 'assistant/message')).toBe(false)
+    const assistant = agent.session.events.find(e => e.type === 'assistant/message')!
+    expect(assistant.type === 'assistant/message' && assistant.data).toEqual({
+      turn: 1,
+      step: 1,
+      content: [],
+      provenance: { provider: 'mock', model: 'mock' },
+    })
+    expect(assistant.sourceEventSeqs?.length).toBe(1)
     expect(agent.session.deriveMessages()).toEqual([{ role: 'user', content: [{ type: 'text', text: 'go' }] }])
   })
 
@@ -784,7 +884,7 @@ describe('agent loop', () => {
       expect(message.content).toEqual([{ type: 'text', text: 'partial text' }])
       return next()
     })
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'go')
     await waitForIdle(ctx, agent)
@@ -793,7 +893,7 @@ describe('agent loop', () => {
     expect(agent.session.events.some(e => e.type === 'tool/call')).toBe(false)
     expect(agent.session.deriveMessages()).toEqual([
       { role: 'user', content: [{ type: 'text', text: 'go' }] },
-      { role: 'assistant', content: [{ type: 'text', text: 'partial text' }] },
+      { role: 'assistant', content: [{ type: 'text', text: 'partial text' }], provenance: { provider: 'mock', model: 'mock' } },
     ])
   })
 
@@ -811,7 +911,7 @@ describe('agent loop', () => {
         return [{ type: 'text', text: String(args.text) }]
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     let threw = false
     // Post-commit session observers cannot control the loop. The tool call still
     // drives the second model request, and the turn completes normally.
@@ -830,7 +930,7 @@ describe('agent loop', () => {
   it('chains queued messages into consecutive turns', async () => {
     const adapter = new MockAdapter([textResponse('first'), textResponse('second')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const turns: number[] = []
     ctx.on('session/event', (_s, event) => { if (event.type === 'turn/start') turns.push(event.data.turn) })
@@ -855,7 +955,7 @@ describe('agent loop', () => {
   it('awaits session/flush at turn end (persistence checkpoint)', async () => {
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let flushed = 0
     let flushedBeforeIdle = false
@@ -875,7 +975,7 @@ describe('agent loop', () => {
   it('errors from the model surface as agent/error and end the turn', async () => {
     const adapter = new MockAdapter([]) // script exhausted → throws
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const errors: Error[] = []
     const reasons: TurnEndReason[] = []
@@ -900,7 +1000,7 @@ describe('agent loop', () => {
 
     let agent!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      agent = inner.agentLoop.create(AgentId('scoped'), { model: 'mock' })
+      agent = inner.agentLoop.create(AgentId('scoped'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
 
     expect(ctx.agents.get(AgentId('scoped'))).toBe(agent)
@@ -925,7 +1025,7 @@ describe('agent loop', () => {
     await ctx.plugin(ToolRegistry)
     await ctx.plugin(AgentRegistry)
     await ctx.plugin(AgentLoop, {
-      agents: [{ id: AgentId('config-agent'), model: 'mock' }],
+      agents: [{ id: AgentId('config-agent'), provider: 'mock', model: 'mock' }],
     })
     ctx.llm.registerAdapter(['mock'], adapter)
 
@@ -948,7 +1048,7 @@ describe('agent loop', () => {
     await ctx.plugin(ToolRegistry)
     await ctx.plugin(AgentRegistry)
     await ctx.plugin(AgentLoop, {
-      agents: [{ id: AgentId('config-agent'), model: 'mock', cwd: '/work/project' }],
+      agents: [{ id: AgentId('config-agent'), provider: 'mock', model: 'mock', cwd: '/work/project' }],
     })
 
     const agent = ctx.agents.get(AgentId('config-agent'))! as ReactLoopAgent
@@ -969,7 +1069,7 @@ describe('agent loop', () => {
         return [{ type: 'text', text: String(args.text) }]
       },
     }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     send(agent, 'run')
     await waitForIdle(ctx, agent)
 
diff --git a/packages/core/agent-loop/tests/properties.spec.ts b/packages/core/agent-loop/tests/properties.spec.ts
index dbc43ad985..6e502187a6 100644
--- a/packages/core/agent-loop/tests/properties.spec.ts
+++ b/packages/core/agent-loop/tests/properties.spec.ts
@@ -90,7 +90,7 @@ describe('agent loop scheduling properties', () => {
       async (texts) => {
         const ctx = await harness()
         try {
-          const agent = ctx.agentLoop.create(AgentId('a'), { model: 'mock' })
+          const agent = ctx.agentLoop.create(AgentId('a'), { provider: 'mock', model: 'mock' })
           const { seen: trace } = recordStatus(ctx, agent)
           const idle = nextIdle(ctx, agent)
           // Send all in one synchronous tick: they queue before the loop wakes.
@@ -115,7 +115,7 @@ describe('agent loop scheduling properties', () => {
       async (texts) => {
         const ctx = await harness()
         try {
-          const agent = ctx.agentLoop.create(AgentId('a'), { model: 'mock' })
+          const agent = ctx.agentLoop.create(AgentId('a'), { provider: 'mock', model: 'mock' })
           for (const text of texts) {
             const idle = nextIdle(ctx, agent)
             agent.send([{ type: 'text', text }])
@@ -140,7 +140,7 @@ describe('agent loop scheduling properties', () => {
       async (steps) => {
         const ctx = await harness()
         try {
-          const agent = ctx.agentLoop.create(AgentId('a'), { model: 'mock' })
+          const agent = ctx.agentLoop.create(AgentId('a'), { provider: 'mock', model: 'mock' })
           // Capture before each send; the last waiter covers the final turn, and
           // awaiting an already-settled earlier waiter is harmless.
           let lastIdle: Promise<void> | undefined
diff --git a/packages/core/agent-loop/tests/request-cache.e2e.ts b/packages/core/agent-loop/tests/request-cache.e2e.ts
index 9b8513d16f..72c2c577ab 100644
--- a/packages/core/agent-loop/tests/request-cache.e2e.ts
+++ b/packages/core/agent-loop/tests/request-cache.e2e.ts
@@ -43,7 +43,7 @@ async function loopHarness(): Promise<Context> {
   await created.plugin(ToolRegistry)
   await created.plugin(AgentRegistry)
   await created.plugin(AgentLoop, { agents: [] })
-  await created.plugin(LlmDeepSeek, { models: ['deepseek-v4-flash'] })
+  await created.plugin(LlmDeepSeek)
   created.tools.register(defineTool({
     name: 'lookup',
     description: 'Look up the stored value for a key.',
@@ -69,7 +69,7 @@ function waitForIdle(context: Context, agent: Agent): Promise<void> {
 describe.skipIf(!process.env.DEEPSEEK_API_KEY)('log-derived request cache hits (real API)', () => {
   it('every request after the first hits the provider prefix cache', async () => {
     ctx = await loopHarness()
-    const agent = ctx.agentLoop.create(AgentId('cache-e2e'), { model: 'deepseek-v4-flash' })
+    const agent = ctx.agentLoop.create(AgentId('cache-e2e'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     // Turn 1: forces a tool call → at least two steps (two model requests).
     agent.send([{ type: 'text', text: 'Look up the key "deploy-color" with the lookup tool and tell me the value.' }])
diff --git a/packages/core/agent-loop/tests/request-log.spec.ts b/packages/core/agent-loop/tests/request-log.spec.ts
index a6befde84e..4a3cfe37a4 100644
--- a/packages/core/agent-loop/tests/request-log.spec.ts
+++ b/packages/core/agent-loop/tests/request-log.spec.ts
@@ -1,9 +1,8 @@
 /**
- * recordRequestHeader unit tests: exactly one of four things per request —
+ * recordRequestHeader unit tests: exactly one of three things per request —
  * an 'initial' snapshot (log has no header yet), a 'resume' snapshot (fresh
- * loop instance over a log that has one), nothing (header unchanged), a
- * round-tripping delta, or a 'fallback' snapshot when the delta encoding
- * cannot express the change (pure tool reordering).
+ * loop instance over a log that has one), nothing (header unchanged), or a
+ * full 'change' snapshot.
  */
 
 import { describe, expect, it } from 'vitest'
@@ -23,14 +22,14 @@ function openSession(id: string): Session {
 }
 
 function headerEvents(session: Session): SessionEvent[] {
-  return session.events.filter(e => e.type === 'request/header' || e.type === 'request/header-delta')
+  return session.events.filter(e => e.type === 'request/header')
 }
 
 describe('recordRequestHeader', () => {
   it("anchors a new conversation with an 'initial' snapshot, then logs nothing while unchanged", () => {
     const session = openSession('rl-initial')
     const state = createTransmissionLog()
-    const header = canonicalHeader({ config: { model: 'm' }, system: 's', tools: [tool('t')] })
+    const header = canonicalHeader({ config: { provider: 'mock', model: 'm' }, system: 's', tools: [tool('t')] })
 
     recordRequestHeader(session, state, header)
     const [first] = headerEvents(session)
@@ -42,7 +41,7 @@ describe('recordRequestHeader', () => {
 
   it("anchors a fresh loop instance over an anchored log with a 'resume' snapshot, even unchanged", () => {
     const session = openSession('rl-resume')
-    const header = canonicalHeader({ config: { model: 'm' }, system: 's' })
+    const header = canonicalHeader({ config: { provider: 'mock', model: 'm' }, system: 's' })
     recordRequestHeader(session, createTransmissionLog(), header)
 
     // A second instance (process restart / fork): the boundary itself is a
@@ -53,33 +52,31 @@ describe('recordRequestHeader', () => {
     expect(events[1]?.type === 'request/header' && events[1].data.reason).toBe('resume')
   })
 
-  it('logs a round-tripping delta for a mid-run change, and the fold reproduces the header', () => {
-    const session = openSession('rl-delta')
+  it("logs a full 'change' snapshot for a mid-run change, and the fold reproduces the header", () => {
+    const session = openSession('rl-change')
     const state = createTransmissionLog()
-    const first = canonicalHeader({ config: { model: 'm' }, system: 'a\nb', tools: [tool('t')] })
+    const first = canonicalHeader({ config: { provider: 'mock', model: 'm' }, system: 'a\nb', tools: [tool('t')] })
     recordRequestHeader(session, state, first)
 
-    const second = canonicalHeader({ config: { model: 'm' }, system: 'a\nc', tools: [tool('t'), tool('u')] })
+    const second = canonicalHeader({ config: { provider: 'mock', model: 'm' }, system: 'a\nc', tools: [tool('t'), tool('u')] })
     recordRequestHeader(session, state, second)
     const events = headerEvents(session)
     expect(events).toHaveLength(2)
-    expect(events[1]?.type).toBe('request/header-delta')
+    expect(events[1]?.type === 'request/header' && events[1].data.reason).toBe('change')
     expect(session.requestHeader()).toEqual(second)
   })
 
-  it("records a change the delta cannot express (pure reordering) as a 'fallback' snapshot", () => {
-    const session = openSession('rl-fallback')
+  it("records a pure tool reordering as a 'change' snapshot", () => {
+    const session = openSession('rl-reorder')
     const state = createTransmissionLog()
-    const first = canonicalHeader({ config: { model: 'm' }, tools: [tool('a'), tool('b')] })
+    const first = canonicalHeader({ config: { provider: 'mock', model: 'm' }, tools: [tool('a'), tool('b')] })
     recordRequestHeader(session, state, first)
 
-    const reordered = canonicalHeader({ config: { model: 'm' }, tools: [tool('b'), tool('a')] })
+    const reordered = canonicalHeader({ config: { provider: 'mock', model: 'm' }, tools: [tool('b'), tool('a')] })
     recordRequestHeader(session, state, reordered)
     const events = headerEvents(session)
     expect(events).toHaveLength(2)
-    expect(events[1]?.type === 'request/header' && events[1].data.reason).toBe('fallback')
-    // The fold still lands on the exact header — deltas are an encoding
-    // optimization, never a correctness dependency.
+    expect(events[1]?.type === 'request/header' && events[1].data.reason).toBe('change')
     expect(session.requestHeader()).toEqual(reordered)
   })
 })
diff --git a/packages/core/agent-loop/tests/request-reconstruction.spec.ts b/packages/core/agent-loop/tests/request-reconstruction.spec.ts
index c4fe471bd0..ab2c18c99d 100644
--- a/packages/core/agent-loop/tests/request-reconstruction.spec.ts
+++ b/packages/core/agent-loop/tests/request-reconstruction.spec.ts
@@ -1,9 +1,8 @@
 /**
  * Loop-level reconstructability: every request the loop sends is a pure function of the
- * session log — messages are the derivation at the step/start boundary, the header is the fold
- * of request/header* events — and every request is an append-extension of its predecessor
- * unless a logged event (compaction replace, header change) explains the difference. Mock-adapter
- * requests are the observable, and the final offline rebuild states the full contract end to end.
+ * session log — messages derive at the step/start boundary and the header is the latest
+ * request/header snapshot. Each request extends its predecessor unless a logged compaction
+ * replacement or header change explains the difference.
  */
 
 import { describe, expect, it } from 'vitest'
@@ -72,7 +71,7 @@ describe('request stability across the loop', () => {
     ])
     const ctx = await harness(adapter)
     registerEcho(ctx)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'go')
     await waitForIdle(ctx, agent)
@@ -85,7 +84,7 @@ describe('request stability across the loop', () => {
       expect(Object.isFrozen(request.messages)).toBe(true)
     }
     // One anchoring header snapshot; no further header events (nothing changed).
-    const headerEvents = agent.session.events.filter(e => e.type === 'request/header' || e.type === 'request/header-delta')
+    const headerEvents = agent.session.events.filter(e => e.type === 'request/header')
     expect(headerEvents).toHaveLength(1)
     expect(headerEvents[0]?.type === 'request/header' && headerEvents[0].data.reason).toBe('initial')
   })
@@ -93,7 +92,7 @@ describe('request stability across the loop', () => {
   it('a later turn append-extends the previous turn (one conversation, one log)', async () => {
     const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'first')
     await waitForIdle(ctx, agent)
@@ -107,7 +106,7 @@ describe('request stability across the loop', () => {
   it('a compaction replace rewrites the resend, and the log explains it', async () => {
     const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'first')
     await waitForIdle(ctx, agent)
@@ -122,8 +121,8 @@ describe('request stability across the loop', () => {
         content: [{ type: 'text', text: '[summary of turn 1]' }],
         source: { kind: 'plugin', plugin: 'test-compact' },
       }, {
-        surfaceOp: { op: 'replace', start: nodes[0]!.seq, end: nodes[1]!.seq },
-        sourceEventSeqs: [nodes[0]!.seq, nodes[1]!.seq],
+        surfaceOp: { op: 'replace', start: nodes[0]!, end: nodes[1]! },
+        sourceEventSeqs: [nodes[0]!, nodes[1]!],
       })
     })
 
@@ -137,24 +136,25 @@ describe('request stability across the loop', () => {
     expect(agent.session.events.filter(e => e.type === 'request/header')).toHaveLength(1)
   })
 
-  it('a real system-prompt change is a logged header delta; a stable prompt logs nothing', async () => {
+  it('a real system-prompt change is a full changed-header snapshot; a stable prompt logs nothing', async () => {
     const adapter = new MockAdapter([textResponse('one'), textResponse('two'), textResponse('three')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'first')
     await waitForIdle(ctx, agent)
     send(agent, 'second')
     await waitForIdle(ctx, agent)
     // Identical assembly re-rendered per step is NOT a change.
-    expect(agent.session.events.filter(e => e.type === 'request/header-delta')).toHaveLength(0)
+    expect(agent.session.events.filter(e => e.type === 'request/header')).toHaveLength(1)
 
     ctx.systemPrompt.section({ name: 'extra', order: 2, text: 'new guidance' })
     send(agent, 'third')
     await waitForIdle(ctx, agent)
 
-    const deltas = agent.session.events.filter(e => e.type === 'request/header-delta')
-    expect(deltas).toHaveLength(1)
+    const snapshots = agent.session.events.filter(e => e.type === 'request/header')
+    expect(snapshots).toHaveLength(2)
+    expect(snapshots[1]?.data.reason).toBe('change')
     expect(adapter.requests[2]!.system).toContain('new guidance')
     // History is preserved across the change — only the header moved.
     expect(adapter.requests[2]!.messages.length).toBeGreaterThan(adapter.requests[1]!.messages.length)
@@ -163,7 +163,7 @@ describe('request stability across the loop', () => {
   it('an inject() during the agent/request waterfall joins the NEXT request (the step/start boundary)', async () => {
     const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     let injected = false
     ctx.on('agent/request', async (_agent, _turn, _step, _config, next) => {
@@ -191,7 +191,7 @@ describe('request stability across the loop', () => {
   it('a mutation attempt on the frozen request content throws into the step (loud, not silent)', async () => {
     const adapter = new MockAdapter([textResponse('one')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     const errors: Error[] = []
     ctx.on('agent/error', (_agent, _turn, _step, error) => void errors.push(error))
@@ -212,7 +212,7 @@ describe('request stability across the loop', () => {
   it('a fresh loop instance over a seeded log anchors with a resume snapshot and stays cache-aligned', async () => {
     const adapter = new MockAdapter([textResponse('one')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('gen1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('gen1'), { provider: 'mock', model: 'mock' })
     send(agent, 'first')
     await waitForIdle(ctx, agent)
 
@@ -224,7 +224,7 @@ describe('request stability across the loop', () => {
       agentId: AgentId('gen2'),
       sessionId: SessionId('gen2-session'),
       seed: [...agent.session.events],
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     const agent2 = handle.agent as ReactLoopAgent
     send(agent2, 'second')
@@ -241,7 +241,7 @@ describe('request stability across the loop', () => {
   it('a delegating listener cannot mutate the seed through next() — the fold stays log-true', async () => {
     const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     ctx.on('agent/request', async (_agent, _turn, _step, _config, next) => {
       const config = await next()
@@ -260,9 +260,9 @@ describe('request stability across the loop', () => {
     send(agent, 'second')
     await waitForIdle(ctx, agent)
 
-    // No delta was logged (nothing really changed), and the session's own
+    // No changed snapshot was logged (nothing really changed), and the session's own
     // fold is immutable state.
-    expect(agent.session.events.filter(e => e.type === 'request/header-delta')).toHaveLength(0)
+    expect(agent.session.events.filter(e => e.type === 'request/header')).toHaveLength(1)
     expect(Object.isFrozen(agent.session.requestHeader())).toBe(true)
     expect(adapter.requests[1]!.temperature).toBeUndefined()
   })
@@ -275,7 +275,7 @@ describe('request stability across the loop', () => {
     ])
     const ctx = await harness(adapter)
     registerEcho(ctx)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
 
     send(agent, 'go')
     await waitForIdle(ctx, agent)
@@ -296,7 +296,7 @@ describe('request stability across the loop', () => {
       const rebuilt = new Session(SessionId(`rebuild-${index}`), structuredClone(events.slice(0, stepStart.seq)))
       expect(structuredClone(request.messages)).toEqual(rebuilt.deriveMessages())
 
-      // Header: the fold of request/header* events up to this step's dispatch
+      // Header: the latest request/header snapshot up to this step's dispatch
       // (its header event sits between step/start and the first chunk).
       const firstChunk = events.find(e => e.type === 'assistant/chunk' && e.seq > stepStart.seq)!
       const header = foldRequestHeader(events.slice(0, firstChunk.seq))!
diff --git a/packages/core/agent-loop/tests/resume.spec.ts b/packages/core/agent-loop/tests/resume.spec.ts
index 3747824231..aa131009c6 100644
--- a/packages/core/agent-loop/tests/resume.spec.ts
+++ b/packages/core/agent-loop/tests/resume.spec.ts
@@ -201,7 +201,7 @@ describe('the session-persistence RFC: AgentLoop factory create/resume', () => {
     const resuming = ctx.agents.resume({
       agentId: AgentId('resumed-atomic'),
       resumeSessionId: sessionId,
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       setup: async (agentCtx) => {
         expect(agentCtx.agent?.id).toBe(AgentId('resumed-atomic'))
         expect(agentCtx.agent?.session.events).toHaveLength(2)
@@ -242,7 +242,7 @@ describe('the session-persistence RFC: AgentLoop factory create/resume', () => {
     const handle = await ctx.agents.resume({
       agentId,
       resumeSessionId: sessionId,
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     const transactionLabels = [
       `agentLoop.owner(${agentId})`,
@@ -267,7 +267,7 @@ describe('the session-persistence RFC: AgentLoop factory create/resume', () => {
     await expect(ctx.agents.resume({
       agentId: AgentId('resume-reject'),
       resumeSessionId: sessionId,
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       setup: async () => {
         await Promise.resolve()
         throw new Error('resume setup failed')
@@ -280,7 +280,7 @@ describe('the session-persistence RFC: AgentLoop factory create/resume', () => {
     const retry = await ctx.agents.resume({
       agentId: AgentId('resume-reject'),
       resumeSessionId: sessionId,
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     await retry.dispose()
     await ctx.fiber.dispose()
@@ -301,7 +301,7 @@ describe('the session-persistence RFC: AgentLoop factory create/resume', () => {
       resuming = inner.agents.resume({
         agentId: AgentId('resume-owner-race'),
         resumeSessionId: sessionId,
-        agentOptions: { model: 'mock' },
+        agentOptions: { provider: 'mock', model: 'mock' },
         setup: async () => {
           setupStarted.resolve(undefined)
           await gate.promise
@@ -348,7 +348,7 @@ describe('the session-persistence RFC: AgentLoop factory create/resume', () => {
 
     let resuming!: ReturnType<typeof ctx.agents.resume>
     const owner = await ctx.plugin(Object.assign((inner: Context) => {
-      resuming = inner.agents.resume({ agentId, resumeSessionId: sessionId, agentOptions: { model: 'mock' } })
+      resuming = inner.agents.resume({ agentId, resumeSessionId: sessionId, agentOptions: { provider: 'mock', model: 'mock' } })
     }, { inject: ['agents'] }))
     await loadStarted.promise
 
@@ -360,7 +360,7 @@ describe('the session-persistence RFC: AgentLoop factory create/resume', () => {
 
     // owner.dispose() awaited transaction settlement, so the same identities
     // can be reused before awaiting the public rejection.
-    const retry = await promptly(ctx.agents.resume({ agentId, resumeSessionId: sessionId, agentOptions: { model: 'mock' } }))
+    const retry = await promptly(ctx.agents.resume({ agentId, resumeSessionId: sessionId, agentOptions: { provider: 'mock', model: 'mock' } }))
     await rejection
     expect(loads).toBe(2)
     expect(published).toEqual(['session/created', 'agent/created', 'agent/session-start'])
@@ -404,7 +404,7 @@ describe('the session-persistence RFC: AgentLoop factory create/resume', () => {
     ctx.on('session/created', () => void published.push('session/created'))
     ctx.on('agent/created', () => void published.push('agent/created'))
 
-    const resuming = ctx.agents.resume({ agentId, resumeSessionId: sessionId, agentOptions: { model: 'mock' } })
+    const resuming = ctx.agents.resume({ agentId, resumeSessionId: sessionId, agentOptions: { provider: 'mock', model: 'mock' } })
     await loadStarted.promise
     const rejection = expect(promptly(resuming)).rejects.toThrow(/agent loop is not active/)
     await promptly(loopFiber.dispose())
diff --git a/packages/core/agent-loop/tests/scope-lifecycle.spec.ts b/packages/core/agent-loop/tests/scope-lifecycle.spec.ts
index 40272ac42b..472cb8cd8a 100644
--- a/packages/core/agent-loop/tests/scope-lifecycle.spec.ts
+++ b/packages/core/agent-loop/tests/scope-lifecycle.spec.ts
@@ -144,7 +144,7 @@ describe('agent scope lifecycle', () => {
 
   it('wires agent.ctx: tagged with the agent, DX field set, ctx.agent safe elsewhere', async () => {
     const ctx = await harness()
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     expect(scopeOf(agent.ctx)).toBe(agent)
     expect(agent.ctx.agent).toBe(agent)
     // The root accessor default: a plain context answers undefined, not a throw.
@@ -154,7 +154,7 @@ describe('agent scope lifecycle', () => {
 
   it('scoped registrations live in the agent world and die with the agent', async () => {
     const ctx = await harness()
-    const handle = await ctx.agents.create({ agentId: AgentId('a1'), sessionId: SessionId('s1'), agentOptions: { model: 'mock' } })
+    const handle = await ctx.agents.create({ agentId: AgentId('a1'), sessionId: SessionId('s1'), agentOptions: { provider: 'mock', model: 'mock' } })
     const { agent } = handle
     agent.ctx.systemPrompt.section({ name: 'deployment:persona', order: 0, text: 'You run tests.' })
     agent.ctx.tools.register({
@@ -179,8 +179,8 @@ describe('agent scope lifecycle', () => {
 
   it('agent.ctx listeners hear only their own agent (scoped dispatch end to end)', async () => {
     const ctx = await harness(new MockAdapter([textResponse('one'), textResponse('two')]))
-    const a = ctx.agentLoop.create(AgentId('a'), { model: 'mock' })
-    const b = ctx.agentLoop.create(AgentId('b'), { model: 'mock' })
+    const a = ctx.agentLoop.create(AgentId('a'), { provider: 'mock', model: 'mock' })
+    const b = ctx.agentLoop.create(AgentId('b'), { provider: 'mock', model: 'mock' })
 
     const heard: string[] = []
     a.ctx.on('agent/status', (subject, status) => void heard.push(`a-sees:${subject.id}:${status}`))
@@ -212,7 +212,7 @@ describe('agent scope lifecycle', () => {
     const handle = await ctx.agents.create({
       agentId: AgentId('child'),
       sessionId: SessionId('child-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       setup: async (agentCtx) => {
         order.push('setup')
         await Promise.resolve()
@@ -236,7 +236,7 @@ describe('agent scope lifecycle', () => {
     })
     ctx.on('agent/created', () => void order.push('agent/created'))
     ctx.on('agent/session-start', () => void order.push('agent/session-start'))
-    const acceptedOptions = { model: 'mock' }
+    const acceptedOptions = { provider: 'mock', model: 'mock' }
 
     const creating = ctx.agents.create({
       agentId: AgentId('atomic'),
@@ -285,13 +285,13 @@ describe('agent scope lifecycle', () => {
     const first = ctx.agents.create({
       agentId,
       sessionId: SessionId('concurrent-final-enter-a'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       setup,
     })
     const second = ctx.agents.create({
       agentId,
       sessionId: SessionId('concurrent-final-enter-b'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       setup,
     })
     await bothStarted.promise
@@ -320,7 +320,7 @@ describe('agent scope lifecycle', () => {
     const pending = ctx.agents.create({
       agentId: AgentId('signal-pending'),
       sessionId: SessionId('signal-pending-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       signal: pendingController.signal,
       setup: async () => {
         setupStarted.resolve(undefined)
@@ -337,7 +337,7 @@ describe('agent scope lifecycle', () => {
     const live = await ctx.agents.create({
       agentId: AgentId('signal-live'),
       sessionId: SessionId('signal-live-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       signal: liveController.signal,
     })
     liveController.abort(new Error('too late'))
@@ -360,7 +360,7 @@ describe('agent scope lifecycle', () => {
       creating = inner.agents.create({
         agentId: AgentId('owner-race'),
         sessionId: SessionId('owner-race-s'),
-        agentOptions: { model: 'mock' },
+        agentOptions: { provider: 'mock', model: 'mock' },
         setup: async () => {
           setupStarted.resolve(undefined)
           await gate.promise
@@ -388,7 +388,7 @@ describe('agent scope lifecycle', () => {
       creating2 = inner.agents.create({
         agentId: AgentId('owner-race-2'),
         sessionId: SessionId('owner-race-s-2'),
-        agentOptions: { model: 'mock' },
+        agentOptions: { provider: 'mock', model: 'mock' },
         setup: async () => {
           setupStarted2.resolve(undefined)
           await gate2.promise
@@ -415,7 +415,7 @@ describe('agent scope lifecycle', () => {
     const creating = ctx.agents.create({
       agentId: AgentId('factory-setup-race'),
       sessionId: SessionId('factory-setup-race-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       setup: async () => {
         setupStarted.resolve(undefined)
         await gate.promise
@@ -446,7 +446,7 @@ describe('agent scope lifecycle', () => {
     const creating = ctx.agents.create({
       agentId: AgentId('factory-scope-race'),
       sessionId: SessionId('factory-scope-race-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       setup: () => { setupCalls += 1 },
     })
     await expect(creating).rejects.toThrow(/agent loop is not active/)
@@ -481,7 +481,7 @@ describe('agent scope lifecycle', () => {
       creating = inner.agents.create({
         agentId: AgentId('caller-scope-race'),
         sessionId: SessionId('caller-scope-race-s'),
-        agentOptions: { model: 'mock' },
+        agentOptions: { provider: 'mock', model: 'mock' },
       })
     }, { inject: ['agents'] }))
 
@@ -511,7 +511,7 @@ describe('agent scope lifecycle', () => {
       void loopFiber.dispose()
     })
 
-    expect(() => ctx.agentLoop.create(AgentId('config-scope-race'), { model: 'mock' }))
+    expect(() => ctx.agentLoop.create(AgentId('config-scope-race'), { provider: 'mock', model: 'mock' }))
       .toThrow(/agent loop is not active/)
     await loopFiber.dispose()
     expect(ctx.agents.get(AgentId('config-scope-race'))).toBeUndefined()
@@ -523,9 +523,9 @@ describe('agent scope lifecycle', () => {
     const ctx = await harness()
     const id = AgentId('config-prepare-failure')
 
-    expect(() => ctx.agentLoop.create(id, { model: 'mock' }, { cwd: 'relative' }))
+    expect(() => ctx.agentLoop.create(id, { provider: 'mock', model: 'mock' }, { cwd: 'relative' }))
       .toThrow(/absolute path/)
-    const replacement = ctx.agentLoop.create(id, { model: 'mock' }, { cwd: '/recovered' })
+    const replacement = ctx.agentLoop.create(id, { provider: 'mock', model: 'mock' }, { cwd: '/recovered' })
     expect(ctx.agents.get(id)).toBe(replacement)
     await replacement.whenIdle()
     await ctx.fiber.dispose()
@@ -544,7 +544,7 @@ describe('agent scope lifecycle', () => {
     await expect(ctx.agents.create({
       agentId: AgentId('factory-scope-throw'),
       sessionId: SessionId('factory-scope-throw-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })).rejects.toThrow('scope preparation failed')
     await loopFiber.dispose()
     expect(ctx.agents.get(AgentId('factory-scope-throw'))).toBeUndefined()
@@ -560,7 +560,7 @@ describe('agent scope lifecycle', () => {
     const handle = await ctx.agents.create({
       agentId,
       sessionId: SessionId('factory-live-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
 
     await loopFiber.dispose()
@@ -585,7 +585,7 @@ describe('agent scope lifecycle', () => {
       creating = inner.agents.create({
         agentId: AgentId('dependency-origin'),
         sessionId: SessionId('dependency-origin-s'),
-        agentOptions: { model: 'mock' },
+        agentOptions: { provider: 'mock', model: 'mock' },
         setup: (agentCtx) => {
           agentCtx.tools.register({
             name: 'dependency-origin-tool',
@@ -640,7 +640,7 @@ describe('agent scope lifecycle', () => {
       creating = inner.agents.create({
         agentId: AgentId('session-created-barrier'),
         sessionId: SessionId('session-created-barrier-s'),
-        agentOptions: { model: 'mock' },
+        agentOptions: { provider: 'mock', model: 'mock' },
       })
     }, { inject: ['agents'] }))
 
@@ -689,7 +689,7 @@ describe('agent scope lifecycle', () => {
       creating = inner.agents.create({
         agentId: AgentId('agent-created-barrier'),
         sessionId: SessionId('agent-created-barrier-s'),
-        agentOptions: { model: 'mock' },
+        agentOptions: { provider: 'mock', model: 'mock' },
       })
     }, { inject: ['agents'] }))
 
@@ -723,7 +723,7 @@ describe('agent scope lifecycle', () => {
       creating = inner.agents.create({
         agentId: AgentId('listener-dispose'),
         sessionId: SessionId('listener-dispose-s'),
-        agentOptions: { model: 'mock' },
+        agentOptions: { provider: 'mock', model: 'mock' },
       })
     }, { inject: ['agents'] }))
 
@@ -764,7 +764,7 @@ describe('agent scope lifecycle', () => {
       creating = inner.agents.create({
         agentId: AgentId('session-start-dispose'),
         sessionId: SessionId('session-start-dispose-s'),
-        agentOptions: { model: 'mock' },
+        agentOptions: { provider: 'mock', model: 'mock' },
       })
     }, { inject: ['agents'] }))
 
@@ -789,7 +789,7 @@ describe('agent scope lifecycle', () => {
     await expect(ctx.agents.create({
       agentId: AgentId('bad'),
       sessionId: SessionId('bad-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       setup: async () => {
         await Promise.resolve()
         throw new Error('boom setup')
@@ -800,7 +800,7 @@ describe('agent scope lifecycle', () => {
     expect(published).toEqual([])
     expect(ctx.agents.get(AgentId('bad'))).toBeUndefined()
     expect(ctx.sessions.get(SessionId('bad-s'))).toBeUndefined()
-    const retry = await ctx.agents.create({ agentId: AgentId('bad'), sessionId: SessionId('bad-s'), agentOptions: { model: 'mock' } })
+    const retry = await ctx.agents.create({ agentId: AgentId('bad'), sessionId: SessionId('bad-s'), agentOptions: { provider: 'mock', model: 'mock' } })
     await retry.dispose()
   })
 
@@ -819,7 +819,7 @@ describe('agent scope lifecycle', () => {
     await expect(ctx.agents.create({
       agentId: AgentId('exotic-seed'),
       sessionId: SessionId('exotic-seed-session'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       seed,
     })).rejects.toThrow(/seed event at index 0 is not losslessly JSON-serializable/)
 
@@ -829,7 +829,7 @@ describe('agent scope lifecycle', () => {
     const retry = await ctx.agents.create({
       agentId: AgentId('exotic-seed'),
       sessionId: SessionId('exotic-seed-session'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     await retry.dispose()
   })
@@ -843,13 +843,13 @@ describe('agent scope lifecycle', () => {
       if (boom) { boom = false; throw new Error('boom created') }
     })
     await expect(ctx.agents.create({
-      agentId: AgentId('bad'), sessionId: SessionId('bad-s'), agentOptions: { model: 'mock' },
+      agentId: AgentId('bad'), sessionId: SessionId('bad-s'), agentOptions: { provider: 'mock', model: 'mock' },
     })).rejects.toThrow('boom created')
     expect(ctx.agents.get(AgentId('bad'))).toBeUndefined()
     expect(ctx.sessions.get(SessionId('bad-s'))).toBeUndefined()
     expect(disposed).toEqual([]) // inserted but never announced: no impossible disposed edge
     // The rollback also disposed the scope fiber: re-creating works cleanly.
-    const retry = await ctx.agents.create({ agentId: AgentId('bad'), sessionId: SessionId('bad-s'), agentOptions: { model: 'mock' } })
+    const retry = await ctx.agents.create({ agentId: AgentId('bad'), sessionId: SessionId('bad-s'), agentOptions: { provider: 'mock', model: 'mock' } })
     expect(scopeOf(retry.agent.ctx)).toBe(retry.agent)
     await retry.dispose()
   })
@@ -868,7 +868,7 @@ describe('agent scope lifecycle', () => {
     await expect(ctx.agents.create({
       agentId: AgentId('partial-agent'),
       sessionId: SessionId('partial-session'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })).rejects.toThrow('agent observer failed')
 
     expect(lifecycle).toEqual([
@@ -892,7 +892,7 @@ describe('agent scope lifecycle', () => {
       }
     })
 
-    expect(() => ctx.agentLoop.create(AgentId('config-bad'), { model: 'mock' }))
+    expect(() => ctx.agentLoop.create(AgentId('config-bad'), { provider: 'mock', model: 'mock' }))
       .toThrow('config publish failed')
     expect(ctx.agents.get(AgentId('config-bad'))).toBeUndefined()
     expect(ctx.sessions.list()).toHaveLength(sessionsBefore)
@@ -900,15 +900,15 @@ describe('agent scope lifecycle', () => {
 
   it('registrations through a disposed agent ctx throw INACTIVE_EFFECT', async () => {
     const ctx = await harness()
-    const handle = await ctx.agents.create({ agentId: AgentId('a1'), sessionId: SessionId('s1'), agentOptions: { model: 'mock' } })
+    const handle = await ctx.agents.create({ agentId: AgentId('a1'), sessionId: SessionId('s1'), agentOptions: { provider: 'mock', model: 'mock' } })
     await handle.dispose()
     expect(() => handle.agent.ctx.on('agent/status', () => {})).toThrow(/inactive context/)
   })
 
   it('agentEvents fuses carrier and subject for custom drivers', async () => {
     const ctx = await harness()
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    const other = ctx.agentLoop.create(AgentId('a2'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+    const other = ctx.agentLoop.create(AgentId('a2'), { provider: 'mock', model: 'mock' })
     const heard: string[] = []
     agent.ctx.on('agent/error', (subject: Agent, turn: number) => void heard.push(`${subject.id}:${turn}`))
 
@@ -921,7 +921,7 @@ describe('agent scope lifecycle', () => {
     const ctx = await harness()
     let handle!: Awaited<ReturnType<typeof ctx.agents.create>>
     const owner = await ctx.plugin(Object.assign(async (inner: Context) => {
-      handle = await inner.agents.create({ agentId: AgentId('o1'), sessionId: SessionId('o1-s'), agentOptions: { model: 'mock' } })
+      handle = await inner.agents.create({ agentId: AgentId('o1'), sessionId: SessionId('o1-s'), agentOptions: { provider: 'mock', model: 'mock' } })
     }, { inject: ['agents'] }))
     const { agent } = handle
 
@@ -953,7 +953,7 @@ describe('agent scope lifecycle', () => {
     const ctx = await harness()
     let handle!: Awaited<ReturnType<typeof ctx.agents.create>>
     const owner = await ctx.plugin(Object.assign(async (inner: Context) => {
-      handle = await inner.agents.create({ agentId: AgentId('h1'), sessionId: SessionId('h1-s'), agentOptions: { model: 'mock' } })
+      handle = await inner.agents.create({ agentId: AgentId('h1'), sessionId: SessionId('h1-s'), agentOptions: { provider: 'mock', model: 'mock' } })
     }, { inject: ['agents'] }))
 
     const teardownDone: string[] = []
@@ -976,7 +976,7 @@ describe('agent scope lifecycle', () => {
     const handle = await ctx.agents.create({
       agentId,
       sessionId: SessionId('retired-owner-effect-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
 
     expect(ctx.fiber.getEffects().map(effect => effect.label)).toContain(`agentLoop.owner(${agentId})`)
@@ -994,7 +994,7 @@ describe('agent scope lifecycle', () => {
       handle = await inner.agents.create({
         agentId: AgentId('manual-first'),
         sessionId: SessionId('manual-first-s'),
-        agentOptions: { model: 'mock' },
+        agentOptions: { provider: 'mock', model: 'mock' },
         setup(agentCtx) {
           agentCtx.effect(() => async () => {
             cleanupStarted.resolve(undefined)
@@ -1030,7 +1030,7 @@ describe('agent scope lifecycle', () => {
     const first = await ctx.agents.create({
       agentId,
       sessionId,
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
       setup(agentCtx) {
         agentCtx.effect(() => async () => {
           cleanupStarted.resolve(undefined)
@@ -1043,7 +1043,7 @@ describe('agent scope lifecycle', () => {
     await Promise.all([sessionDisposed.promise, cleanupStarted.promise])
     expect(ctx.agents.get(agentId)).toBeUndefined()
     expect(ctx.sessions.get(sessionId)).toBeUndefined()
-    const replacement = await ctx.agents.create({ agentId, sessionId, agentOptions: { model: 'mock' } })
+    const replacement = await ctx.agents.create({ agentId, sessionId, agentOptions: { provider: 'mock', model: 'mock' } })
     expect(ctx.agents.get(agentId)).toBe(replacement.agent)
     expect(ctx.sessions.get(sessionId)).toBe(replacement.agent.session)
 
@@ -1058,7 +1058,7 @@ describe('agent scope lifecycle', () => {
     const handle = await ctx.agents.create({
       agentId: AgentId('idle-flush'),
       sessionId: SessionId('idle-flush-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     const gate = Promise.withResolvers<undefined>()
     let flushStarted = false
diff --git a/packages/core/agent-loop/tests/tool-calls.spec.ts b/packages/core/agent-loop/tests/tool-calls.spec.ts
new file mode 100644
index 0000000000..8feb20bd96
--- /dev/null
+++ b/packages/core/agent-loop/tests/tool-calls.spec.ts
@@ -0,0 +1,571 @@
+/**
+ * Exercises scheduler ordering and cancellation with deterministic gated tools.
+ * ACP goldens own transcript-facing coverage.
+ */
+
+import { describe, expect, it } from 'vitest'
+import { Context } from 'cordis'
+import { CallId, StreamChunk } from '@deepseek-ai/dsh-llm'
+import SessionStore, { SessionEvent } from '@deepseek-ai/dsh-session'
+import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import LlmService from '@deepseek-ai/dsh-llm'
+import ToolRegistry, { defineTool, type PostToolDecision, type PreToolDecision } from '@deepseek-ai/dsh-tools'
+import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import AgentLoop, { ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { MockAdapter, textResponse } from './mock-adapter.ts'
+
+async function harness(adapter: MockAdapter, maxParallelToolCalls?: number) {
+  const ctx = new Context()
+  await ctx.plugin(LlmService)
+  await ctx.plugin(SessionStore)
+  await ctx.plugin(SystemPrompt, { persona: '' })
+  await ctx.plugin(ToolRegistry)
+  await ctx.plugin(AgentRegistry)
+  await ctx.plugin(AgentLoop, {
+    agents: [],
+    ...maxParallelToolCalls === undefined ? {} : { maxParallelToolCalls },
+  })
+  ctx.llm.registerAdapter(['mock'], adapter)
+  return ctx
+}
+
+function waitForIdle(ctx: Context, agent: ReactLoopAgent): Promise<void> {
+  return new Promise((resolve) => {
+    const dispose = ctx.on('agent/status', (subject, status) => {
+      if (subject === agent && status === 'idle') { dispose(); resolve() }
+    })
+  })
+}
+
+function events(agent: ReactLoopAgent): SessionEvent[] {
+  return [...agent.session.events]
+}
+
+/** Build one assistant response containing the supplied tool calls. */
+function multiCall(calls: { id: string; name: string; args: object }[]): StreamChunk[] {
+  const chunks: StreamChunk[] = []
+  calls.forEach((call, index) => {
+    chunks.push(
+      { type: 'block-start', index, blockType: 'tool-call' },
+      { type: 'block-end', index, block: { type: 'tool-call', id: CallId(call.id), name: call.name, arguments: JSON.stringify(call.args) } },
+    )
+  })
+  chunks.push(
+    { type: 'usage', usage: { inputTokens: 5, outputTokens: 5 } },
+    { type: 'finish', reason: { kind: 'tool-calls' } },
+  )
+  return chunks
+}
+
+/** A tool whose calls block until the test releases them by callId. */
+function gatedTool(name: string, parallel: boolean) {
+  const gates = new Map<string, () => void>()
+  const started: string[] = []
+  const tool = defineTool({
+    name,
+    description: `gated ${name}`,
+    parameters: { id: { type: 'string', required: true } },
+    ...parallel ? { isConcurrencySafe: () => true } : {},
+    async execute(args) {
+      started.push(args.id)
+      await new Promise<void>((resolve) => { gates.set(args.id, resolve) })
+      return [{ type: 'text', text: `done-${args.id}` }]
+    },
+  })
+  return {
+    tool,
+    started,
+    release(id: string) { gates.get(id)?.(); gates.delete(id) },
+    pending() { return [...gates.keys()] },
+  }
+}
+
+function gatedParallelTool(name: string) {
+  return gatedTool(name, true)
+}
+
+function gatedExclusiveTool(name: string) {
+  return gatedTool(name, false)
+}
+
+/** Poll until `predicate` holds, letting microtasks/timers drain between checks. */
+async function until(predicate: () => boolean): Promise<void> {
+  for (let i = 0; i < 1000 && !predicate(); i++) await new Promise(r => setTimeout(r, 0))
+  if (!predicate()) throw new Error('until: condition never held')
+}
+
+describe('tool-call scheduler: grouping and barriers', () => {
+  it('runs parallel-safe siblings concurrently (all start before any completes)', async () => {
+    const adapter = new MockAdapter([
+      multiCall([{ id: 'c1', name: 'p', args: { id: '1' } }, { id: 'c2', name: 'p', args: { id: '2' } }, { id: 'c3', name: 'p', args: { id: '3' } }]),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 3)
+    expect(gated.started).toEqual(['1', '2', '3'])
+    gated.release('1'); gated.release('2'); gated.release('3')
+    await waitForIdle(ctx, agent)
+  })
+
+  it('an exclusive call between two parallel-safe calls forms a barrier (3 groups)', async () => {
+    const order: string[] = []
+    const adapter = new MockAdapter([
+      multiCall([
+        { id: 'c1', name: 'r', args: { id: 'A1' } },
+        { id: 'c2', name: 'w', args: { id: 'A2' } },
+        { id: 'c3', name: 'r', args: { id: 'A3' } },
+      ]),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter)
+    ctx.tools.register(defineTool({
+      name: 'r', description: 'read', parameters: { id: { type: 'string', required: true } },
+      isConcurrencySafe: () => true,
+      async execute(args) { order.push(`r-start-${args.id}`); order.push(`r-end-${args.id}`); return [{ type: 'text', text: 'r' }] },
+    }))
+    ctx.tools.register(defineTool({
+      name: 'w', description: 'write', parameters: { id: { type: 'string', required: true } },
+      async execute(args) { order.push(`w-${args.id}`); return [{ type: 'text', text: 'w' }] },
+    }))
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+    agent.send([{ type: 'text', text: 'go' }])
+    await waitForIdle(ctx, agent)
+
+    expect(order).toEqual(['r-start-A1', 'r-end-A1', 'w-A2', 'r-start-A3', 'r-end-A3'])
+  })
+
+  it('reclassifies pending calls after an exclusive barrier replaces their tool', async () => {
+    const adapter = new MockAdapter([
+      multiCall([
+        { id: 'c1', name: 'replace', args: { id: '0' } },
+        { id: 'c2', name: 'x', args: { id: '1' } },
+        { id: 'c3', name: 'x', args: { id: '2' } },
+      ]),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter)
+    const replacement = gatedExclusiveTool('x')
+    const disposeSafe = ctx.tools.register(defineTool({
+      name: 'x',
+      description: 'initially safe',
+      parameters: { id: { type: 'string', required: true } },
+      isConcurrencySafe: () => true,
+      async execute(args) { return [{ type: 'text', text: `old-${args.id}` }] },
+    }))
+    ctx.tools.register(defineTool({
+      name: 'replace',
+      description: 'replace x',
+      parameters: { id: { type: 'string', required: true } },
+      async execute() {
+        disposeSafe()
+        ctx.tools.register(replacement.tool)
+        return [{ type: 'text', text: 'replaced' }]
+      },
+    }))
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => replacement.started.length === 1)
+    await new Promise(r => setTimeout(r, 5))
+    expect(replacement.started).toEqual(['1'])
+    replacement.release('1')
+    await until(() => replacement.started.length === 2)
+    expect(replacement.started).toEqual(['1', '2'])
+    replacement.release('2')
+    await waitForIdle(ctx, agent)
+  })
+
+  it('stops replenishing when a result observer makes the next call exclusive', async () => {
+    const adapter = new MockAdapter([
+      multiCall([
+        { id: 'c1', name: 'x', args: { id: '1' } },
+        { id: 'c2', name: 'x', args: { id: '2' } },
+        { id: 'c3', name: 'x', args: { id: '3' } },
+      ]),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter, 2)
+    const initial = gatedParallelTool('x')
+    const replacement = gatedExclusiveTool('x')
+    const disposeInitial = ctx.tools.register(initial.tool)
+    ctx.on('tools/result', (exec) => {
+      if (exec.callId !== CallId('c1')) return
+      disposeInitial()
+      ctx.tools.register(replacement.tool)
+    })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => initial.started.length === 2)
+    initial.release('1')
+    await until(() => events(agent).some(event =>
+      event.type === 'tool/result' && event.data.callId === CallId('c1')))
+    await new Promise(r => setTimeout(r, 5))
+    expect(replacement.started).toEqual([])
+    initial.release('2')
+    await until(() => replacement.started.length === 1)
+    expect(replacement.started).toEqual(['3'])
+    replacement.release('3')
+    await waitForIdle(ctx, agent)
+  })
+})
+
+describe('tool-call scheduler: model-order results despite out-of-order settlement', () => {
+  it('commits tool/result in model order even when a later call settles first', async () => {
+    const adapter = new MockAdapter([
+      multiCall([{ id: 'c1', name: 'p', args: { id: '1' } }, { id: 'c2', name: 'p', args: { id: '2' } }]),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 2)
+    gated.release('2')
+    await new Promise(r => setTimeout(r, 5))
+    const beforeFirst = events(agent).filter(e => e.type === 'tool/result')
+    expect(beforeFirst).toEqual([])
+    gated.release('1')
+    await waitForIdle(ctx, agent)
+
+    const results = events(agent).filter(e => e.type === 'tool/result')
+    expect(results.map(e => e.data.callId)).toEqual([CallId('c1'), CallId('c2')])
+  })
+
+  it('derived history pairs calls in model order regardless of tool/call log interleaving', async () => {
+    const adapter = new MockAdapter([
+      multiCall([{ id: 'c1', name: 'p', args: { id: '1' } }, { id: 'c2', name: 'p', args: { id: '2' } }]),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 2)
+    gated.release('2'); gated.release('1')
+    await waitForIdle(ctx, agent)
+
+    const messages = agent.session.deriveMessages()
+    const toolResults = messages.flatMap(m => m.content.filter(b => b.type === 'tool-result'))
+    expect(toolResults.map(b => b.toolCallId)).toEqual([CallId('c1'), CallId('c2')])
+  })
+})
+
+describe('tool-call scheduler: rolling pool honors maxParallelToolCalls', () => {
+  it('rejects invalid global maxParallelToolCalls config at plugin load', async () => {
+    await expect(harness(new MockAdapter([]), 0)).rejects.toThrow()
+    await expect(harness(new MockAdapter([]), 1.5)).rejects.toThrow()
+  })
+
+  it('defensively rejects invalid caps when direct construction bypasses the config schema', () => {
+    expect(() => new AgentLoop(new Context(), { agents: [], maxParallelToolCalls: 0 }))
+      .toThrow('maxParallelToolCalls must be a positive integer')
+    expect(() => new AgentLoop(new Context(), { agents: [], maxParallelToolCalls: 1.5 }))
+      .toThrow('maxParallelToolCalls must be a positive integer')
+  })
+
+  it('defaults the cap when direct construction bypasses the config schema', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    await ctx.plugin(SessionStore)
+    await ctx.plugin(SystemPrompt, { persona: '' })
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(AgentRegistry)
+
+    expect(() => new AgentLoop(ctx, { agents: [] })).not.toThrow()
+    await ctx.fiber.dispose()
+  })
+
+  it('starts at most the cap, replenishing as calls settle', async () => {
+    const adapter = new MockAdapter([
+      multiCall([1, 2, 3, 4].map(n => ({ id: `c${n}`, name: 'p', args: { id: String(n) } }))),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter, 2)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 2)
+    await new Promise(r => setTimeout(r, 5))
+    expect(gated.started).toEqual(['1', '2'])
+    gated.release('1')
+    await until(() => gated.started.length === 3)
+    expect(gated.started).toEqual(['1', '2', '3'])
+    expect(events(agent)
+      .filter(e => e.type === 'tool/call' || e.type === 'tool/result')
+      .map(e => `${e.type}:${String(e.data.callId)}`)
+      .slice(0, 4))
+      .toEqual(['tool/call:c1', 'tool/call:c2', 'tool/result:c1', 'tool/call:c3'])
+    gated.release('2'); gated.release('3')
+    await until(() => gated.started.length === 4)
+    gated.release('4')
+    await waitForIdle(ctx, agent)
+    expect(events(agent).filter(e => e.type === 'tool/result').map(e => e.data.callId))
+      .toEqual([CallId('c1'), CallId('c2'), CallId('c3'), CallId('c4')])
+  })
+
+  it('maxParallelToolCalls: 1 is fully serial (no second start before the first settles)', async () => {
+    const adapter = new MockAdapter([
+      multiCall([{ id: 'c1', name: 'p', args: { id: '1' } }, { id: 'c2', name: 'p', args: { id: '2' } }]),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter, 1)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 1)
+    await new Promise(r => setTimeout(r, 5))
+    expect(gated.started).toEqual(['1'])
+    gated.release('1')
+    await until(() => gated.started.length === 2)
+    gated.release('2')
+    await waitForIdle(ctx, agent)
+  })
+
+  it('applies the configured cap to every factory-created agent', async () => {
+    const adapter = new MockAdapter([
+      multiCall([{ id: 'c1', name: 'p', args: { id: '1' } }, { id: 'c2', name: 'p', args: { id: '2' } }]),
+      textResponse('done'),
+    ])
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    await ctx.plugin(SessionStore)
+    await ctx.plugin(SystemPrompt, { persona: '' })
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(AgentRegistry)
+    await ctx.plugin(AgentLoop, { agents: [], maxParallelToolCalls: 1 })
+    ctx.llm.registerAdapter(['mock'], adapter)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 1)
+    await new Promise(r => setTimeout(r, 5))
+    expect(gated.started).toEqual(['1'])
+    gated.release('1')
+    await until(() => gated.started.length === 2)
+    gated.release('2')
+    await waitForIdle(ctx, agent)
+  })
+
+})
+
+describe('tool-call scheduler: ordered middleware and additional contexts', () => {
+  it('tools/pre-execute and tools/post-execute observe model call order', async () => {
+    const adapter = new MockAdapter([
+      multiCall([{ id: 'c1', name: 'p', args: { id: '1' } }, { id: 'c2', name: 'p', args: { id: '2' } }, { id: 'c3', name: 'p', args: { id: '3' } }]),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const pre: string[] = []
+    const post: string[] = []
+    ctx.on('tools/pre-execute', async (exec, next): Promise<PreToolDecision> => { pre.push(String(exec.callId)); return next() })
+    ctx.on('tools/post-execute', async (exec, _result, next): Promise<PostToolDecision> => { post.push(String(exec.callId)); return next() })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 3)
+    gated.release('3'); gated.release('2'); gated.release('1')
+    await waitForIdle(ctx, agent)
+
+    expect(pre).toEqual([CallId('c1'), CallId('c2'), CallId('c3')].map(String))
+    expect(post).toEqual([CallId('c1'), CallId('c2'), CallId('c3')].map(String))
+  })
+
+  it('injects additional contexts in model call order, not settlement order', async () => {
+    const adapter = new MockAdapter([
+      multiCall([{ id: 'c1', name: 'p', args: { id: '1' } }, { id: 'c2', name: 'p', args: { id: '2' } }]),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter, 2)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    ctx.on('tools/post-execute', async (exec, _result): Promise<PostToolDecision> =>
+      ({ kind: 'accept', additionalContexts: [{ content: [{ type: 'text', text: `ctx-${exec.callId}` }], source: { kind: 'plugin', plugin: 'p' } }] }))
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 2)
+    gated.release('2'); gated.release('1')
+    await waitForIdle(ctx, agent)
+
+    const log = events(agent)
+    const contextTexts = log.filter(e => e.type === 'context/message')
+      .map(e => (e.data.content[0] as { text: string }).text)
+    expect(contextTexts).toEqual(['ctx-c1', 'ctx-c2'])
+    const lastResult = log.findLastIndex(e => e.type === 'tool/result')
+    const firstContext = log.findIndex(e => e.type === 'context/message')
+    expect(lastResult).toBeLessThan(firstContext)
+  })
+
+  it('orders pre-execute denials and errors without dispatching them', async () => {
+    const adapter = new MockAdapter([
+      multiCall([
+        { id: 'c1', name: 'p', args: { id: '1' } },
+        { id: 'c2', name: 'p', args: { id: '2' } },
+        { id: 'c3', name: 'p', args: { id: '3' } },
+      ]),
+      textResponse('done'),
+    ])
+    const ctx = await harness(adapter)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const post: string[] = []
+    ctx.on('tools/pre-execute', async (exec, next): Promise<PreToolDecision> => {
+      if (exec.callId === CallId('c2')) return { kind: 'deny', reason: 'blocked by policy' }
+      if (exec.callId === CallId('c3')) throw new Error('pre exploded')
+      return next()
+    })
+    ctx.on('tools/post-execute', async (exec, _result, next): Promise<PostToolDecision> => {
+      post.push(String(exec.callId))
+      return next()
+    })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 1)
+    gated.release('1')
+    await waitForIdle(ctx, agent)
+
+    expect(gated.started).toEqual(['1'])
+    expect(post).toEqual(['c1', 'c2'])
+    const results = events(agent).filter(e => e.type === 'tool/result')
+    expect(results.map(e => e.data.callId)).toEqual([CallId('c1'), CallId('c2'), CallId('c3')])
+    expect((results[1]!.data.content[0] as { text: string }).text).toContain('blocked by policy')
+    expect((results[2]!.data.content[0] as { text: string }).text).toContain('pre exploded')
+  })
+})
+
+describe('tool-call scheduler: abort handling', () => {
+  it('starts no calls when the signal is already aborted before a parallel group', async () => {
+    const adapter = new MockAdapter([
+      multiCall([{ id: 'c1', name: 'p', args: { id: '1' } }, { id: 'c2', name: 'p', args: { id: '2' } }]),
+      textResponse('should never be requested'),
+    ])
+    const ctx = await harness(adapter)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+    ctx.on('session/event', (session, event) => {
+      if (session === agent.session && event.type === 'assistant/message') {
+        ;(agent as unknown as { currentAbort?: AbortController }).currentAbort?.abort('already aborted')
+      }
+    })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await waitForIdle(ctx, agent)
+
+    expect(gated.started).toEqual([])
+    expect(events(agent).filter(e => e.type === 'tool/call')).toEqual([])
+    expect(events(agent).filter(e => e.type === 'tool/result')).toEqual([])
+  })
+
+  it('stops starting siblings when abort fires during ordered pre-execute', async () => {
+    const adapter = new MockAdapter([
+      multiCall([{ id: 'c1', name: 'p', args: { id: '1' } }, { id: 'c2', name: 'p', args: { id: '2' } }]),
+      textResponse('should never be requested'),
+    ])
+    const ctx = await harness(adapter)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+    ctx.on('tools/pre-execute', async (exec, next): Promise<PreToolDecision> => {
+      if (exec.callId === CallId('c1')) {
+        ;(agent as unknown as { currentAbort?: AbortController }).currentAbort?.abort('pre cancelled')
+      }
+      return next()
+    })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 1)
+    await new Promise(r => setTimeout(r, 5))
+    expect(gated.started).toEqual(['1'])
+    gated.release('1')
+    await waitForIdle(ctx, agent)
+
+    expect(events(agent).filter(e => e.type === 'tool/call').map(e => e.data.callId))
+      .toEqual([CallId('c1')])
+    expect(events(agent).filter(e => e.type === 'tool/result').map(e => e.data.callId))
+      .toEqual([CallId('c1')])
+  })
+
+  it('stops replenishing after abort, commits started results, and drains accepted additional contexts', async () => {
+    const adapter = new MockAdapter([
+      multiCall([1, 2, 3, 4].map(n => ({ id: `c${n}`, name: 'p', args: { id: String(n) } }))),
+      textResponse('should never be requested'),
+    ])
+    const ctx = await harness(adapter, 2)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    ctx.on('tools/post-execute', async (exec, _result, next): Promise<PostToolDecision> => ({
+      ...await next(),
+      additionalContexts: [{ content: [{ type: 'text', text: `ctx-${exec.callId}` }], source: { kind: 'plugin', plugin: 'p' } }],
+    }))
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 2)
+    ;(agent as unknown as { currentAbort?: AbortController }).currentAbort?.abort('stop now')
+    gated.release('1')
+    gated.release('2')
+    await waitForIdle(ctx, agent)
+
+    expect(gated.started).toEqual(['1', '2'])
+    expect(events(agent).filter(e => e.type === 'tool/call').map(e => e.data.callId))
+      .toEqual([CallId('c1'), CallId('c2')])
+    expect(events(agent).filter(e => e.type === 'tool/result').map(e => e.data.callId))
+      .toEqual([CallId('c1'), CallId('c2')])
+    const settled = events(agent).filter(e => e.type === 'tool/result' || e.type === 'context/message')
+    expect(settled.map(e => e.type))
+      .toEqual(['tool/result', 'tool/result', 'context/message', 'context/message'])
+    expect(settled.filter(e => e.type === 'context/message')
+      .map(e => (e.data.content[0] as { text: string }).text))
+      .toEqual(['ctx-c1', 'ctx-c2'])
+  })
+
+  it('does not run an exclusive barrier after a parallel group aborts', async () => {
+    const adapter = new MockAdapter([
+      multiCall([
+        { id: 'c1', name: 'p', args: { id: '1' } },
+        { id: 'c2', name: 'p', args: { id: '2' } },
+        { id: 'c3', name: 'x', args: { id: '3' } },
+      ]),
+      textResponse('should never be requested'),
+    ])
+    const ctx = await harness(adapter, 2)
+    const gated = gatedParallelTool('p')
+    const exclusive: string[] = []
+    ctx.tools.register(gated.tool)
+    ctx.tools.register(defineTool({
+      name: 'x',
+      description: 'exclusive',
+      parameters: { id: { type: 'string', required: true } },
+      async execute(args) { exclusive.push(args.id); return [{ type: 'text', text: 'x' }] },
+    }))
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 2)
+    ;(agent as unknown as { currentAbort?: AbortController }).currentAbort?.abort('stop before barrier')
+    gated.release('1')
+    gated.release('2')
+    await waitForIdle(ctx, agent)
+
+    expect(exclusive).toEqual([])
+    expect(events(agent).filter(e => e.type === 'tool/call').map(e => e.data.callId))
+      .toEqual([CallId('c1'), CallId('c2')])
+  })
+})
diff --git a/packages/core/agent-loop/tests/tool-order.spec.ts b/packages/core/agent-loop/tests/tool-order.spec.ts
index 13e731b7b8..940b7fdabe 100644
--- a/packages/core/agent-loop/tests/tool-order.spec.ts
+++ b/packages/core/agent-loop/tests/tool-order.spec.ts
@@ -56,7 +56,7 @@ async function runTurn(registrationOrder: string[], toolOrder?: SystemPromptConf
   const adapter = new MockAdapter([textResponse('done')])
   const ctx = await harness(adapter, toolOrder)
   for (const name of registrationOrder) registerNamed(ctx, name)
-  const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+  const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
   agent.send([{ type: 'text', text: 'go' }])
   await waitForIdle(ctx, agent)
   return { ctx, agent, adapter }
@@ -98,7 +98,7 @@ describe('loop-level canonical tool order', () => {
     registerNamed(ctx, 'alpha')
     const errors: Error[] = []
     ctx.on('agent/error', (_agent, _turn, _step, error) => void errors.push(error))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
     expect(adapter.requests).toHaveLength(0)
diff --git a/packages/core/agent-loop/tests/turn-stop.spec.ts b/packages/core/agent-loop/tests/turn-stop.spec.ts
index c275fee88c..5e979988d7 100644
--- a/packages/core/agent-loop/tests/turn-stop.spec.ts
+++ b/packages/core/agent-loop/tests/turn-stop.spec.ts
@@ -45,7 +45,7 @@ describe('agent/turn-stop', () => {
       textResponse('must not be requested'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('terminal-steering'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('terminal-steering'), { provider: 'mock', model: 'mock' })
     agent.ctx.on('agent/turn-stop', (): ContinuationStop => ({ action: 'stop' }))
 
     let steered = false
@@ -72,7 +72,7 @@ describe('agent/turn-stop', () => {
       textResponse('must not become a late-steering turn'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('terminal-flush-steering'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('terminal-flush-steering'), { provider: 'mock', model: 'mock' })
     agent.ctx.on('agent/turn-stop', (): ContinuationStop => ({ action: 'stop' }))
 
     let injected = false
@@ -98,7 +98,7 @@ describe('agent/turn-stop', () => {
       textResponse('queued follow-up answer'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('terminal-flush-send'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('terminal-flush-send'), { provider: 'mock', model: 'mock' })
     agent.ctx.on('agent/turn-stop', (): ContinuationStop => ({ action: 'stop' }))
 
     let queued = false
@@ -124,8 +124,8 @@ describe('agent/turn-stop', () => {
     ])
     const ctx = await harness(adapter)
     registerEcho(ctx)
-    const stopped = ctx.agentLoop.create(AgentId('stopped'), { model: 'mock' })
-    const ordinary = ctx.agentLoop.create(AgentId('ordinary'), { model: 'mock' })
+    const stopped = ctx.agentLoop.create(AgentId('stopped'), { provider: 'mock', model: 'mock' })
+    const ordinary = ctx.agentLoop.create(AgentId('ordinary'), { provider: 'mock', model: 'mock' })
     stopped.ctx.on('agent/turn-stop', (): ContinuationStop => ({ action: 'stop' }))
 
     await send(stopped)
@@ -145,7 +145,7 @@ describe('agent/turn-stop', () => {
     ])
     const ctx = await harness(adapter)
     registerEcho(ctx)
-    const agent = ctx.agentLoop.create(AgentId('owned-listener'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('owned-listener'), { provider: 'mock', model: 'mock' })
     const disposeStop = agent.ctx.on('agent/turn-stop', (): ContinuationStop => ({ action: 'stop' }))
 
     await send(agent, 'first turn')
@@ -162,7 +162,7 @@ describe('agent/turn-stop', () => {
       textResponse('healthy later turn'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('bad-policy'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('bad-policy'), { provider: 'mock', model: 'mock' })
     const reasons: TurnEndReason[] = []
     const errors: string[] = []
     ctx.on('session/event', (session, event) => {
diff --git a/packages/core/agent/README.md b/packages/core/agent/README.md
index 5639f30daf..1255731484 100644
--- a/packages/core/agent/README.md
+++ b/packages/core/agent/README.md
@@ -33,6 +33,8 @@ The loop plugin registers `AgentFactory`, keeping consumers independent of its c
 
 Most interception points are cooperative waterfalls returning seam-specific decisions. `agent/pre-step` is a serial surface-mutation checkpoint, while `agent/turn-stop` is the terminal serial fold: it runs after ordinary continuation and steering folding, and a returned stop remains in force through turn close and flush so later steering cannot create an extra step or turn. Ordinary queued prompts remain intact. The full rationale is in the [agent-scope runtime-design RFC](../../../docs/rfc/implemented/architecture/2026-07-12-agent-scope-runtime-design.md#three-execution-boundaries-are-deliberately-one-way).
 
+`PromptDecision.additionalContexts` is an array so every injected context keeps its own source, envelope, and metadata. A `ContinuationDecision` reason is narrower: it becomes a `steering/message`, not a `context/message`, and therefore carries only content and source.
+
 Turn and step boundaries and the model token stream are durable `session/event` facts rather than mirrored `agent/*` notifications. Consumers read `turn/*`, `step/*`, and `assistant/chunk` from the session feed; tool policy and outcome observation belong to the complete pipeline documented by [`dsh-tools`](../tools/README.md).
 
 ### Agent interface (`types.ts`)
@@ -41,7 +43,7 @@ The handle every plugin programs against:
 
 - `agent.send(content, options?)` — queue a message; starts a turn when idle. Content and resolved source become one detached, deeply frozen lossless-JSON record before `agent/queued` and enqueue; invalid data throws synchronously, and caller or notification-listener in-place mutation cannot change the log or model input (`agent/prompt-submit` still rewrites by returning replacement content).
 - `agent.steer(content, options?)` — steer a running turn (inject between steps); uses the same owned acceptance boundary and behaves like `send` when idle
-- `agent.inject(content, options?)` — inject in-session context (context/message event); the next request sees it. Does not run the model. While a turn is open it joins that turn; while idle it is wrapped in a one-shot `injection` turn so every event stays turn-enclosed ([the turn-enclosure invariant](../../../docs/rfc/implemented/architecture/2026-06-15-turn-enclosure-invariant.md))
+- `agent.inject(content, options?)` — accept detached in-session context without running the model; the next request sees its `context/message`. `options.envelope` defaults to the canonical `<context>` framing and may be `'raw'` when the caller owns a complete familiar frame; `options.meta` persists opaque JSON state without rendering it. While a turn is open it joins that turn, deferring FIFO while the current tool batch executes and draining before turn close if execution is interrupted; while idle it is wrapped in a one-shot `injection` turn and durability checkpoint ([the turn-enclosure invariant](../../../docs/rfc/implemented/architecture/2026-06-15-turn-enclosure-invariant.md)).
 - `agent.cancel(reason?)` — cancel ALL pending work: clears the queued + steering FIFOs, aborts the in-flight step, and drops a turn about to start (the pre-step window) so a queued-but-not-started prompt never runs. A UI/ACP `session/cancel` maps to this. The single public stop primitive. Idle with nothing pending → a safe no-op.
 - `agent.whenIdle()` — resolve once the agent reaches quiescence after settling out of `running` (idle → immediately; disposed → awaits the loop exit). A non-owner's quiescence-observation hook: it observes the work settling WITHOUT tearing the agent down. Teardown is separate — a lifecycle owner stops and unregisters via `AgentHandle.dispose()`, which awaits the loop exit directly.
 - `agent.session`, `agent.status`, `agent.options`, `agent.id`
diff --git a/packages/core/agent/src/types.ts b/packages/core/agent/src/types.ts
index 3aad65a70e..5c709c828b 100644
--- a/packages/core/agent/src/types.ts
+++ b/packages/core/agent/src/types.ts
@@ -22,7 +22,7 @@ export type AgentId = Branded<'AgentId'>
 export function AgentId(id: string): AgentId {
   return id as AgentId
 }
-import type { Session } from '@deepseek-ai/dsh-session'
+import type { ContextEnvelope, JsonValue, Session } from '@deepseek-ai/dsh-session'
 
 declare module '@deepseek-ai/dsh-system-prompt' {
   interface AssembleContext {
@@ -33,7 +33,9 @@ declare module '@deepseek-ai/dsh-system-prompt' {
 
 /** Merge-extensible agent creation options. Persona belongs to system-prompt sections. */
 export interface AgentOptions {
-  /** Model name (must have a registered adapter at call time). */
+  /** Provider route (must have a registered adapter at call time). */
+  provider?: string
+  /** Model id interpreted by the selected provider adapter. */
   model?: string
 }
 
@@ -42,6 +44,14 @@ export interface SendOptions {
   source?: MessageSource
 }
 
+/** Options specific to durable synthetic context injection. */
+export interface InjectOptions extends SendOptions {
+  /** Keep the canonical context tag, or send caller-owned framing verbatim. */
+  envelope?: ContextEnvelope
+  /** Opaque JSON state retained in the session event but hidden from the model. */
+  meta?: JsonValue
+}
+
 /**
  * An agent's lifecycle state, emitted on every transition as `agent/status`:
  * `idle` (parked, waiting for queued work), `running` (a turn is in progress),
@@ -54,21 +64,25 @@ export type AgentStatus = 'idle' | 'running' | 'disposed'
 export interface HookContext {
   content: ContentBlock[]
   source: MessageSource
+  /** Keep the canonical context tag, or use caller-owned framing verbatim. */
+  envelope?: ContextEnvelope
+  /** Opaque JSON state retained in the session event but hidden from the model. */
+  meta?: JsonValue
 }
 
 /**
- * Prompt interception result. `allow.content` replaces the prompt and
- * `additionalContext` becomes a separate context message. `block` records a
+ * Prompt interception result. `allow.content` replaces the prompt and each
+ * `additionalContexts` entry becomes a separate context message. `block` records a
  * durable `prompt/blocked`; an all-blocked batch ends a zero-step rejected turn.
  */
 export type PromptDecision =
-  | { kind: 'allow'; content?: ContentBlock[]; additionalContext?: HookContext }
+  | { kind: 'allow'; content?: ContentBlock[]; additionalContexts?: HookContext[] }
   | { kind: 'block'; reason: string }
 
 /** Turn continuation override; a continue reason is recorded as next-step steering in the same turn. */
 export type ContinuationDecision =
   | { action: 'stop' }
-  | { action: 'continue'; reason?: HookContext }
+  | { action: 'continue'; reason?: { content: ContentBlock[]; source: MessageSource } }
 
 /**
  * The terminal subset of {@link ContinuationDecision}. A listener on
@@ -103,12 +117,13 @@ export interface Agent {
   steer(content: ContentBlock[], options?: SendOptions): void
 
   /**
-   * Append model-facing context without running the model. Idle injection uses
-   * a one-shot turn and durability checkpoint, while injection during an open
-   * turn joins it at the current log position. Disposal awaits idle checkpoints;
-   * flush failures are reported through `agent/error`, not thrown to the caller.
+   * Append detached model-facing context without running the model. An open-turn
+   * injection joins at the current log position unless the current tool batch is
+   * executing; then it waits FIFO until that batch settles and drains before turn
+   * close even when interrupted. Idle injection uses a one-shot turn and durability
+   * checkpoint. Disposal awaits idle checkpoints; flush failures report through `agent/error`.
    */
-  inject(content: ContentBlock[], options?: SendOptions): void
+  inject(content: ContentBlock[], options?: InjectOptions): void
 
   /**
    * Clear queued and steering work, including work waiting to start, and abort
diff --git a/packages/core/agent/tests/verify-export-jsdoc.spec.ts b/packages/core/agent/tests/verify-export-jsdoc.spec.ts
index 2761351d10..cc361a956e 100644
--- a/packages/core/agent/tests/verify-export-jsdoc.spec.ts
+++ b/packages/core/agent/tests/verify-export-jsdoc.spec.ts
@@ -30,6 +30,22 @@ function fixture(files: Record<string, string>): string {
 const make = (content: string): string => fixture({ 'index.ts': content })
 
 describe('verify-export-jsdoc functions and consts', () => {
+  it('limits packages without src/* exports to declarations reachable from package entrypoints', () => {
+    const root = fixture({
+      'index.ts': "export { publicFn } from './internal.ts'\n",
+      'internal.ts': `
+export function publicFn(value: string): string { return value }
+export function hiddenFn(value: string): string { return value }
+`,
+    })
+    writeFileSync(join(root, 'packages/group/fix/package.json'), JSON.stringify({
+      exports: { '.': { types: './lib/types/index.d.ts', default: './lib/index.js' } },
+    }))
+    const violations = collectExportJsdocViolations(root)
+    expect(violations).toHaveLength(1)
+    expect(violations.every(violation => violation.includes('publicFn'))).toBe(true)
+  })
+
   it('accepts a fully documented surface', () => {
     expect(collectExportJsdocViolations(make(`
 /**
diff --git a/packages/core/session/README.md b/packages/core/session/README.md
index 7e201ab647..46b6ecb703 100644
--- a/packages/core/session/README.md
+++ b/packages/core/session/README.md
@@ -1,6 +1,6 @@
 # dsh-session
 
-Event-sourced session log and in-memory store. A `Session` is the append-only source of truth for an agent's whole interaction history — the LLM message history is *derived* from it. A **surface** layer (a linked list of message-producing events) is maintained on top of the raw log for efficient derivation and compaction.
+Event-sourced session log and in-memory store. A `Session` is the append-only source of truth for an agent's whole interaction history — the LLM message history is *derived* from it. A **surface** layer (an ordered projection of message-producing events) is maintained on top of the raw log for efficient derivation and compaction.
 
 ## Service: `SessionStore` (ctx key: `sessions`)
 
@@ -32,8 +32,8 @@ The store pairs announced creation with disposal, publishes post-commit append n
 
 Plain class (not a Cordis Service). Create via `ctx.sessions.create()`.
 
-- `session.append(type, data, opts?)` snapshots and freezes durable data and surface metadata, commits synchronously, then notifies observers with independent failure containment. Reentrant attached-session appends reject, and runtime checks cover widened unions and loaded logs.
-- `session.deriveMessages()` incrementally projects each new surface node once and returns a fresh array over shared frozen messages. A surface rewrite rebuilds the projection; there is no raw-log fallback.
+- `session.append(type, data, opts?)` snapshots and freezes durable data and surface metadata, validates marker shape, provenance, and complete replacement coverage, commits synchronously, then notifies observers with independent failure containment. Reentrant attached-session appends reject, and runtime checks cover widened unions and loaded logs.
+- `session.deriveMessages()` incrementally projects each new surface entry once and returns a fresh array over shared frozen messages. Assistant projections preserve provider/model provenance and adapter-private replay state. A surface rewrite rebuilds the projection; there is no raw-log fallback.
 - `session.deriveEventMessage(event)` is the canonical per-event projection used by reconstruction and invariants.
 - `session.surface` lazily folds only new `surfaceOp` markers; `replaceGeneration` changes on every rewrite.
 - `session.events` is a cached frozen snapshot invalidated by append; accepted events remain deeply frozen.
@@ -46,19 +46,20 @@ Durable values need one accepted representation, not a check followed by a secon
 
 ### Surface types
 
-- `SurfaceOp` — how a surface node entered the linked list: `'append'` (normal tail append) or `{ op: 'replace', start, end }` (replace nodes from `start` through `end` inclusive — both must be valid surface node seqs; `start === end` replaces a single node). Used by compaction to shadow old nodes without deleting them.
+- `SurfaceOp` — how an event entered the ordered surface: `'append'` (normal tail append) or `{ op: 'replace', start, end }` (replace entries from `start` through `end` inclusive — both must be valid surface seqs; `start === end` replaces one entry). Used by compaction to shadow old events without deleting them.
 - `SurfaceIntent` — `{ surfaceOp: SurfaceOp; sourceEventSeqs?: number[] }`, the required third parameter to `session.append()` for surface-eligible types.
-- `SurfaceNode` — `{ seq: number; prev: number | null; next: number | null }`, one node in the surface linked list.
-- `foldSurface(events)` — replay the canonical surface transitions into detached current nodes and actual replacement ranges, rejecting surface-eligible events that lack their mandatory marker. `SurfaceManager` shares the same transitions while retaining its incremental cache.
-- `isSurfaceEvent(event)` / `isSurfaceEligibleType(type)` — the first narrows a `SessionEvent` to a fully-formed surface node (type is surface-eligible AND `surfaceOp` present); the second is the type-only check (is this one of the five `SurfaceEventType` values?), used to detect a surface-eligible event MISSING its marker — e.g. when validating a seed/load log.
+- `foldSurface(events)` — replay the canonical surface contract into detached current event sequences and actual replacement ranges. The same pass rejects non-contiguous seqs, misplaced or malformed metadata, empty or duplicate provenance, non-earlier sources, invalid positional ranges, and replacements that fail to cite every shadowed surface entry; `SurfaceManager` shares the atomic transition while retaining only its incremental sequence cache.
+- `isSurfaceEvent(event)` / `isSurfaceEligibleType(type)` — the first narrows a `SessionEvent` to a fully formed surface event; the second detects a surface-eligible event missing its marker when validating a seed or loaded log.
 
 ### Request-header reconstruction (`request-header.ts`)
 
-`request/header` and `request/header-delta` make the non-history request envelope reconstructable from the log. `foldRequestHeader()` reconstructs the active header, `diffHeader()` encodes changes, and `applyHeaderDelta()` replays them; unsupported deltas fall back to a full snapshot. `messagePrefix` remains separate from derived history. See the [reconstructable-requests RFC](../../../docs/rfc/implemented/architecture/2026-07-05-reconstructable-requests.md).
+`request/header` records a full canonical snapshot of the non-history request envelope with reason `initial`, `resume`, or `change`. `foldRequestHeader()` selects the latest snapshot; legacy delta events and the removed `fallback` reason are rejected. `messagePrefix` remains separate from derived history. See the [reconstructable-requests RFC](../../../docs/rfc/implemented/architecture/2026-07-05-reconstructable-requests.md).
+
+`context/message` defaults to the canonical tagged context projection. A producer may set `envelope: 'raw'` when its `content` already contains the complete model-facing frame, and may attach JSON `meta` for replayable plugin state; metadata remains durable but is excluded from `deriveMessages()`.
 
 ### Session event vocabulary (`types.ts`)
 
-The append-only log's event types, enumerated member by member — payloads, surface badges, provenance — in the generated [persistence log event catalog](../../../docs/persistence-catalog.md). Token usage rides on `assistant/message.usage`; an operational error's step is on `turn/end.reason` for `kind: 'error'`.
+The append-only log's event types, enumerated member by member — payloads, surface badges, provenance — in the generated [persistence log event catalog](../../../docs/persistence-catalog.md). Token usage and provider/model/replay provenance ride on `assistant/message`; an operational error's step is on `turn/end.reason` for `kind: 'error'`.
 
 Merge-extensible via `SessionEventMap` — a plugin declaration-merges its own types (the compaction seam's `compact/*`, the hook bridges' `hook/*`); merged members appear in the same catalog.
 
@@ -66,7 +67,7 @@ Also defines `TurnTriggerMap` and `TurnEndReasonMap` (merge-extensible sum types
 
 Every `SessionEvent` carries two optional top-level fields (structural metadata):
 
-- `sourceEventSeqs?: number[]` — seq numbers of provenance sources (e.g., the `assistant/chunk` seqs behind an `assistant/message`, or the shadowed nodes behind a compaction replace node).
+- `sourceEventSeqs?: number[]` — seq numbers of provenance sources (e.g., the `assistant/chunk` seqs behind an `assistant/message`, or the shadowed entries behind a compaction replacement entry). On `assistant/message`, a present `[]` records a known empty provider stream, while omission means legacy or otherwise unrecorded provenance; other surface events require a non-empty list when this field is present.
 - `surfaceOp?: SurfaceOp` — how this event entered the surface. Absent for non-surface events (boundaries, chunks, usage, errors).
 
 ### Metadata types (`types.ts`)
@@ -76,16 +77,16 @@ Every `SessionEvent` carries two optional top-level fields (structural metadata)
 ### Extension points
 
 - Persistence plugins: subscribe to `session/event` (write-behind) and drain on `session/flush` (awaited) and fiber dispose. A durable backend reads the log and reloads it into a live session; the metadata seam (`SessionHeader`, `session.header`) is what such a backend stores beside the log.
-- Replay/fork: `create(id, { seed })` validates and freezes a contiguous log and rebuilds its surface. `fork(source, boundary?, childSessionId?)` selects a completed-turn prefix and records lineage.
-- Compaction: the `dsh-compact-basic` plugin appends a `user/message` with `surfaceOp: { op: 'replace', start, end }` to shadow old surface nodes behind a summary checkpoint.
+- Replay/fork: `create(id, { seed })` validates and freezes a contiguous current-format log and rebuilds its surface; request headers require provider/model and assistant messages require provider/model provenance. `fork(source, boundary?, childSessionId?)` selects a completed-turn prefix and records lineage.
+- Compaction: the `dsh-compact-basic` plugin appends a `user/message` with `surfaceOp: { op: 'replace', start, end }` to shadow old surface entries behind a summary checkpoint. Tool-pairing boundary policy and its cache belong to the [`dsh-compact` seam](../../compact/compact/README.md), while this package owns ordered surface membership and `replaceGeneration`.
 
 ## Model Experience
 
 ### Derived message history
 
-**What the model sees**: The model receives projections of `user/message`, `assistant/message`, and `tool/result` surface nodes verbatim. A `context/message` is a user-role message containing exactly `<context source="<source-kind>">`, its content blocks, and `</context>`; `steering/message` uses the identical `<steering source="<source-kind>">` / `</steering>` wrapper. Tool calls live inside assistant messages. Chunks, boundaries, usage, hook records, todo records, and other log-only events add no message.
+**What the model sees**: The model receives projections of `user/message`, `assistant/message`, and `tool/result` surface entries verbatim. A `context/message` is a user-role message containing exactly `<context source="<source-kind>">`, its content blocks, and `</context>`; `steering/message` uses the identical `<steering source="<source-kind>">` / `</steering>` wrapper. Tool calls live inside assistant messages. Chunks, boundaries, usage, hook records, todo records, and other log-only events add no message.
 
-**Token effect**: Appended surface nodes are resent on later steps. A `replace` surface operation removes the shadowed nodes from future inputs without deleting their raw log records.
+**Token effect**: Appended surface entries are resent on later steps. A `replace` surface operation removes the shadowed entries from future inputs without deleting their raw log records.
 
 ### Crash-repair result
 
diff --git a/packages/core/session/src/index.ts b/packages/core/session/src/index.ts
index 89f1627445..985d475101 100644
--- a/packages/core/session/src/index.ts
+++ b/packages/core/session/src/index.ts
@@ -13,19 +13,18 @@ import { scopeOf, scopeTarget } from '@deepseek-ai/dsh-scope'
 import type { Scoped } from '@deepseek-ai/dsh-scope'
 import type { ContentBlock, Message, MessageSource } from '@deepseek-ai/dsh-llm'
 import { SESSION_FORMAT_VERSION, SessionId } from './types.ts'
-import type { CreateSessionOptions, EpochHeader, SessionEvent, SessionEventMap, SessionEventType, SessionHeader, SurfaceIntent, SurfaceEventType } from './types.ts'
+import type { ContextEnvelope, CreateSessionOptions, EpochHeader, SessionEvent, SessionEventMap, SessionEventType, SessionHeader, SurfaceIntent, SurfaceEventType } from './types.ts'
 import { snapshotJsonValue } from './json.ts'
-import { SurfaceManager, isSurfaceEligibleType } from './surface.ts'
+import { SurfaceManager } from './surface.ts'
 import { foldRequestHeader } from './request-header.ts'
 
 export * from './types.ts'
 export { isJsonValue, snapshotJsonValue } from './json.ts'
 export type { JsonValue } from './json.ts'
 export { interruptedTurnClosers } from './repair.ts'
-export type { SurfaceFoldReplacement, SurfaceFoldResult, SurfaceNode } from './surface.ts'
+export type { SurfaceFoldReplacement, SurfaceFoldResult } from './surface.ts'
 export { foldSurface, isSurfaceEvent, isSurfaceEligibleType } from './surface.ts'
-export { isToolPairingBalanced } from './tool-pairing.ts'
-export { applyHeaderDelta, canonicalHeader, diffHeader, foldRequestHeader, headerEquals } from './request-header.ts'
+export { canonicalHeader, foldRequestHeader, headerEquals } from './request-header.ts'
 
 declare module 'cordis' {
   interface Context {
@@ -131,46 +130,12 @@ function snapshotSessionHeader(id: SessionId, source?: SessionHeader): SessionHe
   return deepFreeze(record as unknown as SessionHeader)
 }
 
-/** Validate the runtime shape of surface metadata after its JSON snapshot. */
-function assertSurfaceMetadataShape(
-  type: string,
-  surfaceOp: unknown,
-  sourceEventSeqs: unknown,
-): void {
-  const eligible = isSurfaceEligibleType(type)
-  if (!eligible) {
-    if (surfaceOp !== undefined || sourceEventSeqs !== undefined) {
-      throw new Error(`session event "${type}" is not surface-eligible and cannot carry surface metadata`)
-    }
-    return
-  }
-  if (surfaceOp === undefined) {
-    throw new Error(`session event "${type}" is surface-eligible and requires a surfaceOp marker`)
-  }
-  if (surfaceOp !== 'append') {
-    if (surfaceOp === null || typeof surfaceOp !== 'object' || Array.isArray(surfaceOp)) {
-      throw new Error(`session event "${type}" carries an invalid surfaceOp`)
-    }
-    const op = surfaceOp as Record<string, unknown>
-    const keys = Object.keys(op)
-    if (keys.length !== 3 || !Object.hasOwn(op, 'op') || !Object.hasOwn(op, 'start') || !Object.hasOwn(op, 'end')
-      || op['op'] !== 'replace'
-      || typeof op['start'] !== 'number' || !Number.isSafeInteger(op['start']) || op['start'] < 0
-      || typeof op['end'] !== 'number' || !Number.isSafeInteger(op['end']) || op['end'] < 0) {
-      throw new Error(`session event "${type}" carries an invalid replace surfaceOp`)
-    }
-  }
-  if (sourceEventSeqs !== undefined) {
-    if (!Array.isArray(sourceEventSeqs)
-      || sourceEventSeqs.some(seq => typeof seq !== 'number' || !Number.isSafeInteger(seq) || seq < 0)) {
-      throw new Error(`session event "${type}" sourceEventSeqs must contain non-negative safe integers`)
-    }
-  }
-}
-
 /** Validate the fixed event envelope after one-pass JSON materialization. */
 function assertSessionEventEnvelope(value: Record<string, unknown>, index: number): asserts value is SessionEvent {
   const event = value
+  if (event['type'] === 'request/header-delta') {
+    throw new Error(`seed event at index ${index} uses unsupported legacy request/header-delta format`)
+  }
   const allowed = new Set(['type', 'seq', 'time', 'data', 'surfaceOp', 'sourceEventSeqs'])
   if (Object.keys(event).some(key => !allowed.has(key))
     || !Object.hasOwn(event, 'type') || typeof event['type'] !== 'string'
@@ -181,6 +146,42 @@ function assertSessionEventEnvelope(value: Record<string, unknown>, index: numbe
     || !Object.hasOwn(event, 'data')) {
     throw new Error(`seed event at index ${index} has an invalid event envelope`)
   }
+  assertCurrentLlmShape(event, index)
+}
+
+/** Reject pre-provider request headers and assistant messages at the seed/load boundary. */
+function assertCurrentLlmShape(event: Record<string, unknown>, index: number): void {
+  const data = event['data']
+  if (typeof data !== 'object' || data === null) return
+  const record = data as Record<string, unknown>
+  if (event['type'] === 'request/header') {
+    const header = record['header']
+    const config = typeof header === 'object' && header !== null ? (header as Record<string, unknown>)['config'] : undefined
+    if (!hasProviderModel(config)) throw new Error(`seed request/header at index ${index} lacks provider/model`)
+  }
+  if (event['type'] === 'assistant/message' && !hasProviderModel(record['provenance'])) {
+    throw new Error(`seed assistant/message at index ${index} lacks provider/model provenance`)
+  }
+}
+
+/** Whether an unknown value carries the current provider/model pair. */
+function hasProviderModel(value: unknown): boolean {
+  if (typeof value !== 'object' || value === null) return false
+  const pair = value as Record<string, unknown>
+  return typeof pair['provider'] === 'string' && pair['provider'].length > 0
+    && typeof pair['model'] === 'string' && pair['model'].length > 0
+}
+
+/** Reject request-header vocabulary removed with the legacy delta codec. */
+function assertSupportedRequestHeader(type: string, data: unknown, location: string): void {
+  if (type === 'request/header-delta') {
+    throw new Error(`${location} uses unsupported legacy request/header-delta format`)
+  }
+  if (type === 'request/header'
+    && data !== null && typeof data === 'object' && !Array.isArray(data)
+    && (data as Record<string, unknown>)['reason'] === 'fallback') {
+    throw new Error(`${location} uses unsupported legacy request/header reason "fallback"`)
+  }
 }
 
 type SessionCallback = (...args: unknown[]) => unknown
@@ -226,6 +227,22 @@ interface SessionEntry {
 /** Store attachment for the append path; module-private to keep Session store-agnostic publicly. */
 const attachments = new WeakMap<Session, SessionEntry>()
 
+/**
+ * Render one context contribution exactly as it will appear in model history.
+ * @param content - content blocks supplied by the context producer.
+ * @param source - attribution used by the canonical context envelope.
+ * @param envelope - canonical tagged framing or caller-owned raw framing.
+ * @returns a detached block list ready for the derived model transcript.
+ */
+export function renderContextContent(
+  content: ContentBlock[],
+  source: MessageSource,
+  envelope: ContextEnvelope = 'context',
+): ContentBlock[] {
+  const cloned = structuredClone(content)
+  return envelope === 'raw' ? cloned : renderTagged('context', cloned, source)
+}
+
 /**
  * An event-sourced session: an append-only log of {@link SessionEvent}s.
  *
@@ -234,17 +251,19 @@ const attachments = new WeakMap<Session, SessionEntry>()
  */
 export class Session {
   private log: SessionEvent[] = []
+  /** Incremental acceptance state, kept separate from the public lazy view. */
+  private readonly surfaceValidator = new SurfaceManager(this.log)
 
   /**
-   * Derived surface — a cached linked list of message-producing events.
+   * Derived surface — a cached order of message-producing event sequences.
    * Lazily rebuilt from `surfaceOp` markers in the log; processes only new
    * events (delta) on each access — the log is append-only, so prior events
    * never change.
-   * `append`. Undefined until first accessed (including after fork/seed).
+   * Undefined until first accessed (including after fork/seed).
    */
   private _surface: SurfaceManager | undefined
 
-  /** The surface linked list over this session's event log. */
+  /** The ordered surface over this session's event log. */
   get surface(): SurfaceManager {
     if (!this._surface) this._surface = new SurfaceManager(this.log)
     return this._surface
@@ -269,7 +288,7 @@ export class Session {
       // `seq = log.length` contract the whole system relies on). Without this,
       // a bad seed would surface only later as a backend rejection or a silent
       // divergence between the live log and disk.
-      this.log = Array.from(seed, (source, index) => {
+      for (const [index, source] of seed.entries()) {
         // The seed is a persistence/replay boundary: validate and detach the
         // complete event in one lossless-JSON pass.
         const snapshot = snapshotJsonValue(source)
@@ -277,23 +296,20 @@ export class Session {
           throw new Error(`seed event at index ${index} is not losslessly JSON-serializable`)
         }
         assertSessionEventEnvelope(snapshot, index)
+        assertSupportedRequestHeader(snapshot.type, snapshot.data, `seed event at index ${index}`)
         if (snapshot.seq !== index) {
           throw new Error(`seed event at index ${index} has seq ${snapshot.seq} (expected ${index}); seed must be contiguous from 0`)
         }
-        // Surface-eligible events MUST carry a surfaceOp marker — the surface is
-        // the sole source of derived history, so a marker-less message event
-        // would load fine yet vanish from deriveMessages(). `append` enforces
-        // this at compile time via its typed overload; a seed arrives as raw
-        // SessionEvent[] (replay/fork/load), bypassing that, so re-check at
-        // runtime here rather than silently resuming with empty history.
-        const structural = snapshot as SessionEvent & { surfaceOp?: unknown; sourceEventSeqs?: unknown }
+        // A seed is accepted incrementally through the same transition as a
+        // live append and a full-log fold. The candidate is planned before it
+        // enters `log`, so a failure cannot partially mutate the surface.
         try {
-          assertSurfaceMetadataShape(snapshot.type, structural.surfaceOp, structural.sourceEventSeqs)
+          this.surfaceValidator.validateNext(snapshot)
         } catch (error: unknown) {
           throw new Error(`invalid seed event at index ${index}: ${error instanceof Error ? error.message : 'invalid surface metadata'}`)
         }
-        return deepFreeze(snapshot)
-      })
+        this.log.push(deepFreeze(snapshot))
+      }
     }
     this.header = snapshotSessionHeader(id, header)
   }
@@ -328,7 +344,7 @@ export class Session {
    * @param type - The event type (key of {@link SessionEventMap}).
    * @param data - The event payload; must be JSON-serializable.
    * @param opts - Surface metadata: `surfaceOp` controls how the event enters
-   *   the surface linked list; `sourceEventSeqs` records provenance (the seq
+   *   the ordered surface; `sourceEventSeqs` records provenance (the seq
    *   numbers of events this one derives from). REQUIRED for
    *   {@link SurfaceEventType} events (every message-producing event must
    *   declare how it joins the surface, the sole source of derived history) and
@@ -340,7 +356,10 @@ export class Session {
    * @throws if `data` or surface metadata is not losslessly JSON-serializable
    *   (BigInt, function, symbol, undefined, negative zero, non-finite number,
    *   circular reference, sparse array, or an exotic object such as
-   *   Map/Set/Date/class instance). One recursive pass reads, validates, and
+   *   Map/Set/Date/class instance), or when the candidate violates the
+   *   canonical surface contract (marker shape and eligibility, unique
+   *   earlier provenance, positional replacement validity, and complete
+   *   shadowed-node coverage). One recursive pass reads, validates, and
    *   copies each nested value once, so a stateful getter cannot supply one value
    *   to validation and another to storage. The event log is the durable source
    *   of truth, so a bad event fails at the append site rather than later during
@@ -362,29 +381,26 @@ export class Session {
     if (dataSnapshot === undefined) {
       throw new Error(`session event "${type}" carries non-JSON-serializable data`)
     }
+    assertSupportedRequestHeader(type, dataSnapshot, `session event "${type}"`)
     const surfaceMetadataSnapshot = snapshotJsonValue(surfaceMetadata)
     if (surfaceMetadataSnapshot === undefined) {
       throw new Error(`session event "${type}" carries non-JSON-serializable surface metadata`)
     }
-    assertSurfaceMetadataShape(
-      type,
-      (surfaceMetadataSnapshot as { surfaceOp?: unknown }).surfaceOp,
-      (surfaceMetadataSnapshot as { sourceEventSeqs?: unknown }).sourceEventSeqs,
-    )
-
     const entry = attachments.get(this)
     if (entry?.appending) {
       throw new Error('session append cannot reenter while another append is being published')
     }
+    const event = deepFreeze({
+      type,
+      seq: this.log.length,
+      time: Date.now(),
+      data: dataSnapshot,
+      ...(surfaceMetadataSnapshot as { surfaceOp?: unknown; sourceEventSeqs?: unknown }),
+    } as unknown as SessionEvent<T>)
+    this.surfaceValidator.validateNext(event as SessionEvent)
+
     if (entry !== undefined) entry.appending = true
     try {
-      const event = deepFreeze({
-        type,
-        seq: this.log.length,
-        time: Date.now(),
-        data: dataSnapshot,
-        ...surfaceMetadataSnapshot,
-      } as unknown as SessionEvent<T>)
       let callbacks: SessionCallback[] | undefined
       const callbackArgs: unknown[] = [this, event]
       if (entry !== undefined) {
@@ -437,8 +453,8 @@ export class Session {
   private derivedGeneration = 0
 
   /**
-   * Derive the LLM message history by walking the session surface — the linked
-   * list of message-producing events maintained by `surfaceOp` markers. The
+   * Derive the LLM message history by walking the ordered sequences of
+   * message-producing events maintained by `surfaceOp` markers. The
    * surface is the single source of derived history: every message-producing
    * append records its `surfaceOp`, so a raw event with no marker (a chunk, a
    * turn boundary) is correctly absent, and a compaction `replace` deletes the
@@ -462,11 +478,11 @@ export class Session {
       this.derivedNodes = 0
       this.derivedGeneration = generation
     }
-    for (const node of nodes.slice(this.derivedNodes)) {
-      // Surface nodes are built from this.log — node.seq is always a valid
+    for (const seq of nodes.slice(this.derivedNodes)) {
+      // Surface sequences are built from this.log — seq is always a valid
       // index by construction. The non-null assertion expresses that invariant.
       // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-      const msg = this.deriveEventMessage(this.log[node.seq]!)
+      const msg = this.deriveEventMessage(this.log[seq]!)
       // A surface node is one of the five message-producing types, but an
       // empty-content assistant/message (a max-tokens step that hosts only
       // usage) derives to null and must not enter the transcript.
@@ -504,7 +520,7 @@ export class Session {
         // max-tokens step's usage and must not inject a content-less assistant
         // turn into the provider transcript.
         if (event.data.content.length === 0) return null
-        return { role: 'assistant', content: event.data.content }
+        return { role: 'assistant', content: event.data.content, provenance: event.data.provenance }
       }
       case 'tool/result': {
         const { callId, content, isError } = event.data
@@ -514,8 +530,8 @@ export class Session {
         }
       }
       case 'context/message': {
-        const { content, source } = event.data
-        return { role: 'user', content: renderTagged('context', content, source) }
+        const { content, source, envelope } = event.data
+        return { role: 'user', content: renderContextContent(content, source, envelope) }
       }
       case 'steering/message': {
         const { content, source } = event.data
diff --git a/packages/core/session/src/request-header.ts b/packages/core/session/src/request-header.ts
index dd7b2a758d..8dd61ee884 100644
--- a/packages/core/session/src/request-header.ts
+++ b/packages/core/session/src/request-header.ts
@@ -1,28 +1,20 @@
 /**
- * Request-header reconstruction utilities over `request/header` snapshots and
- * `request/header-delta` events. Writers round-trip each proposed delta and use
- * a full snapshot when the encoding cannot represent the change.
+ * Request-header reconstruction utilities over full `request/header` session
+ * events. Anyone holding a session log reconstructs the {@link EpochHeader}
+ * any request was built under by taking the latest canonical snapshot; the
+ * loop uses the same equality helper to avoid logging unchanged headers.
+ *
  * @module dsh-session/request-header
  */
 
 import { callConfigEquals } from '@deepseek-ai/dsh-llm'
-import type { LlmCallConfig, Message, ToolSchema } from '@deepseek-ai/dsh-llm'
-import type { EpochHeader, SessionEvent, SystemDelta, ToolsDelta } from './types.ts'
-
-/** The `request/header-delta` payload shape: each present field amends the folded header. */
-type HeaderDelta = {
-  system?: SystemDelta
-  tools?: ToolsDelta
-  config?: LlmCallConfig
-  messagePrefix?: Message[]
-}
+import type { Message, ToolSchema } from '@deepseek-ai/dsh-llm'
+import type { EpochHeader, SessionEvent } from './types.ts'
 
 /**
- * Normalize a header to canonical form: an empty system prompt, an empty
- * tool list, and an empty session prefix become ABSENT fields, matching how
- * requests are built (the request-build spreads skip empty values). Diff,
- * fold, and comparison all operate on canonical headers, so "no system
- * prompt" (and "no session prefix") has exactly one representation.
+ * Normalize a header to canonical form: an empty system prompt, an empty tool
+ * list, and an empty session prefix become absent fields, matching how requests
+ * are built. Logging, folding, and comparison use this one representation.
  * @param header - the header to normalize (not mutated).
  * @returns the canonical header.
  */
@@ -35,85 +27,22 @@ export function canonicalHeader(header: EpochHeader): EpochHeader {
   }
 }
 
-/** Split a canonical (possibly absent) system prompt into lines; absence is zero lines. */
-function systemLines(system: string | undefined): string[] {
-  return system === undefined ? [] : system.split('\n')
-}
-
-/** Join lines back into a canonical system value; zero lines is absence. */
-function joinSystem(lines: string[]): string | undefined {
-  return lines.length === 0 ? undefined : lines.join('\n')
-}
-
-/**
- * Compute the line-level {@link SystemDelta} between two canonical system
- * prompts: trim the common prefix and (non-overlapping) common suffix, and
- * carry the replacement lines between them. Deterministic and library-free;
- * with nothing shared it degenerates to a full replacement.
- */
-function diffSystem(prev: string | undefined, next: string | undefined): SystemDelta {
-  const a = systemLines(prev)
-  const b = systemLines(next)
-  let keepStart = 0
-  while (keepStart < a.length && keepStart < b.length && a[keepStart] === b[keepStart]) keepStart += 1
-  let keepEnd = 0
-  while (
-    keepEnd < a.length - keepStart &&
-    keepEnd < b.length - keepStart &&
-    a[a.length - 1 - keepEnd] === b[b.length - 1 - keepEnd]
-  ) keepEnd += 1
-  return { keepStart, keepEnd, insert: b.slice(keepStart, b.length - keepEnd) }
-}
-
-/** Apply a {@link SystemDelta} to a canonical system prompt. */
-function applySystem(prev: string | undefined, delta: SystemDelta): string | undefined {
-  const a = systemLines(prev)
-  return joinSystem([...a.slice(0, delta.keepStart), ...delta.insert, ...a.slice(a.length - delta.keepEnd)])
-}
-
-/** Canonical JSON equality for tool schemas — sound because schemas are
- * JSON-serializable by construction and both sides come from the same
- * assembly path, so key insertion order matches when the values do. */
+/** Canonical JSON equality for tool schemas assembled through the same path. */
 function sameSchema(a: ToolSchema, b: ToolSchema): boolean {
   return JSON.stringify(a) === JSON.stringify(b)
 }
 
-/**
- * Compute the name-keyed {@link ToolsDelta} between two canonical tool lists.
- * A pure reordering produces an empty delta — the writer's round-trip guard
- * catches that case and records a snapshot instead.
- */
-function diffTools(prev: readonly ToolSchema[], next: readonly ToolSchema[]): ToolsDelta {
-  const prevByName = new Map(prev.map(tool => [tool.name, tool]))
-  const nextNames = new Set(next.map(tool => tool.name))
-  return {
-    added: next.filter(tool => !prevByName.has(tool.name)),
-    removed: prev.filter(tool => !nextNames.has(tool.name)).map(tool => tool.name),
-    changed: next.filter((tool) => {
-      const before = prevByName.get(tool.name)
-      return before !== undefined && !sameSchema(before, tool)
-    }),
-  }
-}
-
-/** Apply a {@link ToolsDelta} to a canonical tool list: drop removed, replace changed in place, append added. */
-function applyTools(prev: readonly ToolSchema[], delta: ToolsDelta): ToolSchema[] {
-  const removed = new Set(delta.removed)
-  const changedByName = new Map(delta.changed.map(tool => [tool.name, tool]))
-  const kept = prev
-    .filter(tool => !removed.has(tool.name))
-    .map(tool => changedByName.get(tool.name) ?? tool)
-  return [...kept, ...delta.added]
+/** Canonical JSON equality over session-prefix arrays; absence equals empty. */
+function sameMessages(a: readonly Message[] | undefined, b: readonly Message[] | undefined): boolean {
+  return JSON.stringify(a ?? []) === JSON.stringify(b ?? [])
 }
 
 /**
- * Field-wise equality over canonical headers — the cheap comparison the writer's round-trip
- * guard runs (`applyHeaderDelta(prev, delta)` must equal the intended header) and the loop
- * runs to skip logging an unchanged header.
- *
+ * Field-wise equality over canonical headers. Tool schemas compare in order;
+ * the session prefix compares as canonical JSON.
  * @param a - one canonical header.
  * @param b - the other.
- * @returns whether config, system, tools (in order), and the session prefix all match.
+ * @returns whether config, system, tools, and session prefix all match.
  */
 export function headerEquals(a: EpochHeader, b: EpochHeader): boolean {
   if (!callConfigEquals(a.config, b.config) || a.system !== b.system) return false
@@ -123,74 +52,19 @@ export function headerEquals(a: EpochHeader, b: EpochHeader): boolean {
   return at.length === bt.length && at.every((tool, i) => sameSchema(tool, bt[i] as ToolSchema))
 }
 
-/** Canonical JSON equality over session-prefix arrays; absence equals the empty array. */
-function sameMessages(a: readonly Message[] | undefined, b: readonly Message[] | undefined): boolean {
-  return JSON.stringify(a ?? []) === JSON.stringify(b ?? [])
-}
-
 /**
- * Compute the `request/header-delta` payload between two canonical headers, or
- * `undefined` when they are equal. The encoding cannot represent every change,
- * including pure tool reordering, so callers must apply and compare the result
- * before logging it and fall back to a full snapshot on mismatch. The session
- * prefix is replaced whole; an empty array removes it.
- *
- * @param prev - the folded header the log currently implies.
- * @param next - the header the next request will actually use.
- * @returns the delta payload, or undefined when nothing changed.
- */
-export function diffHeader(prev: EpochHeader, next: EpochHeader): HeaderDelta | undefined {
-  const delta: HeaderDelta = {}
-  if (prev.system !== next.system) delta.system = diffSystem(prev.system, next.system)
-  const prevTools = prev.tools ?? []
-  const nextTools = next.tools ?? []
-  if (JSON.stringify(prevTools) !== JSON.stringify(nextTools)) delta.tools = diffTools(prevTools, nextTools)
-  if (!callConfigEquals(prev.config, next.config)) delta.config = next.config
-  if (!sameMessages(prev.messagePrefix, next.messagePrefix)) delta.messagePrefix = next.messagePrefix ?? []
-  return Object.keys(delta).length > 0 ? delta : undefined
-}
-
-/**
- * Apply a `request/header-delta` payload to a canonical header, producing the
- * canonical header it encodes. Total for well-formed logs (the writer only
- * appends round-trip-verified deltas).
- * @param prev - the folded header before the delta.
- * @param delta - the logged delta payload.
- * @returns the canonical header after the delta.
- */
-export function applyHeaderDelta(prev: EpochHeader, delta: HeaderDelta): EpochHeader {
-  const system = delta.system !== undefined ? applySystem(prev.system, delta.system) : prev.system
-  const tools = delta.tools !== undefined ? applyTools(prev.tools ?? [], delta.tools) : prev.tools
-  const messagePrefix = delta.messagePrefix ?? prev.messagePrefix
-  return canonicalHeader({
-    config: delta.config ?? prev.config,
-    ...system !== undefined ? { system } : {},
-    ...tools !== undefined ? { tools } : {},
-    ...messagePrefix !== undefined ? { messagePrefix } : {},
-  })
-}
-
-/**
- * Fold the header events of a log (or any prefix of one) into the {@link EpochHeader} in
- * force after the last of them: each `request/header` snapshot replaces the state, each
- * `request/header-delta` amends it.
- *
- * @param events - session events in log order (non-header events are skipped).
- * @param from - a previously folded state to continue from (the live session's incremental
- *   cursor); omit to fold from nothing.
- * @returns the folded header, or undefined when no header event exists yet.
+ * Fold the header events of a log (or any prefix) into the
+ * {@link EpochHeader} in force after the last snapshot. Non-header events are
+ * skipped. This is the pure offline reconstruction path; the live session
+ * tracks the same fold incrementally.
+ * @param events - session events in log order.
+ * @param from - a previously folded state to continue from.
+ * @returns the latest canonical header, or undefined when none exists yet.
  */
 export function foldRequestHeader(events: readonly SessionEvent[], from?: EpochHeader): EpochHeader | undefined {
-  let state: EpochHeader | undefined = from
+  let state = from
   for (const event of events) {
-    if (event.type === 'request/header') {
-      state = canonicalHeader(event.data.header)
-    } else if (event.type === 'request/header-delta') {
-      if (state === undefined) {
-        throw new Error(`request/header-delta at seq ${event.seq} before any request/header snapshot: corrupt log`)
-      }
-      state = applyHeaderDelta(state, event.data)
-    }
+    if (event.type === 'request/header') state = canonicalHeader(event.data.header)
   }
   return state
 }
diff --git a/packages/core/session/src/surface.ts b/packages/core/session/src/surface.ts
index 37eebbf7c3..d4f43bc303 100644
--- a/packages/core/session/src/surface.ts
+++ b/packages/core/session/src/surface.ts
@@ -1,19 +1,13 @@
 /**
- * Surface layer on top of the session event log: a derived, cached linked list
- * of events that produce LLM messages. Rebuilt deterministically from
- * `surfaceOp` markers in the log — the log is the source of truth; the surface
- * is a view.
+ * Surface layer on top of the session event log: an ordered view of events
+ * that produce LLM messages. The append-only log remains the source of truth.
  *
  * @module @deepseek-ai/dsh-session/surface
  */
 
 import type { SessionEvent, SurfaceEvent, SurfaceEventType, SurfaceOp } from './types.ts'
 
-/**
- * The set of event type strings that are eligible for the surface linked list.
- * Mirrors the {@link SurfaceEventType} union; kept as a runtime set so the
- * type guard can check membership without a chain of string comparisons.
- */
+/** Runtime counterpart of the message-producing event union. */
 const SURFACE_EVENT_TYPES = new Set<string>([
   'user/message',
   'assistant/message',
@@ -23,39 +17,22 @@ const SURFACE_EVENT_TYPES = new Set<string>([
 ])
 
 /**
- * Check only whether a type may enter the message surface; it does not require `surfaceOp`. This
- * detects eligible seed/load events missing their mandatory marker. Use {@link isSurfaceEvent} to
- * narrow a fully formed event whose marker is present.
- * @param type - the event type string to test.
- * @returns true when the type is one of the five message-producing types.
+ * Whether an event type can join the model-visible surface.
+ * @param type - event type to test.
+ * @returns true for one of the five message-producing event types.
  */
 export function isSurfaceEligibleType(type: string): boolean {
   return SURFACE_EVENT_TYPES.has(type)
 }
 
 /**
- * Narrow a {@link SessionEvent} to {@link SurfaceEvent}: checks that the
- * event's `type` is surface-eligible AND that `surfaceOp` is present.
- * The narrowed type has mandatory {@link SurfaceOp}.
- * @param event - the event to narrow.
- * @returns true when the event is surface-eligible and carries its `surfaceOp` marker.
+ * Narrow an event to a surface-eligible event carrying its required marker.
+ * @param event - event to test.
+ * @returns true when both the type and marker identify a surface event.
  */
 export function isSurfaceEvent(event: SessionEvent): event is SurfaceEvent {
   if (!SURFACE_EVENT_TYPES.has(event.type)) return false
-  // surfaceOp is optional on SessionEvent (even for surface-eligible types)
-  // but mandatory on SurfaceEvent — this check is the narrowing gate.
-  if ((event as SessionEvent<SurfaceEventType>).surfaceOp === undefined) return false
-  return true
-}
-
-/** One node in the surface linked list. */
-export interface SurfaceNode {
-  /** The event seq of this surface node. */
-  seq: number
-  /** The previous surface node's seq, or null if this is the head. */
-  prev: number | null
-  /** The next surface node's seq, or null if this is the tail. */
-  next: number | null
+  return (event as SessionEvent<SurfaceEventType>).surfaceOp !== undefined
 }
 
 /** One replacement operation observed while folding a session surface. */
@@ -66,31 +43,165 @@ export interface SurfaceFoldReplacement {
   start: number
   /** Declared inclusive end seq of the replaced surface range. */
   end: number
-  /** Actual surface nodes removed by the operation, in surface order. */
+  /** Actual surface entries removed by the operation, in surface order. */
   shadowedSeqs: number[]
 }
 
 /** Complete result of replaying the surface operations in a session log. */
 export interface SurfaceFoldResult {
-  /** Current surface nodes in linked-list order. */
-  nodes: SurfaceNode[]
+  /** Current surface event sequences in model-visible order. */
+  nodes: number[]
   /** Replacement operations in event order. */
   replacements: SurfaceFoldReplacement[]
 }
 
-/** Mutable state shared by the incremental manager and the full-log fold. */
+/** Mutable state shared by complete and incremental folds. */
 interface SurfaceFoldState {
-  nodes: SurfaceNode[]
-  nodeBySeq: Map<number, SurfaceNode>
+  nodes: number[]
   replaceGeneration: number
 }
 
+/** A validated replacement transition that has not mutated fold state yet. */
+interface SurfaceReplacePlan extends SurfaceFoldReplacement {
+  kind: 'replace'
+  startIdx: number
+  endIdx: number
+}
+
+/** One validated surface transition that has not mutated fold state yet. */
+type SurfacePlan =
+  | { kind: 'append'; seq: number }
+  | SurfaceReplacePlan
+
 /** Create an empty surface fold state. */
-function createFoldState(replaceGeneration = 0): SurfaceFoldState {
+function createFoldState(): SurfaceFoldState {
+  return { nodes: [], replaceGeneration: 0 }
+}
+
+/** Whether a runtime value is a non-negative safe event sequence. */
+function isEventSeq(value: unknown): value is number {
+  return typeof value === 'number' && Number.isSafeInteger(value) && value >= 0
+}
+
+/** Whether a runtime value is the exact positional-replacement shape. */
+function isReplaceOp(value: object): value is Extract<SurfaceOp, { op: 'replace' }> {
+  const op = value as Record<string, unknown>
+  return Object.keys(op).length === 3
+    && Object.hasOwn(op, 'op')
+    && Object.hasOwn(op, 'start')
+    && Object.hasOwn(op, 'end')
+    && op['op'] === 'replace'
+    && isEventSeq(op['start'])
+    && isEventSeq(op['end'])
+}
+
+/** Validate event-local surface eligibility and return its operation. */
+function surfaceOpOf(event: SessionEvent): SurfaceOp | undefined {
+  const raw = event as SessionEvent & { surfaceOp?: unknown; sourceEventSeqs?: unknown }
+  if (!isSurfaceEligibleType(event.type)) {
+    if (raw.surfaceOp !== undefined) {
+      throw new Error(`session event "${event.type}" is not surface-eligible and cannot carry surfaceOp`)
+    }
+    if (raw.sourceEventSeqs !== undefined) {
+      throw new Error(`session event "${event.type}" is not surface-eligible and cannot carry sourceEventSeqs`)
+    }
+    return
+  }
+  const op = raw.surfaceOp
+  if (op === undefined) {
+    throw new Error(`session event "${event.type}" is surface-eligible and requires a surfaceOp marker`)
+  }
+  if (op === 'append') return op
+  if (op === null || typeof op !== 'object' || Array.isArray(op)) {
+    throw new Error(`session event "${event.type}" carries an invalid surfaceOp`)
+  }
+  if (!isReplaceOp(op)) {
+    throw new Error(`session event "${event.type}" carries an invalid replace surfaceOp`)
+  }
+  return op
+}
+
+/** Validate provenance against prior log entries and the replacement range. */
+function assertProvenance(
+  event: SessionEvent,
+  shadowedSeqs: readonly number[],
+): void {
+  const raw = (event as SessionEvent & { sourceEventSeqs?: unknown }).sourceEventSeqs
+  const sources = new Set<number>()
+  if (raw !== undefined) {
+    if (!Array.isArray(raw)) {
+      throw new Error(`sourceEventSeqs on event at seq ${event.seq} must be an array when present`)
+    }
+    if (raw.length === 0 && event.type !== 'assistant/message') {
+      throw new Error('sourceEventSeqs must not be empty except on assistant/message')
+    }
+    let nonEarlierSource: number | undefined
+    for (const source of raw) {
+      if (!isEventSeq(source)) {
+        throw new Error(`session event "${event.type}" sourceEventSeqs must densely contain non-negative safe integers`)
+      }
+      sources.add(source)
+      if (nonEarlierSource === undefined && source >= event.seq) nonEarlierSource = source
+    }
+    if (sources.size !== raw.length) {
+      throw new Error('sourceEventSeqs must not contain duplicates')
+    }
+    if (nonEarlierSource !== undefined) {
+      throw new Error(`sourceEventSeqs must reference earlier events: ${nonEarlierSource} >= current seq ${event.seq}`)
+    }
+  }
+  const missing = shadowedSeqs.filter(seq => !sources.has(seq))
+  if (missing.length > 0) {
+    throw new Error(`surface replace: sourceEventSeqs must include every shadowed surface node; missing ${missing.join(', ')}`)
+  }
+}
+
+/** Locate one replacement range without mutating the current fold state. */
+function replacementRange(
+  state: SurfaceFoldState,
+  op: Extract<SurfaceOp, { op: 'replace' }>,
+): Pick<SurfaceReplacePlan, 'startIdx' | 'endIdx' | 'shadowedSeqs'> {
+  const startIdx = state.nodes.indexOf(op.start)
+  if (startIdx === -1) {
+    throw new Error(`surface replace: start seq ${op.start} not found in surface`)
+  }
+  const endIdx = state.nodes.indexOf(op.end)
+  if (endIdx === -1) {
+    throw new Error(`surface replace: end seq ${op.end} not found in surface`)
+  }
+  if (startIdx > endIdx) {
+    throw new Error(`surface replace: start seq ${op.start} (index ${startIdx}) is after end seq ${op.end} (index ${endIdx})`)
+  }
   return {
-    nodes: [],
-    nodeBySeq: new Map(),
-    replaceGeneration,
+    startIdx,
+    endIdx,
+    shadowedSeqs: state.nodes.slice(startIdx, endIdx + 1),
+  }
+}
+
+/** Validate one event at its replay boundary and prepare its atomic fold transition. */
+function planSurfaceEvent(
+  state: SurfaceFoldState,
+  event: SessionEvent,
+  expectedSeq: number,
+): SurfacePlan | undefined {
+  if (event.seq !== expectedSeq) {
+    throw new Error(`session event seq ${event.seq} is not contiguous; expected ${expectedSeq}`)
+  }
+  const surfaceOp = surfaceOpOf(event)
+  if (surfaceOp === undefined) return
+  if (surfaceOp === 'append') {
+    assertProvenance(event, [])
+    return { kind: 'append', seq: event.seq }
+  }
+  const range = replacementRange(state, surfaceOp)
+  assertProvenance(event, range.shadowedSeqs)
+  return {
+    kind: 'replace',
+    seq: event.seq,
+    start: surfaceOp.start,
+    end: surfaceOp.end,
+    ...range,
   }
 }
 
@@ -98,137 +209,76 @@ function createFoldState(replaceGeneration = 0): SurfaceFoldState {
 function applySurfaceEvent(
   state: SurfaceFoldState,
   event: SessionEvent,
+  expectedSeq: number,
 ): SurfaceFoldReplacement | undefined {
-  if (!isSurfaceEligibleType(event.type)) return
-  if (!isSurfaceEvent(event)) {
-    throw new Error(`surface event "${event.type}" (seq ${event.seq}) carries no surfaceOp marker`)
+  const plan = planSurfaceEvent(state, event, expectedSeq)
+  if (plan?.kind === 'append') {
+    state.nodes.push(plan.seq)
+  } else if (plan?.kind === 'replace') {
+    state.nodes.splice(plan.startIdx, plan.endIdx - plan.startIdx + 1, plan.seq)
+    state.replaceGeneration += 1
   }
-
-  if (event.surfaceOp === 'append') {
-    const tail = state.nodes.length > 0 ? state.nodes[state.nodes.length - 1] : undefined
-    const node: SurfaceNode = { seq: event.seq, prev: tail?.seq ?? null, next: null }
-    if (tail) tail.next = event.seq
-    state.nodes.push(node)
-    state.nodeBySeq.set(event.seq, node)
-    return
-  }
-
+  if (plan?.kind !== 'replace') return
   return {
-    seq: event.seq,
-    start: event.surfaceOp.start,
-    end: event.surfaceOp.end,
-    shadowedSeqs: replaceSurface(state, event.seq, event.surfaceOp),
+    seq: plan.seq,
+    start: plan.start,
+    end: plan.end,
+    shadowedSeqs: plan.shadowedSeqs,
   }
 }
 
-/** Apply one positional replacement and return the nodes it removed. */
-function replaceSurface(
-  state: SurfaceFoldState,
-  newSeq: number,
-  op: Extract<SurfaceOp, { op: 'replace' }>,
-): number[] {
-  const startNode = state.nodeBySeq.get(op.start)
-  if (!startNode) {
-    throw new Error(`surface replace: start seq ${op.start} not found in surface`)
-  }
-  const endNode = state.nodeBySeq.get(op.end)
-  if (!endNode) {
-    throw new Error(`surface replace: end seq ${op.end} not found in surface`)
-  }
-  const startIdx = state.nodes.indexOf(startNode)
-  const endIdx = state.nodes.indexOf(endNode)
-  if (startIdx > endIdx) {
-    throw new Error(`surface replace: start seq ${op.start} (index ${startIdx}) is after end seq ${op.end} (index ${endIdx})`)
-  }
-
-  const removed = state.nodes.splice(startIdx, endIdx - startIdx + 1)
-  for (const node of removed) state.nodeBySeq.delete(node.seq)
-
-  const prevNode = startIdx > 0 ? state.nodes[startIdx - 1] : undefined
-  const nextNode = startIdx < state.nodes.length ? state.nodes[startIdx] : undefined
-  const newNode: SurfaceNode = {
-    seq: newSeq,
-    prev: prevNode?.seq ?? null,
-    next: nextNode?.seq ?? null,
-  }
-  if (prevNode) prevNode.next = newSeq
-  if (nextNode) nextNode.prev = newSeq
-  state.nodes.splice(startIdx, 0, newNode)
-  state.nodeBySeq.set(newSeq, newNode)
-  state.replaceGeneration += 1
-  return removed.map(node => node.seq)
-}
-
 /**
  * Replay a complete session log through the canonical surface fold.
- *
- * The returned arrays and nodes are detached snapshots. The incremental
- * {@link SurfaceManager} uses the same transition functions, so query read
- * models cannot disagree with `deriveMessages()` about replacement ranges.
  * @param events - session events in contiguous seq order.
- * @returns the current surface and every positional replacement.
- * @throws when a surface-eligible event lacks its mandatory `surfaceOp`, or a
- * replacement names nodes that are absent or reversed on the current surface.
+ * @returns detached current sequences and replacement history.
+ * @throws when an event violates surface metadata, provenance, or range rules.
  */
 export function foldSurface(events: readonly SessionEvent[]): SurfaceFoldResult {
   const state = createFoldState()
   const replacements: SurfaceFoldReplacement[] = []
-  for (const event of events) {
-    const replacement = applySurfaceEvent(state, event)
+  for (const [index, event] of events.entries()) {
+    const replacement = applySurfaceEvent(state, event, index)
     if (replacement !== undefined) replacements.push(replacement)
   }
-  return {
-    nodes: state.nodes.map(node => ({ ...node })),
-    replacements,
-  }
+  return { nodes: [...state.nodes], replacements }
 }
 
-/**
- * Maintains a cached linked list of surface nodes, rebuilt lazily from
- * `surfaceOp` markers in the event log. Because the log is append-only, it
- * processes only the delta since the last rebuild — new events are folded
- * into the existing surface in O(new events) rather than rescanning the
- * whole log.
- */
+/** Incremental ordered surface view and append-boundary validator. */
 export class SurfaceManager {
-  /** Incremental state shared with the complete surface fold. */
+  /** Shared transition state; replacement history is not retained. */
   private _state = createFoldState()
-  /** The last processed seq. -1 folds the seeded log on first access. */
+  /** Last processed seq; -1 folds a seeded log on first access. */
   private _lastProcessedSeq = -1
 
   constructor(private log: readonly SessionEvent[]) {}
 
   /**
-   * The surface's rewrite generation, bumped by every folded `replace` op.
-   * A replace is the ONE operation that rewrites the
-   * surface non-monotonically, so an incremental consumer of {@link nodes}
-   * (the session's derived-message cache) compares this between visits — an
-   * unchanged generation guarantees every node it has not seen is a pure tail
-   * append; a changed one means its view must rebuild. Monotonic: it never
-   * moves backwards, so comparisons cannot be fooled by a re-fold.
+   * Validate the next candidate without mutating the committed surface.
+   * @param event - candidate event that has not entered the log yet.
    */
+  validateNext(event: SessionEvent): void {
+    if (this._lastProcessedSeq < this.log.length - 1) this._processDelta()
+    planSurfaceEvent(this._state, event, this.log.length)
+  }
+
+  /** Monotonic count of folded positional replacements. */
   get replaceGeneration(): number {
     if (this._lastProcessedSeq < this.log.length - 1) this._processDelta()
     return this._state.replaceGeneration
   }
 
-  /** The surface nodes in linked-list order (head to tail). */
-  get nodes(): readonly SurfaceNode[] {
+  /** Surface event sequences in model-visible order. */
+  get nodes(): readonly number[] {
     if (this._lastProcessedSeq < this.log.length - 1) this._processDelta()
     return this._state.nodes
   }
 
-  /**
-   * Process events from `_lastProcessedSeq + 1` through the end of the log,
-   * folding new surface markers into the existing linked list.
-   */
+  /** Fold events appended since the previous access. */
   private _processDelta(): void {
     for (let i = this._lastProcessedSeq + 1; i < this.log.length; i++) {
-      // Index is bounded by i < this.log.length — never undefined.
-      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-      const event = this.log[i]!
-      applySurfaceEvent(this._state, event)
+      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion -- bounded by the loop condition
+      applySurfaceEvent(this._state, this.log[i]!, i)
+      this._lastProcessedSeq = i
     }
-    this._lastProcessedSeq = this.log.length - 1
   }
 }
diff --git a/packages/core/session/src/tool-pairing.ts b/packages/core/session/src/tool-pairing.ts
deleted file mode 100644
index ce9dc639c7..0000000000
--- a/packages/core/session/src/tool-pairing.ts
+++ /dev/null
@@ -1,56 +0,0 @@
-/**
- * Tool-pairing balance over a session surface. Compaction changes surface
- * positions, so safe cuts are derived from tool-call/result content on the
- * surface rather than step markers in the append-only log.
- * @module @deepseek-ai/dsh-session/tool-pairing
- */
-
-import type { SessionEvent } from './types.ts'
-import type { SurfaceNode } from './surface.ts'
-
-/**
- * The tool-pairing delta of a surface node: how it shifts the count of
- * unanswered tool calls. An `assistant/message` opens one bracket per
- * `tool-call` block; a `tool/result` closes one; every other surface node
- * (`user/message`, `context/message`, `steering/message`, a usage-only
- * `assistant/message` with no tool-call blocks) is pairing-neutral.
- */
-function nodeDelta(event: SessionEvent): number {
-  switch (event.type) {
-    case 'assistant/message':
-      return event.data.content.filter(block => block.type === 'tool-call').length
-    case 'tool/result':
-      return -1
-    // Non-pairing surface nodes and every non-surface event contribute nothing.
-    default:
-      return 0
-  }
-}
-
-/**
- * Check that a surface cut does not split a tool call from its result. A region
- * is safe to collapse only when the cuts before its first node and after its
- * last node both return `true`.
- * @param nodes - the surface linked list in head→tail order.
- * @param events - the session log each node's `seq` indexes into.
- * @param beforeSeq - node immediately after the cut; `null` or a seq absent from the surface means after-tail.
- * @returns whether every call before the cut has its result before the cut.
- * @throws if a result appears without a preceding open call.
- */
-export function isToolPairingBalanced(
-  nodes: readonly SurfaceNode[],
-  events: readonly SessionEvent[],
-  beforeSeq: number | null,
-): boolean {
-  let depth = 0
-  for (const node of nodes) {
-    if (node.seq === beforeSeq) return depth === 0
-    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
-    depth += nodeDelta(events[node.seq]!)
-    if (depth < 0) {
-      throw new Error(`tool-pairing balance: tool/result at surface seq ${node.seq} has no matching tool-call (corrupt surface)`)
-    }
-  }
-  // A missing cut node means the after-tail boundary.
-  return depth === 0
-}
diff --git a/packages/core/session/src/types.ts b/packages/core/session/src/types.ts
index f4f42062fd..dc34760e9c 100644
--- a/packages/core/session/src/types.ts
+++ b/packages/core/session/src/types.ts
@@ -1,5 +1,9 @@
 import type { Branded } from '@deepseek-ai/dsh-brand'
-import type { CallId, ContentBlock, LlmCallConfig, Message, MessageSource, StreamChunk, TokenUsage, ToolSchema } from '@deepseek-ai/dsh-llm'
+import type { AssistantProvenance, CallId, ContentBlock, LlmCallConfig, Message, MessageSource, StreamChunk, TokenUsage, ToolSchema } from '@deepseek-ai/dsh-llm'
+import type { JsonValue } from './json.ts'
+
+/** Canonical context-tag framing, or caller-owned framing rendered verbatim. */
+export type ContextEnvelope = 'context' | 'raw'
 
 /** Identifies one session in the store (and its persistence artifacts). */
 export type SessionId = Branded<'SessionId'>
@@ -139,11 +143,11 @@ export interface TodoItem {
 
 /**
  * Logged request state outside derived history: call config, system prompt,
- * tools, and session prefix. Header snapshots and deltas reconstruct it;
+ * tools, and prefix. The latest full `request/header` snapshot reconstructs it;
  * canonical empty optional fields are absent.
  */
 export interface EpochHeader {
-  /** The conversation's call configuration (model + sampling scalars). */
+  /** The conversation's call configuration (provider, model, and sampling scalars). */
   config: LlmCallConfig
   /** Rendered system prompt text; absent for a system-less request. */
   system?: string
@@ -163,43 +167,9 @@ export interface EpochHeader {
  * Why a `request/header` snapshot was appended: `'initial'` — the log's first
  * header (a new conversation); `'resume'` — a loop instance's first request
  * over a log that already has header events (process restart, fork seed);
- * `'fallback'` — a mid-run change the delta encoding could not round-trip
- * (e.g. a pure tool reordering), recorded whole instead.
+ * `'change'` — a later request used a different header.
  */
-export type RequestHeaderReason = 'initial' | 'resume' | 'fallback'
-
-/**
- * Line-level edit of the system prompt: keep the first `keepStart` and last
- * `keepEnd` lines of the previous text, with `insert` replacing everything
- * between. Computed as a common-prefix/common-suffix trim — deterministic,
- * library-free, degenerating to a full replacement when nothing is shared.
- * Absence is encoded as zero lines (the canonical form has no empty-string
- * system), so a transition to or from "no system prompt" round-trips.
- */
-export interface SystemDelta {
-  /** Lines kept from the start of the previous system prompt. */
-  keepStart: number
-  /** Lines kept from the end of the previous system prompt. */
-  keepEnd: number
-  /** Lines replacing everything between the kept edges. */
-  insert: string[]
-}
-
-/**
- * Tool-set edit keyed by tool name (names are unique — the registry rejects
- * duplicates): `removed` names drop, `changed` schemas replace their
- * predecessor in place, `added` schemas append at the end. A change this
- * encoding cannot express (a pure reordering) fails the writer's round-trip
- * guard and is recorded as a `'fallback'` snapshot instead.
- */
-export interface ToolsDelta {
-  /** Schemas appended to the end of the tool list. */
-  added: ToolSchema[]
-  /** Names of schemas dropped from the tool list. */
-  removed: string[]
-  /** Schemas replacing the same-named predecessor in place. */
-  changed: ToolSchema[]
-}
+export type RequestHeaderReason = 'initial' | 'resume' | 'change'
 
 /**
  * The merge-extensible, append-only source of truth for an agent interaction.
@@ -235,9 +205,16 @@ export interface SessionEventMap {
   /**
    * In-session context injection (file-change notices, subdir AGENTS.md,
    * skill content, cron notifications, …). Rendered into the derived history
-   * as tagged synthetic context — NOT a user prompt.
+   * as synthetic context — NOT a user prompt. `envelope: 'raw'` lets a caller
+   * own the complete model-facing frame; `meta` is durable JSON state omitted
+   * from the model projection.
    */
-  'context/message': { content: ContentBlock[]; source: MessageSource }
+  'context/message': {
+    content: ContentBlock[]
+    source: MessageSource
+    envelope?: ContextEnvelope
+    meta?: JsonValue
+  }
   /** Raw stream chunk — token-level replay fidelity. */
   'assistant/chunk': { turn: number; step: number; chunk: StreamChunk }
   /**
@@ -246,7 +223,7 @@ export interface SessionEventMap {
    * the model output and its accounting travel together (there is no separate
    * usage record). `usage` is absent when the adapter reported none.
    */
-  'assistant/message': { turn: number; step: number; content: ContentBlock[]; usage?: TokenUsage }
+  'assistant/message': { turn: number; step: number; content: ContentBlock[]; provenance: AssistantProvenance; usage?: TokenUsage }
   /**
    * The model requested one tool invocation: `name` with the raw `arguments`
    * JSON string exactly as the model produced it (unparsed). `callId` pairs the
@@ -265,22 +242,13 @@ export interface SessionEventMap {
   'tool/result': { turn: number; step: number; callId: CallId; content: ContentBlock[]; isError: boolean; error?: { name: string; code: string }; meta?: unknown }
   /** Steering content injected between steps of a running turn. */
   'steering/message': { turn: number; content: ContentBlock[]; source: MessageSource }
-  /**
-   * Whole-list snapshot; the latest write wins on replay. It is log-only UI
-   * state and never enters derived model history.
-   */
+  /** Whole-list snapshot; latest write wins on replay. Log-only UI state; never derived history. */
   'todo/write': { todos: TodoItem[] }
   /**
-   * Full {@link EpochHeader} for the next request, appended inside its step
-   * before dispatch. It is log-only and anchors subsequent deltas.
+   * Full header for the next request, appended inside its step before dispatch.
+   * It is log-only; the latest snapshot reconstructs the request header.
    */
   'request/header': { header: EpochHeader; reason: RequestHeaderReason }
-  /**
-   * Log-only amendment to the folded {@link EpochHeader}. System and tools use
-   * their delta codecs; config and prefix replace whole, with an empty prefix
-   * encoding removal. Writers verify round-trip equality or log a fallback snapshot.
-   */
-  'request/header-delta': { system?: SystemDelta; tools?: ToolsDelta; config?: LlmCallConfig; messagePrefix?: Message[] }
 }
 
 /** The appendable event-type keys of {@link SessionEventMap}, plugin-merged extensions included. */
@@ -288,7 +256,7 @@ export type SessionEventType = keyof SessionEventMap
 
 /**
  * The subset of {@link SessionEventType} values whose events produce LLM
- * messages and are eligible to appear on the surface linked list. Only these
+ * messages and are eligible to appear on the ordered surface. Only these
  * event types may carry {@link SurfaceOp} and {@link SessionEvent.sourceEventSeqs}.
  */
 export type SurfaceEventType =
@@ -299,7 +267,7 @@ export type SurfaceEventType =
   | 'steering/message'
 
 /**
- * A {@link SessionEvent} that is **on** the surface linked list — its
+ * A {@link SessionEvent} that is **on** the ordered surface — its
  * `surfaceOp` is guaranteed present (mandatory), narrowed from a
  * surface-eligible {@link SessionEvent} by checking both `type` and
  * `surfaceOp` at runtime.
@@ -310,7 +278,7 @@ export type SurfaceEventType =
 export type SurfaceEvent = SessionEvent<SurfaceEventType> & { surfaceOp: SurfaceOp }
 
 /**
- * How a session event entered the surface linked list. Only valid on
+ * How a session event entered the ordered surface. Only valid on
  * {@link SurfaceEventType} events.
  *
  * - `'append'`: added to the tail — normal path for user/assistant/tool/context
@@ -331,6 +299,12 @@ export type SurfaceOp =
  */
 export interface SurfaceIntent {
   surfaceOp: SurfaceOp
+  /**
+   * Complete known provenance source set. `assistant/message` may use a
+   * present empty array for a known empty provider stream; omission means its
+   * provenance was not recorded. Other surface events require a non-empty set
+   * when this field is present.
+   */
   sourceEventSeqs?: number[]
 }
 
@@ -359,7 +333,9 @@ export type SessionEvent<T extends SessionEventType = SessionEventType> = {
     /**
      * Seq numbers of events that are provenance sources of this event
      * (e.g. the `assistant/chunk` seqs that built an `assistant/message`,
-     * or the surface nodes shadowed by a compaction replace node).
+     * or the surface nodes shadowed by a compaction replace node). An
+     * `assistant/message` may carry a present empty array for a known empty
+     * provider stream; omission means unrecorded provenance.
      */
     sourceEventSeqs?: number[]
     /** How this event entered the surface; absent for non-surface events. */
diff --git a/packages/core/session/tests/derived-cache.spec.ts b/packages/core/session/tests/derived-cache.spec.ts
index 5314c5e88f..96d1f7048c 100644
--- a/packages/core/session/tests/derived-cache.spec.ts
+++ b/packages/core/session/tests/derived-cache.spec.ts
@@ -23,9 +23,9 @@ describe('derived-message cache', () => {
     userText(session, 'one')
     expect(session.deriveMessages()).toEqual(scratch(session))
     userText(session, 'two')
-    session.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'reply' }] }, { surfaceOp: 'append' })
+    session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'reply' }] }, { surfaceOp: 'append' })
     expect(session.deriveMessages()).toEqual(scratch(session))
-    session.append('assistant/message', { turn: 1, step: 2, content: [], usage: { inputTokens: 1, outputTokens: 0 } }, { surfaceOp: 'append' })
+    session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 2, content: [], usage: { inputTokens: 1, outputTokens: 0 } }, { surfaceOp: 'append' })
     expect(session.deriveMessages()).toEqual(scratch(session))
   })
 
@@ -40,7 +40,7 @@ describe('derived-message cache', () => {
     const nodes = session.surface.nodes
     session.append('context/message', {
       content: [{ type: 'text', text: 'summary' }], source: { kind: 'plugin', plugin: 'compact' },
-    }, { surfaceOp: { op: 'replace', start: nodes[0]!.seq, end: nodes[1]!.seq }, sourceEventSeqs: [nodes[0]!.seq, nodes[1]!.seq] })
+    }, { surfaceOp: { op: 'replace', start: nodes[0]!, end: nodes[1]! }, sourceEventSeqs: [nodes[0]!, nodes[1]!] })
 
     expect(session.deriveMessages()).toHaveLength(1)
     expect(session.deriveMessages()).toEqual(scratch(session))
@@ -89,7 +89,7 @@ describe('Session.deriveEventMessage — the per-event projection', () => {
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
     const boundary = session.append('step/start', { turn: 1, step: 1 })
     expect(session.deriveEventMessage(boundary)).toBeNull()
-    const empty = session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append' })
+    const empty = session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: 'append' })
     expect(session.deriveEventMessage(empty)).toBeNull()
   })
 })
diff --git a/packages/core/session/tests/fork.spec.ts b/packages/core/session/tests/fork.spec.ts
index af143ea5ee..5344449078 100644
--- a/packages/core/session/tests/fork.spec.ts
+++ b/packages/core/session/tests/fork.spec.ts
@@ -195,14 +195,14 @@ describe('SessionStore.fork', () => {
       ['assistant/message', (session) => {
         session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
         session.append('step/start', { turn: 1, step: 1 })
-        session.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'partial' }] }, { surfaceOp: 'append' })
+        session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'partial' }] }, { surfaceOp: 'append' })
         return lastSeq(session)
       }],
       ['tool/call', (session) => {
         const callId = CallId('call-open')
         session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
         session.append('step/start', { turn: 1, step: 1 })
-        session.append('assistant/message', {
+        session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' },
           turn: 1,
           step: 1,
           content: [{ type: 'tool-call', id: callId, name: 'bash', arguments: '{}' }],
diff --git a/packages/core/session/tests/gen-persistence-catalog.spec.ts b/packages/core/session/tests/gen-persistence-catalog.spec.ts
index cca9cbfa61..9f18e75bbe 100644
--- a/packages/core/session/tests/gen-persistence-catalog.spec.ts
+++ b/packages/core/session/tests/gen-persistence-catalog.spec.ts
@@ -9,6 +9,7 @@ import { join } from 'node:path'
 import { afterEach, describe, expect, it } from 'vitest'
 import {
   annotateSurface,
+  collectEventEnvelopeTypes,
   collectLogEvents,
   collectSurfaceEventTypes,
   render,
@@ -56,6 +57,7 @@ describe('gen-persistence-catalog collectLogEvents', () => {
       scope: 'fix',
       doc: 'A thing was recorded.',
       payload: '{ turn: number }',
+      declaration: '/** A thing was recorded. */\n\'fix/happened\': { turn: number }',
       source: 'packages/core/fix/src/types.ts:3',
     })
   })
@@ -102,10 +104,13 @@ describe('gen-persistence-catalog collectLogEvents', () => {
   it('collapses a newline-separated multi-line payload to a valid one-line fragment', () => {
     const events = collectLogEvents(make({
       'packages/group/fix/src/types.ts': merge(
-        '    /** Wide payload. */\n    \'fix/wide\': {\n      alpha: string[]\n      range: { start: number; end: number }\n      count: number\n    }',
+        '    /** Wide payload. */\n    \'fix/wide\': {\n      /** Alpha values. */\n      alpha: string[]\n      range: { start: number; end: number }\n      count: number\n    }',
       ),
     }))
     expect(events[0]?.payload).toBe('{ alpha: string[]; range: { start: number; end: number }; count: number }')
+    expect(events[0]?.declaration).toBe(
+      '/** Wide payload. */\n\'fix/wide\': {\n  /** Alpha values. */\n  alpha: string[]\n  range: { start: number; end: number }\n  count: number\n}',
+    )
   })
 
   it('hard-errors on a member with no description prose', () => {
@@ -158,6 +163,59 @@ describe('gen-persistence-catalog collectLogEvents', () => {
   })
 })
 
+describe('gen-persistence-catalog collectEventEnvelopeTypes', () => {
+  const declarations = `/** Event keys. */
+export type SessionEventType = keyof SessionEventMap
+/** Surface-producing event keys. */
+export type SurfaceEventType = 'fix/message'
+/** Surface placement. */
+export type SurfaceOp = 'append'
+/** One persisted event. */
+export type SessionEvent<T extends SessionEventType = SessionEventType> = { type: T }
+`
+
+  it('extracts the envelope declarations with their complete JSDoc in canonical order', () => {
+    const entries = collectEventEnvelopeTypes(make({
+      'packages/core/fix/package.json': OWNER_MANIFEST,
+      'packages/core/fix/src/types.ts': declarations,
+    }))
+    expect(entries.map(entry => entry.name)).toEqual([
+      'SessionEventType',
+      'SurfaceEventType',
+      'SurfaceOp',
+      'SessionEvent',
+    ])
+    expect(entries[3]).toMatchObject({
+      declaration: '/** One persisted event. */\nexport type SessionEvent<T extends SessionEventType = SessionEventType> = { type: T }',
+      source: 'packages/core/fix/src/types.ts:8',
+    })
+  })
+
+  it('hard-errors when an envelope declaration is missing', () => {
+    expect(() => collectEventEnvelopeTypes(make({
+      'packages/core/fix/package.json': OWNER_MANIFEST,
+      'packages/core/fix/src/types.ts': declarations.replace('/** Surface placement. */\nexport type SurfaceOp = \'append\'\n', ''),
+    }))).toThrow(/missing event-envelope declaration\(s\): SurfaceOp/)
+  })
+
+  it('hard-errors on duplicate, unexported, undocumented, or mistagged envelope declarations', () => {
+    const violations = new RegExp([
+      '4 JSDoc completeness violation\\(s\\)',
+      '[\\s\\S]*not exported',
+      '[\\s\\S]*@mode tag',
+      '[\\s\\S]*SurfaceOp.*no description prose',
+      '[\\s\\S]*SessionEvent.*already declared',
+    ].join(''))
+    expect(() => collectEventEnvelopeTypes(make({
+      'packages/core/fix/package.json': OWNER_MANIFEST,
+      'packages/core/fix/src/types.ts': declarations
+        .replace('/** Event keys. */\nexport type SessionEventType', '/** Event keys.\n * @mode emit\n */\ntype SessionEventType')
+        .replace('/** Surface placement. */\n', '')
+        + '/** Duplicate event. */\nexport type SessionEvent = { type: never }\n',
+    }))).toThrow(violations)
+  })
+})
+
 describe('gen-persistence-catalog collectSurfaceEventTypes', () => {
   it('parses the literal union', () => {
     const types = collectSurfaceEventTypes(make({
@@ -192,9 +250,21 @@ describe('gen-persistence-catalog annotateSurface + render', () => {
     scope: name.split('/')[0] ?? name,
     payload: '{ turn: number }',
     doc: `Records ${name}.`,
+    declaration: `/** Records ${name}. */\n'${name}': { turn: number }`,
     source: 'packages/core/fix/src/types.ts:3',
   })
 
+  const envelopeTypes = [
+    'SessionEventType',
+    'SurfaceEventType',
+    'SurfaceOp',
+    'SessionEvent',
+  ].map(name => ({
+    name: name as 'SessionEventType' | 'SurfaceEventType' | 'SurfaceOp' | 'SessionEvent',
+    declaration: `/** ${name}. */\nexport type ${name} = never`,
+    source: 'packages/core/fix/src/types.ts:1',
+  }))
+
   it('badges union members surface and everything else log-only', () => {
     const annotated = annotateSurface([entry('fix/message'), entry('fix/marker')], ['fix/message'])
     expect(annotated.map(e => [e.name, e.surface])).toEqual([['fix/message', true], ['fix/marker', false]])
@@ -205,11 +275,13 @@ describe('gen-persistence-catalog annotateSurface + render', () => {
       .toThrow(/'fix\/ghost' name no declared log event/)
   })
 
-  it('renders badges, payload fences, and the generated-file header', () => {
-    const out = render(annotateSurface([entry('fix/message'), entry('fix/marker')], ['fix/message']))
+  it('renders badges, declaration fences, and the generated-file header', () => {
+    const out = render(annotateSurface([entry('fix/message'), entry('fix/marker')], ['fix/message']), envelopeTypes)
     expect(out).toContain('Generated by scripts/gen-persistence-catalog.ts')
+    expect(out).toContain('# Session Persistence Event Catalog')
+    expect(out).toContain('```ts persistence-catalog\n/** SessionEventType. */\nexport type SessionEventType = never')
     expect(out).toContain('#### `fix/message` — surface')
     expect(out).toContain('#### `fix/marker` — log-only')
-    expect(out).toContain('```ts persistence-catalog\n\'fix/marker\': { turn: number }\n```')
+    expect(out).toContain('```ts persistence-catalog\n/** Records fix/marker. */\n\'fix/marker\': { turn: number }\n```')
   })
 })
diff --git a/packages/core/session/tests/properties.spec.ts b/packages/core/session/tests/properties.spec.ts
index d599536db5..9cb744c312 100644
--- a/packages/core/session/tests/properties.spec.ts
+++ b/packages/core/session/tests/properties.spec.ts
@@ -28,8 +28,8 @@ const textContentArb = fc.array(
 // explicit `surfaceOp: 'append'` intent — the marker the real loop passes.
 const messageEventArb: fc.Arbitrary<Appendable> = fc.oneof(
   textContentArb.map((content): Appendable => ({ type: 'user/message', data: { content, source: { kind: 'user' } }, intent: { surfaceOp: 'append' } })),
-  textContentArb.map((content): Appendable => ({ type: 'assistant/message', data: { turn: 1, step: 1, content }, intent: { surfaceOp: 'append' } })),
-  textContentArb.map((content): Appendable => ({ type: 'assistant/message', data: { turn: 1, step: 1, content, usage: { inputTokens: 1, outputTokens: 1 } }, intent: { surfaceOp: 'append' } })),
+  textContentArb.map((content): Appendable => ({ type: 'assistant/message', data: { turn: 1, step: 1, content, provenance: { provider: 'mock', model: 'mock' } }, intent: { surfaceOp: 'append' } })),
+  textContentArb.map((content): Appendable => ({ type: 'assistant/message', data: { turn: 1, step: 1, content, provenance: { provider: 'mock', model: 'mock' }, usage: { inputTokens: 1, outputTokens: 1 } }, intent: { surfaceOp: 'append' } })),
   fc.record({ id: fc.string({ minLength: 1 }), content: textContentArb, isError: fc.boolean() })
     .map((r): Appendable => ({ type: 'tool/result', data: { turn: 1, step: 1, callId: CallId(r.id), content: r.content, isError: r.isError }, intent: { surfaceOp: 'append' } })),
 )
diff --git a/packages/core/session/tests/repair.spec.ts b/packages/core/session/tests/repair.spec.ts
index e0f6a5bb6e..765502b8ce 100644
--- a/packages/core/session/tests/repair.spec.ts
+++ b/packages/core/session/tests/repair.spec.ts
@@ -56,7 +56,7 @@ describe('interruptedTurnClosers', () => {
       { type: 'assistant/message', seq: 2, time: 2, data: { turn: 2, step: 1, content: [
         { type: 'text', text: 'calling a tool' },
         { type: 'tool-call', id: CallId('call-1'), name: 'bash', arguments: '{}' },
-      ] } },
+      ], provenance: { provider: 'mock', model: 'mock' } } },
     ]
     const closers = interruptedTurnClosers(events)
     // tool/result (for the orphaned call) → step/end → turn/end, contiguous seqs.
@@ -74,7 +74,7 @@ describe('interruptedTurnClosers', () => {
       { type: 'step/start', seq: 1, time: 1, data: { turn: 2, step: 1 } },
       { type: 'assistant/message', seq: 2, time: 2, data: { turn: 2, step: 1, content: [
         { type: 'tool-call', id: CallId('call-1'), name: 'bash', arguments: '{}' },
-      ] } },
+      ], provenance: { provider: 'mock', model: 'mock' } } },
       { type: 'tool/result', seq: 3, time: 3, data: { turn: 2, step: 1, callId: CallId('call-1'), content: [{ type: 'text', text: 'ok' }], isError: false } },
     ]
     // The call is answered, so only the open step + turn need closing.
@@ -88,7 +88,7 @@ describe('interruptedTurnClosers', () => {
       { type: 'step/start', seq: 1, time: 1, data: { turn: 2, step: 1 } },
       { type: 'assistant/message', seq: 2, time: 2, data: { turn: 2, step: 1, content: [
         { type: 'tool-call', id: CallId('call-1'), name: 'bash', arguments: '{}' },
-      ] } },
+      ], provenance: { provider: 'mock', model: 'mock' } } },
       { type: 'step/end', seq: 3, time: 3, data: { turn: 2, step: 1 } },
     ]
 
@@ -105,7 +105,7 @@ describe('interruptedTurnClosers', () => {
       { type: 'step/start', seq: 1, time: 1, data: { turn: 1, step: 1 } },
       { type: 'assistant/message', seq: 2, time: 2, data: { turn: 1, step: 1, content: [
         { type: 'tool-call', id: CallId('old-call'), name: 'bash', arguments: '{}' },
-      ] } },
+      ], provenance: { provider: 'mock', model: 'mock' } } },
       { type: 'tool/result', seq: 3, time: 3, data: { turn: 1, step: 1, callId: CallId('old-call'), content: [], isError: false } },
       { type: 'step/end', seq: 4, time: 4, data: { turn: 1, step: 1 } },
       { type: 'turn/end', seq: 5, time: 5, data: { turn: 1, reason: { kind: 'completed' } } },
@@ -113,7 +113,7 @@ describe('interruptedTurnClosers', () => {
       { type: 'step/start', seq: 7, time: 7, data: { turn: 2, step: 1 } },
       { type: 'assistant/message', seq: 8, time: 8, data: { turn: 2, step: 1, content: [
         { type: 'tool-call', id: CallId('new-call'), name: 'bash', arguments: '{}' },
-      ] } },
+      ], provenance: { provider: 'mock', model: 'mock' } } },
     ]
     const closers = interruptedTurnClosers(events)
     expect(closers.map(e => e.type)).toEqual(['tool/result', 'step/end', 'turn/end'])
@@ -128,7 +128,7 @@ describe('interruptedTurnClosers', () => {
       { type: 'assistant/message', seq: 2, time: 2, data: { turn: 1, step: 1, content: [
         { type: 'tool-call', id: CallId('call-a'), name: 'bash', arguments: '{}' },
         { type: 'tool-call', id: CallId('call-b'), name: 'bash', arguments: '{}' },
-      ] } },
+      ], provenance: { provider: 'mock', model: 'mock' } } },
       // call-a got answered before the crash; call-b did not.
       { type: 'tool/result', seq: 3, time: 3, data: { turn: 1, step: 1, callId: CallId('call-a'), content: [], isError: false } },
     ]
@@ -144,7 +144,7 @@ describe('interruptedTurnClosers', () => {
       { type: 'step/start', seq: 1, time: 1, data: { turn: 1, step: 1 } },
       { type: 'assistant/message', seq: 2, time: 2, data: { turn: 1, step: 1, content: [
         { type: 'tool-call', id: CallId('call-1'), name: 'bash', arguments: '{}' },
-      ] } },
+      ], provenance: { provider: 'mock', model: 'mock' } } },
       { type: 'tool/call', seq: 3, time: 3, data: { turn: 1, step: 1, callId: CallId('call-1'), name: 'bash', arguments: '{}' } },
     ]
     const closers = interruptedTurnClosers(events)
diff --git a/packages/core/session/tests/request-header.spec.ts b/packages/core/session/tests/request-header.spec.ts
index 8a5af819c3..b185cf6e56 100644
--- a/packages/core/session/tests/request-header.spec.ts
+++ b/packages/core/session/tests/request-header.spec.ts
@@ -1,18 +1,11 @@
-/**
- * Request-header utility tests: canonical form, the system line-diff
- * (prefix/suffix trim), the name-keyed tools delta, config replacement, the
- * round-trip contract (including the reorder case the encoding cannot
- * express), and the log fold. These pin the reconstruction algebra: for every
- * logged delta, apply(prev, delta) === next, and folding a log prefix yields
- * the header its next request was built under.
- */
+/** Request-header canonicalization, equality, snapshot folding, and format rejection. */
 
 import { describe, expect, it } from 'vitest'
-import { Session, SessionId, applyHeaderDelta, canonicalHeader, diffHeader, foldRequestHeader, headerEquals } from '@deepseek-ai/dsh-session'
+import { Session, SessionId, canonicalHeader, foldRequestHeader, headerEquals } from '@deepseek-ai/dsh-session'
 import type { EpochHeader, SessionEvent } from '@deepseek-ai/dsh-session'
 import type { Message, ToolSchema } from '@deepseek-ai/dsh-llm'
 
-const CONFIG = { model: 'm' }
+const CONFIG = { provider: 'mock', model: 'm' }
 
 function tool(name: string, description = 'd'): ToolSchema {
   return { name, description, parameters: { type: 'object' } }
@@ -22,165 +15,77 @@ function msg(text: string): Message {
   return { role: 'user', content: [{ type: 'text', text }] }
 }
 
-/** Round-trip helper: diff must reproduce `next` from `prev` exactly. */
-function roundTrip(prev: EpochHeader, next: EpochHeader): ReturnType<typeof diffHeader> {
-  const delta = diffHeader(prev, next)
-  if (delta !== undefined) {
-    expect(applyHeaderDelta(prev, delta)).toEqual(canonicalHeader(next))
-  }
-  return delta
-}
-
 describe('canonicalHeader', () => {
-  it('normalizes empty system and empty tools to absent fields', () => {
-    expect(canonicalHeader({ config: CONFIG, system: '', tools: [] })).toEqual({ config: CONFIG })
-    const full = canonicalHeader({ config: CONFIG, system: 's', tools: [tool('a')] })
-    expect(full.system).toBe('s')
-    expect(full.tools).toHaveLength(1)
+  it('normalizes empty optional fields to absence and preserves populated fields', () => {
+    expect(canonicalHeader({ config: CONFIG, system: '', tools: [], messagePrefix: [] })).toEqual({ config: CONFIG })
+    const full = canonicalHeader({ config: CONFIG, system: 's', tools: [tool('a')], messagePrefix: [msg('p')] })
+    expect(full).toEqual({ config: CONFIG, system: 's', tools: [tool('a')], messagePrefix: [msg('p')] })
   })
 })
 
-describe('diffHeader / applyHeaderDelta', () => {
-  it('returns undefined for equal headers', () => {
-    const header = canonicalHeader({ config: CONFIG, system: 'a\nb', tools: [tool('t')] })
-    expect(diffHeader(header, header)).toBeUndefined()
+describe('headerEquals', () => {
+  const base = canonicalHeader({ config: CONFIG, system: 's', tools: [tool('a')], messagePrefix: [msg('p')] })
+
+  it('compares every canonical field and preserves tool order', () => {
+    expect(headerEquals(base, structuredClone(base))).toBe(true)
+    expect(headerEquals(base, { ...base, config: { provider: 'mock', model: 'other' } })).toBe(false)
+    expect(headerEquals(base, { ...base, system: 'other' })).toBe(false)
+    expect(headerEquals(base, { ...base, messagePrefix: [msg('other')] })).toBe(false)
+    expect(headerEquals(base, { ...base, tools: [] })).toBe(false)
+    expect(headerEquals(base, { ...base, tools: [tool('a', 'changed')] })).toBe(false)
+    expect(headerEquals({ config: CONFIG, tools: [tool('a'), tool('b')] }, { config: CONFIG, tools: [tool('b'), tool('a')] })).toBe(false)
   })
 
-  it('encodes a mid-prompt line change as a prefix/suffix trim', () => {
-    const prev = canonicalHeader({ config: CONFIG, system: 'keep1\nold\nkeep2\nkeep3' })
-    const next = canonicalHeader({ config: CONFIG, system: 'keep1\nnew A\nnew B\nkeep2\nkeep3' })
-    const delta = roundTrip(prev, next)
-    expect(delta?.system).toEqual({ keepStart: 1, keepEnd: 2, insert: ['new A', 'new B'] })
-    expect(delta?.tools).toBeUndefined()
-    expect(delta?.config).toBeUndefined()
-  })
-
-  it('degenerates to a full replacement when nothing is shared, and round-trips absence transitions', () => {
-    const none = canonicalHeader({ config: CONFIG })
-    const some = canonicalHeader({ config: CONFIG, system: 'x\ny' })
-    const gained = roundTrip(none, some)
-    expect(gained?.system).toEqual({ keepStart: 0, keepEnd: 0, insert: ['x', 'y'] })
-    const lost = roundTrip(some, none)
-    expect(lost?.system).toEqual({ keepStart: 0, keepEnd: 0, insert: [] })
-  })
-
-  it('does not double-count overlapping prefix and suffix (repeated lines)', () => {
-    const prev = canonicalHeader({ config: CONFIG, system: 'a\na' })
-    const next = canonicalHeader({ config: CONFIG, system: 'a\na\na' })
-    roundTrip(prev, next)
-  })
-
-  it('encodes tool addition, removal, and in-place schema change by name', () => {
-    const prev = canonicalHeader({ config: CONFIG, tools: [tool('keep'), tool('drop'), tool('edit', 'before')] })
-    const next = canonicalHeader({ config: CONFIG, tools: [tool('keep'), tool('edit', 'after'), tool('new')] })
-    const delta = roundTrip(prev, next)
-    expect(delta?.tools?.added.map(t => t.name)).toEqual(['new'])
-    expect(delta?.tools?.removed).toEqual(['drop'])
-    expect(delta?.tools?.changed.map(t => t.name)).toEqual(['edit'])
-  })
-
-  it('round-trips a tool set gained from a tool-less header and lost back to one', () => {
-    const none = canonicalHeader({ config: CONFIG })
-    const some = canonicalHeader({ config: CONFIG, tools: [tool('t')] })
-    const gained = roundTrip(none, some)
-    expect(gained?.tools?.added.map(t => t.name)).toEqual(['t'])
-    const lost = roundTrip(some, none)
-    expect(lost?.tools?.removed).toEqual(['t'])
-  })
-
-  it('cannot express a pure reordering — the writer detects it via the round-trip check', () => {
-    const prev = canonicalHeader({ config: CONFIG, tools: [tool('a'), tool('b')] })
-    const next = canonicalHeader({ config: CONFIG, tools: [tool('b'), tool('a')] })
-    const delta = diffHeader(prev, next)
-    // A delta IS produced (the lists differ)…
-    expect(delta).toBeDefined()
-    // …but applying it cannot reproduce the new order — exactly the case the
-    // writer's guard turns into a 'fallback' snapshot.
-    expect(applyHeaderDelta(prev, delta!)).not.toEqual(next)
-  })
-
-  it('replaces the config whole and leaves untouched parts alone', () => {
-    const prev = canonicalHeader({ config: { model: 'm' }, system: 's', tools: [tool('t')] })
-    const next = canonicalHeader({ config: { model: 'm2', temperature: 0.1 }, system: 's', tools: [tool('t')] })
-    const delta = roundTrip(prev, next)
-    expect(delta).toEqual({ config: { model: 'm2', temperature: 0.1 } })
-  })
-})
-
-describe('the session prefix (messagePrefix)', () => {
-  it('canonicalHeader normalizes an empty prefix to an absent field', () => {
-    expect(canonicalHeader({ config: CONFIG, messagePrefix: [] })).toEqual({ config: CONFIG })
-    const full = canonicalHeader({ config: CONFIG, messagePrefix: [msg('p')] })
-    expect(full.messagePrefix).toEqual([msg('p')])
-  })
-
-  it('headerEquals treats absence and empty as one representation, content differences as unequal', () => {
-    expect(headerEquals(canonicalHeader({ config: CONFIG }), { config: CONFIG, messagePrefix: [] })).toBe(true)
-    expect(headerEquals({ config: CONFIG, messagePrefix: [msg('a')] }, { config: CONFIG, messagePrefix: [msg('b')] })).toBe(false)
-    expect(headerEquals({ config: CONFIG, messagePrefix: [msg('a')] }, { config: CONFIG })).toBe(false)
-  })
-
-  it('replaces a changed prefix whole and leaves untouched parts alone', () => {
-    const prev = canonicalHeader({ config: CONFIG, system: 'keep', messagePrefix: [msg('old')] })
-    const next = canonicalHeader({ config: CONFIG, system: 'keep', messagePrefix: [msg('new'), msg('more')] })
-    const delta = roundTrip(prev, next)
-    expect(delta).toEqual({ messagePrefix: [msg('new'), msg('more')] })
-  })
-
-  it('round-trips a prefix gained from a bare header and lost back to one (empty array encodes absence)', () => {
-    const none = canonicalHeader({ config: CONFIG })
-    const some = canonicalHeader({ config: CONFIG, messagePrefix: [msg('p')] })
-    const gained = roundTrip(none, some)
-    expect(gained).toEqual({ messagePrefix: [msg('p')] })
-    const lost = roundTrip(some, none)
-    expect(lost).toEqual({ messagePrefix: [] })
-  })
-
-  it('folds prefix deltas over the log like any other header amendment', () => {
-    const session = new Session(SessionId('fold-prefix'))
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    const first = canonicalHeader({ config: CONFIG, messagePrefix: [msg('catalog v1')] })
-    session.append('request/header', { header: first, reason: 'initial' })
-    const second = canonicalHeader({ config: CONFIG, messagePrefix: [msg('catalog v2')] })
-    session.append('request/header-delta', diffHeader(first, second)!)
-    expect(foldRequestHeader(session.events)).toEqual(second)
-    session.append('request/header-delta', diffHeader(second, canonicalHeader({ config: CONFIG }))!)
-    expect(foldRequestHeader(session.events)).toEqual({ config: CONFIG })
+  it('treats absent and empty prefix/tool arrays as equivalent canonical absence', () => {
+    expect(headerEquals({ config: CONFIG }, { config: CONFIG, tools: [], messagePrefix: [] })).toBe(true)
   })
 })
 
 describe('foldRequestHeader', () => {
-  function headerEvents(session: Session): readonly SessionEvent[] {
-    return session.events
-  }
-
-  it('returns undefined on a log with no header events', () => {
-    const session = new Session(SessionId('fold-none'))
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    expect(foldRequestHeader(headerEvents(session))).toBeUndefined()
+  it('returns the supplied baseline when no snapshot follows', () => {
+    const from: EpochHeader = { config: CONFIG, system: 'baseline' }
+    const unrelated: SessionEvent[] = [
+      { type: 'turn/start', seq: 0, time: 1, data: { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } } },
+    ]
+    expect(foldRequestHeader(unrelated)).toBeUndefined()
+    expect(foldRequestHeader(unrelated, from)).toBe(from)
   })
 
-  it('folds snapshot then deltas into the header in force, skipping unrelated events', () => {
+  it('takes the latest full snapshot and skips unrelated events', () => {
     const session = new Session(SessionId('fold'))
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    const first = canonicalHeader({ config: { model: 'm' }, system: 'a\nb', tools: [tool('t')] })
-    session.append('request/header', { header: first, reason: 'initial' })
+    session.append('request/header', { header: { config: CONFIG, system: 'first' }, reason: 'initial' })
     session.append('user/message', { content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-
-    const second = canonicalHeader({ config: { model: 'm' }, system: 'a\nc', tools: [tool('t')] })
-    session.append('request/header-delta', diffHeader(first, second)!)
-    expect(foldRequestHeader(headerEvents(session))).toEqual(second)
-
-    // A later snapshot replaces the state wholesale (the 'resume'/'fallback' anchor).
-    const third = canonicalHeader({ config: { model: 'other' } })
-    session.append('request/header', { header: third, reason: 'resume' })
-    expect(foldRequestHeader(headerEvents(session))).toEqual(third)
-  })
-
-  it('throws on a delta before any snapshot (corrupt log)', () => {
-    const session = new Session(SessionId('fold-corrupt'))
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    session.append('request/header-delta', { config: { model: 'x' } })
-    expect(() => foldRequestHeader(headerEvents(session))).toThrow(/before any request\/header snapshot/)
+    session.append('request/header', { header: { config: { provider: 'mock', model: 'other' }, tools: [] }, reason: 'change' })
+    expect(foldRequestHeader(session.events)).toEqual({ config: { provider: 'mock', model: 'other' } })
+  })
+})
+
+describe('legacy request-header format', () => {
+  it('rejects request/header-delta in seeds and untyped appends', () => {
+    const legacy = [{
+      type: 'request/header-delta', seq: 0, time: 1, data: { config: CONFIG },
+    }] as unknown as SessionEvent[]
+    expect(() => new Session(SessionId('legacy'), legacy)).toThrow(/unsupported legacy request\/header-delta/)
+
+    const session = new Session(SessionId('legacy-append-delta'))
+    const appendLegacy = session.append.bind(session) as (type: string, data: unknown) => SessionEvent
+    expect(() => appendLegacy('request/header-delta', { config: CONFIG }))
+      .toThrow(/unsupported legacy request\/header-delta/)
+    expect(session.events).toHaveLength(0)
+  })
+
+  it('rejects the removed fallback reason in seeds and untyped appends', () => {
+    const legacy = [{
+      type: 'request/header', seq: 0, time: 1, data: { header: { config: CONFIG }, reason: 'fallback' },
+    }] as unknown as SessionEvent[]
+    expect(() => new Session(SessionId('legacy-seed-reason'), legacy))
+      .toThrow('unsupported legacy request/header reason "fallback"')
+
+    const session = new Session(SessionId('legacy-append-reason'))
+    const appendLegacy = session.append.bind(session) as (type: string, data: unknown) => SessionEvent
+    expect(() => appendLegacy('request/header', { header: { config: CONFIG }, reason: 'fallback' }))
+      .toThrow('unsupported legacy request/header reason "fallback"')
+    expect(session.events).toHaveLength(0)
   })
 })
diff --git a/packages/core/session/tests/session.spec.ts b/packages/core/session/tests/session.spec.ts
index 2326e7b212..16a74283be 100644
--- a/packages/core/session/tests/session.spec.ts
+++ b/packages/core/session/tests/session.spec.ts
@@ -10,7 +10,7 @@ describe('Session', () => {
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
     session.append('user/message', { content: [{ type: 'text', text: 'hello' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
     session.append('assistant/chunk', { turn: 1, step: 1, chunk: { type: 'text-delta', index: 0, text: 'hi' } })
-    session.append('assistant/message', {
+    session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' },
       turn: 1, step: 1,
       content: [
         { type: 'text', text: 'let me check' },
@@ -59,11 +59,33 @@ describe('Session', () => {
     expect(steeringMessage!.content[0]).toMatchObject({ type: 'text', text: '<steering source="user">' })
   })
 
+  it('renders raw context without a generic envelope while preserving structured metadata', () => {
+    const session = new Session(SessionId('s2-raw'))
+    const meta = {
+      kind: 'workspace-instructions',
+      version: 1,
+      changes: [{ action: 'set', scope: 'pkg', path: 'pkg/AGENTS.md', digest: 'abc123' }],
+    }
+    session.append('context/message', {
+      content: [{ type: 'text', text: '<system-reminder>Additional instructions from: pkg/AGENTS.md</system-reminder>' }],
+      source: { kind: 'plugin', plugin: 'workspace-context' },
+      envelope: 'raw',
+      meta,
+    }, { surfaceOp: 'append' })
+
+    expect(session.deriveMessages()).toEqual([{
+      role: 'user',
+      content: [{ type: 'text', text: '<system-reminder>Additional instructions from: pkg/AGENTS.md</system-reminder>' }],
+    }])
+    const event = session.events[0]
+    expect(event?.type === 'context/message' && event.data.meta).toEqual(meta)
+  })
+
   it('replays identically from a seeded event log', () => {
     const original = new Session(SessionId('s3'))
     original.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
     original.append('user/message', { content: [{ type: 'text', text: 'q' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    original.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'a' }] }, { surfaceOp: 'append' })
+    original.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'a' }] }, { surfaceOp: 'append' })
     original.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
 
     const replayed = new Session(SessionId('s3-replay'), [...original.events])
@@ -71,6 +93,36 @@ describe('Session', () => {
     expect(replayed.seq).toBe(original.seq)
   })
 
+  it('rejects pre-provider request headers and assistant messages on seed/load', () => {
+    const requestHeader = {
+      type: 'request/header', seq: 0, time: 1,
+      data: { header: { config: { model: 'old-model' } }, reason: 'initial' },
+    } as unknown as SessionEvent
+    expect(() => new Session(SessionId('old-header'), [requestHeader]))
+      .toThrow('seed request/header at index 0 lacks provider/model')
+
+    const assistantMessage = {
+      type: 'assistant/message', seq: 0, time: 1,
+      data: { turn: 1, step: 1, content: [{ type: 'text', text: 'old' }] },
+      surfaceOp: 'append',
+    } as unknown as SessionEvent
+    expect(() => new Session(SessionId('old-assistant'), [assistantMessage]))
+      .toThrow('seed assistant/message at index 0 lacks provider/model provenance')
+
+    const malformedHeader = {
+      type: 'request/header', seq: 0, time: 1,
+      data: { header: 'old-header' },
+    } as unknown as SessionEvent
+    expect(() => new Session(SessionId('malformed-header'), [malformedHeader]))
+      .toThrow('seed request/header at index 0 lacks provider/model')
+
+    const unrelatedPrimitiveData = {
+      type: 'plugin/event', seq: 0, time: 1, data: null,
+    } as unknown as SessionEvent
+    expect(new Session(SessionId('primitive-plugin-data'), [unrelatedPrimitiveData]).events)
+      .toEqual([unrelatedPrimitiveData])
+  })
+
   it('isolates the log from mutation through a derived message (append-only contract)', () => {
     const session = new Session(SessionId('s4'))
     session.append('user/message', { content: [{ type: 'text', text: 'original' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
@@ -295,35 +347,52 @@ describe('Session', () => {
       type: 'user/message',
       seq: 0,
       time: 1,
+      data: { content: [{ type: 'text', text: 'source' }], source: { kind: 'user' } },
+      surfaceOp: 'append',
+    }, {
+      type: 'user/message',
+      seq: 1,
+      time: 2,
       data: { content: [{ type: 'text', text: 'hello' }], source: { kind: 'user' } },
       surfaceOp,
+      sourceEventSeqs: [0],
     }] as unknown as SessionEvent[]
 
     const session = new Session(SessionId('seed-unstable-metadata'), seed)
-    const event = session.events[0]!
+    const event = session.events[1]!
     if (event.type !== 'user/message') throw new Error('test fixture must remain a user/message')
 
     expect(reads).toBe(1)
     expect(event.surfaceOp).toEqual({ op: 'replace', start: 0, end: 0 })
   })
 
-  it('adds seed context when surface validation throws a non-Error value', () => {
+  it.each([
+    ['an Error', new Error('validator failed'), 'validator failed'],
+    ['a non-Error value', 'validator failed', 'invalid surface metadata'],
+  ] as const)('adds seed context when surface validation throws %s', (_name, failure, expected) => {
     const originalHasOwn = Object.hasOwn
     const hasOwn = vi.spyOn(Object, 'hasOwn').mockImplementation((object: object, property: PropertyKey): boolean => {
-      if ((object as Record<string, unknown>)['op'] === 'replace') throw 'validator failed'
+      if ((object as Record<string, unknown>)['op'] === 'replace') throw failure
       return originalHasOwn(object, property)
     })
     const seed = [{
       type: 'user/message',
       seq: 0,
       time: 1,
+      data: { content: [{ type: 'text', text: 'source' }], source: { kind: 'user' } },
+      surfaceOp: 'append',
+    }, {
+      type: 'user/message',
+      seq: 1,
+      time: 2,
       data: { content: [{ type: 'text', text: 'hello' }], source: { kind: 'user' } },
       surfaceOp: { op: 'replace', start: 0, end: 0 },
+      sourceEventSeqs: [0],
     }] as unknown as SessionEvent[]
 
     try {
       expect(() => new Session(SessionId('seed-non-error-metadata-failure'), seed))
-        .toThrow('invalid seed event at index 0: invalid surface metadata')
+        .toThrow(`invalid seed event at index 1: ${expected}`)
     } finally {
       hasOwn.mockRestore()
     }
@@ -409,6 +478,11 @@ describe('Session', () => {
 
   it('reads a nested append-metadata getter once and stores its first JSON value', () => {
     const session = new Session(SessionId('append-unstable-metadata'))
+    const source = session.append(
+      'user/message',
+      { content: [{ type: 'text', text: 'source' }], source: { kind: 'user' } },
+      { surfaceOp: 'append' },
+    )
     let reads = 0
     const surfaceOp = Object.defineProperty({ op: 'replace', end: 0 }, 'start', {
       enumerable: true,
@@ -421,12 +495,12 @@ describe('Session', () => {
     const event = session.append(
       'user/message',
       { content: [{ type: 'text', text: 'hello' }], source: { kind: 'user' } },
-      { surfaceOp } as never,
+      { surfaceOp, sourceEventSeqs: [0] } as never,
     )
 
     expect(reads).toBe(1)
     expect(event.surfaceOp).toEqual({ op: 'replace', start: 0, end: 0 })
-    expect(session.events).toEqual([event])
+    expect(session.events).toEqual([source, event])
   })
 
   it('rejects invalid plain surface metadata shapes at append', () => {
@@ -462,7 +536,7 @@ describe('Session', () => {
       'turn/start',
       { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } },
       { surfaceOp: 'append' },
-    )).toThrow(/not surface-eligible and cannot carry surface metadata/)
+    )).toThrow(/not surface-eligible and cannot carry surfaceOp/)
     expect(() => new Session(SessionId('non-surface-metadata-seed'), [{
       type: 'turn/start',
       seq: 0,
@@ -1163,8 +1237,8 @@ describe('todo/write event', () => {
     session.append('todo/write', { todos: [{ content: 'a task', status: 'pending' }] })
     // The todo event must not add a message to the derived history…
     expect(session.deriveMessages()).toHaveLength(before)
-    // …and must not appear on the surface linked list.
-    expect(session.surface.nodes.some(node => node.seq === session.seq - 1)).toBe(false)
+    // …and must not appear on the ordered surface.
+    expect(session.surface.nodes).not.toContain(session.seq - 1)
   })
 
   it('round-trips through a seeded replay identically (durable, no surfaceOp needed)', () => {
diff --git a/packages/core/session/tests/surface.spec.ts b/packages/core/session/tests/surface.spec.ts
index eb15cf3317..fe222f8f2f 100644
--- a/packages/core/session/tests/surface.spec.ts
+++ b/packages/core/session/tests/surface.spec.ts
@@ -1,6 +1,12 @@
 import { describe, expect, it } from 'vitest'
 import type { SessionEvent, SurfaceEvent, SurfaceEventType } from '@deepseek-ai/dsh-session'
-import { Session, SessionId, foldSurface, isSurfaceEligibleType, isSurfaceEvent } from '@deepseek-ai/dsh-session'
+import {
+  Session,
+  SessionId,
+  foldSurface,
+  isSurfaceEligibleType,
+  isSurfaceEvent,
+} from '@deepseek-ai/dsh-session'
 import { CallId } from '@deepseek-ai/dsh-llm'
 
 /** Build a minimal session with turn boundaries and a single user message. */
@@ -8,18 +14,93 @@ function surfaceSession(): Session {
   const s = new Session(SessionId('ss'))
   s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
   s.append('user/message', { content: [{ type: 'text', text: 'hello' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-  s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'hi' }] }, { surfaceOp: 'append' })
+  s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'hi' }] }, { surfaceOp: 'append' })
   s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
   return s
 }
 
+function provenanceEvent(seq: number, sourceEventSeqs: unknown): SessionEvent {
+  return {
+    type: 'user/message',
+    seq,
+    time: seq,
+    data: { content: [], source: { kind: 'user' } },
+    surfaceOp: 'append',
+    ...sourceEventSeqs === undefined ? {} : { sourceEventSeqs },
+  } as unknown as SessionEvent
+}
+
+describe('foldSurface provenance', () => {
+  it('accepts absent or valid provenance and complete replacement coverage', () => {
+    const events = [
+      provenanceEvent(0, undefined),
+      provenanceEvent(1, undefined),
+      {
+        ...provenanceEvent(2, [0, 1]),
+        surfaceOp: { op: 'replace', start: 0, end: 1 },
+      },
+    ] as SessionEvent[]
+    expect(() => foldSurface(events)).not.toThrow()
+  })
+
+  it('rejects provenance on a non-surface event', () => {
+    const event = {
+      type: 'turn/start',
+      seq: 0,
+      time: 1,
+      data: { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } },
+      sourceEventSeqs: [0],
+    } as unknown as SessionEvent
+    expect(() => foldSurface([event])).toThrow(/cannot carry sourceEventSeqs/)
+  })
+
+  it('accepts explicit empty provenance on an assistant message', () => {
+    const event = {
+      type: 'assistant/message',
+      seq: 0,
+      time: 0,
+      data: {
+        provenance: { provider: 'mock', model: 'mock' },
+        turn: 1,
+        step: 1,
+        content: [],
+      },
+      surfaceOp: 'append',
+      sourceEventSeqs: [],
+    } as SessionEvent
+    expect(() => foldSurface([event])).not.toThrow()
+  })
+
+  it.each([
+    ['a non-array', [{ ...provenanceEvent(0, undefined), sourceEventSeqs: 'invalid' }], /must be an array/],
+    ['an empty array', [provenanceEvent(0, [])], /must not be empty/],
+    ['duplicates', [provenanceEvent(0, undefined), provenanceEvent(1, [0, 0])], /must not contain duplicates/],
+    ['a sparse array', [provenanceEvent(0, Array<number>(1))], /densely contain/],
+    ['a non-number', [{ ...provenanceEvent(0, undefined), sourceEventSeqs: ['0'] }], /non-negative safe integers/],
+    ['a fractional number', [provenanceEvent(0, [0.5])], /non-negative safe integers/],
+    ['a negative number', [provenanceEvent(0, [-1])], /non-negative safe integers/],
+    ['a self reference', [provenanceEvent(0, [0])], /must reference earlier events/],
+    ['a non-contiguous event seq', [provenanceEvent(0, undefined), provenanceEvent(2, [1])], /seq 2 is not contiguous; expected 1/],
+    ['incomplete replacement coverage', [
+      provenanceEvent(0, undefined),
+      provenanceEvent(1, undefined),
+      { ...provenanceEvent(2, [0]), surfaceOp: { op: 'replace', start: 0, end: 1 } },
+    ], /missing 1/],
+  ] as const)(
+    'rejects %s',
+    (_name, events, expected) => {
+      expect(() => foldSurface(events as unknown as SessionEvent[])).toThrow(expected)
+    },
+  )
+})
+
 describe('SurfaceManager', () => {
-  it('shares exact nodes and nested replacement ranges with foldSurface', () => {
+  it('shares ordered entries and nested replacement ranges with foldSurface', () => {
     const s = new Session(SessionId('shared-fold'))
     s.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
     s.append('user/message', { content: [{ type: 'text', text: 'b' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'summary' }] }, { surfaceOp: { op: 'replace', start: 0, end: 0 }, sourceEventSeqs: [0] })
-    s.append('assistant/message', { turn: 1, step: 2, content: [{ type: 'text', text: 'summary 2' }] }, { surfaceOp: { op: 'replace', start: 2, end: 1 }, sourceEventSeqs: [2, 1] })
+    s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'summary' }] }, { surfaceOp: { op: 'replace', start: 0, end: 0 }, sourceEventSeqs: [0] })
+    s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 2, content: [{ type: 'text', text: 'summary 2' }] }, { surfaceOp: { op: 'replace', start: 2, end: 1 }, sourceEventSeqs: [2, 1] })
 
     const folded = foldSurface(s.events)
     expect(folded.nodes).toEqual(s.surface.nodes)
@@ -27,18 +108,19 @@ describe('SurfaceManager', () => {
       { seq: 2, start: 0, end: 0, shadowedSeqs: [0] },
       { seq: 3, start: 2, end: 1, shadowedSeqs: [2, 1] },
     ])
-    folded.nodes[0]!.next = 99
+    folded.nodes[0] = 99
     folded.replacements[0]!.shadowedSeqs.push(99)
-    expect(s.surface.nodes).toEqual([{ seq: 3, prev: null, next: null }])
+    expect(s.surface.nodes).toEqual([3])
+    expect(foldSurface(s.events).nodes).toEqual([3])
     expect(foldSurface(s.events).replacements[0]!.shadowedSeqs).toEqual([0])
   })
 
   it('does not retain fold-only replacement history in incremental state', () => {
     const s = new Session(SessionId('incremental-state'))
     s.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'b' }] }, { surfaceOp: { op: 'replace', start: 0, end: 0 } })
+    s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'b' }] }, { surfaceOp: { op: 'replace', start: 0, end: 0 }, sourceEventSeqs: [0] })
 
-    expect(s.surface.nodes).toEqual([{ seq: 1, prev: null, next: null }])
+    expect(s.surface.nodes).toEqual([1])
     const manager = s.surface as unknown as { _state: object }
     expect(Object.hasOwn(manager._state, 'replacements')).toBe(false)
     expect(foldSurface(s.events).replacements).toEqual([
@@ -47,12 +129,29 @@ describe('SurfaceManager', () => {
   })
 
   it('foldSurface reports the same invalid replacement failures as the incremental manager', () => {
-    const s = new Session(SessionId('shared-fold-invalid'))
-    s.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 42, end: 0 }, sourceEventSeqs: [0] })
+    const events = [
+      provenanceEvent(0, undefined),
+      { ...provenanceEvent(1, [0]), surfaceOp: { op: 'replace', start: 42, end: 0 } },
+    ] as SessionEvent[]
 
-    expect(() => foldSurface(s.events)).toThrow(/start seq 42 not found/)
-    expect(() => s.surface.nodes).toThrow(/start seq 42 not found/)
+    expect(() => foldSurface(events)).toThrow(/start seq 42 not found/)
+    expect(() => new Session(SessionId('shared-fold-invalid'), events))
+      .toThrow(/start seq 42 not found/)
+  })
+
+  it('leaves incremental state unchanged when candidate validation fails', () => {
+    const s = new Session(SessionId('atomic-validation'))
+    s.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
+
+    expect(() => s.append(
+      'assistant/message',
+      { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'invalid' }] },
+      { surfaceOp: { op: 'replace', start: 0, end: 0 } },
+    )).toThrow(/missing 0/)
+
+    expect(s.events).toHaveLength(1)
+    s.append('user/message', { content: [{ type: 'text', text: 'b' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
+    expect(s.surface.nodes).toEqual([0, 1])
   })
 
   it('foldSurface rejects a surface-eligible event without its mandatory marker', () => {
@@ -64,21 +163,28 @@ describe('SurfaceManager', () => {
     }
 
     expect(() => foldSurface([malformed]))
-      .toThrow(/surface event "user\/message" \(seq 0\) carries no surfaceOp marker/)
+      .toThrow(/surface-eligible and requires a surfaceOp marker/)
   })
 
-  it('rebuilds a linked list from surfaceOp: append markers', () => {
+  it('foldSurface rejects surfaceOp on a non-surface event', () => {
+    const malformed = {
+      type: 'turn/start',
+      seq: 0,
+      time: 1,
+      data: { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } },
+      surfaceOp: 'append',
+    } as unknown as SessionEvent
+
+    expect(() => foldSurface([malformed]))
+      .toThrow(/not surface-eligible and cannot carry surfaceOp/)
+  })
+
+  it('folds an ordered sequence list from surfaceOp: append markers', () => {
     const s = surfaceSession()
     const nodes = s.surface.nodes
     // Only the user/message and assistant/message carry surfaceOp: 'append'.
     // The turn boundaries do not have surface markers.
-    expect(nodes.length).toBe(2)
-    expect(nodes[0]!.seq).toBe(1) // user/message (turn/start is seq 0)
-    expect(nodes[0]!.prev).toBeNull()
-    expect(nodes[0]!.next).toBe(2) // assistant/message (seq 2)
-    expect(nodes[1]!.seq).toBe(2)
-    expect(nodes[1]!.prev).toBe(1)
-    expect(nodes[1]!.next).toBeNull()
+    expect(nodes).toEqual([1, 2])
   })
 
   it('empty surface yields empty nodes', () => {
@@ -99,9 +205,7 @@ describe('SurfaceManager', () => {
     // Append another surface node
     s.append('tool/result', { turn: 1, step: 1, callId: CallId('c1'), content: [{ type: 'text', text: 'ok' }], isError: false }, { surfaceOp: 'append' })
     expect(s.surface.nodes.length).toBe(3)
-    expect(s.surface.nodes[2]!.seq).toBe(4) // seq 4: after turn/end at seq 3
-    expect(s.surface.nodes[2]!.prev).toBe(2)
-    expect(s.surface.nodes[1]!.next).toBe(4)
+    expect(s.surface.nodes[2]!).toBe(4) // seq 4: after turn/end at seq 3
   })
 
   it('replays identically from a seeded log with surface markers', () => {
@@ -109,21 +213,17 @@ describe('SurfaceManager', () => {
     original.append('tool/result', { turn: 1, step: 1, callId: CallId('c1'), content: [{ type: 'text', text: 'ok' }], isError: false }, { surfaceOp: 'append' })
     const replayed = new Session(SessionId('replay'), [...original.events])
     // Surface rebuilds from the seeded log's markers.
-    expect(replayed.surface.nodes.map(n => n.seq)).toEqual([1, 2, 4])
+    expect(replayed.surface.nodes).toEqual([1, 2, 4])
     expect(replayed.deriveMessages()).toEqual(original.deriveMessages())
   })
 
   it('rebuild with replace operation splices out shadowed nodes', () => {
     const s = surfaceSession()
-    // Replace surface seqs 1 (user) and 2 (assistant) with the summary.
     s.append('assistant/message',
-      { turn: 2, step: 1, content: [{ type: 'text', text: 'summary' }] },
+      { provenance: { provider: 'mock', model: 'mock' }, turn: 2, step: 1, content: [{ type: 'text', text: 'summary' }] },
       { surfaceOp: { op: 'replace', start: 1, end: 2 }, sourceEventSeqs: [1, 2] },
     )
-    expect(s.surface.nodes.length).toBe(1)
-    expect(s.surface.nodes[0]!.seq).toBe(4) // seq of the compaction marker
-    expect(s.surface.nodes[0]!.prev).toBeNull()
-    expect(s.surface.nodes[0]!.next).toBeNull()
+    expect(s.surface.nodes).toEqual([4])
   })
 
   it('replace with both ends at real nodes splices only the range', () => {
@@ -133,15 +233,10 @@ describe('SurfaceManager', () => {
     s.append('user/message', { content: [{ type: 'text', text: 'c' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 2
     // Replace seq 0 through 1 inclusive: shadow a and b, keep c.
     s.append('assistant/message',
-      { turn: 1, step: 1, content: [{ type: 'text', text: 'summary' }] },
+      { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'summary' }] },
       { surfaceOp: { op: 'replace', start: 0, end: 1 }, sourceEventSeqs: [0, 1] },
     ) // seq 3
-    expect(s.surface.nodes.map(n => n.seq)).toEqual([3, 2])
-    // Links: 3 ↔ 2
-    expect(s.surface.nodes[0]!.prev).toBeNull()
-    expect(s.surface.nodes[0]!.next).toBe(2)
-    expect(s.surface.nodes[1]!.prev).toBe(3)
-    expect(s.surface.nodes[1]!.next).toBeNull()
+    expect(s.surface.nodes).toEqual([3, 2])
   })
 
   it('single-node replacement (start === end)', () => {
@@ -150,32 +245,28 @@ describe('SurfaceManager', () => {
     s.append('user/message', { content: [{ type: 'text', text: 'b' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 1
     // Replace only seq 1 (single node).
     s.append('assistant/message',
-      { turn: 1, step: 1, content: [{ type: 'text', text: 'x' }] },
+      { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'x' }] },
       { surfaceOp: { op: 'replace', start: 1, end: 1 }, sourceEventSeqs: [1] },
     ) // seq 2
-    expect(s.surface.nodes.map(n => n.seq)).toEqual([0, 2])
-    expect(s.surface.nodes[0]!.next).toBe(2)
-    expect(s.surface.nodes[1]!.prev).toBe(0)
+    expect(s.surface.nodes).toEqual([0, 2])
   })
 
   it('throws when replace start is not found', () => {
     const s = new Session(SessionId('bad-start'))
     s.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 0
-    s.append('assistant/message',
-      { turn: 1, step: 1, content: [{ type: 'text', text: 'y' }] },
-      { surfaceOp: { op: 'replace', start: 5, end: 0 }, sourceEventSeqs: [5, 0] },
-    )
-    expect(() => s.surface.nodes).toThrow(/surface replace: start seq 5 not found/)
+    expect(() => s.append('assistant/message',
+      { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'y' }] },
+      { surfaceOp: { op: 'replace', start: 5, end: 0 }, sourceEventSeqs: [0] },
+    )).toThrow(/surface replace: start seq 5 not found/)
   })
 
   it('throws when replace end is not found', () => {
     const s = new Session(SessionId('bad-end'))
     s.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 0
-    s.append('assistant/message',
-      { turn: 1, step: 1, content: [{ type: 'text', text: 'y' }] },
+    expect(() => s.append('assistant/message',
+      { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'y' }] },
       { surfaceOp: { op: 'replace', start: 0, end: 99 }, sourceEventSeqs: [0] },
-    )
-    expect(() => s.surface.nodes).toThrow(/surface replace: end seq 99 not found/)
+    )).toThrow(/surface replace: end seq 99 not found/)
   })
 
   it('throws when start is after end', () => {
@@ -183,49 +274,42 @@ describe('SurfaceManager', () => {
     s.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 0
     s.append('user/message', { content: [{ type: 'text', text: 'b' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 1
     // start=1, end=0 would be reversed order.
-    s.append('assistant/message',
-      { turn: 1, step: 1, content: [{ type: 'text', text: 'y' }] },
+    expect(() => s.append('assistant/message',
+      { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'y' }] },
       { surfaceOp: { op: 'replace', start: 1, end: 0 }, sourceEventSeqs: [1, 0] },
-    )
-    expect(() => s.surface.nodes).toThrow(/start seq 1.*after end seq 0/)
+    )).toThrow(/start seq 1.*after end seq 0/)
   })
 
   it('sourceEventSeqs is snapshot so caller mutation does not affect logged event', () => {
     const s = new Session(SessionId('immutable'))
-    const sources = [10, 20]
-    s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'h' }] }, { surfaceOp: 'append', sourceEventSeqs: sources })
+    s.append('user/message', { content: [{ type: 'text', text: 'source' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
+    const sources = [0]
+    s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'h' }] }, { surfaceOp: 'append', sourceEventSeqs: sources })
     // Mutate caller's array after append.
-    sources.push(30)
+    sources.push(1)
     sources[0] = 99
-    const logged = s.events[0]! as SurfaceEvent
-    expect(logged.sourceEventSeqs).toEqual([10, 20])
+    const logged = s.events[1]! as SurfaceEvent
+    expect(logged.sourceEventSeqs).toEqual([0])
   })
 
-  it('replace starting at non-head position links to previous node correctly', () => {
+  it('replace starting at non-head position preserves surrounding order', () => {
     const s = new Session(SessionId('mid-replace'))
     s.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 0
     s.append('user/message', { content: [{ type: 'text', text: 'b' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 1
     s.append('user/message', { content: [{ type: 'text', text: 'c' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 2
     // Replace the middle node (seq 1) only, keeping seq 0 and seq 2.
     s.append('assistant/message',
-      { turn: 1, step: 1, content: [{ type: 'text', text: 'x' }] },
+      { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'x' }] },
       { surfaceOp: { op: 'replace', start: 1, end: 1 }, sourceEventSeqs: [1] },
     ) // seq 3
-    expect(s.surface.nodes.map(n => n.seq)).toEqual([0, 3, 2])
-    // Links: 0 → 3 → 2
-    expect(s.surface.nodes[0]!.prev).toBeNull()
-    expect(s.surface.nodes[0]!.next).toBe(3)
-    expect(s.surface.nodes[1]!.prev).toBe(0)
-    expect(s.surface.nodes[1]!.next).toBe(2)
-    expect(s.surface.nodes[2]!.prev).toBe(3)
-    expect(s.surface.nodes[2]!.next).toBeNull()
+    expect(s.surface.nodes).toEqual([0, 3, 2])
   })
 
   it('surfaceOp replace object is snapshot so caller mutation is isolated', () => {
     const s = new Session(SessionId('immutable-op'))
     s.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
     const op = { op: 'replace' as const, start: 0, end: 0 }
-    s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 's' }] }, { surfaceOp: op, sourceEventSeqs: [0] })
+    s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 's' }] }, { surfaceOp: op, sourceEventSeqs: [0] })
     // Mutate caller's object after append.
     op.start = 99
     const logged = s.events[1]! as SurfaceEvent
@@ -250,7 +334,7 @@ describe('deriveMessages with surface', () => {
     s.append('assistant/chunk', { turn: 1, step: 1, chunk: { type: 'text-delta', index: 0, text: 'h' } })
     s.append('assistant/chunk', { turn: 1, step: 1, chunk: { type: 'text-delta', index: 1, text: 'i' } })
     s.append('user/message', { content: [{ type: 'text', text: 'hello' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'hi' }] }, { surfaceOp: 'append' })
+    s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'hi' }] }, { surfaceOp: 'append' })
     s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
     // Chunks and boundaries are NOT in the surface, so only 2 messages.
     expect(s.deriveMessages()).toHaveLength(2)
@@ -259,7 +343,7 @@ describe('deriveMessages with surface', () => {
   it('deriveMessages via surface respects replace (shadowed nodes are excluded)', () => {
     const s = new Session(SessionId('compacted'))
     s.append('user/message', { content: [{ type: 'text', text: 'original' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'compacted' }] }, { surfaceOp: { op: 'replace', start: 0, end: 0 }, sourceEventSeqs: [0] })
+    s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'compacted' }] }, { surfaceOp: { op: 'replace', start: 0, end: 0 }, sourceEventSeqs: [0] })
     // Only the compaction node is visible.
     const messages = s.deriveMessages()
     expect(messages).toHaveLength(1)
@@ -280,15 +364,17 @@ describe('deriveMessages with surface', () => {
 describe('Session.append surface opts', () => {
   it('records sourceEventSeqs and surfaceOp on the event', () => {
     const s = new Session(SessionId('opts'))
+    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
+    s.append('step/start', { turn: 1, step: 1 })
     const event = s.append('assistant/message',
-      { turn: 1, step: 1, content: [{ type: 'text', text: 'h' }] },
-      { surfaceOp: 'append', sourceEventSeqs: [3, 5, 7] },
+      { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'h' }] },
+      { surfaceOp: 'append', sourceEventSeqs: [0, 1] },
     )
-    expect(event.sourceEventSeqs).toEqual([3, 5, 7])
+    expect(event.sourceEventSeqs).toEqual([0, 1])
     expect(event.surfaceOp).toBe('append')
     // The logged event matches the returned event.
-    expect((s.events[0]! as SurfaceEvent).sourceEventSeqs).toEqual([3, 5, 7])
-    expect((s.events[0]! as SurfaceEvent).surfaceOp).toBe('append')
+    expect((s.events[2]! as SurfaceEvent).sourceEventSeqs).toEqual([0, 1])
+    expect((s.events[2]! as SurfaceEvent).surfaceOp).toBe('append')
   })
 
   it('deriveMessages skips a surface node that derives to null (empty assistant/message)', () => {
@@ -298,7 +384,7 @@ describe('Session.append surface opts', () => {
     const seed: SessionEvent[] = [
       { type: 'turn/start', seq: 0, time: 1, data: { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } } },
       { type: 'step/start', seq: 1, time: 2, data: { turn: 1, step: 1 } },
-      { type: 'assistant/message', seq: 2, time: 3, data: { turn: 1, step: 1, content: [] }, surfaceOp: 'append' },
+      { type: 'assistant/message', seq: 2, time: 3, data: { turn: 1, step: 1, content: [], provenance: { provider: 'mock', model: 'mock' } }, surfaceOp: 'append' },
       { type: 'step/end', seq: 3, time: 4, data: { turn: 1, step: 1 } },
       { type: 'turn/end', seq: 4, time: 5, data: { turn: 1, reason: { kind: 'completed' } } },
     ]
@@ -316,7 +402,7 @@ describe('Session.append surface opts', () => {
 
   it('surfaceOp primitives are not cloned (they are immutable)', () => {
     const s = new Session(SessionId('prim'))
-    const event = s.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append' })
+    const event = s.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: 'append' })
     // The string 'append' is a primitive — identity-preserving is fine.
     expect(event.surfaceOp).toBe('append')
   })
@@ -390,7 +476,7 @@ describe('SurfaceManager.replaceGeneration', () => {
     const nodes = s.surface.nodes
     s.append('context/message', {
       content: [{ type: 'text', text: 'summary' }], source: { kind: 'plugin', plugin: 'compact' },
-    }, { surfaceOp: { op: 'replace', start: nodes[0]!.seq, end: nodes[1]!.seq }, sourceEventSeqs: [nodes[0]!.seq, nodes[1]!.seq] })
+    }, { surfaceOp: { op: 'replace', start: nodes[0]!, end: nodes[1]! }, sourceEventSeqs: [nodes[0]!, nodes[1]!] })
     expect(s.surface.replaceGeneration).toBe(1)
   })
 })
diff --git a/packages/core/session/tests/tool-pairing.spec.ts b/packages/core/session/tests/tool-pairing.spec.ts
deleted file mode 100644
index eb7b1a6203..0000000000
--- a/packages/core/session/tests/tool-pairing.spec.ts
+++ /dev/null
@@ -1,292 +0,0 @@
-import { describe, expect, it } from 'vitest'
-import { CallId } from '@deepseek-ai/dsh-llm'
-import { Session, SessionId, isToolPairingBalanced } from '../src/index.ts'
-import type { SessionEvent, SurfaceNode } from '../src/index.ts'
-
-/**
- * Unit coverage for compaction-cut safety: a cut is balanced only when it
- * separates no assistant tool call from its result. Non-step nodes are neutral,
- * and replace operations prove surface order—not raw log order—is authoritative.
- */
-
-const SURFACE = { surfaceOp: 'append' as const }
-
-/** Surface nodes + log for a session, the two args the balance check takes. */
-function surfaceOf(session: Session): { nodes: readonly SurfaceNode[]; events: readonly SessionEvent[] } {
-  return { nodes: session.surface.nodes, events: session.events }
-}
-
-/** The cut BEFORE the surface node at `seq` is balanced (safe region start). */
-function startBalanced(session: Session, seq: number): boolean {
-  const { nodes, events } = surfaceOf(session)
-  return isToolPairingBalanced(nodes, events, seq)
-}
-
-/** The cut AFTER the surface node at `seq` is balanced (safe region end). */
-function endBalanced(session: Session, seq: number): boolean {
-  const { nodes, events } = surfaceOf(session)
-  const node = nodes.find(n => n.seq === seq)
-  if (!node) throw new Error(`seq ${seq} is not a surface node`)
-  return isToolPairingBalanced(nodes, events, node.next)
-}
-
-/** Surface seq of the nth (0-based) event of a given type. */
-function seqOf(s: Session, type: SessionEvent['type'], nth = 0): number {
-  return s.events.filter(e => e.type === type)[nth]!.seq
-}
-
-/** A closed turn with one closed step holding an assistant + its tool result. */
-function toolStepSession(): Session {
-  const s = new Session(SessionId('tool-step'))
-  s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-  s.append('user/message', { content: [{ type: 'text', text: 'go' }], source: { kind: 'user' } }, SURFACE)
-  s.append('step/start', { turn: 1, step: 1 })
-  s.append('assistant/message', {
-    turn: 1, step: 1,
-    content: [
-      { type: 'text', text: 'calling' },
-      { type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{}' },
-    ],
-  }, SURFACE)
-  s.append('tool/call', { turn: 1, step: 1, callId: CallId('c1'), name: 'bash', arguments: '{}' })
-  s.append('tool/result', { turn: 1, step: 1, callId: CallId('c1'), content: [{ type: 'text', text: 'out' }], isError: false }, SURFACE)
-  s.append('step/end', { turn: 1, step: 1 })
-  s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-  return s
-}
-
-describe('isToolPairingBalanced — region START (cut before a node)', () => {
-  it('is true for a pre-step user/message (belongs to no step)', () => {
-    const s = toolStepSession()
-    expect(startBalanced(s, seqOf(s, 'user/message'))).toBe(true)
-  })
-
-  it('is true for the first surface node of a step (the assistant/message)', () => {
-    // The cut before the assistant is balanced — nothing unanswered precedes it.
-    const s = toolStepSession()
-    expect(startBalanced(s, seqOf(s, 'assistant/message'))).toBe(true)
-  })
-
-  it('is false for a tool/result whose assistant/message precedes it in the same step', () => {
-    // The cut before the tool/result has one unanswered tool-call (the
-    // assistant's) → starting the region here would orphan that call.
-    const s = toolStepSession()
-    expect(startBalanced(s, seqOf(s, 'tool/result'))).toBe(false)
-  })
-
-  it('is true at the surface head (nothing precedes)', () => {
-    const s = new Session(SessionId('lone'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('user/message', { content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } }, SURFACE)
-    expect(startBalanced(s, seqOf(s, 'user/message'))).toBe(true)
-  })
-})
-
-describe('isToolPairingBalanced — region END (cut after a node)', () => {
-  it('is true for the last surface node of a closed step (the tool/result)', () => {
-    // After the tool/result the assistant's single call is answered → balanced.
-    const s = toolStepSession()
-    expect(endBalanced(s, seqOf(s, 'tool/result'))).toBe(true)
-  })
-
-  it('is false for an assistant/message with a later tool/result in the same step', () => {
-    // After the assistant its tool-call is still unanswered → ending here strands
-    // the result.
-    const s = toolStepSession()
-    expect(endBalanced(s, seqOf(s, 'assistant/message'))).toBe(false)
-  })
-
-  it('is true for a pre-step user/message', () => {
-    const s = toolStepSession()
-    expect(endBalanced(s, seqOf(s, 'user/message'))).toBe(true)
-  })
-
-  it('is false at the tail when the node is inside an open (unclosed) step', () => {
-    // step/start then an assistant tool-call, but no tool/result yet (mid-flight).
-    // The after-tail cut still has one unanswered call → not balanced.
-    const s = new Session(SessionId('open-step'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('assistant/message', {
-      turn: 1, step: 1,
-      content: [{ type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{}' }],
-    }, SURFACE)
-    expect(endBalanced(s, seqOf(s, 'assistant/message'))).toBe(false)
-  })
-
-  it('is true at the tail when the node is a trailing inter-step node (step already closed)', () => {
-    // A steering message appended after step/end, at the tail. The prior step's
-    // pair is balanced and steering is neutral → the after-tail cut is balanced.
-    const s = new Session(SessionId('trailing-steer'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'a' }] }, SURFACE)
-    s.append('step/end', { turn: 1, step: 1 })
-    s.append('steering/message', { turn: 1, content: [{ type: 'text', text: 's' }], source: { kind: 'user' } }, SURFACE)
-    expect(endBalanced(s, seqOf(s, 'steering/message'))).toBe(true)
-  })
-
-  it('is true at the tail when no step ever opened', () => {
-    const s = new Session(SessionId('no-step'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('user/message', { content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } }, SURFACE)
-    expect(endBalanced(s, seqOf(s, 'user/message'))).toBe(true)
-  })
-})
-
-describe('isToolPairingBalanced — multiple tool calls in one assistant message', () => {
-  // An assistant message with two tool-calls needs BOTH results before the cut
-  // after it is balanced — depth +2, then -1, -1.
-  function twoCallStep(): Session {
-    const s = new Session(SessionId('two-call'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('assistant/message', {
-      turn: 1, step: 1,
-      content: [
-        { type: 'tool-call', id: CallId('c1'), name: 'a', arguments: '{}' },
-        { type: 'tool-call', id: CallId('c2'), name: 'b', arguments: '{}' },
-      ],
-    }, SURFACE)
-    s.append('tool/result', { turn: 1, step: 1, callId: CallId('c1'), content: [{ type: 'text', text: '1' }], isError: false }, SURFACE)
-    s.append('tool/result', { turn: 1, step: 1, callId: CallId('c2'), content: [{ type: 'text', text: '2' }], isError: false }, SURFACE)
-    s.append('step/end', { turn: 1, step: 1 })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    return s
-  }
-
-  it('is unbalanced after the first of two results (one call still open)', () => {
-    const s = twoCallStep()
-    expect(endBalanced(s, seqOf(s, 'tool/result', 0))).toBe(false)
-  })
-
-  it('is balanced after the second result (both calls answered)', () => {
-    const s = twoCallStep()
-    expect(endBalanced(s, seqOf(s, 'tool/result', 1))).toBe(true)
-  })
-})
-
-describe('isToolPairingBalanced — a mid-step injection context/message', () => {
-  // The injected context is pairing-neutral, but both adjacent cuts remain
-  // unbalanced because the tool call is still open across them.
-  function midStepInjection(): Session {
-    const s = new Session(SessionId('mid-inject'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('assistant/message', {
-      turn: 1, step: 1,
-      content: [{ type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{}' }],
-    }, SURFACE)
-    s.append('context/message', { content: [{ type: 'text', text: 'bg task done' }], source: { kind: 'plugin', plugin: 'tool-bash' } }, SURFACE)
-    s.append('tool/result', { turn: 1, step: 1, callId: CallId('c1'), content: [{ type: 'text', text: 'out' }], isError: false }, SURFACE)
-    s.append('step/end', { turn: 1, step: 1 })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    return s
-  }
-
-  it('start cut before the mid-step context/message is unbalanced (call still open)', () => {
-    const s = midStepInjection()
-    expect(startBalanced(s, seqOf(s, 'context/message'))).toBe(false)
-  })
-
-  it('end cut after the mid-step context/message is unbalanced (call still open)', () => {
-    const s = midStepInjection()
-    expect(endBalanced(s, seqOf(s, 'context/message'))).toBe(false)
-  })
-})
-
-describe('isToolPairingBalanced on an injection turn (no step)', () => {
-  // An idle inject() wraps a context/message in a bare turn/start →
-  // context/message → turn/end with NO step. The context node is a free boundary
-  // both ways (pairing-neutral, nothing open around it).
-  function injectionSession(): Session {
-    const s = new Session(SessionId('injection'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'injection', source: { kind: 'user' } } })
-    s.append('context/message', { content: [{ type: 'text', text: 'ctx' }], source: { kind: 'user' } }, SURFACE)
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    return s
-  }
-
-  it('start: balanced', () => {
-    const s = injectionSession()
-    expect(startBalanced(s, seqOf(s, 'context/message'))).toBe(true)
-  })
-
-  it('end: balanced', () => {
-    const s = injectionSession()
-    expect(endBalanced(s, seqOf(s, 'context/message'))).toBe(true)
-  })
-})
-
-describe('isToolPairingBalanced — CBR-001: a head checkpoint left by a replace op', () => {
-  // A replacement checkpoint has a high log seq but sits at the surface head;
-  // its cuts are balanced regardless of later raw-log neighbors.
-  function checkpointHeadedSession(): Session {
-    const s = new Session(SessionId('checkpoint'))
-    // A closed turn with a tool step → surface [u1, asst(call), result].
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('user/message', { content: [{ type: 'text', text: 'u1' }], source: { kind: 'user' } }, SURFACE)
-    s.append('assistant/message', {
-      turn: 1, step: 1,
-      content: [{ type: 'tool-call', id: CallId('c1'), name: 'bash', arguments: '{}' }],
-    }, SURFACE)
-    s.append('tool/result', { turn: 1, step: 1, callId: CallId('c1'), content: [{ type: 'text', text: 'out' }], isError: false }, SURFACE)
-    s.append('step/end', { turn: 1, step: 1 })
-    s.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-    // An OPEN turn whose step is in progress (loop fires compaction here).
-    s.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 2, step: 1 })
-    // Compaction replaces the whole turn-1 surface ([u1, asst, result]) with one
-    // summary user/message — appended now, so it carries a high log seq.
-    const u1 = seqOf(s, 'user/message')
-    const result = s.events.find(e => e.type === 'tool/result')!.seq
-    s.append('user/message', {
-      content: [{ type: 'text', text: 'CHECKPOINT' }],
-      source: { kind: 'plugin', plugin: 'compact' },
-    }, { surfaceOp: { op: 'replace', start: u1, end: result } })
-    // The step's own assistant/message lands AFTER the checkpoint in the log,
-    // still inside the open step.
-    s.append('assistant/message', { turn: 2, step: 1, content: [{ type: 'text', text: 'a2' }] }, SURFACE)
-    return s
-  }
-
-  it('the head checkpoint sits at the surface head while a later surface node follows it in the log', () => {
-    const s = checkpointHeadedSession()
-    const nodes = s.surface.nodes
-    const checkpointSeq = nodes[0]!.seq
-    // The checkpoint heads the surface, yet a surface node (the open step's
-    // assistant) follows it in LOG order — the exact split between surface
-    // position and log position that the log-position scan tripped on.
-    const laterSurfaceInLog = s.events.find(
-      e => e.seq > checkpointSeq && nodes.some(n => n.seq === e.seq),
-    )
-    expect(laterSurfaceInLog).toBeDefined()
-    expect(nodes[0]!.seq).toBe(checkpointSeq)
-  })
-
-  it('start cut before the head checkpoint is balanced (it is the head)', () => {
-    const s = checkpointHeadedSession()
-    expect(startBalanced(s, s.surface.nodes[0]!.seq)).toBe(true)
-  })
-
-  it('end cut after the head checkpoint is balanced (it carries no tool pair)', () => {
-    // This is the exact assertion the log-position scan failed: the forward log scan from the
-    // checkpoint reached the open step's assistant/message and wrongly reported mid-step.
-    const s = checkpointHeadedSession()
-    expect(endBalanced(s, s.surface.nodes[0]!.seq)).toBe(true)
-  })
-})
-
-describe('isToolPairingBalanced — corrupt surface guard', () => {
-  it('throws when a tool/result has no preceding tool-call (depth goes negative)', () => {
-    // A surface that opens with a tool/result (no assistant call before it) is
-    // structurally corrupt — surfaced loudly rather than mis-classified.
-    const s = new Session(SessionId('corrupt'))
-    s.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    s.append('step/start', { turn: 1, step: 1 })
-    s.append('tool/result', { turn: 1, step: 1, callId: CallId('c1'), content: [{ type: 'text', text: 'x' }], isError: false }, SURFACE)
-    const { nodes, events } = surfaceOf(s)
-    expect(() => isToolPairingBalanced(nodes, events, null)).toThrow(/no matching tool-call/)
-  })
-})
diff --git a/packages/core/tools/README.md b/packages/core/tools/README.md
index 0f7974e0f9..cfe2fdf2fa 100644
--- a/packages/core/tools/README.md
+++ b/packages/core/tools/README.md
@@ -21,6 +21,7 @@ tools:
 - `ctx.tools.schemas(scope?: ScopeKey): ToolSchema[]` Schemas of everything the scope can see (without the `execute` functions). The shipped tools' schemas are catalogued in [docs/tool-catalog.md](../../../docs/tool-catalog.md), generated by booting each tool plugin and harvesting this method (see [the tool-schema-catalog RFC](../../../docs/rfc/implemented/process/2026-07-02-tool-schema-catalog.md)).
 - `ctx.tools.guard(guard: ToolGuard): () => void` Register a monotonic synchronous execution guard after `tools/pre-execute`: returning a reason denies the call, while `undefined` leaves it unchanged. A plain-context guard applies globally; an `agent.ctx` guard applies only to that agent. Later waterfall listeners cannot turn a guard denial back into permission. Disposed with the calling fiber.
 - `ctx.tools.execute(exec)` losslessly snapshots and freezes arguments, assigns an opaque token, runs the complete policy/dispatch/result pipeline, then independently snapshots the authoritative outcome before final observation. Invalid arguments use the same result path without reaching policy or the body; around wrappers may replace only `signal`.
+- `ctx.tools.executionMode(exec)` returns `parallel` only when the visible definition's `isConcurrencySafe(exec.arguments)` classifier returns exactly `true`; unknown, hidden, undeclared, invalid, or throwing classifications are exclusive.
 
 ### Injected services
 
@@ -32,13 +33,14 @@ The live registry pipeline has three transformable waterfalls followed by the ob
 
 ### Key types
 
-- `ToolDefinition` — `ToolSchema` + `execute(args, exec)`, optional presentation callbacks, and cooperative `timeoutMs`.
+- `ToolDefinition` — `ToolSchema` + `execute(args, exec)`, optional presentation callbacks, cooperative `timeoutMs`, and optional per-call `isConcurrencySafe(args)` classification.
 - `ToolExecutionInput` — the caller-supplied call description: `{ callId, name, arguments, agent?, parent?, signal? }`; callers may pass an enclosing execution's opaque token as `parent` but never choose the new execution's own token.
 - `ToolExecutionToken` — a fresh branded `Symbol` assigned by the registry. It supports equality correlation only and never crosses a model, log, or worker boundary.
 - `ToolExecution` — the pipeline-owned call: immutable `{ token, callId, name, arguments, agent?, parent? }` identity plus optional operational `signal`, which an around wrapper may add, replace, remove, and restore. A nested call's `parent` is a `ToolExecutionToken`, not an execution object.
-- `ToolExecutionResult` — losslessly JSON-serializable outcome: `{ content, isError, error?, additionalContext?, meta? }`. Call identity stays on the immutable `ToolExecution` supplied alongside the result instead of being duplicated on the outcome. The registry materializes and freezes the complete post-policy value before final observation. On failure with a `HarnessError`, `error: { name, code }` carries the structured failure class alongside the model-facing text.
+- `ToolRunContext` — the execution passed to a tool body, extending `ToolExecution` with `deferContext(context)`. Composite tools use it to ferry context produced by nested dispatches to the outer result even when the tool later throws; it never injects immediately.
+- `ToolExecutionResult` — losslessly JSON-serializable outcome: `{ content, isError, error?, additionalContexts?, meta? }`. Call identity stays on the immutable `ToolExecution` supplied alongside the result instead of being duplicated on the outcome. The registry materializes and freezes the complete post-policy value before final observation. On failure with a `HarnessError`, `error: { name, code }` carries the structured failure class alongside the model-facing text. `additionalContexts` preserves each deferred or post-execute `HookContext` with its own source, envelope, and durable JSON metadata; the loop buffers the array and appends each entry as a `context/message` after all `tool/result`s in the step.
 - `PreToolDecision` — `{kind:'allow'}` | `{kind:'deny', reason}` | `{kind:'ask', reason?}`. Input rewrite is deliberately not offered; `ask` is serviced by [`ctx.approval`](../../ui/user-approval/README.md) when mounted and otherwise degrades to deny.
-- `PostToolDecision` — `{kind:'accept', content?, additionalContext?}` (keep the call successful, optionally replacing the model-facing content) | `{kind:'block', feedback, additionalContext?}` (turn it into an `isError` whose content is the corrective feedback). Output replacement is clean because `tool/result` is logged AFTER `execute()` returns.
+- `PostToolDecision` — `{kind:'accept', content?, additionalContexts?}` (keep the call successful, optionally replacing the model-facing content) | `{kind:'block', feedback, additionalContexts?}` (turn it into an `isError` whose content is the corrective feedback). Accept preserves tool-deferred contexts before decision contexts; block discards tool-deferred contexts and exposes only contexts explicitly supplied by the blocking decision.
 - `ToolGuard` — `(execution) => string | undefined`; the returned string is a final monotonic denial reason evaluated after the reorderable pre-execute waterfall and before dispatch.
 - `ToolCallView` / `ToolResultView` — provider-neutral `card`-tagged render intents a tool returns from `presentCall` / `presentResult` to own how a UI renders ITS calls (see "Tool-owned UI presentation").
 
@@ -47,7 +49,7 @@ The live registry pipeline has three transformable waterfalls followed by the ob
 - Tool plugins call `ctx.tools.register()` — schemas flow into the assembly automatically.
 - `tools/pre-execute` is the reorderable allow/deny/ask gate; `ctx.tools.guard()` adds monotonic owner policy after it.
 - `tools/execute` wraps normalized core dispatch for timeout, retry, or metrics. Wrappers may replace only the operational signal.
-- `tools/post-execute` may replace content, block with feedback, or attach context; `tools/result` observes the immutable final outcome.
+- `tools/post-execute` may replace content, block with feedback, or attach ordered contexts; `tools/result` observes the immutable final outcome.
 - Exact signatures and ordering live in the generated [event catalog](../../../docs/cordis-catalog/events.md) and [pipeline](../../../docs/tool-execution-pipeline.md).
 - MCP servers: one plugin per server, discover tools, call `ctx.tools.register()` with the server's schemas.
 
@@ -86,6 +88,8 @@ See `defineTool`, `validateArgs`, `ToolArgsError`, `SchemaSpec`, `InferArgs`, an
 
 Optional `timeoutMs` must be positive and finite; it is policy metadata, not model-visible schema.
 
+Optional `isConcurrencySafe(args)` receives typed, softly validated arguments. Exact `true` permits concurrent dispatch/body execution; invalid input and all other outcomes remain exclusive. Opted-in bodies do not mutate parent-owned state, and shared-state races must commute or fail closed. The [parallel tool-call RFC](../../../docs/rfc/implemented/feature/2026-07-10-parallel-tool-call-execution.md) owns the full safety contract.
+
 ### Structured-output schema subset
 
 `StructuredOutputSchema` is the object-rooted raw JSON Schema subset used by subagents and workflows for machine-readable results. It accepts one scalar `type`, object `properties`/`required`/boolean `additionalProperties`, array `items`, and scalar `enum`/`const`. The annotations `description`, `title`, `default`, and `examples` are ignored but must remain JSON data. Type arrays, undeclared required keys, and unsupported keywords fail through `OutputSchemaError` rather than being ignored; `validateStructuredValue()` returns path-qualified violations without throwing.
@@ -101,7 +105,15 @@ Returning `undefined` selects generic fallback. Presenters depend only on their
 
 ### Code Mode
 
-Under `code` or `both`, the registry exposes the reserved `run_code` transport and a deterministic TypeScript SDK for the current scope; only program output re-enters model context. Each JSON-normalized binding re-enters the complete tool pipeline sequentially with logged correlation to the outer call. Denials reject that binding, ordinary side effects are not rolled back, and mid-run `additionalContext` is omitted to preserve call/result adjacency. Run settlement aborts and drains outstanding bindings; failures surface as `CodeRunFailedError`. See the [Code Mode RFC](../../../docs/rfc/implemented/feature/2026-06-15-code-mode.md) and [code-runtime seam](../../code-runtime/README.md). Try `pnpm run demo:code-mode`.
+Under `code` or `both`, the registry exposes the reserved `run_code` transport and a deterministic TypeScript SDK for the current scope; only program output re-enters model context. Each JSON-normalized binding re-enters the complete tool pipeline sequentially with logged correlation to the outer call. Denials reject that binding, ordinary side effects are not rolled back, and sub-call `additionalContexts` are deferred through the parent result to preserve call/result adjacency. Run settlement aborts and drains outstanding bindings; failures surface as `CodeRunFailedError`. See the [Code Mode RFC](../../../docs/rfc/implemented/feature/2026-06-15-code-mode.md) and [code-runtime seam](../../code-runtime/README.md). Try `pnpm run demo:code-mode`.
+
+- **The SDK section** (`tools:sdk`, order 150): a lazy prompt section regenerating, at each assembly, a `declare const tools: {...}` TypeScript declaration of the calling scope's visible end capabilities (exotic names via quoted keys), plus fixed usage instructions. Deterministic — lexicographic tool order, byte-identical text for an unchanged tool set (prefix-cache-friendly). The codegen (`jsonSchemaToTs`, exported) is total: constructs outside the `defineTool` subset degrade to `unknown`, never throw.
+- **The dispatch bridge** (`run_code`'s execute): every binding call is JSON-normalized before dispatch (a value that does not survive — `BigInt`, circulars — rejects that one call, so the dispatched form and logged form are the same JSON value by construction), serialized through a per-run queue (even `Promise.all` executes underlying calls one at a time in submission order), given the outer execution's opaque token as `parent`, and run through the complete pre-execute → guards → execute → post-execute → result pipeline. A denial reaches the program as a binding rejection, and each sub-call is logged as a `tool/code-dispatch` session event with deterministic id `<parent>:code:<n>`; `deriveMessages()` does not surface that event. Token correlation lets commit-style observers defer an inner success until the final `run_code` result without exposing the live outer execution; ordinary tool side effects are not rolled back. Every sub-call `additionalContexts` entry is deferred through the outer `ToolRunContext` in dispatch order; the loop appends those contexts only after the parent `run_code` result, preserving adjacency and retaining each source/envelope/meta even when the program later fails.
+- **Settlement discipline**: the bridge owns a run-scoped abort that follows the outer signal in and fires when the run settles for any reason, so a budget expiry aborts an in-flight sub-tool instead of orphaning it; the bridge then drains its queue BEFORE returning, so every `tool/code-dispatch` lands inside the open turn. A failed run throws `CodeRunFailedError` (`code: 'CODE_RUN_FAILED'`, message = the failure kind + captured logs), which the pipeline converts to a structured `isError` the model self-corrects from.
+
+### Parallel execution
+
+The agent loop groups consecutive `parallel` calls into a bounded rolling pool and treats each `exclusive` call as an ordering barrier. Only dispatch/body overlaps; policy, durable results, and context retain model order. Code Mode bindings remain serial. The [parallel tool-call RFC](../../../docs/rfc/implemented/feature/2026-07-10-parallel-tool-call-execution.md) owns the shipped declarations and rationale.
 
 ## Model Experience
 
@@ -140,7 +152,7 @@ The available tools:
 
 ## Known Limitations and Deferred Work
 
-- **Native tool calls execute sequentially** — `ToolDefinition` carries no concurrency-safety metadata; adding it (and parallel execution in the loop) waits on the deferred tool-shapes review (`TODO(review)`).
+- **Concurrency policy is not an event seam** — `executionMode()` reads the resolved tool definition directly; plugins can only declare a classifier on definitions they own.
 - **`tools/pre-execute` deliberately cannot rewrite `exec.arguments`** — logged and rendered args would desync from what ran; the rewrite design is [a proposed RFC](../../../docs/rfc/proposed/feature/2026-06-30-pre-tool-input-rewrite.md).
 - **`defineTool`'s schema DSL is a deliberate subset** — string/number/boolean/object/array with string-only `enum`; `validateArgs` tolerates extra keys and never applies `default` (`XXX(unused-default)` flags removing that field); raw-registered JSON-Schema tools validate their own input.
 - **`timeoutMs` on a definition is declarative only** — the registry never enforces deadlines; enforcement requires the `@deepseek-ai/dsh-timeout-policy` wrapper.
diff --git a/packages/core/tools/src/code-mode.ts b/packages/core/tools/src/code-mode.ts
index 012182db1b..0a65c9e434 100644
--- a/packages/core/tools/src/code-mode.ts
+++ b/packages/core/tools/src/code-mode.ts
@@ -5,6 +5,7 @@
  * @module @deepseek-ai/dsh-tools/src/code-mode
  */
 
+import { parse } from 'node:path'
 import { inspect } from 'node:util'
 import { CallId, HarnessError } from '@deepseek-ai/dsh-llm'
 import type { ContentBlock } from '@deepseek-ai/dsh-llm'
@@ -16,11 +17,19 @@ import type { ToolDefinition, ToolRegistry } from './index.ts'
 declare module '@deepseek-ai/dsh-session' {
   interface SessionEventMap {
     /**
-     * One bridged sub-dispatch from a `run_code` program: the parent `run_code` call id, the
-     * deterministic sub-call id (`<parent>:code:<n>`), the tool `name` with its
-     * JSON-normalized `arguments` — the exact value dispatched, normalized before dispatch, so
-     * this append can never fail on payload shape — whether the sub-call errored, and a
-     * bounded `resultSummary` of its model-facing text.
+     * One bridged sub-dispatch from a `run_code` program: the parent
+     * `run_code` call id, the deterministic sub-call id
+     * (`<parent>:code:<n>`), the tool `name` with its JSON-normalized
+     * `arguments` — the exact value dispatched, normalized BEFORE dispatch,
+     * so this append can never fail on payload shape — whether the sub-call
+     * errored, and a bounded `resultSummary` of its model-facing text. Before
+     * bounding, occurrences of a non-root session workspace path are
+     * normalized to `.` so host-specific absolute path lengths cannot change
+     * the summary.
+     * Log-only: `deriveMessages()` ignores it, so sub-calls never re-enter
+     * model context; persistence and UIs get every call. Appended inside the
+     * parent `run_code`'s execution (the bridge drains its queue before
+     * returning), so the turn-enclosure invariant holds by construction.
      */
     'tool/code-dispatch': { parentCallId: CallId; subCallId: CallId; name: string; arguments: unknown; isError: boolean; resultSummary: string }
   }
@@ -70,9 +79,12 @@ function textOf(content: ContentBlock[]): string {
     .join('\n')
 }
 
-/** Bound a sub-call's model-facing text for the log event's `resultSummary`. */
-function summarize(text: string): string {
-  return text.length > SUMMARY_MAX_CHARS ? `${text.slice(0, SUMMARY_MAX_CHARS)}…` : text
+/** Normalize workspace paths, then bound a sub-call's model-facing text for its durable log summary. */
+function summarize(text: string, cwd: string | undefined): string {
+  const stableText = cwd === undefined || cwd === parse(cwd).root
+    ? text
+    : text.replaceAll(cwd, '.')
+  return stableText.length > SUMMARY_MAX_CHARS ? `${stableText.slice(0, SUMMARY_MAX_CHARS)}…` : stableText
 }
 
 /**
@@ -189,10 +201,10 @@ export function createRunCodeTool(registry: ToolRegistry, requireRuntime: () =>
             parent: exec.token,
             signal: runController.signal,
           })
+          for (const context of result.additionalContexts ?? []) {
+            exec.deferContext(context)
+          }
           const text = textOf(result.content)
-          // Sub-call `additionalContext` is deliberately DROPPED here: the loop's buffering
-          // (append after the step's tool/results) has no safe analogue from inside a running
-          // run_code — injecting now would break tool-call/result adjacency.
           exec.agent?.session.append('tool/code-dispatch', {
             parentCallId: exec.callId,
             subCallId,
@@ -202,7 +214,7 @@ export function createRunCodeTool(registry: ToolRegistry, requireRuntime: () =>
             // this record from what it actually received.
             arguments: normalized.logged,
             isError: result.isError,
-            resultSummary: summarize(text),
+            resultSummary: summarize(text, exec.agent.session.header.cwd),
           })
           return { text, isError: result.isError }
         })
diff --git a/packages/core/tools/src/index.ts b/packages/core/tools/src/index.ts
index d0c75e8035..5fc045ace2 100644
--- a/packages/core/tools/src/index.ts
+++ b/packages/core/tools/src/index.ts
@@ -117,15 +117,12 @@ declare module 'cordis' {
   }
 }
 
-// TODO(concurrency): revisit these shapes when concurrency metadata becomes useful
-// (for example, a read-only hint that would permit safe parallel execution).
-
 /** Tool output, optionally with lossless-JSON presentation metadata persisted for replay. */
 export type ToolExecuteReturn = ContentBlock[] | { content: ContentBlock[]; meta?: unknown }
 
 /** A registered tool: its schema plus the execution function. */
 export interface ToolDefinition extends ToolSchema {
-  execute(args: unknown, exec: ToolExecution): Promise<ToolExecuteReturn>
+  execute(args: unknown, exec: ToolRunContext): Promise<ToolExecuteReturn>
   /**
    * Cooperative tool-call timeout budget in milliseconds. Omit for no deadline.
    * Enforced by `@deepseek-ai/dsh-timeout-policy` (a `tools/execute` wrapper); it
@@ -134,6 +131,20 @@ export interface ToolDefinition extends ToolSchema {
    * cooperative implementation that can reach quiescence when the signal aborts.
    */
   timeoutMs?: number
+  /**
+   * Pure synchronous classifier for overlap with sibling tool calls. Only
+   * `true` opts in; omission, exceptions, non-`true` returns, and invalid
+   * `defineTool` arguments are exclusive. This metadata is never model-visible.
+   *
+   * Opted-in executions must not mutate parent-owned state. Shared state must
+   * tolerate concurrent dispatch; recorder races are permitted only when they
+   * commute or fail closed. See the
+   * [parallel-tool-call RFC](../../../../docs/rfc/implemented/feature/2026-07-10-parallel-tool-call-execution.md)
+   * for the full contract.
+   * @param args - parsed arguments; `defineTool` validates before calling.
+   * @returns Whether this call may join a parallel group.
+   */
+  isConcurrencySafe?(args: unknown): boolean
   /**
    * Optional: how to present the PENDING state of one call in a UI, derived from
    * the call's `args` (parsed arguments, `unknown` — the tool validates/narrows
@@ -195,6 +206,14 @@ export interface ToolExecutionInput {
   signal?: AbortSignal
 }
 
+/**
+ * Scheduling mode for one pending call. `parallel` may overlap with siblings;
+ * `exclusive` runs alone and forms an ordering barrier.
+ */
+export type ToolExecutionMode =
+  | { kind: 'parallel' }
+  | { kind: 'exclusive' }
+
 /**
  * One pending tool call inside the registry pipeline. Parsed arguments cross
  * one lossless-JSON materialization boundary before policy and are deep-frozen;
@@ -207,6 +226,62 @@ export interface ToolExecution extends ToolExecutionInput {
   readonly token: ToolExecutionToken
 }
 
+/**
+ * Runtime context handed to a tool implementation after the registry has
+ * accepted a {@link ToolExecution}. A composite tool uses
+ * {@link deferContext} to ferry context produced by nested dispatches back to
+ * the outer result; the loop appends it only after the outer `tool/result`.
+ */
+export interface ToolRunContext extends ToolExecution {
+  /**
+   * Defer one nested-dispatch context until this tool's final result reaches
+   * the agent loop. Contexts retain their individual source, envelope, and
+   * metadata and are emitted in call order.
+   */
+  deferContext(context: HookContext): void
+}
+
+/**
+ * Scheduler-only result after ordered pre-execute and guards. A `post-result`
+ * still receives post-execute; a `final-result` bypasses it.
+ * @internal
+ */
+export type ScheduledToolPreparation =
+  | { kind: 'dispatch'; exec: ToolRunContext }
+  | { kind: 'post-result'; exec: ToolRunContext; result: ToolExecutionResult }
+  | { kind: 'final-result'; exec: ToolRunContext; result: ToolExecutionResult }
+
+/**
+ * Scheduler-only dispatch result. A `post-result` still receives post-execute;
+ * a `final-result` already matches {@link ToolRegistry.execute} failure semantics.
+ * @internal
+ */
+export type ScheduledToolDispatch =
+  | { kind: 'post-result'; result: ToolExecutionResult }
+  | { kind: 'final-result'; result: ToolExecutionResult }
+
+/**
+ * Symbol-keyed scheduler view that keeps pre/post policy ordered while
+ * overlapping dispatch. Ordinary callers use {@link ToolRegistry.execute};
+ * this is not a plugin seam.
+ * @internal
+ */
+export interface ToolRegistryScheduler {
+  /** Materialize input, run the ordered pre-execute/guard gate, and decide what stage follows. */
+  prepare(exec: ToolExecutionInput): Promise<ScheduledToolPreparation>
+  /** Run only the around-dispatch/body stage. */
+  dispatch(exec: ToolRunContext): Promise<ScheduledToolDispatch>
+  /** Run ordered post-execute finalization, then materialize and notify the final outcome. */
+  finalize(exec: ToolRunContext, result: ToolExecutionResult): Promise<ToolExecutionResult>
+  /** Materialize and notify a final outcome that must bypass post-execute. */
+  finish(exec: ToolRunContext, result: ToolExecutionResult): ToolExecutionResult
+}
+
+/**
+ * Scheduler entry point omitted from the generated named service API.
+ * @internal
+ */
+export const TOOL_REGISTRY_SCHEDULER: unique symbol = Symbol('@deepseek-ai/dsh-tools.scheduler')
 /** Structured error metadata for a failed tool call (alongside the model-facing text). */
 export interface ToolErrorInfo {
   name: string
@@ -237,10 +312,10 @@ export interface ToolExecutionResult {
    */
   error?: ToolErrorInfo
   /**
-   * Model-facing context for the next request, separate from this tool result.
-   * The loop buffers it until all step results are logged, preserving pairing.
+   * Model-facing context for the next request, separate from this tool result. The loop
+   * accepts it into the active-batch FIFO, then appends after recorded results even if interrupted.
    */
-  additionalContext?: HookContext
+  additionalContexts?: HookContext[]
   /**
    * The tool-private presentation payload from a successful `execute` (the object
    * return form). Threaded onto the `tool/result` session event and back into
@@ -266,8 +341,8 @@ export type PreToolDecision =
  * request, or block by turning corrective feedback into an error result.
  */
 export type PostToolDecision =
-  | { kind: 'accept'; content?: ContentBlock[]; additionalContext?: HookContext }
-  | { kind: 'block'; feedback: ContentBlock[]; additionalContext?: HookContext }
+  | { kind: 'accept'; content?: ContentBlock[]; additionalContexts?: HookContext[] }
+  | { kind: 'block'; feedback: ContentBlock[]; additionalContexts?: HookContext[] }
 
 /**
  * Best-effort human-readable message from an arbitrary thrown value: Error
@@ -367,6 +442,16 @@ export class ToolRegistry extends Service {
     mode: z.union(['native', 'code', 'both'] as const).default('native'),
   })
 
+  /** Internal staged view consumed by `dsh-agent-loop`'s parallel scheduler. */
+  readonly [TOOL_REGISTRY_SCHEDULER]: ToolRegistryScheduler = {
+    prepare: exec => this.prepareScheduledExecution(exec),
+    dispatch: exec => this.dispatchScheduledExecution(exec),
+    finalize: (exec, result) => this.finalizeScheduledExecution(exec, result),
+    finish: (exec, result) => this.finishScheduledExecution(exec, result),
+  }
+
+  /** Context deferred by a running tool body, keyed by its scheduler-owned execution. */
+  private deferredContexts = new WeakMap<ToolRunContext, HookContext[]>()
   private global = new Map<string, ToolDefinition>()
   private scoped = new Map<ScopeKey, Map<string, ToolDefinition>>()
   /** Compiled restriction filters, per scope (see {@link restrict}). */
@@ -667,6 +752,24 @@ export class ToolRegistry extends Service {
     }
   }
 
+  /**
+   * Classify a pending call through the caller's visible tool definition. Only
+   * an exact `true` is parallel; unknown, hidden, undeclared, invalid, or
+   * throwing classifiers are exclusive.
+   * @param exec - call name, parsed arguments, and optional agent scope.
+   * @returns the fail-closed scheduling mode.
+   */
+  executionMode(exec: ToolExecutionInput): ToolExecutionMode {
+    const tool = this.get(exec.name, exec.agent)
+    if (!tool?.isConcurrencySafe) return { kind: 'exclusive' }
+    try {
+      const concurrencySafe: unknown = tool.isConcurrencySafe(exec.arguments)
+      return concurrencySafe === true ? { kind: 'parallel' } : { kind: 'exclusive' }
+    } catch {
+      return { kind: 'exclusive' }
+    }
+  }
+
   /**
    * Execute through pre-policy, guards, around-dispatch, post-policy, and final
    * notification. Tool and listener failures resolve as materialized error
@@ -677,6 +780,29 @@ export class ToolRegistry extends Service {
    * @returns the materialized final result.
    */
   async execute(exec: ToolExecutionInput): Promise<ToolExecutionResult> {
+    return this.prepareExecution(exec, prepared => this.completeScheduledExecution(prepared))
+  }
+
+  private async completeScheduledExecution(prepared: ScheduledToolPreparation): Promise<ToolExecutionResult> {
+    switch (prepared.kind) {
+      case 'dispatch': {
+        const dispatched = await this.dispatchScheduledExecution(prepared.exec)
+        return dispatched.kind === 'post-result'
+          ? await this.finalizeScheduledExecution(prepared.exec, dispatched.result)
+          : this.finishScheduledExecution(prepared.exec, dispatched.result)
+      }
+      case 'post-result':
+        return await this.finalizeScheduledExecution(prepared.exec, prepared.result)
+      case 'final-result':
+        return this.finishScheduledExecution(prepared.exec, prepared.result)
+      /* v8 ignore next -- closed-union exhaustiveness guard */
+      default:
+        return assertNever(prepared, 'scheduled tool preparation')
+    }
+  }
+
+  private createExecution(exec: ToolExecutionInput): ScheduledToolPreparation | { kind: 'ready'; exec: ToolRunContext } {
+    const deferredContexts: HookContext[] = []
     const token = createExecutionToken()
     const callId = exec.callId
     const name = exec.name
@@ -690,97 +816,147 @@ export class ToolRegistry extends Service {
       ...agent !== undefined ? { agent } : {},
       ...parent !== undefined ? { parent } : {},
       ...signal !== undefined ? { signal } : {},
+      deferContext(context: HookContext): void {
+        deferredContexts.push(context)
+      },
     }
-    let execution: ToolExecution
     try {
       const detached = snapshotJsonValue(exec.arguments)
       if (detached === undefined) {
         throw new TypeError('tool execution arguments must be losslessly JSON-serializable')
       }
-      execution = {
-        ...base,
-        arguments: deepFreeze(detached),
-      }
+      const execution: ToolRunContext = { ...base, arguments: deepFreeze(detached) }
+      this.deferredContexts.set(execution, deferredContexts)
+      return { kind: 'ready', exec: execution }
     } catch (error: unknown) {
-      execution = { ...base, arguments: undefined }
-      const result = this.materializeFinalResult(toolErrorResult(error))
-      this.notifyResult(execution, result)
-      return result
+      const execution: ToolRunContext = { ...base, arguments: undefined }
+      return { kind: 'final-result', exec: execution, result: toolErrorResult(error) }
     }
-    let result: ToolExecutionResult
+  }
+
+  /**
+   * Run the ordered pre-execute and monotonic guard stages for the scheduler.
+   * @param input - the caller-supplied execution input.
+   * @returns the prepared execution plus the next scheduler stage.
+   * @internal
+   */
+  private async prepareScheduledExecution(input: ToolExecutionInput): Promise<ScheduledToolPreparation> {
+    return this.prepareExecution(input, prepared => prepared)
+  }
+
+  private async prepareExecution<T>(
+    input: ToolExecutionInput,
+    next: (prepared: ScheduledToolPreparation) => T | PromiseLike<T>,
+  ): Promise<T> {
+    const created = this.createExecution(input)
+    if (created.kind !== 'ready') return next(created)
+    const exec = created.exec
     try {
-      result = this.materializeFinalResult(await this.executePipeline(execution))
-    } catch (error: unknown) {
-      // Outer backstop: a throwing pre/post-execute listener, guard, or the
-      // waterfall machinery becomes an isError result, never a turn failure.
-      result = this.materializeFinalResult(toolErrorResult(error))
-    }
-    this.notifyResult(execution, result)
-    return result
-  }
-
-  /** Run the transformable pipeline; {@link execute} owns final normalization and notification. */
-  private async executePipeline(exec: ToolExecution): Promise<ToolExecutionResult> {
-    // --- Gate: tools/pre-execute. An `ask` resolves through the optional
-    // approval seam (or degrades to deny) before the monotonic guards run. The
-    // carrier keys dispatch by exec.agent, so an `agent.ctx` listener gates only
-    // its own agent's calls (agent-less calls are subject-less).
-    const carrier = scopeTarget(this, exec.agent)
-    const gate = await this.ctx.waterfall(
-      carrier, 'tools/pre-execute', exec,
-      () => Promise.resolve<PreToolDecision>({ kind: 'allow' }),
-    )
-    const decision = gate.kind === 'ask' ? await this.serviceAsk(exec, gate) : gate
-    const denialReason = decision.kind === 'allow'
-      ? this.guardReason(exec)
-      : decision.reason
-    if (denialReason !== undefined) {
-      // Every non-grant, including a failed/unavailable approval request, takes
-      // the same deny path and still reaches post-policy plus result observers.
-      const denied: ToolExecutionResult = {
-        content: [{ type: 'text', text: `Error: ${denialReason}` }],
-        isError: true,
+      const carrier = scopeTarget(this, exec.agent)
+      const gate = await this.ctx.waterfall(
+        carrier, 'tools/pre-execute', exec,
+        () => Promise.resolve<PreToolDecision>({ kind: 'allow' }),
+      )
+      const decision = gate.kind === 'ask' ? await this.serviceAsk(exec, gate) : gate
+      const denialReason = decision.kind === 'allow'
+        ? this.guardReason(exec)
+        : decision.reason
+      if (denialReason !== undefined) {
+        return await next({
+          kind: 'post-result',
+          exec,
+          result: {
+            content: [{ type: 'text', text: `Error: ${denialReason}` }],
+            isError: true,
+          },
+        })
       }
-      return await this.postExecute(exec, denied)
+      return await next({ kind: 'dispatch', exec })
+    } catch (error: unknown) {
+      return next({ kind: 'final-result', exec, result: toolErrorResult(error) })
     }
-
-    // --- Around-dispatch: tools/execute. The base `next` is the dispatch-
-    // with-normalization thunk — the tool body's own try/catch turns a throw
-    // into an isError result so a wrapper (and post-execute) can inspect it;
-    // an unknown tool routes through the same catch. A `tools/execute` listener
-    // (e.g. a timeout plugin) wraps this thunk: it may replace `exec.signal`
-    // before delegating and inspect the normalized result after. Dispatched with the
-    // same carrier as the gate, so an `agent.ctx` wrapper wraps only its own
-    // agent's calls. ---
-    const result = await this.ctx.waterfall(
-      carrier, 'tools/execute', exec,
-      async (): Promise<ToolExecutionResult> => {
-        try {
-          // Resolve through the CALLER's visible view ({@link get}): a scoped
-          // tool shadows its global name-twin for that agent, and a
-          // restricted-away global tool is exactly as absent as a nonexistent
-          // one — same UNKNOWN_TOOL result, no capability leak in the error.
-          const tool = this.get(exec.name, exec.agent)
-          if (!tool) throw new ToolNotFoundError(exec.name)
-          // Normalize the two `execute` return shapes: a bare ContentBlock[] (no
-          // meta) or a { content, meta } object (a tool attaching a private
-          // presentation payload). An array IS the content; the object carries it.
-          const returned = await tool.execute(exec.arguments, exec)
-          const content = Array.isArray(returned) ? returned : returned.content
-          const meta = Array.isArray(returned) ? undefined : returned.meta
-          return { content, isError: false, ...meta !== undefined ? { meta } : {} }
-        } catch (error: unknown) {
-          return toolErrorResult(error)
-        }
-      },
-    )
-    return await this.postExecute(exec, result)
   }
 
-  /** Notify final-result observers without giving them a mutation/error channel into the outcome. */
+  /**
+   * Run around-dispatch and the tool body. Tool and unknown-tool failures still
+   * receive post-execute; pipeline failures are already final.
+   * @param exec - the prepared execution.
+   * @returns whether the result still needs post-execute.
+   * @internal
+   */
+  private async dispatchScheduledExecution(exec: ToolRunContext): Promise<ScheduledToolDispatch> {
+    try {
+      const carrier = scopeTarget(this, exec.agent)
+      const result = await this.ctx.waterfall(
+        carrier, 'tools/execute', exec,
+        async (): Promise<ToolExecutionResult> => {
+          try {
+            const tool = this.get(exec.name, exec.agent)
+            if (!tool) throw new ToolNotFoundError(exec.name)
+            const returned = await tool.execute(exec.arguments, exec)
+            const content = Array.isArray(returned) ? returned : returned.content
+            const meta = Array.isArray(returned) ? undefined : returned.meta
+            return { content, isError: false, ...meta !== undefined ? { meta } : {} }
+          } catch (error: unknown) {
+            return toolErrorResult(error)
+          }
+        },
+      )
+      const deferredContexts = this.deferredContexts.get(exec)
+      /* v8 ignore next -- dispatch only receives executions minted by this registry's prepare stage */
+      if (deferredContexts === undefined) throw new Error('tool registry scheduler invariant violated: unprepared execution')
+      const resultWithDeferredContexts: ToolExecutionResult = deferredContexts.length === 0
+        ? result
+        : {
+          ...result,
+          additionalContexts: [
+            ...deferredContexts,
+            ...result.additionalContexts ?? [],
+          ],
+        }
+      return { kind: 'post-result', result: resultWithDeferredContexts }
+    } catch (error: unknown) {
+      return { kind: 'final-result', result: toolErrorResult(error) }
+    }
+  }
+
+  /**
+   * Run ordered post-execute, then materialize and notify the final outcome.
+   * @param exec - the prepared execution.
+   * @param result - dispatch/pre result that still needs post-execute.
+   * @returns the materialized final result.
+   * @internal
+   */
+  private async finalizeScheduledExecution(exec: ToolRunContext, result: ToolExecutionResult): Promise<ToolExecutionResult> {
+    try {
+      return this.finishScheduledExecution(exec, await this.postExecute(exec, result))
+    } catch (error: unknown) {
+      return this.finishScheduledExecution(exec, toolErrorResult(error))
+    }
+  }
+
+  /**
+   * Materialize and notify a final result that must bypass post-execute.
+   * @param exec - the prepared execution.
+   * @param result - final result.
+   * @returns the materialized final result.
+   * @internal
+   */
+  private finishScheduledExecution(exec: ToolRunContext, result: ToolExecutionResult): ToolExecutionResult {
+    let finalResult: ToolExecutionResult
+    try {
+      finalResult = this.materializeFinalResult(result)
+    } catch (error: unknown) {
+      finalResult = this.materializeFinalResult(toolErrorResult(error))
+    }
+    this.notifyResult(exec, finalResult)
+    return finalResult
+  }
+
+  /** Notify observers without exposing a mutation or error channel into the outcome. */
   private notifyResult(exec: ToolExecution, result: ToolExecutionResult): void {
-    // The pipeline is over: freeze the remaining mutable signal slot so every
-    // observer sees the SAME WeakMap-keyable execution without a mutation race.
+    // Freeze the remaining mutable signal slot before observers receive the
+    // shared WeakMap-keyable execution object.
     Object.freeze(exec)
     const callbacks = this.ctx.events.dispatch('emit', [
       scopeTarget(this, exec.agent), 'tools/result', exec, result,
@@ -836,8 +1012,11 @@ export class ToolRegistry extends Service {
    * Run the `tools/post-execute` waterfall over a dispatched `result` and apply
    * its {@link PostToolDecision}: `accept` keeps the call successful (replacing
    * `content` when given), `block` turns it into an `isError` whose content is
-   * the corrective `feedback`. Either decision may attach `additionalContext`,
-   * which is ferried on the returned result for the loop's per-step buffer.
+   * the corrective `feedback`. Either decision may attach `additionalContexts`,
+   * which are ferried on the returned result for the loop's active-batch FIFO.
+   * Context deferred by the tool body survives an accepted result but is
+   * discarded when the outer call is blocked; a block exposes only context the
+   * blocking decision explicitly supplied.
    * Runs inside `execute`'s outer try/catch (a throwing listener → isError).
    */
   private async postExecute(exec: ToolExecution, result: ToolExecutionResult): Promise<ToolExecutionResult> {
@@ -845,19 +1024,24 @@ export class ToolRegistry extends Service {
       scopeTarget(this, exec.agent), 'tools/post-execute', exec, result,
       () => Promise.resolve<PostToolDecision>({ kind: 'accept' }),
     )
-    const additionalContext = decision.additionalContext
+    const decisionContexts = decision.additionalContexts ?? []
     if (decision.kind === 'block') {
       return {
         content: decision.feedback,
         isError: true,
-        ...additionalContext ? { additionalContext } : {},
+        ...decisionContexts.length > 0 ? { additionalContexts: decisionContexts } : {},
       }
     }
-    // Accept: replace content if supplied and preserve the dispatched outcome.
+    // Accept: replace content if supplied, preserve the dispatched outcome, and
+    // append decision contexts after contexts deferred by the tool body.
+    const additionalContexts = [
+      ...result.additionalContexts ?? [],
+      ...decisionContexts,
+    ]
     return {
       ...result,
       ...decision.content ? { content: decision.content } : {},
-      ...additionalContext ? { additionalContext } : {},
+      ...additionalContexts.length > 0 ? { additionalContexts } : {},
     }
   }
 
diff --git a/packages/core/tools/src/schema.ts b/packages/core/tools/src/schema.ts
index 9e671f6c73..c61c819618 100644
--- a/packages/core/tools/src/schema.ts
+++ b/packages/core/tools/src/schema.ts
@@ -1,7 +1,7 @@
 /** Typed tool-parameter DSL with argument inference and JSON Schema output. @module dsh-tools/schema */
 
 import { assertNever, HarnessError } from '@deepseek-ai/dsh-llm'
-import type { ToolDefinition, ToolExecuteReturn, ToolExecution, ToolResult } from './index.ts'
+import type { ToolDefinition, ToolExecuteReturn, ToolRunContext, ToolResult } from './index.ts'
 import type { ToolCallView, ToolResultView } from './presentation.ts'
 
 // ---------------------------------------------------------------------------
@@ -283,13 +283,21 @@ export interface DefineToolOptions<S extends SchemaSpec> {
    * is never sent to the model.
    */
   readonly timeoutMs?: number
+  /**
+   * Optional pure synchronous classifier for sibling overlap. It receives typed
+   * arguments after soft validation; invalid input returns `false` without
+   * invoking it. See {@link ToolDefinition.isConcurrencySafe}.
+   * @param args - typed validated arguments.
+   * @returns whether this call may join a parallel group.
+   */
+  isConcurrencySafe?(args: InferArgs<S>): boolean
   /**
    * Tool execution function. `args` is typed as {@link InferArgs<S>} — zero
    * casts needed. Returns either a bare {@link ContentBlock}`[]` (model-facing
    * content only) or a `{ content, meta }` object to also attach a tool-private
    * presentation payload (see {@link ToolExecuteReturn}).
    */
-  execute(args: InferArgs<S>, exec: ToolExecution): Promise<ToolExecuteReturn>
+  execute(args: InferArgs<S>, exec: ToolRunContext): Promise<ToolExecuteReturn>
   /**
    * Optional: how to present the PENDING state of one call in a UI (an editor
    * tool-call card, a CLI log line). `args` is the typed, schema-validated
@@ -315,7 +323,7 @@ export interface DefineToolOptions<S extends SchemaSpec> {
  * @param options - the tool's name, description, typed parameter schema,
  *   execute body, and optional presenters.
  * @returns a registry-ready definition with strict execution validation and
- *   soft presenter validation for replay compatibility.
+ *   soft presenter and classifier validation for replay compatibility.
  */
 export function defineTool<S extends SchemaSpec>(options: DefineToolOptions<S>): ToolDefinition {
   // Object-literal execute methods don't use `this`; the reference is safe.
@@ -325,6 +333,8 @@ export function defineTool<S extends SchemaSpec>(options: DefineToolOptions<S>):
   const userPresentCall = options.presentCall
   // eslint-disable-next-line @typescript-eslint/unbound-method
   const userPresentResult = options.presentResult
+  // eslint-disable-next-line @typescript-eslint/unbound-method
+  const userIsConcurrencySafe = options.isConcurrencySafe
   if (options.timeoutMs !== undefined && (!Number.isFinite(options.timeoutMs) || options.timeoutMs <= 0)) {
     throw new Error(`defineTool(${options.name}): timeoutMs must be a positive finite number`)
   }
@@ -333,7 +343,7 @@ export function defineTool<S extends SchemaSpec>(options: DefineToolOptions<S>):
     description: options.description,
     parameters: schemaSpecToJsonSchema(options.parameters) as unknown as Record<string, unknown>,
     ...(options.timeoutMs !== undefined ? { timeoutMs: options.timeoutMs } : {}),
-    async execute(args: unknown, exec: ToolExecution): Promise<ToolExecuteReturn> {
+    async execute(args: unknown, exec: ToolRunContext): Promise<ToolExecuteReturn> {
       // Validate the model-generated args before the typed body runs. On
       // mismatch we throw ToolArgsError; the registry turns it into an
       // isError result so the model can self-correct. After this guard, the
@@ -359,5 +369,12 @@ export function defineTool<S extends SchemaSpec>(options: DefineToolOptions<S>):
       return userPresentResult(args as InferArgs<S>, result)
     }
   }
+  // Invalid arguments fail closed without invoking the typed classifier.
+  if (userIsConcurrencySafe) {
+    tool.isConcurrencySafe = (args: unknown): boolean => {
+      if (validateArgs(options.parameters, args).length > 0) return false
+      return userIsConcurrencySafe(args as InferArgs<S>)
+    }
+  }
   return tool
 }
diff --git a/packages/core/tools/tests/code-mode.spec.ts b/packages/core/tools/tests/code-mode.spec.ts
index 54575403cf..d4660f7ed7 100644
--- a/packages/core/tools/tests/code-mode.spec.ts
+++ b/packages/core/tools/tests/code-mode.spec.ts
@@ -82,10 +82,11 @@ function registerEcho(ctx: Context, name = 'echo'): unknown[] {
 }
 
 /** A structural fake of the owning agent: captures session appends. */
-function fakeAgent(): { agent: Agent; events: { type: string; data: unknown }[] } {
+function fakeAgent(options: { cwd?: string } = { cwd: '/workspace' }): { agent: Agent; events: { type: string; data: unknown }[] } {
   const events: { type: string; data: unknown }[] = []
   const agent = {
     session: {
+      header: options.cwd === undefined ? {} : { cwd: options.cwd },
       append: (type: string, data: unknown) => { events.push({ type, data }) },
     },
   } as unknown as Agent
@@ -477,27 +478,71 @@ describe('the run_code dispatch bridge', () => {
     expect(dispatch.arguments).toEqual({ value: 'x', when: '1970-01-01T00:00:00.000Z' })
   })
 
-  it('suppresses sub-call additionalContext (deliberately; pinned)', async () => {
+  it('defers sub-call additionalContexts onto the outer run_code result', async () => {
     const { ctx, runtime } = await setup({ mode: 'code' })
     registerEcho(ctx)
     ctx.on('tools/post-execute', (exec, _result, next): Promise<PostToolDecision> => {
       if (exec.name === 'echo') {
         return Promise.resolve({
           kind: 'accept' as const,
-          additionalContext: { content: [{ type: 'text' as const, text: 'context for the next request' }], source: { kind: 'plugin' as const, plugin: 'test' } },
+          additionalContexts: [{
+            content: [{ type: 'text' as const, text: `context for ${exec.callId}` }],
+            source: { kind: 'plugin' as const, plugin: 'test' },
+            envelope: 'raw' as const,
+            meta: { callId: exec.callId },
+          }],
         })
       }
       return next()
     })
     runtime.behavior = async (request) => {
       await request.bindings[0]!.functions.echo!({ value: 'x' })
+      await request.bindings[0]!.functions.echo!({ value: 'y' })
       return { logs: [], value: 'done' }
     }
     const result = await runCode(ctx, 'program')
     expect(result.isError).toBe(false)
-    // The sub-call's context has no safe outlet mid-run; the parent result
-    // must not carry it either.
-    expect(result.additionalContext).toBeUndefined()
+    expect(result.additionalContexts).toEqual([
+      {
+        content: [{ type: 'text', text: 'context for call-1:code:1' }],
+        source: { kind: 'plugin', plugin: 'test' },
+        envelope: 'raw',
+        meta: { callId: 'call-1:code:1' },
+      },
+      {
+        content: [{ type: 'text', text: 'context for call-1:code:2' }],
+        source: { kind: 'plugin', plugin: 'test' },
+        envelope: 'raw',
+        meta: { callId: 'call-1:code:2' },
+      },
+    ])
+  })
+
+  it('keeps sub-call contexts when run_code fails after the nested dispatch', async () => {
+    const { ctx, runtime } = await setup({ mode: 'both' })
+    registerEcho(ctx)
+    ctx.on('tools/post-execute', (exec, _result, next): Promise<PostToolDecision> => {
+      if (exec.name !== 'echo') return next()
+      return Promise.resolve({
+        kind: 'accept',
+        additionalContexts: [{
+          content: [{ type: 'text', text: 'nested context' }],
+          source: { kind: 'plugin', plugin: 'test' },
+        }],
+      })
+    })
+    runtime.behavior = async (request) => {
+      await request.bindings[0]!.functions.echo!({ value: 'x' })
+      return { logs: [], error: { kind: 'exception', message: 'program failed later' } }
+    }
+
+    const result = await runCode(ctx, 'program')
+
+    expect(result.isError).toBe(true)
+    expect(result.additionalContexts).toEqual([{
+      content: [{ type: 'text', text: 'nested context' }],
+      source: { kind: 'plugin', plugin: 'test' },
+    }])
   })
 
   it('converts a failed run into a structured isError result carrying kind, message, and captured logs', async () => {
@@ -671,6 +716,52 @@ describe('the run_code dispatch bridge', () => {
     expect(dispatch.resultSummary.endsWith('…')).toBe(true)
   })
 
+  it('normalizes the session workspace root before bounding durable result summaries', async () => {
+    const { ctx, runtime } = await setup({ mode: 'code' })
+    ctx.tools.register(defineTool({
+      name: 'workspace_path',
+      description: 'Return a path beneath the session workspace.',
+      parameters: {},
+      execute(_args, exec) {
+        const cwd = exec.agent?.session.header.cwd ?? ''
+        return Promise.resolve([{ type: 'text' as const, text: `<path>${cwd}/nested/task.txt</path>\n${'x'.repeat(240)}` }])
+      },
+    }))
+    runtime.behavior = async request => ({
+      logs: [],
+      value: await request.bindings[0]!.functions.workspace_path!({}),
+    })
+
+    const short = fakeAgent({ cwd: '/tmp/workspace' })
+    const long = fakeAgent({ cwd: `/tmp/${'long-segment/'.repeat(30)}workspace` })
+    const shortResult = await runCode(ctx, 'program', { agent: short.agent })
+    const longResult = await runCode(ctx, 'program', { agent: long.agent })
+    const shortDispatch = short.events[0]!.data as SessionEventMap['tool/code-dispatch']
+    const longDispatch = long.events[0]!.data as SessionEventMap['tool/code-dispatch']
+
+    expect(shortResult.content).not.toEqual(longResult.content)
+    expect(shortDispatch.resultSummary).toBe(longDispatch.resultSummary)
+    expect(shortDispatch.resultSummary).toHaveLength(201)
+    expect(shortDispatch.resultSummary).toMatch(/^<path>\.\/nested\/task\.txt<\/path>\n.+…$/)
+  })
+
+  it('leaves result summaries unchanged when a session cwd is absent or is the filesystem root', async () => {
+    const { ctx, runtime } = await setup({ mode: 'code' })
+    registerEcho(ctx)
+    runtime.behavior = async request => ({
+      logs: [],
+      value: await request.bindings[0]!.functions.echo!({ value: '/workspace/value' }),
+    })
+
+    const absent = fakeAgent({})
+    const root = fakeAgent({ cwd: '/' })
+    await runCode(ctx, 'program', { agent: absent.agent })
+    await runCode(ctx, 'program', { agent: root.agent })
+
+    expect((absent.events[0]!.data as SessionEventMap['tool/code-dispatch']).resultSummary).toBe('echo:/workspace/value')
+    expect((root.events[0]!.data as SessionEventMap['tool/code-dispatch']).resultSummary).toBe('echo:/workspace/value')
+  })
+
   it('rejects undefined, JSON-throwing, and JSON-unrepresentable binding arguments BEFORE dispatch', async () => {
     const { ctx, runtime } = await setup({ mode: 'code' })
     const calls = registerEcho(ctx)
diff --git a/packages/core/tools/tests/execution-mode.spec.ts b/packages/core/tools/tests/execution-mode.spec.ts
new file mode 100644
index 0000000000..9a12f33a51
--- /dev/null
+++ b/packages/core/tools/tests/execution-mode.spec.ts
@@ -0,0 +1,136 @@
+/** Covers fail-closed per-call classification and model-schema isolation. */
+
+import { describe, expect, expectTypeOf, it } from 'vitest'
+import { Context } from 'cordis'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import ToolRegistry, {
+  defineTool,
+  type ToolDefinition,
+  type ToolExecutionInput,
+  type ToolExecutionMode,
+} from '@deepseek-ai/dsh-tools'
+
+async function setup() {
+  const ctx = new Context()
+  await ctx.plugin(SystemPrompt)
+  await ctx.plugin(ToolRegistry)
+  return ctx
+}
+
+function exec(name: string, args: unknown): ToolExecutionInput {
+  return { callId: CallId('c1'), name, arguments: args }
+}
+
+describe('ToolRegistry.executionMode', () => {
+  it('returns parallel only for an explicit true classifier', async () => {
+    const ctx = await setup()
+    ctx.tools.register(defineTool({
+      name: 'safe',
+      description: 'parallel-safe',
+      parameters: {},
+      isConcurrencySafe: () => true,
+      async execute() { return [] },
+    }))
+    expect(ctx.tools.executionMode(exec('safe', {}))).toEqual({ kind: 'parallel' })
+  })
+
+  it('defaults to exclusive for a tool with no isConcurrencySafe declaration', async () => {
+    const ctx = await setup()
+    ctx.tools.register(defineTool({
+      name: 'plain',
+      description: 'no declaration',
+      parameters: {},
+      async execute() { return [] },
+    }))
+    expect(ctx.tools.executionMode(exec('plain', {}))).toEqual({ kind: 'exclusive' })
+  })
+
+  it('returns exclusive for an unknown tool', async () => {
+    const ctx = await setup()
+    expect(ctx.tools.executionMode(exec('nonexistent', {}))).toEqual({ kind: 'exclusive' })
+  })
+
+  it('returns exclusive when the classifier returns false for these args', async () => {
+    const ctx = await setup()
+    ctx.tools.register(defineTool({
+      name: 'rw',
+      description: 'read or write',
+      parameters: { mode: { type: 'string', required: true } },
+      isConcurrencySafe: args => args.mode === 'read',
+      async execute() { return [] },
+    }))
+    expect(ctx.tools.executionMode(exec('rw', { mode: 'read' }))).toEqual({ kind: 'parallel' })
+    expect(ctx.tools.executionMode(exec('rw', { mode: 'write' }))).toEqual({ kind: 'exclusive' })
+  })
+
+  it('classifies invalid defineTool arguments as exclusive without throwing', async () => {
+    const ctx = await setup()
+    ctx.tools.register(defineTool({
+      name: 'needs-mode',
+      description: 'requires mode',
+      parameters: { mode: { type: 'string', required: true } },
+      isConcurrencySafe: () => true,
+      async execute() { return [] },
+    }))
+    expect(ctx.tools.executionMode(exec('needs-mode', {}))).toEqual({ kind: 'exclusive' })
+  })
+
+  it('treats a throwing raw classifier as exclusive', async () => {
+    const ctx = await setup()
+    const raw: ToolDefinition = {
+      name: 'thrower',
+      description: 'classifier throws',
+      parameters: { type: 'object', properties: {} },
+      isConcurrencySafe() { throw new Error('boom') },
+      async execute() { return [] },
+    }
+    ctx.tools.register(raw)
+    expect(ctx.tools.executionMode(exec('thrower', {}))).toEqual({ kind: 'exclusive' })
+  })
+
+  it('treats a truthy non-boolean raw result as exclusive', async () => {
+    const ctx = await setup()
+    const raw = {
+      name: 'truthy',
+      description: 'classifier returns a truthy string',
+      parameters: { type: 'object', properties: {} },
+      isConcurrencySafe() { return 'yes' },
+      async execute() { return [] },
+    } as unknown as ToolDefinition
+    ctx.tools.register(raw)
+    expect(ctx.tools.executionMode(exec('truthy', {}))).toEqual({ kind: 'exclusive' })
+  })
+
+  it('passes parsed arguments directly to a raw definition', async () => {
+    const ctx = await setup()
+    let seen: unknown
+    ctx.tools.register({
+      name: 'raw-safe',
+      description: 'raw',
+      parameters: { type: 'object', properties: {} },
+      isConcurrencySafe(args) { seen = args; return true },
+      async execute() { return [] },
+    })
+    expect(ctx.tools.executionMode(exec('raw-safe', { anything: 1 }))).toEqual({ kind: 'parallel' })
+    expect(seen).toEqual({ anything: 1 })
+  })
+
+  it('isConcurrencySafe never reaches the model-facing schemas() projection', async () => {
+    const ctx = await setup()
+    ctx.tools.register(defineTool({
+      name: 'safe',
+      description: 'parallel-safe',
+      parameters: { x: { type: 'string', required: true } },
+      isConcurrencySafe: () => true,
+      async execute() { return [] },
+    }))
+    const schema = ctx.tools.schemas()[0] as unknown as Record<string, unknown>
+    expect(Object.keys(schema).sort()).toEqual(['description', 'name', 'parameters'])
+    expect(schema.isConcurrencySafe).toBeUndefined()
+  })
+
+  it('ToolExecutionMode is the object-tagged union', () => {
+    expectTypeOf<ToolExecutionMode>().toEqualTypeOf<{ kind: 'parallel' } | { kind: 'exclusive' }>()
+  })
+})
diff --git a/packages/core/tools/tests/gen-tool-catalog.spec.ts b/packages/core/tools/tests/gen-tool-catalog.spec.ts
index ad739173c9..3097f1abeb 100644
--- a/packages/core/tools/tests/gen-tool-catalog.spec.ts
+++ b/packages/core/tools/tests/gen-tool-catalog.spec.ts
@@ -23,7 +23,7 @@ describe('gen-tool-catalog collectToolCatalog', () => {
   it('boots every shipped tool package and harvests its model-facing schemas', async () => {
     const catalog = await collectToolCatalog()
     const names = catalog.flatMap(entry => entry.schemas.map(s => s.name)).sort()
-    expect(names).toEqual(['ask_user_question', 'bash', 'bash_kill', 'bash_output', 'cordis_inspect', 'cordis_mount', 'cordis_unmount', 'edit', 'read', 'run_code', 'skill', 'subagent', 'todo_write', 'web_fetch', 'web_search', 'workflow', 'write'])
+    expect(names).toEqual(['ask_user_question', 'bash', 'cordis_inspect', 'cordis_mount', 'cordis_unmount', 'edit', 'glob', 'grep', 'read', 'run_code', 'skill', 'subagent', 'task_kill', 'task_list', 'task_output', 'todo_write', 'web_fetch', 'web_search', 'workflow', 'write'])
     // Every tool carries a JSON-Schema `parameters` object (what the model sees).
     for (const entry of catalog) {
       for (const schema of entry.schemas) {
diff --git a/packages/core/tools/tests/tools.spec.ts b/packages/core/tools/tests/tools.spec.ts
index 59bd34e7ab..3de73c3326 100644
--- a/packages/core/tools/tests/tools.spec.ts
+++ b/packages/core/tools/tests/tools.spec.ts
@@ -348,7 +348,7 @@ describe('ToolRegistry', () => {
     expect(result.content[0]).toMatchObject({ text: 'output rejected: try again' })
   })
 
-  it('a block decision can ALSO attach additionalContext', async () => {
+  it('a block decision can ALSO attach additionalContexts', async () => {
     const ctx = await setup()
     ctx.tools.register(echoTool)
 
@@ -356,24 +356,95 @@ describe('ToolRegistry', () => {
       ({
         kind: 'block',
         feedback: [{ type: 'text', text: 'rejected' }],
-        additionalContext: { content: [{ type: 'text', text: 'why it was rejected' }], source: { kind: 'plugin', plugin: 'test' } },
+        additionalContexts: [{ content: [{ type: 'text', text: 'why it was rejected' }], source: { kind: 'plugin', plugin: 'test' } }],
       }))
 
     const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'echo', arguments: { text: 'hi' } })
     expect(result.isError).toBe(true)
     expect(result.content[0]).toMatchObject({ text: 'rejected' })
-    expect(result.additionalContext).toMatchObject({ content: [{ text: 'why it was rejected' }], source: { kind: 'plugin', plugin: 'test' } })
+    expect(result.additionalContexts).toMatchObject([{ content: [{ text: 'why it was rejected' }], source: { kind: 'plugin', plugin: 'test' } }])
   })
 
-  it('a post-execute additionalContext rides on the result for the loop to buffer', async () => {
+  it('post-execute additionalContexts ride on the result for the loop to buffer', async () => {
     const ctx = await setup()
     ctx.tools.register(echoTool)
 
     ctx.on('tools/post-execute', async (_exec, _result, _next): Promise<PostToolDecision> =>
-      ({ kind: 'accept', additionalContext: { content: [{ type: 'text', text: 'fyi' }], source: { kind: 'plugin', plugin: 'test' } } }))
+      ({ kind: 'accept', additionalContexts: [{ content: [{ type: 'text', text: 'fyi' }], source: { kind: 'plugin', plugin: 'test' } }] }))
 
     const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'echo', arguments: { text: 'hi' } })
-    expect(result.additionalContext).toMatchObject({ content: [{ text: 'fyi' }], source: { kind: 'plugin', plugin: 'test' } })
+    expect(result.additionalContexts).toMatchObject([{ content: [{ text: 'fyi' }], source: { kind: 'plugin', plugin: 'test' } }])
+  })
+
+  it('preserves tool-deferred, execute-wrapper, and post-execute contexts in order', async () => {
+    const ctx = await setup()
+    ctx.tools.register(defineTool({
+      name: 'composite',
+      description: 'composite',
+      parameters: {},
+      async execute(_args, exec) {
+        exec.deferContext({ content: [{ type: 'text', text: 'nested-1' }], source: { kind: 'plugin', plugin: 'nested-1' }, meta: { n: 1 } })
+        exec.deferContext({ content: [{ type: 'text', text: 'nested-2' }], source: { kind: 'plugin', plugin: 'nested-2' }, envelope: 'raw' })
+        return [{ type: 'text', text: 'done' }]
+      },
+    }))
+    ctx.on('tools/execute', async (_exec, next) => {
+      const result = await next()
+      return {
+        ...result,
+        additionalContexts: [
+          ...result.additionalContexts ?? [],
+          { content: [{ type: 'text', text: 'wrapper' }], source: { kind: 'plugin', plugin: 'wrapper' } },
+        ],
+      }
+    })
+    ctx.on('tools/post-execute', async (_exec, _result, next): Promise<PostToolDecision> => {
+      const downstream = await next()
+      return {
+        ...downstream,
+        additionalContexts: [
+          { content: [{ type: 'text', text: 'post' }], source: { kind: 'plugin', plugin: 'post' } },
+          ...downstream.additionalContexts ?? [],
+        ],
+      }
+    })
+
+    const result = await ctx.tools.execute({ callId: CallId('composite'), name: 'composite', arguments: {} })
+
+    expect(result.additionalContexts?.map(context => context.source)).toEqual([
+      { kind: 'plugin', plugin: 'nested-1' },
+      { kind: 'plugin', plugin: 'nested-2' },
+      { kind: 'plugin', plugin: 'wrapper' },
+      { kind: 'plugin', plugin: 'post' },
+    ])
+    expect(result.additionalContexts?.[0]?.meta).toEqual({ n: 1 })
+    expect(result.additionalContexts?.[1]?.envelope).toBe('raw')
+  })
+
+  it('keeps deferred contexts when a composite tool throws, but drops them when the outer call is blocked', async () => {
+    const ctx = await setup()
+    ctx.tools.register(defineTool({
+      name: 'failing-composite',
+      description: 'failing composite',
+      parameters: {},
+      async execute(_args, exec) {
+        exec.deferContext({ content: [{ type: 'text', text: 'nested' }], source: { kind: 'plugin', plugin: 'nested' } })
+        throw new Error('outer failure')
+      },
+    }))
+
+    const failed = await ctx.tools.execute({ callId: CallId('failed'), name: 'failing-composite', arguments: {} })
+    expect(failed.isError).toBe(true)
+    expect(failed.additionalContexts?.map(context => context.source)).toEqual([{ kind: 'plugin', plugin: 'nested' }])
+
+    ctx.on('tools/post-execute', async (): Promise<PostToolDecision> => ({
+      kind: 'block',
+      feedback: [{ type: 'text', text: 'blocked' }],
+      additionalContexts: [{ content: [{ type: 'text', text: 'block-only' }], source: { kind: 'plugin', plugin: 'blocker' } }],
+    }))
+    const blocked = await ctx.tools.execute({ callId: CallId('blocked'), name: 'failing-composite', arguments: {} })
+    expect(blocked.isError).toBe(true)
+    expect(blocked.additionalContexts?.map(context => context.source)).toEqual([{ kind: 'plugin', plugin: 'blocker' }])
   })
 
   it('composes pre + post waterfalls around dispatch (sandbox-wrap pattern)', async () => {
@@ -532,25 +603,25 @@ describe('ToolRegistry', () => {
     expect(result.content[0]).toMatchObject({ text: 'short-circuited' })
   })
 
-  it('preserves additionalContext supplied by an around-dispatch result', async () => {
+  it('preserves additionalContexts supplied by an around-dispatch result', async () => {
     const ctx = await setup()
     ctx.tools.register(echoTool)
     ctx.on('tools/execute', async () => ({
       content: [{ type: 'text', text: 'short-circuited with context' }],
       isError: false,
-      additionalContext: {
+      additionalContexts: [{
         content: [{ type: 'text', text: 'from around dispatch' }],
         source: { kind: 'plugin', plugin: 'test' },
-      },
+      }],
     }))
 
     const result = await ctx.tools.execute({
       callId: CallId('around-context'), name: 'echo', arguments: {},
     })
-    expect(result.additionalContext).toEqual({
+    expect(result.additionalContexts).toEqual([{
       content: [{ type: 'text', text: 'from around dispatch' }],
       source: { kind: 'plugin', plugin: 'test' },
-    })
+    }])
   })
 
   it('returns an isError result when a tools/execute listener throws', async () => {
diff --git a/packages/examples/README.md b/packages/examples/README.md
index 5ee0206d0f..65dfa40329 100644
--- a/packages/examples/README.md
+++ b/packages/examples/README.md
@@ -4,7 +4,7 @@ Pre-composed plugin bundles a thin leaf `cordis.yml` loads instead of assembling
 
 | Package | npm name | Role |
 |---|---|---|
-| `agent-spine-demo/` | `@deepseek-ai/dsh-agent-spine-demo` | The executor-less/UI-less agent spine as one bundle plugin (`timer` + `llm` + sessions + system-prompt + tools + skills + agents + invariants + `tool-bash` + `tool-skill` + `agent-loop`) |
+| `agent-spine-demo/` | `@deepseek-ai/dsh-agent-spine-demo` | The executor-less/UI-less agent spine as one bundle plugin (`timer` + `llm` + sessions + system-prompt + tools + skills + agents + invariants + `tool-bash` + workspace-context + `tool-skill` + `agent-loop`) |
 | `stdio-demo/` | `@deepseek-ai/dsh-stdio-demo` | Terminal stdio chat app: the spine + console logger + readline UI + a pre-created `main` agent, with a boot `bin` |
 | `acp-demo/` | `@deepseek-ai/dsh-acp-demo` | ACP server app: the spine + JSONL persistence + the [`acp`](../ui/acp/README.md) bridge (no stdout logger), with a boot `bin` |
 | `jsonrpc-demo/` | `@deepseek-ai/dsh-jsonrpc-demo` | Bin-only runtime that boots an external `cordis.yml` for the stdio JSON-RPC SDK client |
diff --git a/packages/examples/acp-demo/README.md b/packages/examples/acp-demo/README.md
index 17e3da78f3..1dbda9a08a 100644
--- a/packages/examples/acp-demo/README.md
+++ b/packages/examples/acp-demo/README.md
@@ -25,11 +25,16 @@ Because the package wires no logger entry, an ACP leaf has **nothing to get wron
 
 | Key | Default | Routed to |
 |---|---|---|
-| `model` | (required) | the per-session agent template the bridge creates agents from |
-| `persona` | — | the deployment persona template (may reference `{{model}}`/`{{cwd}}`), routed to `dsh-system-prompt` |
+| `provider` | (required) | the initial provider route for each per-session agent the bridge creates; ACP model selection may replace it per session |
+| `model` | (required) | the initial model for each per-session agent; ACP clients may switch among adapter-advertised models |
+| `maxParallelToolCalls` | agent-loop default | positive-integer concurrent tool-call cap shared by the bundled loop's agents; `1` is serial |
+| `persona` | — | the deployment persona template (may reference `{{provider}}`/`{{model}}`/`{{cwd}}`), routed to `dsh-system-prompt` |
 | `toolOrder` | — | explicit model-facing tool order (a name list with one `'<unlisted-tools>'` rest entry; absent — lexicographic; an unregistered name fails each turn at prompt assembly), routed to `dsh-system-prompt` |
+| `dshHome` | `$DSH_HOME` or `~/.dsh` | Harness home exposed to model bash and used by local skill discovery |
 | `tools` | `{ mode: 'native' }` | tool-registry presentation config (`native` / `code` / `both`), routed through `dsh-agent-spine-demo` |
 | `skills` | owner defaults | registry-cache, local-provider, and model-facing skill-tool config, routed through `dsh-agent-spine-demo` |
+| `toolBash` | owner defaults | model-facing bash config routed through `dsh-agent-spine-demo`, including bash's producer-local `enableRunInBackground` |
+| `toolTasks` | owner defaults | generic `task_output` wait bounds routed through `dsh-agent-spine-demo` |
 | `persistenceRoot` | `./.sessions` | the JSONL backend's root directory |
 
 The leaf supplies the swappable backends: an LLM adapter (`llm-deepseek` for the real model, `llm-replay` for keyless snapshot replay) and a bash executor.
diff --git a/packages/examples/acp-demo/package.json b/packages/examples/acp-demo/package.json
index 3c4851c961..28d057c0a0 100644
--- a/packages/examples/acp-demo/package.json
+++ b/packages/examples/acp-demo/package.json
@@ -35,6 +35,7 @@
     "@deepseek-ai/dsh-app-boot": "^0.0.1",
     "@deepseek-ai/dsh-acp": "^0.0.1",
     "@deepseek-ai/dsh-agent-spine-demo": "^0.0.1",
+    "@deepseek-ai/dsh-workspace-context": "^0.0.1",
     "@deepseek-ai/dsh-session-persistence-jsonl": "^0.0.1",
     "@deepseek-ai/dsh-tools": "^0.0.1",
     "@deepseek-ai/dsh-user-interaction": "^0.0.1",
@@ -50,6 +51,7 @@
     "@deepseek-ai/dsh-agent-spine-demo": "workspace:^",
     "@deepseek-ai/dsh-tools": "workspace:^",
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-workspace-context": "workspace:^",
     "@deepseek-ai/dsh-session-persistence-jsonl": "workspace:^",
     "@deepseek-ai/dsh-user-interaction": "workspace:^",
     "cordis": "^4.0.0-rc.7",
diff --git a/packages/examples/acp-demo/src/index.ts b/packages/examples/acp-demo/src/index.ts
index 0439363301..caa6ffd582 100644
--- a/packages/examples/acp-demo/src/index.ts
+++ b/packages/examples/acp-demo/src/index.ts
@@ -13,6 +13,7 @@ import type { Context } from 'cordis'
 import z from 'schemastery'
 import * as acp from '@deepseek-ai/dsh-acp'
 import * as agentCore from '@deepseek-ai/dsh-agent-spine-demo'
+import * as workspaceContext from '@deepseek-ai/dsh-workspace-context'
 import ToolRegistry, { type Config as ToolsConfig } from '@deepseek-ai/dsh-tools'
 import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
 import UserInteractionService from '@deepseek-ai/dsh-user-interaction'
@@ -20,7 +21,7 @@ import UserInteractionService from '@deepseek-ai/dsh-user-interaction'
 export const name = 'acp-demo'
 
 /**
- * App config: the swappable per-deployment values. `model` configures the
+ * App config: the swappable per-deployment values. `provider` and `model` configure the
  * agent template the ACP bridge creates each session's agent from (NOT a
  * pre-created agent — ACP creates agents at `session/new`); `persona` is the
  * deployment persona (forwarded to the system-prompt plugin); `toolOrder` is
@@ -29,35 +30,53 @@ export const name = 'acp-demo'
  * through agent-spine-demo); `persistenceRoot` is the JSONL backend's directory.
  */
 export interface Config {
+  /** Provider route for ACP-created agents. */
+  provider: string
   /** Model name for ACP-created agents (must have a registered adapter). */
   model: string
+  /** Bundled agent-loop concurrency cap; `1` is serial and omission uses its default. */
+  maxParallelToolCalls?: number
   /** Deployment persona (the system-prompt plugin's `persona` config). */
   persona?: string
   /** Explicit model-facing tool order (the system-prompt plugin's `toolOrder` config; see dsh-system-prompt). */
   toolOrder?: string[]
   /** Tool-registry config — its presentation `mode` (forwarded through agent-spine-demo; see dsh-tools). */
   tools?: ToolsConfig
+  /** DeepSeek Harness home directory exposed to bash and used for local skill discovery. */
+  dshHome?: string
   /** Directory the JSONL session backend writes under. Defaults to `./.sessions`. */
   persistenceRoot?: string
+  /** Controls automatic AGENTS.md/CLAUDE.md loading; configure a byte budget or set `false`. */
+  workspaceContext: agentCore.Config['workspaceContext']
   /** Skill registry, local-provider, and model-facing consumer config forwarded to agent-spine-demo. */
   skills?: agentCore.SkillConfig
+  /** Model-facing bash tool config forwarded through agent-core. */
+  toolBash?: NonNullable<agentCore.Config['toolBash']>
+  /** Generic background-task control-tool config forwarded through agent-core. */
+  toolTasks?: NonNullable<agentCore.Config['toolTasks']>
 }
 
 // Each front door owns a complete, directly readable config schema; extracting
 // the common fields would make two small app contracts depend on a new facade.
 /* jscpd:ignore-start */
 export const Config: z<Config> = z.object({
+  provider: z.string().required(),
   model: z.string().required(),
+  maxParallelToolCalls: z.number().step(1).min(1),
   persona: z.string(),
   // The array default is forced to undefined: ABSENT means "lexicographic
   // order" (the owning dsh-system-prompt schema does the same), while
   // schemastery's native [] default would read as an invalid configured list.
   toolOrder: z.array(z.string()).default(undefined as unknown as string[]),
   tools: ToolRegistry.Config,
+  dshHome: z.string(),
   // TODO(single-default-literal): share this schema default and the defensive
   // apply() fallback through one named constant while retaining both boundaries.
   persistenceRoot: z.string().default('./.sessions'),
+  workspaceContext: z.union([z.const(false), workspaceContext.Config]).required(),
   skills: agentCore.SkillConfigSchema,
+  toolBash: agentCore.ToolBashConfigSchema,
+  toolTasks: agentCore.ToolTasksConfigSchema,
 })
 /* jscpd:ignore-end */
 
@@ -66,16 +85,11 @@ export const Config: z<Config> = z.object({
  * NO agents (its `agents` list defaults to `[]`) and carries the deployment
  * `persona`; the JSONL backend persists under `persistenceRoot`; the ACP
  * bridge owns stdout for JSON-RPC and creates one agent per `session/new`
- * from `model`. No logger, no `hmr` — stdout stays pure.
+ * from the provider/model pair. No logger, no `hmr` — stdout stays pure.
  */
 export function apply(ctx: Context, config: Config): void {
-  ctx.plugin(agentCore, {
-    ...config.persona !== undefined ? { persona: config.persona } : {},
-    ...config.toolOrder !== undefined ? { toolOrder: config.toolOrder } : {},
-    ...config.tools !== undefined ? { tools: config.tools } : {},
-    ...config.skills !== undefined ? { skills: config.skills } : {},
-  })
+  ctx.plugin(agentCore, agentCore.pickSpineConfig(config))
   ctx.plugin(UserInteractionService)
   ctx.plugin(SessionPersistenceJsonl, { root: config.persistenceRoot ?? './.sessions' })
-  ctx.plugin(acp, { model: config.model })
+  ctx.plugin(acp, { provider: config.provider, model: config.model })
 }
diff --git a/packages/examples/acp-demo/tests/acp-agent.spec.ts b/packages/examples/acp-demo/tests/acp-agent.spec.ts
index f3495c827c..c033056f12 100644
--- a/packages/examples/acp-demo/tests/acp-agent.spec.ts
+++ b/packages/examples/acp-demo/tests/acp-agent.spec.ts
@@ -11,7 +11,7 @@ import * as acpAgent from '../src/index.ts'
 
 /**
  * In-process unit coverage for the @deepseek-ai/dsh-acp-demo composition:
- * mounting it brings up the agent-core spine + JSONL persistence + the ACP
+ * mounting it brings up the agent-spine-demo spine + JSONL persistence + the ACP
  * bridge in one `ctx.plugin`. Unlike the stdio app, this one loads NO
  * Loader-only plugin (no hmr), so it mounts in a plain Context.
  *
@@ -19,8 +19,9 @@ import * as acpAgent from '../src/index.ts'
  * ACP operations end-to-end) is the keyless bin smoke in `load-path.e2e.ts`;
  * this spec asserts the composition and the persistenceRoot default branch.
  */
-async function mount(config: acpAgent.Config): Promise<Context> {
+async function mount(config: acpAgent.Config, withBash = false): Promise<Context> {
   const ctx = new Context()
+  if (withBash) ctx.provide('bash', { sandboxMode: undefined })
   await ctx.plugin(acpAgent, config)
   // The bundle mounts its children inside apply() (not awaited there); let their
   // fibers settle so the spine services are ready.
@@ -69,7 +70,7 @@ async function withIsolatedSkillHomes<T>(run: () => Promise<T>): Promise<T> {
 
 describe('dsh-acp-demo composition', () => {
   it('brings up the spine + persistence + the ACP bridge', async () => {
-    const ctx = await mount({ model: 'mock', persona: 'hi', persistenceRoot: '/tmp/dsh-acp-demo-test', skills: await isolatedSkillsConfig() })
+    const ctx = await mount({ provider: 'mock', model: 'mock', persona: 'hi', persistenceRoot: '/tmp/dsh-acp-demo-test', skills: await isolatedSkillsConfig(), workspaceContext: false })
     expect(ctx.get('agents')).toBeDefined()
     expect(ctx.get('sessions')).toBeDefined()
     expect(ctx.get('sessionPersistence')).toBeDefined()
@@ -88,16 +89,29 @@ describe('dsh-acp-demo composition', () => {
     // persistenceRoot, so the runtime fallback is the one that fires.
     const ctx = new Context()
     // No persona: covers the omitted-persona forwarding branch too.
-    acpAgent.apply(ctx, { model: 'mock', skills: await isolatedSkillsConfig() })
+    acpAgent.apply(ctx, { provider: 'mock', model: 'mock', skills: await isolatedSkillsConfig(), workspaceContext: false })
     await new Promise(resolve => setTimeout(resolve, 50))
     expect(ctx.get('sessionPersistence')).toBeDefined()
     await ctx.fiber.dispose()
   })
 
+  it('forwards explicit project-instruction controls to the bundled spine', async () => {
+    const ctx = await mount({
+      provider: 'mock',
+      model: 'mock',
+      persona: 'hi',
+      persistenceRoot: '/tmp/dsh-acp-demo-workspace-context',
+      workspaceContext: false,
+    })
+    expect(ctx.get('agents')).toBeDefined()
+    expect(ctx.get('agentLoop')).toBeDefined()
+    await ctx.fiber.dispose()
+  })
+
   it('uses default skill config when apply is called directly without skills', async () => {
     await withIsolatedSkillHomes(async () => {
       const ctx = new Context()
-      acpAgent.apply(ctx, { model: 'mock' })
+      acpAgent.apply(ctx, { provider: 'mock', model: 'mock', workspaceContext: false })
       await new Promise(resolve => setTimeout(resolve, 50))
       expect(ctx.skills).toBeDefined()
       expect(await ctx.skills.list()).toEqual([])
@@ -105,23 +119,54 @@ describe('dsh-acp-demo composition', () => {
     })
   })
 
-  it('forwards skill config into agent-core', async () => {
-    const ctx = await mount({ model: 'mock', persona: 'hi', skills: await isolatedSkillsConfig(6) })
+  it('forwards skill config and dshHome into agent-spine-demo', async () => {
+    const skills = await isolatedSkillsConfig(6)
+    const ctx = await mount({ provider: 'mock', model: 'mock', persona: 'hi', dshHome: skills.local!.dshHome!, skills, workspaceContext: false })
     ctx.skills.register({ name: 'acp-skill', description: 'ACP skill', source: 'runtime', content: 'body' })
     expect(JSON.stringify(await composePrefix(ctx))).toContain('- `acp-skill`: ACP...')
     await ctx.fiber.dispose()
   })
 
+  it('forwards maxParallelToolCalls to the bundled agent loop', async () => {
+    const ctx = await mount({
+      provider: 'mock',
+      model: 'mock',
+      maxParallelToolCalls: 3,
+      persistenceRoot: '/tmp/dsh-acp-demo-test-parallel',
+      skills: await isolatedSkillsConfig(),
+      workspaceContext: false,
+    })
+    expect(ctx.get('agentLoop')?.config.maxParallelToolCalls).toBe(3)
+    await ctx.fiber.dispose()
+  })
+
+  it('forwards bundled tool config into agent-core', async () => {
+    const ctx = await mount({
+      provider: 'mock',
+      model: 'mock',
+      workspaceContext: false,
+      toolBash: { enableRunInBackground: false },
+      toolTasks: { waitTimeoutMs: 7, maxWaitTimeoutMs: 11 },
+      skills: await isolatedSkillsConfig(),
+    }, true)
+    const bash = ctx.tools.schemas().find(tool => tool.name === 'bash')
+    expect(Object.keys((bash!.parameters as { properties: Record<string, unknown> }).properties))
+      .not.toContain('run_in_background')
+    await ctx.fiber.dispose()
+  })
+
   it('exposes its plugin shape', () => {
     expect(acpAgent.name).toBe('acp-demo')
     expect(acpAgent.Config).toBeDefined()
   })
 
-  it('forwards toolOrder through agent-core to the system-prompt assembly', async () => {
+  it('forwards toolOrder through agent-spine-demo to the system-prompt assembly', async () => {
     const ctx = await mount({
+      provider: 'mock',
       model: 'mock',
       toolOrder: ['zulu', TOOL_ORDER_REST],
       persistenceRoot: '/tmp/dsh-acp-demo-test-tool-order',
+      workspaceContext: false,
     })
     // The bundle's own bash tools pend on the absent `ctx.bash` executor in
     // this providerless mount, so register two plain tools to order.
@@ -134,7 +179,7 @@ describe('dsh-acp-demo composition', () => {
       })
     }
     const assembly = await ctx.get('systemPrompt')!.assemble()
-    expect(assembly.tools.map(tool => tool.name)).toEqual(['zulu', 'alpha', 'skill'])
+    expect(assembly.tools.map(tool => tool.name)).toEqual(['zulu', 'alpha', 'skill', 'task_kill', 'task_list', 'task_output'])
     await ctx.fiber.dispose()
   })
 
diff --git a/packages/examples/acp-demo/tests/built-bin.e2e.ts b/packages/examples/acp-demo/tests/built-bin.e2e.ts
index 9e799b29db..3f90485679 100644
--- a/packages/examples/acp-demo/tests/built-bin.e2e.ts
+++ b/packages/examples/acp-demo/tests/built-bin.e2e.ts
@@ -30,9 +30,9 @@ const acpBin = join(repoRoot, 'packages/examples/acp-demo/lib/bin.js')
 const dshPackages = [
   'examples/agent-spine-demo', 'core/agent', 'core/session', 'core/system-prompt',
   'core/tools', 'core/agent-loop', 'llm/llm', 'llm/llm-deepseek', 'bash/bash',
-  'bash/bash-local', 'bash/tool-bash', 'support/invariants', 'ui/app-boot',
+  'bash/bash-local', 'bash/tool-bash', 'context/workspace-context', 'support/invariants', 'ui/app-boot',
   'session-persistence/session-persistence',
-  'session-persistence/session-persistence-jsonl', 'ui/acp', 'examples/acp-demo',
+  'session-persistence/session-persistence-jsonl', 'ui/acp', 'examples/acp-demo', 'util/paths',
 ]
 const vendorPackages = [
   'cordis', 'loader', 'include', 'timer', 'hmr', 'logger-console',
@@ -78,14 +78,15 @@ async function makeConsumer(): Promise<string> {
     '  name: \'@deepseek-ai/dsh-llm-deepseek\'',
     '  config:',
     '    apiKey: !!js process.env.DEEPSEEK_API_KEY',
-    '    models: [deepseek-v4-flash]',
     '- id: bash',
     '  name: \'@deepseek-ai/dsh-bash-local\'',
     '- id: acp-agent',
     '  name: \'@deepseek-ai/dsh-acp-demo\'',
     '  config:',
+    '    provider: deepseek',
     '    model: deepseek-v4-flash',
     '    persona: \'test agent\'',
+    '    workspaceContext: false',
     '',
   ].join('\n'))
   return dir
diff --git a/packages/examples/acp-demo/tests/load-path.e2e.ts b/packages/examples/acp-demo/tests/load-path.e2e.ts
index f1edf7832a..982fac3b50 100644
--- a/packages/examples/acp-demo/tests/load-path.e2e.ts
+++ b/packages/examples/acp-demo/tests/load-path.e2e.ts
@@ -36,14 +36,15 @@ const CORDIS_YML = `
   name: '@deepseek-ai/dsh-llm-deepseek'
   config:
     apiKey: !!js process.env.DEEPSEEK_API_KEY
-    models: [deepseek-v4-flash]
 - id: bash
   name: '@deepseek-ai/dsh-bash-local'
 - id: acp-agent
   name: '@deepseek-ai/dsh-acp-demo'
   config:
+    provider: deepseek
     model: deepseek-v4-flash
     persona: 'You are a test agent.'
+    workspaceContext: false
 `
 
 interface Spawned {
diff --git a/packages/examples/acp-demo/tsconfig.json b/packages/examples/acp-demo/tsconfig.json
index 6805405026..b0e537574a 100644
--- a/packages/examples/acp-demo/tsconfig.json
+++ b/packages/examples/acp-demo/tsconfig.json
@@ -29,6 +29,9 @@
     {
       "path": "../agent-spine-demo"
     },
+    {
+      "path": "../../context/workspace-context"
+    },
     {
       "path": "../../ui/user-interaction"
     },
diff --git a/packages/examples/agent-spine-demo/README.md b/packages/examples/agent-spine-demo/README.md
index a85f3e0889..9f1b1012f5 100644
--- a/packages/examples/agent-spine-demo/README.md
+++ b/packages/examples/agent-spine-demo/README.md
@@ -17,9 +17,12 @@ Read this package for the whole plugin tree and its composition order.
 @deepseek-ai/dsh-skill            skill provider registry
 @deepseek-ai/dsh-skill-local      local filesystem skill provider
 @deepseek-ai/dsh-agent            agent registry + agent/* event vocabulary
-@deepseek-ai/dsh-invariants       runtime event-contract assertions
-@deepseek-ai/dsh-tool-bash        the model-facing bash/bash_output/bash_kill schemas
+@deepseek-ai/dsh-tasks            generic background-task registry
+@deepseek-ai/dsh-invariants       dev-mode event-contract assertions
+@deepseek-ai/dsh-tool-bash        the model-facing bash schema
+@deepseek-ai/dsh-workspace-context  AGENTS.md/CLAUDE.md workspace context loader
 @deepseek-ai/dsh-tool-skill       session-prefix skill catalog + model-facing loader schema
+@deepseek-ai/dsh-tool-tasks       task_output/task_list/task_kill schemas + completion notices
 @deepseek-ai/dsh-agent-loop       THE concrete loop (gets the forwarded `agents`)
                                   (dsh-system-prompt gets the forwarded `persona`)
 ```
@@ -39,11 +42,11 @@ This is the [interface/implementation/consumer seam](../../../docs/rfc/implement
 
 ```ts
 import type { Config } from '@deepseek-ai/dsh-agent-spine-demo'
-// { agents?, persona?, toolOrder?, tools?, skills? } — the schema intersects the owner schemas,
-// so validation and defaulting can never drift from the owners.
+// { agents?, maxParallelToolCalls?, persona?, toolOrder?, tools?, dshHome?, skills?, workspaceContext, toolBash?, toolTasks? }
+// workspaceContext requires { maxBytes } or false; the other owner schemas supply defaults.
 ```
 
-The bundle FORWARDS each field to the child that owns it: `agents` to `agent-loop` (default `[]`), so each app supplies its own pre-created agents — a stdio app pre-creates a `main`; the ACP app pre-creates none (it creates agents on demand at `session/new`) — `persona` and `toolOrder` to `dsh-system-prompt`; `tools` to the tool registry for its presentation mode; and `skills.registry`, `skills.local`, and `skills.tool` to the skill registry, local provider, and model-facing consumer. Forwarding is exactly why the owners can live in the shared spine even though the apps disagree on what to configure.
+The bundle FORWARDS each field to the child that owns it: `agents` and `maxParallelToolCalls` to `agent-loop` (`agents` defaults to `[]`; the cap defaults there), so each app supplies its own pre-created agents — a stdio app pre-creates `main`, while the ACP app creates agents on demand at `session/new`; `persona` and `toolOrder` to `dsh-system-prompt`; `tools` to the tool registry for its presentation mode; `skills.registry`, `skills.local`, and `skills.tool` to the skill registry, local provider, and model-facing consumer; the required `workspaceContext` choice to `dsh-workspace-context` (`{ maxBytes }` enables loading and `false` disables it); and `toolBash`/`toolTasks` to the two model-facing tool plugins the bundle owns. It resolves `dshHome` once through [`@deepseek-ai/dsh-home`](../../util/home/README.md) and forwards that absolute value to tool-bash's managed environment and local skill discovery. An absent top-level `dshHome` adopts `skills.local.dshHome`; supplying both with different resolved paths fails loudly. `toolBash.enableRunInBackground` controls only the bash producer, while `toolTasks` controls generic `task_output` wait bounds; independently loaded producers keep their own config. Workspace instructions register before the skill catalog so their session-prefix message renders first. App packages use `pickSpineConfig()` to copy only these bundle-owned fields.
 
 ## Why a code bundle, not a shared YAML include
 
diff --git a/packages/examples/agent-spine-demo/package.json b/packages/examples/agent-spine-demo/package.json
index 049530872a..772d6c059b 100644
--- a/packages/examples/agent-spine-demo/package.json
+++ b/packages/examples/agent-spine-demo/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@deepseek-ai/dsh-agent-spine-demo",
-  "description": "The default executor-less/UI-less agent spine as one Cordis bundle plugin (timer + llm + sessions + system-prompt + tools + skills + agents + invariants + tool-bash + tool-skill + agent-loop)",
+  "description": "The default executor-less/UI-less agent spine as one Cordis bundle plugin (timer + llm + sessions + system-prompt + tools + skills + agents + tasks + invariants + tool-bash + workspace-context + tool-skill + tool-tasks + agent-loop)",
   "version": "0.0.1",
   "private": true,
   "type": "module",
@@ -26,13 +26,17 @@
     "@deepseek-ai/dsh-agent": "^0.0.1",
     "@deepseek-ai/dsh-agent-loop": "^0.0.1",
     "@deepseek-ai/dsh-invariants": "^0.0.1",
+    "@deepseek-ai/dsh-home": "^0.0.1",
     "@deepseek-ai/dsh-llm": "^0.0.1",
+    "@deepseek-ai/dsh-workspace-context": "^0.0.1",
     "@deepseek-ai/dsh-session": "^0.0.1",
     "@deepseek-ai/dsh-skill": "^0.0.1",
     "@deepseek-ai/dsh-skill-local": "^0.0.1",
     "@deepseek-ai/dsh-system-prompt": "^0.0.1",
+    "@deepseek-ai/dsh-tasks": "^0.0.1",
     "@deepseek-ai/dsh-tool-bash": "^0.0.1",
     "@deepseek-ai/dsh-tool-skill": "^0.0.1",
+    "@deepseek-ai/dsh-tool-tasks": "^0.0.1",
     "@deepseek-ai/dsh-tools": "^0.0.1",
     "cordis": "^4.0.0-rc.7"
   },
@@ -40,14 +44,19 @@
     "@cordisjs/plugin-timer": "workspace:^",
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-fs-local": "workspace:^",
     "@deepseek-ai/dsh-invariants": "workspace:^",
+    "@deepseek-ai/dsh-home": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-workspace-context": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
     "@deepseek-ai/dsh-skill": "workspace:^",
     "@deepseek-ai/dsh-skill-local": "workspace:^",
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-tasks": "workspace:^",
     "@deepseek-ai/dsh-tool-bash": "workspace:^",
     "@deepseek-ai/dsh-tool-skill": "workspace:^",
+    "@deepseek-ai/dsh-tool-tasks": "workspace:^",
     "@deepseek-ai/dsh-tools": "workspace:^",
     "cordis": "^4.0.0-rc.7"
   },
diff --git a/packages/examples/agent-spine-demo/src/index.ts b/packages/examples/agent-spine-demo/src/index.ts
index d1aecaa4c0..a4965ca457 100644
--- a/packages/examples/agent-spine-demo/src/index.ts
+++ b/packages/examples/agent-spine-demo/src/index.ts
@@ -1,6 +1,7 @@
 /**
  * Default executor-less, UI-less agent spine. It bundles the common services,
- * concrete loop, local skill provider, and model-facing bash/skill consumers;
+ * background-task registry and controls, concrete loop, local skill and
+ * workspace-context providers, and model-facing bash/skill consumers;
  * deployments still choose the LLM adapter, bash executor, and presentation.
  * The plugin intentionally exposes named exports only because Loader default
  * unwrapping would discard its `Config` schema (see docs/postmortem/0001).
@@ -17,10 +18,14 @@ import ToolRegistry, { type Config as ToolsConfig } from '@deepseek-ai/dsh-tools
 import SkillService, { type Config as SkillRegistryConfig } from '@deepseek-ai/dsh-skill'
 import * as SkillLocal from '@deepseek-ai/dsh-skill-local'
 import AgentRegistry from '@deepseek-ai/dsh-agent'
+import TaskService from '@deepseek-ai/dsh-tasks'
 import * as invariants from '@deepseek-ai/dsh-invariants'
 import * as toolBash from '@deepseek-ai/dsh-tool-bash'
+import * as workspaceContext from '@deepseek-ai/dsh-workspace-context'
 import * as toolSkill from '@deepseek-ai/dsh-tool-skill'
+import * as toolTasks from '@deepseek-ai/dsh-tool-tasks'
 import AgentLoop, { type Config as AgentLoopConfig } from '@deepseek-ai/dsh-agent-loop'
+import { resolveDshHome } from '@deepseek-ai/dsh-home'
 
 export const name = 'agent-spine-demo'
 
@@ -35,25 +40,41 @@ export interface SkillConfig {
 }
 
 /**
- * Bundle config: each field forwarded verbatim to the child that owns it — `agents` to the
- * agent loop (an app that pre-creates no agents, like the ACP bridge, omits it),
- * `persona` and `toolOrder` to the system-prompt plugin (the deployment's persona section and
- * the explicit model-facing tool order), the `tools` object to the tool registry (its
- * presentation `mode`), and `skills` to the skill registry/local provider/tool consumer.
- * The schema intersects the owners' schemas, which supply defaults for every
- * optional input and keep validation from drifting.
+ * Bundle config: each field forwarded verbatim to the child that owns it —
+ * `agents` to the agent loop (an app that pre-creates no agents, like the ACP
+ * bridge, simply omits it), `persona` and `toolOrder` to the system-prompt
+ * plugin (the deployment's persona section and the explicit model-facing tool
+ * order), the `tools` object to the tool registry (its presentation `mode`),
+ * `dshHome` to bash environment and local skill discovery, `skills` to the
+ * skill registry/local provider/tool consumer, `workspaceContext` to the
+ * workspace-context loader, and `toolBash`/`toolTasks` to the model-facing tool
+ * plugins this bundle owns. Owner schemas supply defaults for optional input;
+ * workspace context instead requires an explicit byte budget or `false` because
+ * it changes model-visible input. Producer opt-in stays producer-local:
+ * `toolBash` configures bash only; independently composed producers keep their
+ * own config.
  */
 export interface Config {
   /** The agent-loop `agents` list (see dsh-agent-loop's `Config`). */
   agents?: AgentLoopConfig['agents']
+  /** Agent-loop concurrency cap; `1` is serial. */
+  maxParallelToolCalls?: AgentLoopConfig['maxParallelToolCalls']
   /** The deployment persona (see dsh-system-prompt's `Config`). */
   persona?: SystemPromptConfig['persona']
   /** The explicit model-facing tool order (see dsh-system-prompt's `Config`). */
   toolOrder?: SystemPromptConfig['toolOrder']
   /** The tool registry's config — its presentation `mode` (see dsh-tools' `Config`). */
   tools?: ToolsConfig
+  /** DeepSeek Harness home directory shared by shell context and local skill discovery. */
+  dshHome?: string
+  /** Workspace-context loader controls with an explicit byte budget; set `false` for hermetic prompts. */
+  workspaceContext: workspaceContext.Config | false
   /** Skill registry, local provider, and model-facing consumer config. */
   skills?: SkillConfig
+  /** Model-facing bash tool config, including this producer's background opt-in. */
+  toolBash?: toolBash.Config
+  /** Generic background-task control-tool wait bounds. */
+  toolTasks?: toolTasks.Config
 }
 
 /** The skill config schema exported for app packages that forward `skills`. */
@@ -63,23 +84,63 @@ export const SkillConfigSchema: z<SkillConfig> = z.object({
   tool: toolSkill.Config,
 })
 
+/** The bash-tool config schema exported for app packages that forward `toolBash`. */
+export const ToolBashConfigSchema: z<toolBash.Config> = toolBash.Config
+
+/** The task-control-tool config schema exported for app packages that forward `toolTasks`. */
+export const ToolTasksConfigSchema: z<toolTasks.Config> = toolTasks.Config
+
 /** Intersect the owners' schemas so validation + defaulting stay identical. */
 export const Config = z.intersect([
   AgentLoop.Config,
   SystemPrompt.Config,
-  z.object({ tools: ToolRegistry.Config, skills: SkillConfigSchema }),
+  z.object({
+    tools: ToolRegistry.Config,
+    dshHome: z.string(),
+    skills: SkillConfigSchema,
+    workspaceContext: z.union([z.const(false), workspaceContext.Config]).required(),
+    toolBash: ToolBashConfigSchema,
+    toolTasks: ToolTasksConfigSchema,
+  }) as unknown as z<Pick<Config, 'tools' | 'dshHome' | 'skills' | 'workspaceContext' | 'toolBash' | 'toolTasks'>>,
 ]) as unknown as z<Config>
 
+/**
+ * Copy the bundle-owned fields from an app config without leaking front-door settings.
+ * @param config - App config containing the shared spine fields.
+ * @returns The fields accepted by this bundle, preserving optional absence.
+ */
+export function pickSpineConfig(config: Omit<Config, 'agents'>): Omit<Config, 'agents'> {
+  return {
+    ...config.maxParallelToolCalls !== undefined ? { maxParallelToolCalls: config.maxParallelToolCalls } : {},
+    ...config.persona !== undefined ? { persona: config.persona } : {},
+    ...config.toolOrder !== undefined ? { toolOrder: config.toolOrder } : {},
+    ...config.tools !== undefined ? { tools: config.tools } : {},
+    ...config.dshHome !== undefined ? { dshHome: config.dshHome } : {},
+    workspaceContext: config.workspaceContext,
+    ...config.skills !== undefined ? { skills: config.skills } : {},
+    ...config.toolBash !== undefined ? { toolBash: config.toolBash } : {},
+    ...config.toolTasks !== undefined ? { toolTasks: config.toolTasks } : {},
+  }
+}
+
 /**
  * Load the spine. Each `ctx.plugin(...)` mounts one child of the bundle fiber;
  * `agent-loop` receives the forwarded `agents` list and `system-prompt` the
- * forwarded `persona` and `toolOrder`. Load order is irrelevant (cordis pends
- * each fiber on its `inject` until the services it needs exist), but the
+ * forwarded `persona` and `toolOrder`. Workspace-context receives its own
+ * explicitly forwarded config. Load order is irrelevant (cordis
+ * pends each fiber on its `inject` until the services it needs exist), but the
  * listing mirrors the dependency layering for readability: the LLM vocabulary
- * and core registries first, then the dev tripwire and the bash tool consumer,
- * then the loop that drives them.
+ * and core registries first, then extension plugins that wrap request/tool
+ * seams, then the loop that drives them.
  */
 export function apply(ctx: Context, config: Config): void {
+  const nestedDshHome = config.skills?.local?.dshHome
+  if (config.dshHome !== undefined && nestedDshHome !== undefined
+    && resolveDshHome(config.dshHome) !== resolveDshHome(nestedDshHome)) {
+    throw new Error('agent-core: dshHome and skills.local.dshHome must resolve to the same directory')
+  }
+  const dshHome = resolveDshHome(config.dshHome ?? nestedDshHome)
+
   ctx.plugin(Timer)
   ctx.plugin(LlmService)
   ctx.plugin(SessionStore)
@@ -90,10 +151,20 @@ export function apply(ctx: Context, config: Config): void {
   })
   ctx.plugin(ToolRegistry, config.tools ?? {})
   ctx.plugin(SkillService, config.skills?.registry ?? {})
-  ctx.plugin(SkillLocal, config.skills?.local ?? {})
+  ctx.plugin(SkillLocal, Object.assign({}, config.skills?.local, { dshHome }))
   ctx.plugin(AgentRegistry)
+  ctx.plugin(TaskService)
   ctx.plugin(invariants)
-  ctx.plugin(toolBash)
+  ctx.plugin(toolBash, Object.assign({}, config.toolBash, { dshHome }))
+  if (config.workspaceContext !== false) {
+    ctx.plugin(workspaceContext, config.workspaceContext)
+  }
+  // Both plugins prepend session-prefix messages. Registration order is the
+  // rendered order, so workspace instructions must precede the skill catalog.
   ctx.plugin(toolSkill, config.skills?.tool ?? {})
-  ctx.plugin(AgentLoop, { agents: config.agents ?? [] })
+  ctx.plugin(toolTasks, config.toolTasks ?? {})
+  ctx.plugin(AgentLoop, {
+    agents: config.agents ?? [],
+    ...config.maxParallelToolCalls !== undefined ? { maxParallelToolCalls: config.maxParallelToolCalls } : {},
+  })
 }
diff --git a/packages/examples/agent-spine-demo/tests/agent-core.spec.ts b/packages/examples/agent-spine-demo/tests/agent-core.spec.ts
index ecf997f3cd..fdf17286ee 100644
--- a/packages/examples/agent-spine-demo/tests/agent-core.spec.ts
+++ b/packages/examples/agent-spine-demo/tests/agent-core.spec.ts
@@ -1,5 +1,5 @@
-import { describe, expect, it } from 'vitest'
-import { mkdir, mkdtemp, writeFile } from 'node:fs/promises'
+import { describe, expect, it, vi } from 'vitest'
+import { mkdir, mkdtemp, rm, writeFile } from 'node:fs/promises'
 import { join } from 'node:path'
 import { tmpdir } from 'node:os'
 import { Context } from 'cordis'
@@ -7,7 +7,17 @@ import Loader from '@cordisjs/plugin-loader'
 import { TOOL_ORDER_REST } from '@deepseek-ai/dsh-system-prompt'
 import * as agentCore from '../src/index.ts'
 import { AgentId, agentEvents, type Agent } from '@deepseek-ai/dsh-agent'
-import type { Message } from '@deepseek-ai/dsh-llm'
+import { SessionId } from '@deepseek-ai/dsh-session'
+import LocalFileSystem from '@deepseek-ai/dsh-fs-local'
+import { MockAdapter, textResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
+import { CallId, type Message } from '@deepseek-ai/dsh-llm'
+import type { ToolExecution } from '@deepseek-ai/dsh-tools'
+
+declare module '@deepseek-ai/dsh-tasks' {
+  interface TaskKindMap {
+    probe: 'probe'
+  }
+}
 
 async function composePrefix(ctx: Context, cwd: string): Promise<Message[]> {
   const agent = { session: { header: { cwd } } } as unknown as Agent
@@ -28,12 +38,13 @@ async function composePrefix(ctx: Context, cwd: string): Promise<Message[]> {
  * Loader-path guard (export shape, `unwrapExports`) is the app packages' keyless
  * bin smokes; here we assert the composition + config forwarding.
  */
-async function mount(config?: agentCore.Config): Promise<Context> {
+async function mount(config: agentCore.Config, withBash = false): Promise<Context> {
   const oldDshHome = process.env.DSH_HOME
   const oldAgentsHome = process.env.DSH_AGENTS_HOME
   process.env.DSH_HOME = await mkdtemp(join(tmpdir(), 'dsh-agent-spine-demo-home-'))
   process.env.DSH_AGENTS_HOME = await mkdtemp(join(tmpdir(), 'dsh-agent-spine-demo-agents-'))
   const ctx = new Context()
+  if (withBash) ctx.provide('bash', { sandboxMode: undefined })
   try {
     await ctx.plugin(agentCore, config)
     // The bundle mounts its children inside apply() (not awaited there); let their
@@ -75,9 +86,24 @@ async function withIsolatedSkillHomes<T>(run: () => Promise<T>): Promise<T> {
   }
 }
 
+function waitForMainIdle(ctx: Context): Promise<void> {
+  return new Promise((resolve) => {
+    const dispose = ctx.on('agent/status', (agent, status) => {
+      if (agent.id === 'main' && status === 'idle') {
+        dispose()
+        resolve()
+      }
+    })
+  })
+}
+
+function messageText(message: Message | undefined): string {
+  return message?.content.map(block => block.type === 'text' ? block.text : '').join('\n') ?? ''
+}
+
 describe('dsh-agent-spine-demo bundle', () => {
   it('brings up the full default spine', async () => {
-    const ctx = await mount()
+    const ctx = await mount({ workspaceContext: false })
     // One service from each layer of the spine proves the children loaded.
     expect(ctx.get('timer')).toBeDefined()
     expect(ctx.get('llm')).toBeDefined()
@@ -86,12 +112,13 @@ describe('dsh-agent-spine-demo bundle', () => {
     expect(ctx.get('tools')).toBeDefined()
     expect(ctx.get('skills')).toBeDefined()
     expect(ctx.get('agents')).toBeDefined()
+    expect(ctx.get('tasks')).toBeDefined()
     expect(ctx.get('agentLoop')).toBeDefined()
     await ctx.fiber.dispose()
   })
 
   it('includes the skill registry, local provider, and skill tool without builtin skills', async () => {
-    const ctx = await mount()
+    const ctx = await mount({ workspaceContext: false })
 
     expect(ctx.skills).toBeDefined()
     expect(ctx.tools.schemas().map(tool => tool.name)).toContain('skill')
@@ -101,15 +128,16 @@ describe('dsh-agent-spine-demo bundle', () => {
   })
 
   it('defaults the agents list to empty (no pre-created agents)', async () => {
-    const ctx = await mount()
+    const ctx = await mount({ workspaceContext: false })
     expect(ctx.get('agents')?.get(AgentId('main'))).toBeUndefined()
     await ctx.fiber.dispose()
   })
 
   it('forwards a pre-created agent to the loop and the persona to system-prompt', async () => {
     const ctx = await mount({
-      agents: [{ id: AgentId('main'), model: 'mock' }],
+      agents: [{ id: AgentId('main'), provider: 'mock', model: 'mock' }],
       persona: 'You are main.',
+      workspaceContext: false,
     })
     expect(ctx.get('agents')?.get(AgentId('main'))).toBeDefined()
     const assembly = await ctx.get('systemPrompt')!.assemble()
@@ -117,11 +145,21 @@ describe('dsh-agent-spine-demo bundle', () => {
     await ctx.fiber.dispose()
   })
 
+  it('forwards the global maxParallelToolCalls config to agent-loop', async () => {
+    const ctx = await mount({
+      agents: [{ id: AgentId('main'), provider: 'mock', model: 'mock' }],
+      maxParallelToolCalls: 3,
+      workspaceContext: false,
+    })
+    expect(ctx.get('agentLoop')?.config.maxParallelToolCalls).toBe(3)
+    await ctx.fiber.dispose()
+  })
+
   it('tolerates a schema-bypassing direct apply (the ?? fallbacks fire)', async () => {
     // ctx.plugin validates + defaults the bundle config first; a direct apply
     // skips the schema, so the forwarding `?? []` / `?? ''` are what fire.
     const ctx = new Context()
-    agentCore.apply(ctx, {})
+    agentCore.apply(ctx, { workspaceContext: false })
     await new Promise(resolve => setTimeout(resolve, 50))
     expect(ctx.get('agentLoop')).toBeDefined()
     expect(ctx.get('agents')?.list()).toHaveLength(0)
@@ -130,6 +168,64 @@ describe('dsh-agent-spine-demo bundle', () => {
     await ctx.fiber.dispose()
   })
 
+  it('loads workspace instructions into requests through the bundled spine', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-agent-spine-demo-workspace-context-'))
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await writeFile(join(root, 'AGENTS.md'), 'bundled project rule')
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await mount({ workspaceContext: { maxBytes: 65536 } })
+      await ctx.plugin(LocalFileSystem, { cwd: '/' })
+      ctx.llm.registerAdapter(['mock'], adapter)
+      const handle = await ctx.agents.create({
+        agentId: AgentId('main'),
+        sessionId: SessionId('main-session'),
+        meta: { cwd: root },
+        agentOptions: { provider: 'mock', model: 'mock' },
+      })
+      const agent = handle.agent
+
+      agent.send([{ type: 'text', text: 'hi' }])
+      await waitForMainIdle(ctx)
+
+      const sentText = adapter.requests[0]?.messages.map(messageText).join('\n')
+      expect(sentText).toContain('hi')
+      expect(sentText).toContain('bundled project rule')
+      expect(adapter.requests[0]?.system).toContain('You are an AI agent powered by the DeepSeek Harness SDK.')
+      expect(adapter.requests[0]?.system).not.toContain('bundled project rule')
+      await handle.dispose()
+      await ctx.fiber.dispose()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+    }
+  })
+
+  it('forwards workspace-context config to the bundled loader', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-agent-spine-demo-workspace-context-disabled-'))
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await writeFile(join(root, 'AGENTS.md'), 'must not be injected')
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await mount({ workspaceContext: { maxBytes: 0 } })
+      ctx.llm.registerAdapter(['mock'], adapter)
+      const handle = await ctx.agents.create({
+        agentId: AgentId('main'),
+        sessionId: SessionId('main-disabled-session'),
+        meta: { cwd: root },
+        agentOptions: { provider: 'mock', model: 'mock' },
+      })
+
+      handle.agent.send([{ type: 'text', text: 'hi' }])
+      await waitForMainIdle(ctx)
+
+      expect(adapter.requests[0]?.messages).toEqual([{ role: 'user', content: [{ type: 'text', text: 'hi' }] }])
+      await handle.dispose()
+      await ctx.fiber.dispose()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+    }
+  })
+
   it('forwards skill config to the registry, local provider, and model-facing consumer', async () => {
     const home = await mkdtemp(join(tmpdir(), 'dsh-agent-spine-demo-skill-home-'))
     const agentsHome = await mkdtemp(join(tmpdir(), 'dsh-agent-spine-demo-skill-agents-'))
@@ -138,6 +234,7 @@ describe('dsh-agent-spine-demo bundle', () => {
     await writeFile(join(custom, 'custom-skill.md'), '---\nname: custom-skill\ndescription: Custom skill\n---\n\nCustom body.\n')
     const ctx = await mount({
       agents: [],
+      workspaceContext: false,
       skills: {
         registry: { collectCacheMaxEntries: 4 },
         local: {
@@ -153,10 +250,131 @@ describe('dsh-agent-spine-demo bundle', () => {
     await ctx.fiber.dispose()
   })
 
+  it('shares top-level dshHome between local skills and the managed bash environment', async () => {
+    const home = await mkdtemp(join(tmpdir(), 'dsh-agent-core-shared-home-'))
+    const agentsHome = await mkdtemp(join(tmpdir(), 'dsh-agent-core-shared-agents-'))
+    await mkdir(join(home, 'skills'), { recursive: true })
+    await writeFile(join(home, 'skills', 'shared-skill.md'), '---\nname: shared-skill\ndescription: Shared home skill\n---\n\nShared body.\n')
+
+    const ctx = await mount({
+      dshHome: home,
+      workspaceContext: false,
+      skills: { local: { agentsHome } },
+    }, true)
+
+    expect((await ctx.skills.list()).map(skill => skill.name)).toEqual(['shared-skill'])
+    const execution: ToolExecution = {
+      token: Symbol('agent-core-dsh-home-test') as ToolExecution['token'],
+      callId: CallId('agent-core-dsh-home'),
+      name: 'bash',
+      arguments: { command: 'true' },
+    }
+    expect(ctx.bashEnv.collect(execution)).toMatchObject({ DSH_HOME: home, DSH_SHELL: '1' })
+    await ctx.fiber.dispose()
+  })
+
+  it('rejects conflicting global and nested DSH home directories', () => {
+    expect(() => {
+      agentCore.apply(new Context(), {
+        dshHome: '/global-dsh-home',
+        workspaceContext: false,
+        skills: { local: { dshHome: '/nested-dsh-home' } },
+      })
+    }).toThrow(/must resolve to the same directory/)
+  })
+
+  it('places workspace instructions before the skill catalog in the session prefix', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-agent-spine-demo-prefix-order-'))
+    try {
+      await mkdir(join(root, '.git'), { recursive: true })
+      await writeFile(join(root, 'AGENTS.md'), 'workspace rule before skills')
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await mount({ workspaceContext: { maxBytes: 65536 } })
+      await ctx.plugin(LocalFileSystem, { cwd: '/' })
+      ctx.llm.registerAdapter(['mock'], adapter)
+      ctx.skills.register({
+        name: 'prefix-order-skill',
+        description: 'Skill catalog after workspace rules',
+        source: 'runtime',
+        content: 'body',
+      })
+      const handle = await ctx.agents.create({
+        agentId: AgentId('main'),
+        sessionId: SessionId('prefix-order-session'),
+        meta: { cwd: root },
+        agentOptions: { provider: 'mock', model: 'mock' },
+      })
+
+      handle.agent.send([{ type: 'text', text: 'hi' }])
+      await waitForMainIdle(ctx)
+
+      expect(messageText(adapter.requests[0]?.messages[0])).toContain('workspace rule before skills')
+      expect(messageText(adapter.requests[0]?.messages[1])).toContain('prefix-order-skill')
+      await handle.dispose()
+      await ctx.fiber.dispose()
+    } finally {
+      await rm(root, { recursive: true, force: true })
+    }
+  })
+
+  it('forwards its bundled tool configs to tool-bash and tool-tasks', async () => {
+    const ctx = await mount({
+      workspaceContext: false,
+      toolBash: { enableRunInBackground: false },
+      toolTasks: { waitTimeoutMs: 7, maxWaitTimeoutMs: 11 },
+    }, true)
+
+    const bash = ctx.tools.schemas().find(tool => tool.name === 'bash')
+    expect(bash).toBeDefined()
+    expect(Object.keys((bash!.parameters as { properties: Record<string, unknown> }).properties))
+      .not.toContain('run_in_background')
+
+    const id = ctx.tasks.start({
+      kind: 'probe',
+      label: 'config forwarding probe',
+      run: () => ({ cancel: () => {}, done: Promise.resolve({ status: 'completed' }) }),
+    })
+    const wait = vi.spyOn(ctx.tasks, 'wait')
+    await ctx.tools.execute({
+      callId: CallId('task-config-forwarding'),
+      name: 'task_output',
+      arguments: { task_id: id, wait: true },
+    })
+    expect(wait).toHaveBeenCalledWith(id, 7, undefined, undefined)
+
+    await ctx.fiber.dispose()
+  })
+
+  it('picks shared spine config without leaking front-door fields', () => {
+    const appConfig = {
+      model: 'front-door-only',
+      persona: 'You are merged.',
+      toolOrder: ['zulu'],
+      tools: { mode: 'native' as const },
+      dshHome: '/tmp/dsh-home',
+      workspaceContext: false as const,
+      skills: {},
+      toolBash: { enableRunInBackground: false },
+      toolTasks: { waitTimeoutMs: 7, maxWaitTimeoutMs: 11 },
+    }
+
+    expect(agentCore.pickSpineConfig(appConfig)).toEqual({
+      persona: appConfig.persona,
+      toolOrder: appConfig.toolOrder,
+      tools: appConfig.tools,
+      dshHome: appConfig.dshHome,
+      workspaceContext: false,
+      skills: {},
+      toolBash: appConfig.toolBash,
+      toolTasks: appConfig.toolTasks,
+    })
+    expect(agentCore.pickSpineConfig({ workspaceContext: false })).toEqual({ workspaceContext: false })
+  })
+
   it('uses the default skill config when apply is called directly without skills', async () => {
     await withIsolatedSkillHomes(async () => {
       const ctx = new Context()
-      agentCore.apply(ctx, { agents: [] })
+      agentCore.apply(ctx, { agents: [], workspaceContext: false })
       await new Promise(resolve => setTimeout(resolve, 50))
       expect(ctx.skills).toBeDefined()
       expect(await ctx.skills.list()).toEqual([])
@@ -165,7 +383,7 @@ describe('dsh-agent-spine-demo bundle', () => {
   })
 
   it('forwards toolOrder to the system-prompt assembly', async () => {
-    const ctx = await mount({ toolOrder: ['zulu', TOOL_ORDER_REST] })
+    const ctx = await mount({ toolOrder: ['zulu', TOOL_ORDER_REST], workspaceContext: false })
     // The bundle's own bash tools pend on the absent `ctx.bash` executor in
     // this providerless mount, so register two plain tools to order.
     for (const name of ['alpha', 'zulu']) {
@@ -177,7 +395,17 @@ describe('dsh-agent-spine-demo bundle', () => {
       })
     }
     const assembly = await ctx.get('systemPrompt')!.assemble()
-    expect(assembly.tools.map(tool => tool.name)).toEqual(['zulu', 'alpha', 'skill'])
+    expect(assembly.tools.map(tool => tool.name)).toEqual(['zulu', 'alpha', 'skill', 'task_kill', 'task_list', 'task_output'])
+    await ctx.fiber.dispose()
+  })
+
+  it('supports direct apply with workspace instructions disabled and no forwarded agents', async () => {
+    const ctx = new Context()
+    agentCore.apply(ctx, { workspaceContext: false })
+    await new Promise(resolve => setTimeout(resolve, 50))
+
+    expect(ctx.get('agents')?.list()).toEqual([])
+    expect(ctx.get('systemPrompt')).toBeDefined()
     await ctx.fiber.dispose()
   })
 
diff --git a/packages/examples/agent-spine-demo/tsconfig.json b/packages/examples/agent-spine-demo/tsconfig.json
index 4fd0a81e97..89cb2accd8 100644
--- a/packages/examples/agent-spine-demo/tsconfig.json
+++ b/packages/examples/agent-spine-demo/tsconfig.json
@@ -11,9 +11,6 @@
     {
       "path": "../../../vendor/cordis"
     },
-    {
-      "path": "../../../vendor/schemastery"
-    },
     {
       "path": "../../../vendor/timer"
     },
@@ -44,14 +41,26 @@
     {
       "path": "../../core/agent"
     },
+    {
+      "path": "../../context/workspace-context"
+    },
     {
       "path": "../../core/agent-loop"
     },
     {
       "path": "../../support/invariants"
     },
+    {
+      "path": "../../util/home"
+    },
     {
       "path": "../../bash/tool-bash"
+    },
+    {
+      "path": "../../tasks/tasks"
+    },
+    {
+      "path": "../../tasks/tool-tasks"
     }
   ]
 }
diff --git a/packages/examples/stdio-demo/README.md b/packages/examples/stdio-demo/README.md
index 5087a2fca0..5eb6ea766c 100644
--- a/packages/examples/stdio-demo/README.md
+++ b/packages/examples/stdio-demo/README.md
@@ -11,7 +11,7 @@ A terminal chat always wants the same cluster, so the package owns it rather tha
 | Plugin | Why it is here |
 |---|---|
 | `@cordisjs/plugin-logger-console` | the console logger — stdout is just the terminal here, so logging to it is correct (the ACP app must NOT have this) |
-| `@deepseek-ai/dsh-agent-spine-demo` | the spine, pre-creating a `main` agent from this app's `model` with `process.cwd()` as the fresh session cwd and carrying its `persona` |
+| `@deepseek-ai/dsh-agent-spine-demo` | the spine, pre-creating a `main` agent from this app's provider/model pair with `process.cwd()` as the fresh session cwd and carrying its `persona` |
 | `@deepseek-ai/dsh-session-persistence-jsonl` | durable JSONL session log under `persistenceRoot` |
 | `@deepseek-ai/dsh-user-interaction` | the human question/answer seam used by confirmation tools |
 | `@deepseek-ai/dsh-tool-ask-user` | the model-facing `ask_user_question` tool |
@@ -19,17 +19,22 @@ A terminal chat always wants the same cluster, so the package owns it rather tha
 
 `@cordisjs/plugin-hmr` (the dev/demo edit-reload loop) is deliberately a **leaf** entry, NOT baked in here: it is a Loader-only, subprocess-only dev plugin — its constructor throws without `node --expose-internals` + a live `loader`, and the in-process test tier cannot even import it (so a package whose `apply` statically pulled it in could never carry the per-file coverage gate). Unlike the console logger, a stray `hmr` is not a stdout-purity footgun, so leaving it at the leaf costs no safety. The `demo:echo` / `demo:repl` leaves load it and pass `--expose-internals`.
 
-The leaf `cordis.yml` supplies only the **swappable backends** — an LLM adapter (`llm-deepseek` for the real model, or the mock `mock-llm` for a demo) and a bash executor (`bash-local`) — `hmr`, plus this app's [`Config`](#config). The whole plugin tree a run loads is therefore: this app's cluster, the spine inside `agent-core`, `hmr`, and the two leaf backends.
+The leaf `cordis.yml` supplies only the **swappable backends** — an LLM adapter (`llm-deepseek` for the real model, or the mock `mock-llm` for a demo) and a bash executor (`bash-local`) — `hmr`, plus this app's [`Config`](#config). The whole plugin tree a run loads is therefore: this app's cluster, the spine inside `agent-spine-demo`, `hmr`, and the two leaf backends.
 
 ## Config
 
 | Key | Default | Routed to |
 |---|---|---|
+| `provider` | (required) | the pre-created `main` agent's registered provider route |
 | `model` | (required) | the pre-created `main` agent's model |
-| `persona` | — | the deployment persona template (may reference `{{model}}`/`{{cwd}}`), routed to `dsh-system-prompt` |
+| `maxParallelToolCalls` | agent-loop default | positive-integer concurrent tool-call cap shared by the bundled loop's agents; `1` is serial |
+| `persona` | — | the deployment persona template (may reference `{{provider}}`/`{{model}}`/`{{cwd}}`), routed to `dsh-system-prompt` |
 | `toolOrder` | — | explicit model-facing tool order (a name list with one `'<unlisted-tools>'` rest entry; absent — lexicographic; an unregistered name fails each turn at prompt assembly), routed to `dsh-system-prompt` |
+| `dshHome` | `$DSH_HOME` or `~/.dsh` | Harness home exposed to model bash and used by local skill discovery |
 | `tools` | `{ mode: 'native' }` | tool-registry presentation config (`native` / `code` / `both`), routed through `dsh-agent-spine-demo` |
 | `skills` | owner defaults | registry-cache, local-provider, and model-facing skill-tool config, routed through `dsh-agent-spine-demo` |
+| `toolBash` | owner defaults | model-facing bash config routed through `dsh-agent-spine-demo`, including bash's producer-local `enableRunInBackground` |
+| `toolTasks` | owner defaults | generic `task_output` wait bounds routed through `dsh-agent-spine-demo` |
 | `persistenceRoot` | `./.sessions` | the JSONL backend's root directory |
 | `welcome` | `ready.` | the stdin-chat banner |
 | `resumeSessionId` | — | resume a persisted session id instead of starting fresh (sourced from an env var in the leaf) |
@@ -52,7 +57,6 @@ Fresh stdio sessions use the process launch directory as `session.header.cwd`, s
   name: '@deepseek-ai/dsh-llm-deepseek'
   config:
     apiKey: !!js process.env.DEEPSEEK_API_KEY
-    models: [deepseek-v4-flash]
 - id: bash
   name: '@deepseek-ai/dsh-bash-local'
   config:
@@ -60,6 +64,7 @@ Fresh stdio sessions use the process launch directory as `session.header.cwd`, s
 - id: stdio-agent
   name: '@deepseek-ai/dsh-stdio-demo'
   config:
+    provider: deepseek
     model: deepseek-v4-flash
     persona: 'You are a coding assistant powered by the {{model}} model.'
 ```
diff --git a/packages/examples/stdio-demo/package.json b/packages/examples/stdio-demo/package.json
index 3fe11f47a2..a523f61213 100644
--- a/packages/examples/stdio-demo/package.json
+++ b/packages/examples/stdio-demo/package.json
@@ -37,6 +37,7 @@
     "@deepseek-ai/dsh-agent": "^0.0.1",
     "@deepseek-ai/dsh-llm": "^0.0.1",
     "@deepseek-ai/dsh-agent-spine-demo": "^0.0.1",
+    "@deepseek-ai/dsh-workspace-context": "^0.0.1",
     "@deepseek-ai/dsh-session": "^0.0.1",
     "@deepseek-ai/dsh-session-persistence-jsonl": "^0.0.1",
     "@deepseek-ai/dsh-stdio": "^0.0.1",
@@ -55,6 +56,7 @@
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-agent-spine-demo": "workspace:^",
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-workspace-context": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
     "@deepseek-ai/dsh-session-persistence-jsonl": "workspace:^",
     "@deepseek-ai/dsh-stdio": "workspace:^",
diff --git a/packages/examples/stdio-demo/src/index.ts b/packages/examples/stdio-demo/src/index.ts
index 89f4142af6..91c377e9f7 100644
--- a/packages/examples/stdio-demo/src/index.ts
+++ b/packages/examples/stdio-demo/src/index.ts
@@ -16,6 +16,7 @@ import { AgentId } from '@deepseek-ai/dsh-agent'
 import { SessionId } from '@deepseek-ai/dsh-session'
 import ToolRegistry, { type Config as ToolsConfig } from '@deepseek-ai/dsh-tools'
 import * as agentCore from '@deepseek-ai/dsh-agent-spine-demo'
+import * as workspaceContext from '@deepseek-ai/dsh-workspace-context'
 import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
 import UserInteractionService from '@deepseek-ai/dsh-user-interaction'
 import * as toolAskUser from '@deepseek-ai/dsh-tool-ask-user'
@@ -25,7 +26,7 @@ export const name = 'stdio-demo'
 
 /**
  * App config: the swappable per-demo values, each routed to where the app wires
- * it. `model`/`resumeSessionId` configure the pre-created `main` agent (through
+ * it. `provider`/`model`/`resumeSessionId` configure the pre-created `main` agent (through
  * {@link @deepseek-ai/dsh-agent-spine-demo}'s forwarded `agents` list); `persona` is
  * the deployment persona (forwarded to the system-prompt plugin); `toolOrder`
  * is the explicit model-facing tool order (forwarded to the system-prompt plugin);
@@ -34,42 +35,60 @@ export const name = 'stdio-demo'
  * `welcome` is the UI banner.
  */
 export interface Config {
+  /** Provider route for the `main` agent. */
+  provider: string
   /** Model name for the `main` agent (must have a registered adapter). */
   model: string
+  /** Bundled agent-loop concurrency cap; `1` is serial and omission uses its default. */
+  maxParallelToolCalls?: number
   /** Deployment persona (the system-prompt plugin's `persona` config). */
   persona?: string
   /** Explicit model-facing tool order (the system-prompt plugin's `toolOrder` config; see dsh-system-prompt). */
   toolOrder?: string[]
   /** Tool-registry config — its presentation `mode` (forwarded through agent-spine-demo; see dsh-tools). */
   tools?: ToolsConfig
+  /** DeepSeek Harness home directory exposed to bash and used for local skill discovery. */
+  dshHome?: string
   /** Directory the JSONL session backend writes under. Defaults to `./.sessions`. */
   persistenceRoot?: string
   /** stdin-chat banner printed once on start. Defaults to `'ready.'`. */
   welcome?: string
   /** Skill registry, local-provider, and model-facing consumer config forwarded to agent-spine-demo. */
   skills?: agentCore.SkillConfig
+  /** Model-facing bash tool config forwarded through agent-core. */
+  toolBash?: NonNullable<agentCore.Config['toolBash']>
+  /** Generic background-task control-tool config forwarded through agent-core. */
+  toolTasks?: NonNullable<agentCore.Config['toolTasks']>
   /**
    * If set, the `main` agent RESUMES this persisted session id instead of
    * starting fresh. Sourced from an env var in the leaf `cordis.yml`
    * (`resumeSessionId: !!js process.env.RESUME_SESSION_ID`).
    */
   resumeSessionId?: string
+  /** Controls automatic AGENTS.md/CLAUDE.md loading; configure a byte budget or set `false`. */
+  workspaceContext: agentCore.Config['workspaceContext']
 }
 
 export const Config: z<Config> = z.object({
+  provider: z.string().required(),
   model: z.string().required(),
+  maxParallelToolCalls: z.number().step(1).min(1),
   persona: z.string(),
   // The array default is forced to undefined: ABSENT means "lexicographic
   // order" (the owning dsh-system-prompt schema does the same), while
   // schemastery's native [] default would read as an invalid configured list.
   toolOrder: z.array(z.string()).default(undefined as unknown as string[]),
   tools: ToolRegistry.Config,
+  dshHome: z.string(),
   // TODO(single-default-literal): share these schema defaults and defensive
   // apply() fallbacks through named constants while retaining both boundaries.
   persistenceRoot: z.string().default('./.sessions'),
   welcome: z.string().default('ready.'),
   skills: agentCore.SkillConfigSchema,
+  toolBash: agentCore.ToolBashConfigSchema,
+  toolTasks: agentCore.ToolTasksConfigSchema,
   resumeSessionId: z.string(),
+  workspaceContext: z.union([z.const(false), workspaceContext.Config]).required(),
 })
 
 /**
@@ -82,16 +101,14 @@ export const Config: z<Config> = z.object({
 export function apply(ctx: Context, config: Config): void {
   ctx.plugin(ConsoleExporter)
   ctx.plugin(agentCore, {
-    ...config.persona !== undefined ? { persona: config.persona } : {},
-    ...config.toolOrder !== undefined ? { toolOrder: config.toolOrder } : {},
-    ...config.tools !== undefined ? { tools: config.tools } : {},
+    ...agentCore.pickSpineConfig(config),
     agents: [{
       id: AgentId('main'),
+      provider: config.provider,
       model: config.model,
       cwd: process.cwd(),
       ...config.resumeSessionId !== undefined ? { resumeSessionId: SessionId(config.resumeSessionId) } : {},
     }],
-    ...config.skills !== undefined ? { skills: config.skills } : {},
   })
   ctx.plugin(SessionPersistenceJsonl, { root: config.persistenceRoot ?? './.sessions' })
   ctx.plugin(UserInteractionService)
diff --git a/packages/examples/stdio-demo/tests/built-bin.e2e.ts b/packages/examples/stdio-demo/tests/built-bin.e2e.ts
index ec14440b18..d3e0a32d09 100644
--- a/packages/examples/stdio-demo/tests/built-bin.e2e.ts
+++ b/packages/examples/stdio-demo/tests/built-bin.e2e.ts
@@ -21,9 +21,9 @@ const stdioBin = join(repoRoot, 'packages/examples/stdio-demo/lib/bin.js')
 const dshPackages = [
   'examples/agent-spine-demo', 'core/agent', 'core/session', 'core/system-prompt',
   'core/tools', 'core/agent-loop', 'llm/llm', 'bash/bash', 'bash/bash-local',
-  'bash/tool-bash', 'support/invariants', 'ui/app-boot',
+  'bash/tool-bash', 'context/workspace-context', 'support/invariants', 'ui/app-boot',
   'session-persistence/session-persistence',
-  'session-persistence/session-persistence-jsonl', 'examples/stdio-demo',
+  'session-persistence/session-persistence-jsonl', 'examples/stdio-demo', 'util/paths',
   'ui/stdio', 'ui/tool-ask-user', 'ui/user-interaction',
 ]
 const vendorPackages = [
@@ -36,20 +36,37 @@ async function pkgName(absDir: string): Promise<string> {
   return json.name
 }
 
+async function installWorkspacePackageCopy(absDir: string, target: string): Promise<void> {
+  await mkdir(dirname(target), { recursive: true })
+  await cp(absDir, target, {
+    recursive: true,
+    filter: source => !source.split('/').includes('node_modules'),
+  })
+}
+
 /**
  * Build a temporary external consumer with built workspace/vendor links and a mock-backed config.
  * The optional missing-but-disabled plugin verifies load guards accept intentionally fiber-less
  * entries rather than treating them as import failures.
  */
-async function makeConsumer(welcome: string, disabledBrokenEntry = false): Promise<string> {
+async function makeConsumer(
+  welcome: string,
+  disabledBrokenEntry = false,
+  extraDshPackages: string[] = [],
+  extraEntries: string[] = [],
+): Promise<string> {
   const dir = await mkdtemp(join(tmpdir(), 'stdio-built-bin-'))
   const nm = join(dir, 'node_modules')
-  for (const rel of dshPackages) {
+  for (const rel of [...dshPackages, ...extraDshPackages]) {
     const abs = join(repoRoot, 'packages', rel)
     const name = await pkgName(abs)
     const target = join(nm, name)
-    await mkdir(dirname(target), { recursive: true })
-    await symlink(abs, target)
+    if (extraDshPackages.includes(rel)) {
+      await installWorkspacePackageCopy(abs, target)
+    } else {
+      await mkdir(dirname(target), { recursive: true })
+      await symlink(abs, target)
+    }
   }
   for (const v of vendorPackages) {
     const abs = join(repoRoot, 'vendor', v)
@@ -73,9 +90,12 @@ async function makeConsumer(welcome: string, disabledBrokenEntry = false): Promi
     '- id: stdio-agent',
     '  name: \'@deepseek-ai/dsh-stdio-demo\'',
     '  config:',
+    '    provider: mock',
     '    model: mock-echo',
     '    persona: \'demo\'',
+    '    workspaceContext: false',
     `    welcome: '${welcome}'`,
+    ...extraEntries,
     ...disabledBrokenEntry
       ? ['- id: off', '  name: \'./src/does-not-exist.ts\'', '  disabled: true']
       : [],
@@ -147,6 +167,27 @@ describe.skipIf(!existsSync(stdioBin))('dsh-stdio-demo BUILT bin (node lib/bin.j
     expect(code).toBe(0)
   }, 30_000)
 
+  it('boots when optional spill plugins are loaded from a built consumer install', async () => {
+    consumer = await makeConsumer(
+      'SPILL-OK ready.',
+      false,
+      ['spill/spill', 'spill/spill-local', 'spill/spill-policy', 'util/retention'],
+      [
+        '- id: spill-local',
+        '  name: \'@deepseek-ai/dsh-spill-local\'',
+        '- id: spill-policy',
+        '  name: \'@deepseek-ai/dsh-spill-policy\'',
+        '  config:',
+        '    maxInlineBytes: 50000',
+      ],
+    )
+    const { stdout, code, stderr } = await runBuiltBin(consumer, './cordis.yml', '')
+    expect(stderr).not.toContain('failed to load')
+    expect(stderr).not.toContain('Cannot find package')
+    expect(stdout).toContain('SPILL-OK ready.')
+    expect(code).toBe(0)
+  }, 30_000)
+
   it('fails LOUD (non-zero exit + stderr) on a config whose directory does not exist', async () => {
     // boot() pre-resolves the bootstrap include to an absolute URL, so a nonexistent config
     // directory cannot break its import; the include plugin's own read must fail loud instead.
diff --git a/packages/examples/stdio-demo/tests/stdio-agent.spec.ts b/packages/examples/stdio-demo/tests/stdio-agent.spec.ts
index 20bcaa5704..4c81eb93e4 100644
--- a/packages/examples/stdio-demo/tests/stdio-agent.spec.ts
+++ b/packages/examples/stdio-demo/tests/stdio-agent.spec.ts
@@ -11,12 +11,13 @@ import * as stdioAgent from '../src/index.ts'
 
 /**
  * Unit coverage for app composition and config forwarding: console logger, pre-created main agent,
- * agent-core spine, JSONL backend, and readline UI. HMR is a Loader-only leaf concern covered by the
+ * agent-spine-demo spine, JSONL backend, and readline UI. HMR is a Loader-only leaf concern covered by the
  * keyless echo smoke; this tier pins the export shape because an inject-less app could otherwise
  * survive namespace collapse while silently losing its schema.
  */
-async function mount(config: stdioAgent.Config): Promise<Context> {
+async function mount(config: stdioAgent.Config, withBash = false): Promise<Context> {
   const ctx = new Context()
+  if (withBash) ctx.provide('bash', { sandboxMode: undefined })
   await ctx.plugin(stdioAgent, config)
   // The app mounts its children inside apply() (not awaited there); let their
   // fibers settle so the spine services + the pre-created agent are ready.
@@ -65,8 +66,8 @@ async function withIsolatedSkillHomes<T>(run: () => Promise<T>): Promise<T> {
 
 describe('dsh-stdio-demo app', () => {
   it('composes the spine + front-door cluster and pre-creates the main agent', async () => {
-    const ctx = await mount({ model: 'mock', persona: 'hi', persistenceRoot: '/tmp/dsh-stdio-demo-spec', skills: await isolatedSkillsConfig() })
-    // The spine services (brought up by the agent-core bundle) are all present.
+    const ctx = await mount({ provider: 'mock', model: 'mock', persona: 'hi', persistenceRoot: '/tmp/dsh-stdio-demo-spec', skills: await isolatedSkillsConfig(), workspaceContext: false })
+    // The spine services (brought up by the agent-spine-demo bundle) are all present.
     expect(ctx.get('agents')).toBeDefined()
     expect(ctx.get('agentLoop')).toBeDefined()
     expect(ctx.get('sessionPersistence')).toBeDefined()
@@ -86,17 +87,29 @@ describe('dsh-stdio-demo app', () => {
     // schema-bypassing direct-mount caller.
     const ctx = new Context()
     // No persona: covers the omitted-persona forwarding branch too.
-    stdioAgent.apply(ctx, { model: 'mock', skills: await isolatedSkillsConfig() })
+    stdioAgent.apply(ctx, { provider: 'mock', model: 'mock', skills: await isolatedSkillsConfig(), workspaceContext: false })
     await new Promise(resolve => setTimeout(resolve, 80))
     expect(ctx.get('sessionPersistence')).toBeDefined()
     expect(ctx.get('agents')?.get(AgentId('main'))).toBeDefined()
     await ctx.fiber.dispose()
   })
 
+  it('forwards explicit project-instruction controls to the bundled spine', async () => {
+    const ctx = await mount({
+      provider: 'mock',
+      model: 'mock',
+      persona: 'hi',
+      persistenceRoot: '/tmp/dsh-stdio-demo-spec-workspace-context',
+      workspaceContext: false,
+    })
+    expect(ctx.get('agents')?.get(AgentId('main'))).toBeDefined()
+    await ctx.fiber.dispose()
+  })
+
   it('uses default skill config when apply is called directly without skills', async () => {
     await withIsolatedSkillHomes(async () => {
       const ctx = new Context()
-      stdioAgent.apply(ctx, { model: 'mock' })
+      stdioAgent.apply(ctx, { provider: 'mock', model: 'mock', workspaceContext: false })
       await new Promise(resolve => setTimeout(resolve, 80))
       expect(ctx.skills).toBeDefined()
       expect(await ctx.skills.list()).toEqual([])
@@ -109,33 +122,66 @@ describe('dsh-stdio-demo app', () => {
     // session the resume is contained + logged, so no `main` agent registers —
     // the branch that maps resumeSessionId through is what this covers.
     const ctx = await mount({
+      provider: 'mock',
       model: 'mock',
       persona: 'hi',
       persistenceRoot: '/tmp/dsh-stdio-demo-spec-resume',
       resumeSessionId: 'no-such-session',
       skills: await isolatedSkillsConfig(),
+      workspaceContext: false,
     })
     expect(ctx.get('agents')?.get(AgentId('main'))).toBeUndefined()
     await ctx.fiber.dispose()
   })
 
-  it('forwards skill config into agent-core', async () => {
-    const ctx = await mount({ model: 'mock', persona: 'hi', skills: await isolatedSkillsConfig(6) })
+  it('forwards skill config and dshHome into agent-spine-demo', async () => {
+    const skills = await isolatedSkillsConfig(6)
+    const ctx = await mount({ provider: 'mock', model: 'mock', persona: 'hi', dshHome: skills.local!.dshHome!, skills, workspaceContext: false })
     ctx.skills.register({ name: 'stdio-skill', description: 'Stdio skill', source: 'runtime', content: 'body' })
     expect(JSON.stringify(await composePrefix(ctx))).toContain('- `stdio-skill`: Std...')
     await ctx.fiber.dispose()
   })
 
+  it('forwards maxParallelToolCalls to the bundled agent loop', async () => {
+    const ctx = await mount({
+      provider: 'mock',
+      model: 'mock',
+      maxParallelToolCalls: 3,
+      persistenceRoot: '/tmp/dsh-stdio-demo-spec-parallel',
+      skills: await isolatedSkillsConfig(),
+      workspaceContext: false,
+    })
+    expect(ctx.get('agentLoop')?.config.maxParallelToolCalls).toBe(3)
+    await ctx.fiber.dispose()
+  })
+
+  it('forwards bundled tool config into agent-core', async () => {
+    const ctx = await mount({
+      provider: 'mock',
+      model: 'mock',
+      workspaceContext: false,
+      toolBash: { enableRunInBackground: false },
+      toolTasks: { waitTimeoutMs: 7, maxWaitTimeoutMs: 11 },
+      skills: await isolatedSkillsConfig(),
+    }, true)
+    const bash = ctx.tools.schemas().find(tool => tool.name === 'bash')
+    expect(Object.keys((bash!.parameters as { properties: Record<string, unknown> }).properties))
+      .not.toContain('run_in_background')
+    await ctx.fiber.dispose()
+  })
+
   it('exposes its name and Config schema', () => {
     expect(stdioAgent.name).toBe('stdio-demo')
     expect(stdioAgent.Config).toBeDefined()
   })
 
-  it('forwards toolOrder through agent-core to the system-prompt assembly', async () => {
+  it('forwards toolOrder through agent-spine-demo to the system-prompt assembly', async () => {
     const ctx = await mount({
+      provider: 'mock',
       model: 'mock',
       toolOrder: ['zulu', TOOL_ORDER_REST],
       persistenceRoot: '/tmp/dsh-stdio-demo-spec-tool-order',
+      workspaceContext: false,
     })
     // The bundle's own bash tools pend on the absent `ctx.bash` executor in
     // this providerless mount, so register two plain tools to order.
@@ -148,7 +194,7 @@ describe('dsh-stdio-demo app', () => {
       })
     }
     const assembly = await ctx.get('systemPrompt')!.assemble()
-    expect(assembly.tools.map(tool => tool.name)).toEqual(['zulu', 'alpha', 'ask_user_question', 'skill'])
+    expect(assembly.tools.map(tool => tool.name)).toEqual(['zulu', 'alpha', 'ask_user_question', 'skill', 'task_kill', 'task_list', 'task_output'])
     await ctx.fiber.dispose()
   })
 
diff --git a/packages/examples/stdio-demo/tsconfig.json b/packages/examples/stdio-demo/tsconfig.json
index bb360810a5..be08d28f95 100644
--- a/packages/examples/stdio-demo/tsconfig.json
+++ b/packages/examples/stdio-demo/tsconfig.json
@@ -32,6 +32,9 @@
     {
       "path": "../agent-spine-demo"
     },
+    {
+      "path": "../../context/workspace-context"
+    },
     {
       "path": "../../ui/user-interaction"
     },
diff --git a/packages/fs/README.md b/packages/fs/README.md
index ec3bb62afb..039cb39ae9 100644
--- a/packages/fs/README.md
+++ b/packages/fs/README.md
@@ -1,6 +1,6 @@
 # fs/ - filesystem capability family
 
-The filesystem stack: a provider seam (text IO + atomic mutation with an optional version guard), a local implementation, a policy gate plugin (observed-state + read-before-edit + version-guarded write/edit), and the model-facing file tools + executor. All **product** packages.
+The filesystem stack: a provider seam (text IO + atomic mutation with an optional version guard), a local implementation, a policy gate plugin (observed-state + read-before-edit + version-guarded write/edit), the model-facing file tools + executor, and the bash-backed discovery tools. All **product** packages.
 
 | Package | Role | ctx key |
 |---|---|---|
@@ -8,9 +8,10 @@ The filesystem stack: a provider seam (text IO + atomic mutation with an optiona
 | `fs-local/` | Local-filesystem `FileSystem` implementation | (registers `ctx.fs`) |
 | `fs-policy/` | Policy gate plugin: observed-state + read-before-edit + version-guarded write/edit, via the `fs/*` event gate | (no service — `fs/*` listeners) |
 | `tool-fs/` | Model-facing `read`/`write`/`edit` tools AND the executor (reads via `ctx.fs`, owns read windowing, dispatches `fs/*`) | (registers on `ctx.tools`) |
+| `tool-fs-search/` | Model-facing `glob`/`grep` discovery tools, backed by fixed ripgrep commands through the bash seam (`ctx.bash`), NOT by `ctx.fs` provider methods | (registers on `ctx.tools`) |
 
-The interface lives at `fs/fs/`. A sandboxed, remote, or project-scoped filesystem backend can replace `fs-local` without touching the seam, the policy gate, or the model-facing tool schemas. The policy (`fs-policy/`) is a plugin that participates only through the `fs/*` event gate, not a service the tool injects — so dropping it gracefully loses the policy and leaves the unconstrained bare provider rather than breaking the tool. A deployment that loads `tool-fs/` is expected to also load it.
+The interface lives at `fs/fs/`. A sandboxed, remote, or project-scoped filesystem backend can replace `fs-local` without touching the seam, the policy gate, or the model-facing tool schemas. The policy (`fs-policy/`) is a plugin that participates only through the `fs/*` event gate, not a service the tool injects — so dropping it gracefully loses the policy and leaves the unconstrained bare provider rather than breaking the tool. A deployment that loads `tool-fs/` is expected to also load it. Discovery (`tool-fs-search/`) deliberately does NOT extend the provider seam: search is a process-backed `rg` workflow on the bash executor, so filesystem backends stay free of a universal search contract; its results are follow-up-readable when the bash workdir and the `read` root are the same workspace (the co-located deployment its README documents).
 
 ## No timeouts on file IO
 
-`read`/`write`/`edit` take **no** `timeoutMs`, and the provider seam arms no deadline — unlike bash and web, which consume [`@deepseek-ai/dsh-timeout`](../util/timeout/README.md). A local syscall is best-effort-abortable at most: a timeout could not force an in-progress `fsync`/`rename` to stop, so a deadline here would be a knob that cannot deliver on its promise. Adding one would also be an implicit default in the exact place explicit-over-implicit forbids. Both reference agents (Claude Code, Codex) leave file IO untimed for the same reason; cancellation still propagates through the tool-execution signal for best-effort abort at syscall boundaries.
+`read`/`write`/`edit` take **no** `timeoutMs`, and the provider seam arms no deadline — unlike bash and web (which consume [`@deepseek-ai/dsh-timeout`](../util/timeout/README.md)) and the bash-backed `glob`/`grep` (whose declared `timeoutMs` is enforced by `@deepseek-ai/dsh-timeout-policy`): those are process-backed, where a deadline can really kill the work. A local syscall is best-effort-abortable at most: a timeout could not force an in-progress `fsync`/`rename` to stop, so a deadline here would be a knob that cannot deliver on its promise. Adding one would also be an implicit default in the exact place explicit-over-implicit forbids. Both reference agents (Claude Code, Codex) leave file IO untimed for the same reason; cancellation still propagates through the tool-execution signal for best-effort abort at syscall boundaries.
diff --git a/packages/fs/fs-local/README.md b/packages/fs/fs-local/README.md
index e950385223..65ed76efce 100644
--- a/packages/fs/fs-local/README.md
+++ b/packages/fs/fs-local/README.md
@@ -1,6 +1,6 @@
 # @deepseek-ai/dsh-fs-local
 
-The **local-filesystem implementation** of the `ctx.fs` provider seam ([`@deepseek-ai/dsh-fs`](../fs)). Backs the seven `FileSystem` primitives with the host filesystem; loading it as a plugin populates `ctx.fs`.
+The **local-filesystem implementation** of the `ctx.fs` provider seam ([`@deepseek-ai/dsh-fs`](../fs)). Backs the eight `FileSystem` primitives with the host filesystem; loading it as a plugin populates `ctx.fs`.
 
 ```ts ignore-check
 import { LocalFileSystem } from '@deepseek-ai/dsh-fs-local'
@@ -12,8 +12,8 @@ await ctx.plugin(LocalFileSystem, { cwd: process.cwd() })
 
 ## Behavior
 
-- **`resolve(path, opts?)`** — a relative `path` resolves against `opts.cwd` when the caller supplies one (the model-facing tools pass the calling agent's session cwd — see [the per-session cwd RFC](../../../docs/rfc/implemented/architecture/2026-07-02-fs-per-session-cwd.md)), else `config.cwd` (default `process.cwd()`); an absolute `path` ignores both. The `targetKey` is the file's `realpath`, so two input paths reaching the same file through symlinks share one identity, and writes/edits land on the link target (preserving the link). A not-yet-existing path uses the realpathed parent directory plus basename when the parent exists; only an unresolvable parent falls back to the absolute path. `displayPath` is the absolute (un-resolved) path.
-- **`stat`** — returns `FsInfo` (`version` = `mtimeMs:size`, `type` of `file`/`directory`/`other`, byte `size`) or `undefined` when the target is absent.
+- **`resolve(path, opts?)`** — a relative `path` resolves against `opts.cwd` when the caller supplies one (the model-facing tools pass the calling agent's session cwd — see [the per-session cwd RFC](../../../docs/rfc/implemented/architecture/2026-07-02-fs-per-session-cwd.md)), else `config.cwd` (default `process.cwd()`); an absolute `path` ignores both. `opts.signal` is checked before and after local resolution, while a remote sibling backend may use it to abort its round-trip. The `targetKey` is the file's `realpath`, so two input paths reaching the same file through symlinks share one identity, and writes/edits land on the link target (preserving the link). A not-yet-existing path uses the realpathed parent directory plus basename when the parent exists; only an unresolvable parent falls back to the absolute path. `displayPath` is the absolute (un-resolved) path.
+- **`stat` / `lstat`** — return target metadata or `undefined` when absent. `stat` reports `FsInfo` for an already resolved target (`version` = an opaque token derived from bigint `dev:ino:size:mtimeNs:ctimeNs`, `type` of `file`/`directory`/`other`, byte `size`); path-shaped `lstat` reports `FsPathInfo` without following the final symlink and can therefore return `symlink`. Both check cancellation before and after their asynchronous metadata probe, so an abort that lands in flight reports `FS_ABORTED` rather than stale absence.
 - **`readText` / `streamText`** — UTF-8 only. `readText` reads the whole file; `streamText` streams it in chunks (cross-chunk decoding) so a huge file never has to be held whole in memory. Both reject invalid UTF-8 and NUL-byte binary samples (`FS_NOT_TEXT`) and non-regular targets. The `read` tool (`@deepseek-ai/dsh-tool-fs`) decides which to call by size and owns the line windowing.
 - **`listDir`** — lists one directory level in stable `name.localeCompare()` order. Each entry carries the child basename, type, resolved child target (`displayPath` under the listed directory, `targetKey` as the realpath identity), and cheap stat metadata (`version`, plus `size` for regular files). It never opens or decodes file contents. Missing targets report `FS_NOT_FOUND`, file/special-file targets report `FS_NOT_DIRECTORY`, aborted calls report `FS_ABORTED`, permission failures report `FS_PERMISSION_DENIED`, and other listing or child metadata I/O failures report `FS_IO_ERROR`. Broken/disappeared children are returned as `other` without metadata, but permission/IO failures while resolving a child fail the whole listing with a structured `FsError`.
 - **`writeText`** — atomic: writes to a temp file opened exclusively (`wx`, `0o600`) inside a randomly-named private staging dir (`0o700`) next to the target, fsyncs, then renames over the target. An existing file's mode is preserved, while new files default to `0o600`. The `expected` guard is OPTIONAL: omitting it unconditionally creates-or-overwrites; `createIfAbsent` creates a missing target and rejects an existing one (`FS_NOT_OBSERVED`); `replaceIfVersion` replaces only at the observed version (a missing target or mismatch is `FS_STALE_VERSION`).
diff --git a/packages/fs/fs-local/src/fsio.ts b/packages/fs/fs-local/src/fsio.ts
index fba2a240eb..360145e8c8 100644
--- a/packages/fs/fs-local/src/fsio.ts
+++ b/packages/fs/fs-local/src/fsio.ts
@@ -7,8 +7,8 @@
 
 import { randomUUID } from 'node:crypto'
 import { createReadStream } from 'node:fs'
-import { chmod, mkdir, open, readFile, realpath, readdir, rename, rm, stat } from 'node:fs/promises'
-import type { Dirent, Stats } from 'node:fs'
+import { chmod, lstat, mkdir, open, readFile, realpath, readdir, rename, rm, stat } from 'node:fs/promises'
+import type { BigIntStats, Dirent, Stats } from 'node:fs'
 import { basename, dirname, join, resolve } from 'node:path'
 import { TextDecoder } from 'node:util'
 import { FsError, FsTargetKey, FsVersion } from '@deepseek-ai/dsh-fs'
@@ -63,9 +63,9 @@ async function readFileAbortable(absolutePath: string, verb: 'read' | 'edit', si
   }
 }
 
-/** Opaque version token from a stat: mtime (ns precision) + size. */
-function versionOf(info: Stats): FsVersion {
-  return FsVersion(`${info.mtimeMs}:${info.size}`)
+/** Opaque version token from high-resolution identity and freshness metadata. */
+function versionOf(info: BigIntStats): FsVersion {
+  return FsVersion(`${info.dev}:${info.ino}:${info.size}:${info.mtimeNs}:${info.ctimeNs}`)
 }
 
 /**
@@ -98,6 +98,14 @@ export interface PathInfo {
   size: number
 }
 
+/** Result of probing a path without following the final symlink component. */
+export interface PathLinkInfo {
+  version: FsVersion
+  mode: number
+  type: 'file' | 'directory' | 'symlink' | 'other'
+  size: number
+}
+
 /** One local directory child with a resolved target and cheap metadata. */
 export interface LocalDirEntry {
   name: string
@@ -150,22 +158,62 @@ export async function resolveLocalTarget(cwd: string, path: string): Promise<Loc
   }
 }
 
+function pathType(info: Stats | BigIntStats): PathInfo['type'] {
+  if (info.isFile()) return 'file'
+  if (info.isDirectory()) return 'directory'
+  return 'other'
+}
+
+function pathLinkType(info: Stats | BigIntStats): PathLinkInfo['type'] {
+  if (info.isSymbolicLink()) return 'symlink'
+  return pathType(info)
+}
+
+async function probeStats<T extends Stats | BigIntStats>(
+  absolutePath: string,
+  readStats: (path: string) => Promise<T>,
+): Promise<T | null> {
+  try {
+    return await readStats(absolutePath)
+  } catch (error: unknown) {
+    // ENOENT (no such file) and ENOTDIR (a parent segment is a file) both mean
+    // the target is absent; any other metadata failure is a real permission/IO
+    // fault.
+    /* v8 ignore next -- a non-ENOENT/ENOTDIR metadata failure needs a permission/IO fault; surface it. */
+    if (!isENOENT(error) && !isENOTDIR(error)) throw error
+    return null
+  }
+}
+
 /**
  * Probe a path for its version, mode, type, and size. Null if absent.
  * @param absolutePath - the path to stat (typically a target key; symlinks are followed).
  * @returns the metadata, or null when the path — or a parent segment — does not exist.
  */
 export async function probe(absolutePath: string): Promise<PathInfo | null> {
-  try {
-    const info = await stat(absolutePath)
-    const type = info.isFile() ? 'file' : info.isDirectory() ? 'directory' : 'other'
-    return { version: versionOf(info), mode: info.mode & 0o777, type, size: info.size }
-  } catch (error: unknown) {
-    // ENOENT (no such file) and ENOTDIR (a parent segment is a file) both mean
-    // the target is absent; any other stat failure is a real permission/IO fault.
-    /* v8 ignore next -- a non-ENOENT/ENOTDIR stat failure needs a permission/IO fault; surface it. */
-    if (!isENOENT(error) && !isENOTDIR(error)) throw error
-    return null
+  const info = await probeStats(absolutePath, path => stat(path, { bigint: true }))
+  if (!info) return null
+  return {
+    version: versionOf(info),
+    mode: Number(info.mode & 0o777n),
+    type: pathType(info),
+    size: Number(info.size),
+  }
+}
+
+/**
+ * Probe a path without following the final symlink component.
+ * @param absolutePath - the path entry to inspect with `lstat` semantics.
+ * @returns path-entry metadata, or null when the entry is absent.
+ */
+export async function probeNoFollow(absolutePath: string): Promise<PathLinkInfo | null> {
+  const info = await probeStats(absolutePath, path => lstat(path, { bigint: true }))
+  if (!info) return null
+  return {
+    version: versionOf(info),
+    mode: Number(info.mode & 0o777n),
+    type: pathLinkType(info),
+    size: Number(info.size),
   }
 }
 
diff --git a/packages/fs/fs-local/src/index.ts b/packages/fs/fs-local/src/index.ts
index 43ca988338..bd47b53919 100644
--- a/packages/fs/fs-local/src/index.ts
+++ b/packages/fs/fs-local/src/index.ts
@@ -5,6 +5,7 @@
  */
 
 import { Context } from 'cordis'
+import { resolve } from 'node:path'
 import z from 'schemastery'
 import { FileSystem, FsError, FsVersion } from '@deepseek-ai/dsh-fs'
 import type {
@@ -12,6 +13,7 @@ import type {
   FsEditOutcome,
   FsEditRequest,
   FsInfo,
+  FsPathInfo,
   FsTarget,
   FsWriteIntent,
   FsWriteOutcome,
@@ -21,6 +23,7 @@ import {
   listDirectory,
   normalizeLineEndings,
   probe,
+  probeNoFollow,
   readForEdit,
   readTextForDiff,
   readWholeText,
@@ -80,14 +83,26 @@ export class LocalFileSystem extends FileSystem {
     }
   }
 
-  override async resolve(path: string, opts?: { cwd?: string }): Promise<FsTarget> {
+  override async resolve(path: string, opts?: { cwd?: string; signal?: AbortSignal }): Promise<FsTarget> {
+    if (opts?.signal?.aborted) throw new FsError('resolve aborted', 'FS_ABORTED')
     const local = await resolveLocalTarget(opts?.cwd ?? this.config.cwd, path)
+    if (opts?.signal?.aborted) throw new FsError('resolve aborted', 'FS_ABORTED')
     return { targetKey: local.targetKey, displayPath: local.displayPath }
   }
 
   override async stat(target: FsTarget, signal?: AbortSignal): Promise<FsInfo | undefined> {
     if (signal?.aborted) throw new FsError('stat aborted', 'FS_ABORTED')
     const info = await probe(target.targetKey)
+    if (signal?.aborted) throw new FsError('stat aborted', 'FS_ABORTED')
+    if (!info) return undefined
+    return { version: info.version, type: info.type, size: info.size }
+  }
+
+  override async lstat(path: string, opts?: { cwd?: string }, signal?: AbortSignal): Promise<FsPathInfo | undefined> {
+    if (signal?.aborted) throw new FsError('lstat aborted', 'FS_ABORTED')
+    if (path.trim().length === 0) throw new FsError('file_path must be a non-empty string', 'FS_NOT_FOUND')
+    const info = await probeNoFollow(resolve(opts?.cwd ?? this.config.cwd, path))
+    if (signal?.aborted) throw new FsError('lstat aborted', 'FS_ABORTED')
     if (!info) return undefined
     return { version: info.version, type: info.type, size: info.size }
   }
diff --git a/packages/fs/fs-local/tests/filesystem.spec.ts b/packages/fs/fs-local/tests/filesystem.spec.ts
index 997021d346..61e2c0e999 100644
--- a/packages/fs/fs-local/tests/filesystem.spec.ts
+++ b/packages/fs/fs-local/tests/filesystem.spec.ts
@@ -6,8 +6,8 @@
  * `dsh-fs-policy`, so it is not exercised here.
  */
 
-import { afterEach, beforeEach, describe, expect, it } from 'vitest'
-import { mkdir, mkdtemp, readFile, realpath, rm, stat, symlink, writeFile, unlink } from 'node:fs/promises'
+import { afterEach, beforeEach, describe, expect, it, vi } from 'vitest'
+import { mkdir, mkdtemp, readFile, realpath, rm, stat, symlink, unlink, utimes, writeFile } from 'node:fs/promises'
 import { tmpdir } from 'node:os'
 import { join } from 'node:path'
 import { Context } from 'cordis'
@@ -73,6 +73,18 @@ describe('resolve', () => {
     const target = await fs.resolve(join(dir, 'abs.txt'), { cwd: '/nonexistent-base' })
     expect(await fs.readText(target)).toBe('absolute')
   })
+
+  it('honors a pre-aborted signal', async () => {
+    await expect(fs.resolve('a.txt', { signal: AbortSignal.abort() })).rejects.toMatchObject({ code: 'FS_ABORTED' })
+  })
+
+  it('honors a signal aborted while resolution is in flight', async () => {
+    const controller = new AbortController()
+    const pending = fs.resolve('a.txt', { signal: controller.signal })
+    controller.abort()
+
+    await expect(pending).rejects.toMatchObject({ code: 'FS_ABORTED' })
+  })
 })
 
 describe('stat', () => {
@@ -87,11 +99,104 @@ describe('stat', () => {
     expect(await fs.stat(await fs.resolve('missing.txt'))).toBeUndefined()
   })
 
+  it('changes version after a same-size rewrite even when mtime is restored', async () => {
+    const path = join(dir, 'same-size.txt')
+    await writeFile(path, 'first')
+    const target = await fs.resolve(path)
+    const beforeInfo = await stat(path)
+    const beforeVersion = await versionOf(target)
+
+    await fs.writeText(target, 'other')
+    await utimes(path, beforeInfo.atime, beforeInfo.mtime)
+
+    expect((await stat(path)).size).toBe(beforeInfo.size)
+    expect(await versionOf(target)).not.toBe(beforeVersion)
+  })
+
   it('honors a pre-aborted signal', async () => {
     await expect(fs.stat(await fs.resolve('a.txt'), AbortSignal.abort())).rejects.toMatchObject({ code: 'FS_ABORTED' })
   })
 })
 
+describe('lstat', () => {
+  it('reports path metadata without following the final symlink component', async () => {
+    await writeFile(join(dir, 'real.txt'), 'hello')
+    await symlink(join(dir, 'real.txt'), join(dir, 'link.txt'))
+
+    expect((await fs.lstat('real.txt'))?.type).toBe('file')
+    expect((await fs.lstat('link.txt'))?.type).toBe('symlink')
+    expect(await fs.lstat('missing.txt')).toBeUndefined()
+  })
+
+  it('resolves relative paths against opts.cwd and honors a pre-aborted signal', async () => {
+    const other = await mkdtemp(join(tmpdir(), 'dsh-fs-other-'))
+    try {
+      await writeFile(join(other, 'x.txt'), 'in other')
+      expect((await fs.lstat('x.txt', { cwd: other }))?.type).toBe('file')
+      await expect(fs.lstat('x.txt', { cwd: other }, AbortSignal.abort())).rejects.toMatchObject({ code: 'FS_ABORTED' })
+      await expect(fs.lstat('   ')).rejects.toMatchObject({ code: 'FS_NOT_FOUND' })
+    } finally {
+      await rm(other, { recursive: true, force: true })
+    }
+  })
+})
+
+describe('metadata cancellation', () => {
+  it('rejects stat and lstat when their signals abort while the metadata probes are in flight', async () => {
+    await writeFile(join(dir, 'slow.txt'), 'hello')
+    const statStarted = Promise.withResolvers<undefined>()
+    const statRelease = Promise.withResolvers<undefined>()
+    const lstatStarted = Promise.withResolvers<undefined>()
+    const lstatRelease = Promise.withResolvers<undefined>()
+    let isolatedCtx: Context | undefined
+    vi.resetModules()
+    vi.doMock('node:fs/promises', async (importOriginal) => {
+      const actual = await importOriginal<typeof import('node:fs/promises')>()
+      return {
+        ...actual,
+        async stat(path: string) {
+          statStarted.resolve(undefined)
+          await statRelease.promise
+          return actual.stat(path, { bigint: true })
+        },
+        async lstat(path: string) {
+          lstatStarted.resolve(undefined)
+          await lstatRelease.promise
+          return actual.lstat(path, { bigint: true })
+        },
+      }
+    })
+
+    try {
+      const { LocalFileSystem: IsolatedLocalFileSystem } = await import('../src/index.ts')
+      isolatedCtx = new Context()
+      await isolatedCtx.plugin(IsolatedLocalFileSystem, { cwd: dir })
+      const isolatedFs = isolatedCtx.fs as InstanceType<typeof IsolatedLocalFileSystem>
+      const target = await isolatedFs.resolve('slow.txt')
+      const statController = new AbortController()
+      const lstatController = new AbortController()
+      const pendingStat = isolatedFs.stat(target, statController.signal)
+      const pendingLstat = isolatedFs.lstat('slow.txt', undefined, lstatController.signal)
+
+      await Promise.all([statStarted.promise, lstatStarted.promise])
+      statController.abort()
+      lstatController.abort()
+      const statRejected = expect(pendingStat).rejects.toMatchObject({ code: 'FS_ABORTED' })
+      const lstatRejected = expect(pendingLstat).rejects.toMatchObject({ code: 'FS_ABORTED' })
+      statRelease.resolve(undefined)
+      lstatRelease.resolve(undefined)
+
+      await Promise.all([statRejected, lstatRejected])
+    } finally {
+      statRelease.resolve(undefined)
+      lstatRelease.resolve(undefined)
+      await isolatedCtx?.fiber.dispose()
+      vi.doUnmock('node:fs/promises')
+      vi.resetModules()
+    }
+  })
+})
+
 describe('readText / streamText', () => {
   it('reads whole-file text', async () => {
     await writeFile(join(dir, 'a.txt'), 'one\ntwo\nthree')
@@ -292,9 +397,6 @@ describe('writeText', () => {
     await writeFile(join(dir, 'a.txt'), 'v1')
     const target = await fs.resolve('a.txt')
     const before = await versionOf(target)
-    // Change the byte length so the mtimeMs:size token provably differs (a
-    // same-size same-tick rewrite can collide — the documented version-token
-    // limitation; not what this test is about).
     const outcome = await fs.writeText(target, 'a much longer replacement body', { kind: 'replaceIfVersion', version: before })
     expect(outcome.version).not.toBe(before)
     expect(outcome.version).toBe(await versionOf(target))
diff --git a/packages/fs/fs-local/tests/fsio.spec.ts b/packages/fs/fs-local/tests/fsio.spec.ts
index 6723ae9d9c..199c01f411 100644
--- a/packages/fs/fs-local/tests/fsio.spec.ts
+++ b/packages/fs/fs-local/tests/fsio.spec.ts
@@ -14,6 +14,7 @@ import {
   applyLiteralEdit,
   listDirectory,
   probe,
+  probeNoFollow,
   readForEdit,
   readWholeText,
   resolveLocalTarget,
@@ -146,6 +147,27 @@ describe('probe', () => {
   })
 })
 
+describe('probeNoFollow', () => {
+  it('reports symlinks without following them', async () => {
+    const real = join(dir, 'real.txt')
+    const link = join(dir, 'link.txt')
+    await writeFile(real, 'hi')
+    await symlink(real, link)
+
+    expect((await probeNoFollow(real))?.type).toBe('file')
+    const linkInfo = await probeNoFollow(link)
+    expect(linkInfo?.type).toBe('symlink')
+    expect(typeof linkInfo?.version).toBe('string')
+    expect(linkInfo?.size).toBeGreaterThan(0)
+  })
+
+  it('returns null for a missing path or a file-valued ancestor path segment', async () => {
+    expect(await probeNoFollow(join(dir, 'missing'))).toBeNull()
+    await writeFile(join(dir, 'afile'), 'i am a file')
+    expect(await probeNoFollow(join(dir, 'afile', 'child.txt'))).toBeNull()
+  })
+})
+
 describe('listDirectory', () => {
   it('lists direct children in stable order without reading content', async () => {
     const root = join(dir, 'skills')
diff --git a/packages/fs/fs/README.md b/packages/fs/fs/README.md
index bad4133b72..6dae32d235 100644
--- a/packages/fs/fs/README.md
+++ b/packages/fs/fs/README.md
@@ -1,6 +1,6 @@
 # @deepseek-ai/dsh-fs
 
-The **filesystem provider seam**: an abstract `FileSystem` service (`ctx.fs`) defining the storage primitives a backend provides — resolve a path, stat metadata, read/stream text, list directories, write atomically, and apply a literal edit — without saying HOW. Both mutations take their version guard **optionally**, so `ctx.fs` on its own is a complete, unconstrained text-storage seam. This package also owns the `fs/*` policy event vocabulary the tool dispatches and the policy plugin listens for.
+The **filesystem provider seam**: an abstract `FileSystem` service (`ctx.fs`) defining the storage primitives a backend provides — resolve a path, stat metadata, no-follow path metadata, read/stream text, list directories, write atomically, and apply a literal edit — without saying HOW. Both mutations take their version guard **optionally**, so `ctx.fs` on its own is a complete, unconstrained text-storage seam. This package also owns the `fs/*` policy event vocabulary the tool dispatches and the policy plugin listens for.
 
 This package is the provider-seam layer of the four-layer filesystem stack, split so each concern can evolve (and be swapped) independently (see [the capability-seam RFC](../../../docs/rfc/implemented/architecture/2026-06-13-capability-seams.md), [the filesystem capability-seam RFC](../../../docs/rfc/implemented/architecture/2026-06-17-filesystem-capability-seam.md), [the split-the-filesystem-seam RFC](../../../docs/rfc/implemented/simplification/2026-06-26-fsspec-style-fs-seam.md), and [the file-context event-gate RFC](../../../docs/rfc/implemented/architecture/2026-06-26-file-context-as-event-gate.md)):
 
@@ -15,12 +15,13 @@ A future sandboxed, virtual, or remote backend implements this interface and the
 
 ## Service API (`ctx.fs`)
 
-A backend subclasses `FileSystem` and implements seven primitives.
+A backend subclasses `FileSystem` and implements eight primitives.
 
 | Member | Semantics |
 |---|---|
-| `resolve(path, opts?)` | Resolve a path into a stable `FsTarget` (opaque `targetKey`, `displayPath`). `opts.cwd` is the base a relative `path` resolves against (a caller supplies its session workspace; absolute paths ignore it; omitted ⇒ the backend default). Async — a remote backend may need I/O. The same file via different paths must yield the same `targetKey`. |
+| `resolve(path, opts?)` | Resolve a path into a stable `FsTarget` (opaque `targetKey`, `displayPath`). `opts.cwd` is the base a relative `path` resolves against (a caller supplies its session workspace; absolute paths ignore it; omitted ⇒ the backend default), while `opts.signal` aborts a backend round-trip. Async — a remote backend may need I/O. The same file via different paths must yield the same `targetKey`. |
 | `stat(target, signal?)` | Return `FsInfo` metadata (`version`, `type`, optional `size`), or `undefined` when the target is absent. Never content. |
+| `lstat(path, opts?, signal?)` | Return `FsPathInfo` metadata without following the final path component when it is a symlink. This is path-shaped so consumers can reject repository-owned symlinks before `resolve` follows them into a target. |
 | `readText(target, signal?)` | Read the whole regular text file as one decoded string. Owns regular-file checks, UTF-8 decoding, binary/NUL rejection (`FS_NOT_TEXT`). |
 | `streamText(target, signal?)` | Stream the same text as decoded chunks for large files (cross-chunk UTF-8 decoding stays here). |
 | `listDir(target, signal?)` | List direct directory children in stable name order. Returns entry names, entry types, resolved child targets, and cheap metadata (`version`/file `size` when available); never reads file contents. Missing targets throw `FS_NOT_FOUND`, non-directories throw `FS_NOT_DIRECTORY`, permission failures throw `FS_PERMISSION_DENIED`, and other backend I/O failures throw `FS_IO_ERROR`. Broken/disappeared children may be returned as `other` without metadata; child permission/IO failures fail the whole listing with the same structured codes. |
@@ -41,7 +42,7 @@ This package declares three events (see the generated [events catalog](../../../
 
 ## Vocabulary
 
-`FsTargetKey` / `FsVersion` are branded opaque ids ([the branded-ids RFC](../../../docs/rfc/implemented/architecture/2026-06-20-branded-ids.md)) — consumers must not parse `targetKey` or interpret `version`; only `displayPath` is for model/UI output. `FsWriteIntent` is the explicit GUARDED write intent (`createIfAbsent` creates a missing target and rejects an existing one with `FS_NOT_OBSERVED`; `replaceIfVersion` replaces only at the observed version, else `FS_STALE_VERSION`); omitting it from `writeText` is the third, unconditional state. Failures throw `FsError` (extends `HarnessError`, [the structured error taxonomy RFC](../../../docs/rfc/implemented/architecture/2026-06-11-structured-error-taxonomy.md)) carrying a stable `FsErrorCode` (`FS_NOT_FOUND`, `FS_NOT_DIRECTORY`, `FS_NOT_TEXT`, `FS_NOT_REGULAR_FILE`, `FS_PERMISSION_DENIED`, `FS_IO_ERROR`, `FS_STALE_VERSION`, `FS_NOT_OBSERVED`, `FS_AMBIGUOUS_EDIT`, `FS_EDIT_NOT_FOUND`, `FS_ABORTED`); the tool registry surfaces `{ name, code }` on `isError` results. See `src/types.ts` for the full contracts.
+`FsTargetKey` / `FsVersion` are branded opaque ids ([the branded-ids RFC](../../../docs/rfc/implemented/architecture/2026-06-20-branded-ids.md)) — consumers must not parse `targetKey` or interpret `version`; only `displayPath` is for model/UI output. `FsWriteIntent` is the explicit GUARDED write intent (`createIfAbsent` creates a missing target and rejects an existing one with `FS_NOT_OBSERVED`; `replaceIfVersion` replaces only at the observed version, else `FS_STALE_VERSION`); omitting it from `writeText` is the third, unconditional state. `FsPathInfo` is the no-follow metadata shape that can report `symlink`, unlike target-level `FsInfo`. Failures throw `FsError` (extends `HarnessError`, [the structured error taxonomy RFC](../../../docs/rfc/implemented/architecture/2026-06-11-structured-error-taxonomy.md)) carrying a stable `FsErrorCode` (`FS_NOT_FOUND`, `FS_NOT_DIRECTORY`, `FS_NOT_TEXT`, `FS_NOT_REGULAR_FILE`, `FS_PERMISSION_DENIED`, `FS_IO_ERROR`, `FS_STALE_VERSION`, `FS_NOT_OBSERVED`, `FS_AMBIGUOUS_EDIT`, `FS_EDIT_NOT_FOUND`, `FS_ABORTED`); the tool registry surfaces `{ name, code }` on `isError` results. See `src/types.ts` for the full contracts.
 
 ## Model Experience
 
@@ -50,6 +51,6 @@ Indirectly, through `dsh-tool-fs`, which renders provider text and errors as bou
 ## Known Limitations and Deferred Work
 
 - **Text-only by contract** — backends reject binary/non-UTF-8 content with `FS_NOT_TEXT`; binary-safe operations are a deliberate deferral of [the tool-schemas RFC](../../../docs/rfc/implemented/feature/2026-06-17-filesystem-tool-schemas.md).
-- **Seven primitives only** — no delete, rename/move, copy, or watch; `listDir` is single-level, with recursion, globbing, pagination, and search out of scope per [the directory-listing RFC](../../../docs/rfc/implemented/architecture/2026-07-03-filesystem-directory-listing-seam.md).
+- **Eight primitives only** — no delete, rename/move, copy, or watch; `listDir` is single-level, with recursion, globbing, pagination, and search out of scope per [the directory-listing RFC](../../../docs/rfc/implemented/architecture/2026-07-03-filesystem-directory-listing-seam.md).
 - **No IO deadline** — the seam arms no timeout; cancellation is a best-effort optional `AbortSignal` per primitive (the deliberate [fs-family stance](../README.md)).
 - **Resolve-then-operate costs a remote backend two round-trips per tool call** — folding or caching resolution is left to such a backend.
diff --git a/packages/fs/fs/src/index.ts b/packages/fs/fs/src/index.ts
index 1c70226aff..8466962f5a 100644
--- a/packages/fs/fs/src/index.ts
+++ b/packages/fs/fs/src/index.ts
@@ -12,6 +12,7 @@ import type {
   FsEditOutcome,
   FsEditRequest,
   FsInfo,
+  FsPathInfo,
   FsTarget,
   FsVersion,
   FsWriteIntent,
@@ -29,6 +30,7 @@ export type {
   FsDirEntry,
   FsErrorCode,
   FsInfo,
+  FsPathInfo,
   FsTarget,
   FsWriteIntent,
   FsWriteOutcome,
@@ -86,10 +88,10 @@ export abstract class FileSystem extends Service {
    * async even though the local backend only normalizes + realpaths.
    *
    * @param path - the path to resolve; relative paths resolve against `opts.cwd`.
-   * @param opts - `cwd` overrides the backend's default base for relative paths.
+   * @param opts - optional cwd override and cancellation signal.
    * @returns the stable target; the same file yields the same `targetKey`.
    */
-  abstract resolve(path: string, opts?: { cwd?: string }): Promise<FsTarget>
+  abstract resolve(path: string, opts?: { cwd?: string; signal?: AbortSignal }): Promise<FsTarget>
 
   /**
    * Return target metadata, or `undefined` when the target does not exist.
@@ -99,6 +101,22 @@ export abstract class FileSystem extends Service {
    */
   abstract stat(target: FsTarget, signal?: AbortSignal): Promise<FsInfo | undefined>
 
+  /**
+   * Return path metadata without following the final path component when it is a
+   * symbolic link. This is intentionally path-shaped, not target-shaped:
+   * {@link resolve} follows symlinks to produce the stable identity used by
+   * normal reads/writes, while `lstat` lets a consumer reject the path itself
+   * before that follow happens.
+   *
+   * `opts.cwd` follows {@link resolve}'s cwd rules. `undefined` means the path is
+   * absent.
+   * @param path - the path to inspect; relative paths resolve against `opts.cwd`.
+   * @param opts - `cwd` overrides the backend's default base for relative paths.
+   * @param signal - aborts the metadata round-trip.
+   * @returns metadata only, never content; undefined for an absent path.
+   */
+  abstract lstat(path: string, opts?: { cwd?: string }, signal?: AbortSignal): Promise<FsPathInfo | undefined>
+
   /**
    * Read the whole regular text file as a single decoded string.
    * @param target - the resolved target to read.
diff --git a/packages/fs/fs/src/types.ts b/packages/fs/fs/src/types.ts
index 73c8ff4837..c76e88da4b 100644
--- a/packages/fs/fs/src/types.ts
+++ b/packages/fs/fs/src/types.ts
@@ -27,16 +27,17 @@ export function FsTargetKey(key: string): FsTargetKey {
 
 /**
  * Opaque file-version token — the freshness token a write/edit guards against.
- * The local backend derives it from mtime+size; a remote backend might use a
- * revision id. The policy layer records it for stale checks; consumers may
- * display related metadata but MUST NOT interpret this token.
+ * The local backend derives it from high-resolution stat identity and freshness
+ * fields; a remote backend might use a revision id. The policy layer records it
+ * for stale checks; consumers may display related metadata but MUST NOT
+ * interpret this token.
  */
 export type FsVersion = Branded<'FsVersion'>
 
 /**
  * Brand a string as an {@link FsVersion}. For backend use only — a consumer
  * never manufactures a version, it receives one from `stat`/write/edit outcomes.
- * @param v - the backend's raw version string (the local backend derives it from mtime+size).
+ * @param v - the backend's raw version string.
  * @returns the same string, branded; no validation is performed.
  */
 export function FsVersion(v: string): FsVersion {
@@ -72,6 +73,21 @@ export interface FsInfo {
   size?: number
 }
 
+/**
+ * Metadata about a path without following the final path component when it is a
+ * symbolic link. Unlike {@link FsInfo}, this path-level probe can report
+ * `symlink` so consumers with trust-boundary rules can reject repository-owned
+ * links before resolving a target.
+ */
+export interface FsPathInfo {
+  /** Opaque freshness token of the path entry right now. */
+  version: FsVersion
+  /** Whether the path entry is a regular file, directory, symlink, or other. */
+  type: 'file' | 'directory' | 'symlink' | 'other'
+  /** Byte size of the path entry, when the backend can report it. */
+  size?: number
+}
+
 /**
  * One direct child returned by {@link FileSystem.listDir}. Listing returns
  * metadata and resolved targets only; it must not read file contents.
diff --git a/packages/fs/fs/tests/service.spec.ts b/packages/fs/fs/tests/service.spec.ts
index 86ba782c96..19ee033cce 100644
--- a/packages/fs/fs/tests/service.spec.ts
+++ b/packages/fs/fs/tests/service.spec.ts
@@ -13,12 +13,13 @@ import type {
   FsEditOutcome,
   FsEditRequest,
   FsInfo,
+  FsPathInfo,
   FsTarget,
   FsWriteIntent,
   FsWriteOutcome,
 } from '@deepseek-ai/dsh-fs'
 
-/** A minimal in-memory fake implementing the seven provider primitives. */
+/** A minimal in-memory fake implementing the eight provider primitives. */
 class FakeFileSystem extends FileSystem {
   files = new Map<string, string>()
 
@@ -30,6 +31,11 @@ class FakeFileSystem extends FileSystem {
     if (content === undefined) return undefined
     return { version: FsVersion('v1'), type: 'file', size: content.length }
   }
+  override async lstat(path: string): Promise<FsPathInfo | undefined> {
+    const content = this.files.get(path)
+    if (content === undefined) return undefined
+    return { version: FsVersion('v1'), type: 'file', size: content.length }
+  }
   override async readText(target: FsTarget): Promise<string> {
     const content = this.files.get(target.targetKey)
     if (content === undefined) throw new FsError(`not found: ${target.displayPath}`, 'FS_NOT_FOUND')
@@ -120,6 +126,15 @@ describe('FileSystem provider seam', () => {
     const fs = ctx.fs as FakeFileSystem
     expect(await fs.stat(await fs.resolve('missing.txt'))).toBeUndefined()
   })
+
+  it('lstat returns path metadata before resolving a target', async () => {
+    const ctx = new Context()
+    await ctx.plugin(FakeFileSystem)
+    const fs = ctx.fs as FakeFileSystem
+    fs.files.set('a.txt', 'hi')
+    expect(await fs.lstat('a.txt')).toEqual({ version: 'v1', type: 'file', size: 2 })
+    expect(await fs.lstat('missing.txt')).toBeUndefined()
+  })
 })
 
 describe('branded id factories', () => {
diff --git a/packages/fs/tool-fs-search/README.md b/packages/fs/tool-fs-search/README.md
new file mode 100644
index 0000000000..29afbebfec
--- /dev/null
+++ b/packages/fs/tool-fs-search/README.md
@@ -0,0 +1,90 @@
+# @deepseek-ai/dsh-tool-fs-search
+
+The **model-facing filesystem discovery tools** — `glob`, `grep` — backed by the **bash executor seam**, not by `ctx.fs` provider methods. Each call assembles a fixed ripgrep command (every model-controlled value through one package-private shell-quoting helper), runs it via `ctx.bash.resolve(request)` → `ctx.bash.run(spec)` as an ordinary foreground tool call, parses the raw `rg` output, and returns a bounded, workdir-relative result. The package injects `tools`, `systemPrompt`, and `bash` — deliberately **not** `fs`; `ctx.spillStore` is read opportunistically with `ctx.get()` because formatted-result spill is optional.
+
+```ts ignore-check
+// Default deployment: a bash executor, then the discovery tools.
+await ctx.plugin(LocalBashExecutor, { cwd: process.cwd() }) // @deepseek-ai/dsh-bash-local
+await ctx.plugin(ToolFsSearch)                              // this package — registers glob/grep
+// Optional: a spill backend makes capped results fully recoverable.
+await ctx.plugin(LocalSpillStore)                           // @deepseek-ai/dsh-spill-local
+```
+
+Why bash-backed: local workspace discovery is naturally a process-backed `rg` workflow, and putting search on `ctx.fs` would force every filesystem backend to grow a search API. The bash executor owns request defaulting/capping, subprocess execution, process-group termination, environment scrubbing, raw output capture, and backend substitution (local, sandboxed, remote); this package owns schemas, argument validation, shell quoting, parsing, retention, formatted-result spill, and timeout declaration. The tools never call `ctx.bash.start()` and never expose a bash task id — the call returns only after `rg` exits, times out, is aborted, or fails.
+
+## Deployment requirement: co-located bash + filesystem
+
+Returned paths are displayed relative to the resolved bash workdir (the calling agent's session cwd when present, else the executor's configured default) and are follow-up-readable with `read` only when the bash workdir and the filesystem root are the same workspace. v1 documents that requirement and performs no runtime cross-service validation; remote or virtual filesystem search waits for a shared workspace contract or a provider-specific search backend.
+
+## Config
+
+All keys are optional; the defaults are the shipped search caps.
+
+| Key | Default | Meaning |
+|---|---|---|
+| `globMaxResults` | `100` | Max paths one `glob` call retains inline (matches Claude Code's `GlobTool` limit); later paths go to the formatted spill artifact. |
+| `grepMaxMatches` | `250` | Max flat matches one `grep` call retains inline (matches Claude Code's `GrepTool` `head_limit`); later matches go to the formatted spill artifact. |
+| `grepMaxLineBytes` | `2000` | Byte cap per matched-line preview; the cut preserves UTF-8 boundaries and is marked `(line truncated)`. |
+| `rawOutputMaxBytes` | `20000000` | Max complete raw `rg` stdout a search will parse (matches Claude Code's ripgrep raw buffer); larger raw output fails with `SEARCH_RAW_OUTPUT_OVERFLOW`. |
+| `timeoutMs` | `30000` | Cooperative tool-call budget attached to both tool definitions, enforced by `@deepseek-ai/dsh-timeout-policy` through `exec.signal`; the bash backend's own timeout stays a second safety cap. |
+
+## Tools
+
+| Tool | Arguments | Behavior |
+|---|---|---|
+| `glob` | `pattern`, `path?` | `rg --files --glob <pattern> --sort=modified --no-ignore --hidden` plus VCS metadata excludes (`.git`, `.svn`, `.hg`, `.bzr`, `.jj`, `.sl`). `path` is an optional **directory** search root; omitted means the resolved bash workdir. Returns one path per line, modification-time ordered. |
+| `grep` | `pattern`, `path?`, `include?` | Line-oriented `rg --json` parse (no colon-splitting ambiguity). `pattern` is a ripgrep regex; `path` is an optional **file or directory** target; `include` is ONE positive glob filter — a comma-separated list or a negated (`!…`) value is rejected up front (brace alternation like `*.{ts,tsx}` is fine). Returns matches grouped by file as `Line N: <preview>`. |
+
+Routine budgets stay out of the model-facing schema (no `head_limit`/`offset`/`case_insensitive`/output modes): a model that needs surrounding context reads the matched file with `read`; one that needs later results follows the returned spill locator's retrieval hint.
+
+## Two budgets, two artifacts
+
+Raw `rg` stdout is an internal transport detail. Each search requests `stdoutMaxBytes: rawOutputMaxBytes` from the bash seam and parses only complete retained stdout; if the executor still returns `stdout.truncated`, the search fails with `SEARCH_RAW_OUTPUT_OVERFLOW` and tells the model to narrow the query. The model-facing recovery artifact is different: when a search yields more logical results than the inline cap, the tool saves the COMPLETE formatted result through `ctx.spillStore.saveText()` (suggested names `glob-results.txt` / `grep-results.txt`, owner = the calling session, source = the tool execution identity) and appends a footer naming the returned locator and retrieval hint. This is the first tool-owned spill call in the codebase — deliberate, because retention here is item-level: the generic `@deepseek-ai/dsh-spill-policy` only sees the final text on `tools/post-execute`, by which point a capped search has already omitted later paths/matches. A missing spill backend, a call with no session owner, or a `saveText()` failure keeps the inline page and reports that the complete result could not be saved — never an `isError`.
+
+## Errors
+
+Search failures carry the package-owned `SearchError` (a `HarnessError` subclass), surfaced as `{ name, code }` on `isError` results: `SEARCH_INVALID_PATTERN` (ripgrep rejected the regex/glob), `SEARCH_FAILED` (missing `rg`, inaccessible target, signal kill, malformed `--json` output), `SEARCH_RAW_OUTPUT_OVERFLOW` (raw output over `rawOutputMaxBytes`, or still truncated after the requested stdout capture budget), and `SEARCH_ABORTED` (tool timeout, caller cancellation, or the bash executor's own timeout). ripgrep exit semantics are tool-owned: exit 0 is success with results, exit 1 is a successful empty search (`No files found` / `No matches found`), and only other exits are failures. Model argument mistakes (blank pattern, a list-valued `include`) stay ordinary tool argument errors.
+
+## Model Experience
+
+### System prompt
+
+**What the model sees**: Every request in this plugin's registration scope contains the independently registered glob and grep guidance below. Agent-scoped tool restrictions can hide either schema without removing its prompt section.
+
+**Token effect**: Fixed guidance cost per request while the plugin is active.
+
+#### Glob guidance
+
+```markdown
+Use the glob tool — not shell find or ls — to discover files by path pattern. Results are sorted by modification time and include hidden and ignored files.
+```
+
+#### Grep guidance
+
+```markdown
+Use the grep tool — not shell grep or rg — to search file contents. Use read on a matched file when you need surrounding context.
+```
+
+### Tool schemas
+
+**What the model sees**: The generated [`glob` and `grep` schemas](../../../docs/tool-catalog.md#deepseek-aidsh-tool-fs-search) while this surface is visible.
+
+**Token effect**: Fixed schema cost on every request where the tools are visible.
+
+### Results and spill notices
+
+**What the model sees**: `glob` returns one path per line; `grep` groups `Line <line>: <preview>` matches beneath each path. Empty searches return `No files found` or `No matches found`. A capped result ends with its omission count plus the spill locator and backend retrieval hint, or says the complete result could not be saved.
+
+**Token effect**: Inline paths and matches are bounded by `globMaxResults`, `grepMaxMatches`, and `grepMaxLineBytes`; the call and retained result remain in history until compaction.
+
+### Tool errors
+
+**What the model sees**: Failures are normalized as `Error: <message>` with structured `SEARCH_INVALID_PATTERN`, `SEARCH_FAILED`, `SEARCH_RAW_OUTPUT_OVERFLOW`, or `SEARCH_ABORTED` metadata for callers.
+
+**Token effect**: Only a failing call adds these retained tokens.
+
+## Known Limitations and Deferred Work
+
+- **Search and file access have no shared-workspace proof** — returned paths are follow-up-readable only when the bash workdir and filesystem root denote the same workspace; the package performs no runtime cross-service validation.
+- **Ripgrep is a deployment dependency** — a missing or incompatible `rg` executable fails calls with `SEARCH_FAILED`; remote or virtual filesystems need a co-located executor or another search consumer.
+- **The schemas expose one bounded page** — offset pagination, case-mode switches, alternate output modes, and provider-backed discovery remain outside this package; capped complete output requires a spill backend.
diff --git a/packages/fs/tool-fs-search/package.json b/packages/fs/tool-fs-search/package.json
new file mode 100644
index 0000000000..002d54569a
--- /dev/null
+++ b/packages/fs/tool-fs-search/package.json
@@ -0,0 +1,49 @@
+{
+  "name": "@deepseek-ai/dsh-tool-fs-search",
+  "description": "Model-facing filesystem discovery tools (glob, grep) backed by the DeepSeek Harness bash seam (ctx.bash)",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "dependencies": {
+    "schemastery": "^3.18.0"
+  },
+  "peerDependencies": {
+    "@deepseek-ai/dsh-bash": "^0.0.1",
+    "@deepseek-ai/dsh-llm": "^0.0.1",
+    "@deepseek-ai/dsh-retention": "^0.0.1",
+    "@deepseek-ai/dsh-session": "^0.0.1",
+    "@deepseek-ai/dsh-spill": "^0.0.1",
+    "@deepseek-ai/dsh-system-prompt": "^0.0.1",
+    "@deepseek-ai/dsh-tools": "^0.0.1",
+    "cordis": "^4.0.0-rc.6"
+  },
+  "devDependencies": {
+    "@deepseek-ai/dsh-agent": "workspace:^",
+    "@deepseek-ai/dsh-bash": "workspace:^",
+    "@deepseek-ai/dsh-bash-local": "workspace:^",
+    "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-retention": "workspace:^",
+    "@deepseek-ai/dsh-session": "workspace:^",
+    "@deepseek-ai/dsh-spill": "workspace:^",
+    "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-tools": "workspace:^",
+    "cordis": "^4.0.0-rc.6"
+  }
+}
diff --git a/packages/fs/tool-fs-search/src/glob.ts b/packages/fs/tool-fs-search/src/glob.ts
new file mode 100644
index 0000000000..a3e803fb50
--- /dev/null
+++ b/packages/fs/tool-fs-search/src/glob.ts
@@ -0,0 +1,179 @@
+/**
+ * The model-facing `glob` tool: discover files whose paths match a glob
+ * pattern, sorted by modification time. Execution goes through the bash seam
+ * (`ctx.bash`) with a fixed `rg --files` command — this module owns the
+ * model-facing schema, argument validation, shell-safe command construction,
+ * result parsing, retention, and formatting; process concerns (defaulting,
+ * scrubbing, kill, backend substitution) stay behind `ctx.bash`.
+ *
+ * @module @deepseek-ai/dsh-tool-fs-search/glob
+ */
+
+import type { Context } from 'cordis'
+import { defineTool } from '@deepseek-ai/dsh-tools'
+import type { GenericCallView } from '@deepseek-ai/dsh-tools'
+import type { ContentBlock } from '@deepseek-ai/dsh-llm'
+import { ItemRetainer } from '@deepseek-ai/dsh-retention'
+import type { RetainedItems } from '@deepseek-ai/dsh-retention'
+import type { SpillRef } from '@deepseek-ai/dsh-spill'
+import type {} from '@deepseek-ai/dsh-bash'
+import type {} from '@deepseek-ai/dsh-system-prompt'
+import { runRipgrep, toWorkdirRelative, trySaveFormattedResult } from './search-core.ts'
+import { singleQuote } from './shell-quote.ts'
+
+/**
+ * Default cap on paths retained inline by one `glob` call (the `globMaxResults`
+ * config), matching Claude Code's default `GlobTool` result limit.
+ */
+export const GLOB_MAX_RESULTS = 100
+
+/**
+ * Directory names ripgrep must never descend into for a discovery listing: VCS
+ * metadata stores. `--no-ignore --hidden` would otherwise surface them in every
+ * broad search. Each name is excluded with TWO negated `--glob`s (see
+ * {@link buildGlobCommand}): an any-depth directory glob that matches — and
+ * prunes — the directory during traversal, and a contents glob that still
+ * excludes the internals when the search root itself is at or inside the
+ * directory (an explicit `path` of `.git` or `sub/.git`), where the prune glob
+ * alone never matches.
+ */
+export const GLOB_VCS_EXCLUDES: readonly string[] = ['.git', '.svn', '.hg', '.bzr', '.jj', '.sl']
+
+/** Resolved glob-tool caps — plugin config after defaulting (see `Config` in index.ts). */
+export interface GlobToolCaps {
+  /** Max paths retained inline; later paths go to the formatted spill file. */
+  maxResults: number
+  /** Cap on the complete raw `rg` stdout the tool will parse. */
+  rawOutputMaxBytes: number
+  /** Cooperative tool-call budget (ms) attached as `ToolDefinition.timeoutMs`. */
+  timeoutMs: number
+}
+
+/** Validated `glob` arguments. */
+export interface GlobInput {
+  pattern: string
+  path?: string
+}
+
+/**
+ * Validate value constraints the schema DSL can't express: a non-blank
+ * `pattern`, and a non-blank `path` when given. Throws a plain `Error` (an
+ * ordinary tool argument error) otherwise.
+ *
+ * @param args - the schema-validated `glob` arguments.
+ * @returns the accepted input, unchanged.
+ */
+export function parseGlobArgs(args: { pattern: string; path?: string }): GlobInput {
+  if (args.pattern.trim().length === 0) throw new Error('pattern must be a non-empty string')
+  if (args.path !== undefined && args.path.trim().length === 0) throw new Error('path must be a non-empty string when given')
+  return { pattern: args.pattern, ...args.path !== undefined ? { path: args.path } : {} }
+}
+
+/**
+ * Build the fixed `rg --files` command for one `glob` call. Every
+ * model-controlled value ({@link GlobInput.pattern}, {@link GlobInput.path})
+ * passes through {@link singleQuote}; the search root rides behind `--` so a
+ * leading-dash path can never be parsed as a flag. `--sort=modified` orders by
+ * modification time, `--no-ignore --hidden` searches ignored and hidden files,
+ * and {@link GLOB_VCS_EXCLUDES} keeps VCS metadata out.
+ *
+ * @param input - the validated arguments.
+ * @returns the complete, shell-safe command string.
+ */
+export function buildGlobCommand(input: GlobInput): string {
+  const parts = [
+    'rg --files',
+    `--glob=${singleQuote(input.pattern)}`,
+    '--sort=modified --no-ignore --hidden',
+    // Two negated globs per VCS name: the bare form prunes the directory
+    // during traversal; the /** form still excludes the contents when the
+    // search root is AT or INSIDE the directory (where the bare form,
+    // matched against root-prefixed paths, never fires).
+    ...GLOB_VCS_EXCLUDES.flatMap(name => [
+      `--glob=${singleQuote(`!**/${name}`)}`,
+      `--glob=${singleQuote(`!**/${name}/**`)}`,
+    ]),
+  ]
+  if (input.path !== undefined) parts.push('--', singleQuote(input.path))
+  return parts.join(' ')
+}
+
+/**
+ * Format the model-facing `glob` result: the retained paths, then — when the
+ * result was capped — a footer carrying either the formatted-spill recovery
+ * locator or the could-not-save explanation. The omitted count is a budget fact:
+ * the search itself completed.
+ *
+ * @param retained - the retention outcome over every discovered path.
+ * @param spillRef - the saved complete-result reference, or `undefined` when unsaved.
+ * @returns the model-facing text.
+ */
+export function formatGlobOutput(retained: RetainedItems<string>, spillRef: SpillRef | undefined): string {
+  const body = retained.items.join('\n')
+  if (!retained.truncated) return body
+  const recovery = spillRef !== undefined
+    ? `Full sorted result stored at: ${spillRef.locator}. ${spillRef.retrievalHint}`
+    : 'The complete result could not be saved; narrow pattern or path to see more.'
+  return `${body}\n\n(Showing ${retained.kept} of ${retained.seen} paths. ${recovery})`
+}
+
+/**
+ * Pending-call presentation: a search card titled by the pattern (and root).
+ *
+ * @param args - the raw tool arguments; `pattern` and `path` feed the title.
+ * @returns the generic card view (`kind: 'search'`) shown while the call runs.
+ */
+export function presentGlobCall(args: { pattern: string; path?: string }): GenericCallView {
+  const where = args.path !== undefined ? ` in ${args.path}` : ''
+  return { card: 'generic', title: `Glob ${args.pattern}${where}`, kind: 'search', rawInput: args.pattern }
+}
+
+/**
+ * Register the `glob` tool and its system-prompt guidance.
+ *
+ * @param ctx - the plugin context; registrations are effects scoped to it, and
+ *   execution uses its `bash` service.
+ * @param caps - the deployment's resolved glob caps (plugin config after defaulting).
+ */
+export function applyGlobTool(ctx: Context, caps: GlobToolCaps): void {
+  ctx.systemPrompt.section({
+    name: 'tool:glob',
+    order: 103,
+    text: 'Use the glob tool — not shell find or ls — to discover files by path pattern. Results are sorted by modification time and include hidden and ignored files.',
+  })
+
+  ctx.tools.register(defineTool({
+    name: 'glob',
+    description: 'Find files whose paths match a glob pattern. Returns matching paths sorted by modification time, '
+      + 'including hidden and ignored files (VCS metadata directories are excluded). '
+      + `Returns the first ${caps.maxResults} paths inline; a capped result reports where the complete list was saved.`,
+    parameters: {
+      pattern: { type: 'string', required: true, description: 'Glob pattern to match file paths against (e.g. "**/*.ts", "src/**/*.test.js").' },
+      path: { type: 'string', description: 'Directory to search in. Defaults to the session workspace; a relative path resolves against it.' },
+    },
+    timeoutMs: caps.timeoutMs,
+    async execute(args, exec): Promise<ContentBlock[]> {
+      const input = parseGlobArgs(args)
+      const run = await runRipgrep(ctx, exec, 'glob', buildGlobCommand(input), caps.rawOutputMaxBytes)
+      if (run.noMatches) return [{ type: 'text', text: 'No files found' }]
+
+      const retainer = new ItemRetainer<string>({ kind: 'head', maxItems: caps.maxResults })
+      const all: string[] = []
+      for (const line of run.stdout.split('\n')) {
+        if (line.length === 0) continue
+        const displayPath = toWorkdirRelative(line, run.workdir)
+        all.push(displayPath)
+        retainer.push(displayPath)
+      }
+      const retained = retainer.finish()
+
+      // The complete sorted list is the recovery artifact; save it only when
+      // the inline page omitted paths (an uncapped result needs no spill file).
+      const spillRef = retained.truncated
+        ? await trySaveFormattedResult(ctx, exec, 'glob-results.txt', all.join('\n'))
+        : undefined
+      return [{ type: 'text', text: formatGlobOutput(retained, spillRef) }]
+    },
+    presentCall: presentGlobCall,
+  }))
+}
diff --git a/packages/fs/tool-fs-search/src/grep.ts b/packages/fs/tool-fs-search/src/grep.ts
new file mode 100644
index 0000000000..3935513b73
--- /dev/null
+++ b/packages/fs/tool-fs-search/src/grep.ts
@@ -0,0 +1,315 @@
+/**
+ * The model-facing `grep` tool: search file contents with a ripgrep regular
+ * expression. Execution goes through the bash seam (`ctx.bash`) with a fixed
+ * line-oriented `rg --json` command so file path, line number, and line text
+ * parse without colon-splitting ambiguity — this module owns the model-facing
+ * schema, argument validation, shell-safe command construction, `--json`
+ * record parsing, per-line preview retention, match retention, grouping, and
+ * formatting; process concerns stay behind `ctx.bash`.
+ *
+ * @module @deepseek-ai/dsh-tool-fs-search/grep
+ */
+
+import type { Context } from 'cordis'
+import { defineTool } from '@deepseek-ai/dsh-tools'
+import type { GenericCallView } from '@deepseek-ai/dsh-tools'
+import type { ContentBlock } from '@deepseek-ai/dsh-llm'
+import { ItemRetainer, TextRetainer } from '@deepseek-ai/dsh-retention'
+import type { RetainedItems } from '@deepseek-ai/dsh-retention'
+import type { SpillRef } from '@deepseek-ai/dsh-spill'
+import type {} from '@deepseek-ai/dsh-bash'
+import type {} from '@deepseek-ai/dsh-system-prompt'
+import { SearchError, runRipgrep, toWorkdirRelative, trySaveFormattedResult } from './search-core.ts'
+import { singleQuote } from './shell-quote.ts'
+
+/**
+ * Default cap on flat matches retained inline by one `grep` call (the
+ * `grepMaxMatches` config), matching Claude Code's default `GrepTool`
+ * `head_limit`.
+ */
+export const GREP_MAX_MATCHES = 250
+
+/**
+ * Default cap in bytes on one matched-line preview (the `grepMaxLineBytes`
+ * config); the cut preserves UTF-8 boundaries.
+ */
+export const GREP_MAX_LINE_BYTES = 2000
+
+/** Resolved grep-tool caps — plugin config after defaulting (see `Config` in index.ts). */
+export interface GrepToolCaps {
+  /** Max flat matches retained inline; later matches go to the formatted spill file. */
+  maxMatches: number
+  /** Max bytes retained per matched-line preview. */
+  maxLineBytes: number
+  /** Cap on the complete raw `rg` stdout the tool will parse. */
+  rawOutputMaxBytes: number
+  /** Cooperative tool-call budget (ms) attached as `ToolDefinition.timeoutMs`. */
+  timeoutMs: number
+}
+
+/** Validated `grep` arguments. */
+export interface GrepInput {
+  pattern: string
+  path?: string
+  include?: string
+}
+
+/** One parsed match: the file, the 1-based line number, and the (possibly previewed) line text. */
+export interface GrepMatch {
+  path: string
+  lineNumber: number
+  line: string
+}
+
+/**
+ * Reject an `include` that is not ONE positive glob filter: blank strings,
+ * negated patterns (`!…`), and comma-separated lists. A comma inside a brace
+ * group is fine — `*.{ts,tsx}` is one glob with alternation, not a list.
+ */
+function validateInclude(include: string): void {
+  if (include.trim().length === 0) throw new Error('include must be a non-empty glob when given')
+  if (include.startsWith('!')) throw new Error('include must be a positive glob filter; negated patterns ("!…") are not supported')
+  let braceDepth = 0
+  for (const char of include) {
+    if (char === '{') braceDepth++
+    else if (char === '}') braceDepth = Math.max(0, braceDepth - 1)
+    else if (char === ',' && braceDepth === 0) {
+      throw new Error('include must be one glob, not a comma-separated list (use {a,b} alternation instead)')
+    }
+  }
+}
+
+/**
+ * Validate value constraints the schema DSL can't express: a non-EMPTY
+ * `pattern` (whitespace is a legitimate regex), a non-blank `path` when given,
+ * and a single positive `include` glob ({@link GrepInput}). Throws a plain
+ * `Error` (an ordinary tool argument error) otherwise.
+ *
+ * @param args - the schema-validated `grep` arguments.
+ * @returns the accepted input, unchanged.
+ */
+export function parseGrepArgs(args: { pattern: string; path?: string; include?: string }): GrepInput {
+  if (args.pattern.length === 0) throw new Error('pattern must be a non-empty string')
+  if (args.path !== undefined && args.path.trim().length === 0) throw new Error('path must be a non-empty string when given')
+  if (args.include !== undefined) validateInclude(args.include)
+  return {
+    pattern: args.pattern,
+    ...args.path !== undefined ? { path: args.path } : {},
+    ...args.include !== undefined ? { include: args.include } : {},
+  }
+}
+
+/**
+ * Build the fixed line-oriented `rg --json` command for one `grep` call. Every
+ * model-controlled value ({@link GrepInput.pattern}, {@link GrepInput.path},
+ * {@link GrepInput.include}) passes through {@link singleQuote}; the pattern
+ * and include ride in `--flag=value` form and the target behind `--`, so a
+ * leading-dash value can never be parsed as a flag.
+ *
+ * @param input - the validated arguments.
+ * @returns the complete, shell-safe command string.
+ */
+export function buildGrepCommand(input: GrepInput): string {
+  const parts = ['rg --json', `--regexp=${singleQuote(input.pattern)}`]
+  if (input.include !== undefined) parts.push(`--glob=${singleQuote(input.include)}`)
+  if (input.path !== undefined) parts.push('--', singleQuote(input.path))
+  return parts.join(' ')
+}
+
+/**
+ * The uniform malformed-output failure: raw `rg --json` is an internal
+ * transport, so a shape surprise is a search failure, not a partial result.
+ */
+function malformedRecord(detail: string, cause?: unknown): SearchError {
+  return new SearchError(`grep received malformed ripgrep --json output (${detail})`, 'SEARCH_FAILED', cause !== undefined ? { cause } : undefined)
+}
+
+/**
+ * Parse one `rg --json` NDJSON line into a match, `undefined` for the
+ * non-match record types (`begin`/`end`/`context`/`summary`). A line that is
+ * not JSON, or a `match` record missing its path / line number / line content,
+ * throws {@link SearchError} `SEARCH_FAILED`. A match whose line is not valid
+ * UTF-8 (ripgrep sends base64 `bytes` instead of `text`) yields a placeholder
+ * preview rather than failing the whole search.
+ */
+function parseRecord(line: string): GrepMatch | undefined {
+  let parsed: unknown
+  try {
+    parsed = JSON.parse(line)
+  } catch (error: unknown) {
+    throw malformedRecord('a line is not JSON', error)
+  }
+  if (typeof parsed !== 'object' || parsed === null) throw malformedRecord('a record is not an object')
+  const record = parsed as { type?: unknown; data?: unknown }
+  // Non-match record types (begin/end/context/summary — and any future type)
+  // are transport framing, not results: skipped, not malformed.
+  if (record.type !== 'match') return undefined
+  if (typeof record.data !== 'object' || record.data === null) throw malformedRecord('a match record has no data')
+  const data = record.data as { path?: unknown; line_number?: unknown; lines?: unknown }
+  const pathText = typeof data.path === 'object' && data.path !== null ? (data.path as { text?: unknown }).text : undefined
+  if (typeof pathText !== 'string') throw malformedRecord('a match record has no path text')
+  if (typeof data.line_number !== 'number') throw malformedRecord('a match record has no line number')
+  if (typeof data.lines !== 'object' || data.lines === null) throw malformedRecord('a match record has no line content')
+  const lines = data.lines as { text?: unknown; bytes?: unknown }
+  if (typeof lines.text === 'string') {
+    return { path: pathText, lineNumber: data.line_number, line: lines.text.replace(/\r?\n$/, '') }
+  }
+  if (typeof lines.bytes === 'string') {
+    return { path: pathText, lineNumber: data.line_number, line: '(line is not valid UTF-8)' }
+  }
+  throw malformedRecord('a match record has neither line text nor bytes')
+}
+
+/**
+ * Parse complete `rg --json` stdout into flat matches, in output order (ripgrep
+ * emits one file's matches contiguously). Only `match` records are consumed.
+ *
+ * @param stdout - the complete raw `rg --json` stdout.
+ * @returns the flat matches; empty for output with no match records.
+ */
+export function parseGrepMatches(stdout: string): GrepMatch[] {
+  const matches: GrepMatch[] = []
+  for (const line of stdout.split('\n')) {
+    if (line.length === 0) continue
+    const match = parseRecord(line)
+    if (match !== undefined) matches.push(match)
+  }
+  return matches
+}
+
+/**
+ * Bound one matched-line preview to `maxBytes` (UTF-8 boundary preserved) and
+ * mark the cut. The cap is a per-line budget fact; the complete line stays in
+ * the searched file for `read`.
+ *
+ * @param line - the matched line text (trailing newline already stripped).
+ * @param maxBytes - the preview budget in bytes.
+ * @returns the preview, suffixed with ` (line truncated)` when bytes were cut.
+ */
+export function previewLine(line: string, maxBytes: number): string {
+  const retainer = new TextRetainer({ kind: 'head', maxBytes })
+  retainer.push(line)
+  const kept = retainer.finish()
+  return kept.truncated ? `${kept.text} (line truncated)` : kept.text
+}
+
+/** `match` / `matches` for a count. */
+function matchNoun(count: number): string {
+  return count === 1 ? 'match' : 'matches'
+}
+
+/**
+ * Group flat matches by file (first-seen order) into the model-facing body:
+ * each file's display path, then one `Line N: <text>` row per match.
+ *
+ * @param matches - the flat matches to render.
+ * @returns the grouped body text.
+ */
+export function formatGrepMatches(matches: GrepMatch[]): string {
+  const byFile = new Map<string, GrepMatch[]>()
+  for (const match of matches) {
+    const group = byFile.get(match.path)
+    if (group !== undefined) group.push(match)
+    else byFile.set(match.path, [match])
+  }
+  const sections: string[] = []
+  for (const [path, group] of byFile) {
+    sections.push(`${path}\n${group.map(m => `Line ${m.lineNumber}: ${m.line}`).join('\n')}`)
+  }
+  return sections.join('\n\n')
+}
+
+/**
+ * Format the model-facing `grep` result: a found-count header, the retained
+ * matches grouped by file, then — when the result was capped — a footer
+ * carrying either the formatted-spill recovery locator or the could-not-save
+ * explanation. The omitted count is a budget fact: the search itself completed.
+ *
+ * @param retained - the retention outcome over every parsed match.
+ * @param spillRef - the saved complete-result reference, or `undefined` when unsaved.
+ * @returns the model-facing text.
+ */
+export function formatGrepOutput(retained: RetainedItems<GrepMatch>, spillRef: SpillRef | undefined): string {
+  const header = retained.truncated
+    ? `Found ${retained.kept} of ${retained.seen} matches`
+    : `Found ${retained.seen} ${matchNoun(retained.seen)}`
+  const body = formatGrepMatches(retained.items)
+  if (!retained.truncated) return `${header}\n\n${body}`
+  const recovery = spillRef !== undefined
+    ? `Full grep result stored at: ${spillRef.locator}. ${spillRef.retrievalHint}`
+    : 'The complete result could not be saved; narrow pattern, path, or include to see more.'
+  return `${header}\n\n${body}\n\n(${recovery})`
+}
+
+/**
+ * Pending-call presentation: a search card titled by the pattern (and target /
+ * include filter).
+ *
+ * @param args - the raw tool arguments; `pattern`, `path`, and `include` feed the title.
+ * @returns the generic card view (`kind: 'search'`) shown while the call runs.
+ */
+export function presentGrepCall(args: { pattern: string; path?: string; include?: string }): GenericCallView {
+  const where = args.path !== undefined ? ` in ${args.path}` : ''
+  const filter = args.include !== undefined ? ` (${args.include})` : ''
+  return { card: 'generic', title: `Grep ${args.pattern}${where}${filter}`, kind: 'search', rawInput: args.pattern }
+}
+
+/**
+ * Register the `grep` tool and its system-prompt guidance.
+ *
+ * @param ctx - the plugin context; registrations are effects scoped to it, and
+ *   execution uses its `bash` service.
+ * @param caps - the deployment's resolved grep caps (plugin config after defaulting).
+ */
+export function applyGrepTool(ctx: Context, caps: GrepToolCaps): void {
+  ctx.systemPrompt.section({
+    name: 'tool:grep',
+    order: 104,
+    text: 'Use the grep tool — not shell grep or rg — to search file contents. Use read on a matched file when you need surrounding context.',
+  })
+
+  ctx.tools.register(defineTool({
+    name: 'grep',
+    description: 'Search file contents with a ripgrep regular expression. Returns matching lines with line numbers, grouped by file. '
+      + `Returns the first ${caps.maxMatches} matches inline; a capped result reports where the complete match list was saved. `
+      + 'Use read on a matched file for surrounding context.',
+    parameters: {
+      pattern: { type: 'string', required: true, description: 'Regular expression to search for (ripgrep syntax).' },
+      path: { type: 'string', description: 'File or directory to search. Defaults to the session workspace; a relative path resolves against it.' },
+      include: { type: 'string', description: 'One glob filter for which files to search (e.g. "*.ts", "*.{js,jsx}"). Not a list; negation is not supported.' },
+    },
+    timeoutMs: caps.timeoutMs,
+    async execute(args, exec): Promise<ContentBlock[]> {
+      const input = parseGrepArgs(args)
+      const run = await runRipgrep(ctx, exec, 'grep', buildGrepCommand(input), caps.rawOutputMaxBytes)
+      if (run.noMatches) return [{ type: 'text', text: 'No matches found' }]
+
+      const retainer = new ItemRetainer<GrepMatch>({ kind: 'head', maxItems: caps.maxMatches })
+      const all: GrepMatch[] = []
+      for (const raw of parseGrepMatches(run.stdout)) {
+        const match: GrepMatch = {
+          path: toWorkdirRelative(raw.path, run.workdir),
+          lineNumber: raw.lineNumber,
+          line: previewLine(raw.line, caps.maxLineBytes),
+        }
+        all.push(match)
+        retainer.push(match)
+      }
+      const retained = retainer.finish()
+
+      // The spill file stores the FULL formatted match list (same grouped,
+      // per-line-previewed shape the model saw), so read offset/limit pages the
+      // same logical result; save only when the inline page omitted matches.
+      const spillRef = retained.truncated
+        ? await trySaveFormattedResult(
+          ctx,
+          exec,
+          'grep-results.txt',
+          `Found ${all.length} ${matchNoun(all.length)}\n\n${formatGrepMatches(all)}`,
+        )
+        : undefined
+      return [{ type: 'text', text: formatGrepOutput(retained, spillRef) }]
+    },
+    presentCall: presentGrepCall,
+  }))
+}
diff --git a/packages/fs/tool-fs-search/src/index.ts b/packages/fs/tool-fs-search/src/index.ts
new file mode 100644
index 0000000000..8c33d5770a
--- /dev/null
+++ b/packages/fs/tool-fs-search/src/index.ts
@@ -0,0 +1,110 @@
+/**
+ * The model-facing filesystem discovery tool suite (`glob`, `grep`) over the
+ * bash executor seam (`ctx.bash`). This single plugin registers both tools.
+ *
+ * ## Bash-backed, not a `ctx.fs` provider method
+ *
+ * Local workspace discovery is a process-backed `rg` workflow, so these tools
+ * execute through `ctx.bash.resolve(request)` → `ctx.bash.run(spec)` with fixed
+ * ripgrep command templates — never `ctx.bash.start()`, never a model-visible
+ * background task. The tool layer owns schemas, argument validation, shell
+ * quoting ({@link module:@deepseek-ai/dsh-tool-fs-search/shell-quote}), result
+ * parsing, retention, formatted-result spill, and timeout declaration; the
+ * bash executor owns request defaulting/capping, subprocess execution,
+ * process-group termination, environment scrubbing, raw output capture, and
+ * backend substitution. The package injects `tools`, `systemPrompt`, and
+ * `bash` — deliberately NOT `fs`, and `ctx.spillStore` is read opportunistically
+ * with `ctx.get()` because formatted-result spill is optional.
+ *
+ * Returned paths are displayed relative to the resolved bash workdir and are
+ * follow-up-readable only in co-located deployments where the bash workdir and
+ * the filesystem `read` root are the same workspace — a documented v1
+ * deployment requirement, not runtime-validated.
+ *
+ * @module @deepseek-ai/dsh-tool-fs-search
+ */
+
+import type { Context } from 'cordis'
+import z from 'schemastery'
+import { GLOB_MAX_RESULTS, applyGlobTool } from './glob.ts'
+import { GREP_MAX_LINE_BYTES, GREP_MAX_MATCHES, applyGrepTool } from './grep.ts'
+import { RAW_OUTPUT_MAX_BYTES, SEARCH_TIMEOUT_MS } from './search-core.ts'
+
+export { GLOB_MAX_RESULTS, GLOB_VCS_EXCLUDES, applyGlobTool, buildGlobCommand, formatGlobOutput, parseGlobArgs, presentGlobCall } from './glob.ts'
+export type { GlobInput, GlobToolCaps } from './glob.ts'
+export {
+  GREP_MAX_LINE_BYTES,
+  GREP_MAX_MATCHES,
+  applyGrepTool,
+  buildGrepCommand,
+  formatGrepMatches,
+  formatGrepOutput,
+  parseGrepArgs,
+  parseGrepMatches,
+  presentGrepCall,
+  previewLine,
+} from './grep.ts'
+export type { GrepInput, GrepMatch, GrepToolCaps } from './grep.ts'
+export { RAW_OUTPUT_MAX_BYTES, SEARCH_TIMEOUT_MS, SearchError, runRipgrep, toWorkdirRelative, trySaveFormattedResult } from './search-core.ts'
+export type { RipgrepRun, SearchErrorCode } from './search-core.ts'
+export { singleQuote } from './shell-quote.ts'
+
+/** Cordis plugin name used by loader diagnostics. */
+export const name = 'tool-fs-search'
+
+/** Services required by the search tool suite (`spillStore` is optional, read via `ctx.get()`). */
+export const inject = ['tools', 'systemPrompt', 'bash']
+
+/** Plugin config (all optional — `Config` supplies the defaults). */
+export interface Config {
+  /** Max paths one `glob` call retains inline; later paths go to the formatted spill file. */
+  globMaxResults?: number
+  /** Max flat matches one `grep` call retains inline; later matches go to the formatted spill file. */
+  grepMaxMatches?: number
+  /** Max bytes retained for one matched-line preview (the cut preserves UTF-8 boundaries). */
+  grepMaxLineBytes?: number
+  /** Max complete raw `rg` stdout bytes a search will parse; larger raw output fails with `SEARCH_RAW_OUTPUT_OVERFLOW`. */
+  rawOutputMaxBytes?: number
+  /** Cooperative tool-call timeout budget (ms) on both tools, enforced by `@deepseek-ai/dsh-timeout-policy` through `exec.signal`. */
+  timeoutMs?: number
+}
+
+export const Config: z<Config> = z.object({
+  globMaxResults: z.number().default(GLOB_MAX_RESULTS),
+  grepMaxMatches: z.number().default(GREP_MAX_MATCHES),
+  grepMaxLineBytes: z.number().default(GREP_MAX_LINE_BYTES),
+  rawOutputMaxBytes: z.number().default(RAW_OUTPUT_MAX_BYTES),
+  timeoutMs: z.number().default(SEARCH_TIMEOUT_MS),
+})
+
+/** The shape after schemastery applied the defaults. */
+type ResolvedConfig = Required<Config>
+
+/** Every search cap counts items/bytes/milliseconds — a positive integer, or retention and timeout arithmetic misbehaves silently. */
+function assertPositiveInteger(name: string, value: number): void {
+  if (!Number.isInteger(value) || value < 1) {
+    throw new Error(`tool-fs-search: ${name} must be a positive integer`)
+  }
+}
+
+/** Register the `glob`/`grep` filesystem discovery tool suite. */
+export function apply(ctx: Context, config: Config): void {
+  // schemastery (Config) has already filled every defaulted field.
+  const resolved = config as ResolvedConfig
+  assertPositiveInteger('globMaxResults', resolved.globMaxResults)
+  assertPositiveInteger('grepMaxMatches', resolved.grepMaxMatches)
+  assertPositiveInteger('grepMaxLineBytes', resolved.grepMaxLineBytes)
+  assertPositiveInteger('rawOutputMaxBytes', resolved.rawOutputMaxBytes)
+  assertPositiveInteger('timeoutMs', resolved.timeoutMs)
+  applyGlobTool(ctx, {
+    maxResults: resolved.globMaxResults,
+    rawOutputMaxBytes: resolved.rawOutputMaxBytes,
+    timeoutMs: resolved.timeoutMs,
+  })
+  applyGrepTool(ctx, {
+    maxMatches: resolved.grepMaxMatches,
+    maxLineBytes: resolved.grepMaxLineBytes,
+    rawOutputMaxBytes: resolved.rawOutputMaxBytes,
+    timeoutMs: resolved.timeoutMs,
+  })
+}
diff --git a/packages/fs/tool-fs-search/src/search-core.ts b/packages/fs/tool-fs-search/src/search-core.ts
new file mode 100644
index 0000000000..0682c86e35
--- /dev/null
+++ b/packages/fs/tool-fs-search/src/search-core.ts
@@ -0,0 +1,262 @@
+/**
+ * Shared execution plumbing for the `glob` / `grep` search tools: the
+ * package-owned `SEARCH_*` error vocabulary, one bash-seam run helper that
+ * turns a fixed `rg` command into complete raw stdout, the best-effort
+ * formatted-result spill handoff, and workdir-relative path display.
+ *
+ * Both tools execute through `ctx.bash.resolve(request)` → `ctx.bash.run(spec)`
+ * as ordinary foreground tool calls — never `ctx.bash.start()`, never a
+ * model-visible background task. Raw `rg` stdout is an internal transport
+ * detail: the tools request a per-run stdout capture budget from the bash seam,
+ * parse only complete in-memory stdout within `rawOutputMaxBytes`, and never
+ * read executor spill files. The model-facing recovery artifact is the
+ * formatted result saved through `ctx.spillStore.saveText()`
+ * ({@link trySaveFormattedResult}).
+ *
+ * @module @deepseek-ai/dsh-tool-fs-search/search-core
+ */
+
+import { isAbsolute, relative, sep } from 'node:path'
+import type { Context } from 'cordis'
+import { HarnessError } from '@deepseek-ai/dsh-llm'
+import type { BashRunResult, CollectedOutput } from '@deepseek-ai/dsh-bash'
+import type { SaveTextSpill, SpillRef } from '@deepseek-ai/dsh-spill'
+import type { ToolExecution } from '@deepseek-ai/dsh-tools'
+
+/**
+ * Default cap on the complete raw `rg` stdout the tools will parse (the
+ * `rawOutputMaxBytes` config), matching Claude Code's ripgrep raw buffer.
+ */
+export const RAW_OUTPUT_MAX_BYTES = 20_000_000
+
+/**
+ * Default cooperative tool-call timeout budget in milliseconds (the `timeoutMs`
+ * config), attached to both tool definitions for
+ * `@deepseek-ai/dsh-timeout-policy` to enforce through `exec.signal`.
+ */
+export const SEARCH_TIMEOUT_MS = 30_000
+
+/**
+ * Stable, machine-routable codes for search failures. Package-owned (not
+ * `FsErrorCode`) because these tools are bash-backed discovery, not `ctx.fs`
+ * provider operations: `SEARCH_INVALID_PATTERN` — ripgrep rejected the regex or
+ * glob; `SEARCH_FAILED` — the search could not run or its output could not be
+ * parsed (missing `rg`, inaccessible target, signal kill, malformed `--json`);
+ * `SEARCH_RAW_OUTPUT_OVERFLOW` — raw `rg` output exceeded `rawOutputMaxBytes`
+ * or stayed truncated after that requested stdout budget; `SEARCH_ABORTED` — the tool
+ * timeout, caller cancellation, or the bash executor's own timeout cut the
+ * search short.
+ */
+export type SearchErrorCode =
+  | 'SEARCH_INVALID_PATTERN'
+  | 'SEARCH_FAILED'
+  | 'SEARCH_RAW_OUTPUT_OVERFLOW'
+  | 'SEARCH_ABORTED'
+
+/**
+ * Typed search failure. Extends {@link HarnessError} so it carries a stable
+ * {@link SearchErrorCode} and chains `cause`; the tool registry surfaces
+ * `{ name, code }` on `isError` results so retry/permission/UI layers can
+ * branch without parsing messages.
+ */
+export class SearchError extends HarnessError {
+  override readonly code: SearchErrorCode
+
+  constructor(message: string, code: SearchErrorCode, options?: ErrorOptions) {
+    super(message, code, options)
+    this.code = code
+  }
+}
+
+/** The completed acquisition of one `rg` run: complete stdout plus the resolved workdir. */
+export interface RipgrepRun {
+  /** Complete raw stdout retained by the bash executor within the requested cap. */
+  stdout: string
+  /** True when ripgrep exited 1: a successful search with zero results. */
+  noMatches: boolean
+  /** The resolved working directory the command ran in (the display-relativization base). */
+  workdir: string
+}
+
+/**
+ * The retained stderr tail as a diagnostic excerpt, with a truncation note when
+ * the executor dropped bytes (the tool never reads `stderr.spillPath`).
+ */
+function stderrExcerpt(stderr: CollectedOutput): string {
+  const text = stderr.text.trim()
+  if (text.length === 0) return ''
+  return stderr.truncated ? `${text} [stderr truncated]` : text
+}
+
+/** Classify a nonzero-exit `rg` run into the search error vocabulary (invalid pattern vs missing `rg` vs everything else). */
+function classifyRunFailure(toolName: string, result: BashRunResult): SearchError {
+  const stderr = stderrExcerpt(result.stderr)
+  if (/regex parse error|error parsing glob/i.test(stderr)) {
+    return new SearchError(`${toolName} pattern rejected by ripgrep: ${stderr}`, 'SEARCH_INVALID_PATTERN')
+  }
+  if (result.exitCode === 127 || /command not found/i.test(stderr)) {
+    return new SearchError(`${toolName} requires ripgrep (rg) on the bash executor's PATH${stderr.length > 0 ? `: ${stderr}` : ''}`, 'SEARCH_FAILED')
+  }
+  return new SearchError(`${toolName} search failed (exit ${result.exitCode})${stderr.length > 0 ? `: ${stderr}` : ''}`, 'SEARCH_FAILED')
+}
+
+/**
+ * Acquire the COMPLETE raw stdout of a finished run, enforcing
+ * `rawOutputMaxBytes` on the in-memory transport. A truncated result means the
+ * bash backend could not retain complete stdout within the requested budget, so
+ * the tool fails clearly instead of parsing a silently-partial stream.
+ */
+function completeStdout(toolName: string, result: BashRunResult, rawOutputMaxBytes: number): string {
+  const narrow = 'narrow pattern, path, or include and retry'
+  if (!result.stdout.truncated) {
+    const inlineBytes = Buffer.byteLength(result.stdout.text, 'utf8')
+    if (inlineBytes > rawOutputMaxBytes) {
+      throw new SearchError(
+        `${toolName} produced ${inlineBytes} bytes of raw output, over the ${rawOutputMaxBytes}-byte cap; ${narrow}`,
+        'SEARCH_RAW_OUTPUT_OVERFLOW',
+      )
+    }
+    return result.stdout.text
+  }
+  throw new SearchError(
+    `${toolName} produced more raw output than the bash executor retained within the ${rawOutputMaxBytes}-byte cap; ${narrow}`,
+    'SEARCH_RAW_OUTPUT_OVERFLOW',
+  )
+}
+
+/**
+ * Run one fixed `rg` command through the bash seam and return its complete raw
+ * stdout. The bash request workdir is the calling agent's session cwd
+ * (`exec.agent.session.header.cwd`) when available — mirroring `dsh-tool-bash` /
+ * `dsh-tool-fs` — else omitted so the implementation's `resolve()` applies its
+ * configured default. `exec.signal` is forwarded so the cooperative tool
+ * timeout (`@deepseek-ai/dsh-timeout-policy`) and caller cancellation kill the
+ * command; the bash backend's own timeout stays a second safety cap.
+ *
+ * Exit semantics are tool-owned: exit 0 is success with results, exit 1 is
+ * success with zero results (`noMatches`), anything else throws a
+ * {@link SearchError} (abort/timeout → `SEARCH_ABORTED`, invalid pattern →
+ * `SEARCH_INVALID_PATTERN`, the rest → `SEARCH_FAILED` /
+ * `SEARCH_RAW_OUTPUT_OVERFLOW`). A `run()` REJECTION — the seam's
+ * infrastructure failures (pre-aborted signal, unusable workdir, missing
+ * shell) — is translated into the same taxonomy: a pre-aborted signal becomes
+ * `SEARCH_ABORTED`, everything else `SEARCH_FAILED`, with the original as
+ * `cause`.
+ *
+ * @param ctx - the plugin context; execution uses its `bash` service.
+ * @param exec - the tool-execution context; supplies the session cwd and the abort signal.
+ * @param toolName - `glob` or `grep`, used in error messages.
+ * @param command - the fully-quoted `rg` command string (every model value already through `singleQuote`).
+ * @param rawOutputMaxBytes - cap on the complete raw stdout the tool will parse.
+ * @returns the complete stdout, the zero-result flag, and the resolved workdir.
+ */
+export async function runRipgrep(
+  ctx: Context,
+  exec: ToolExecution,
+  toolName: string,
+  command: string,
+  rawOutputMaxBytes: number,
+): Promise<RipgrepRun> {
+  const cwd = exec.agent?.session.header.cwd
+  const spec = ctx.bash.resolve({
+    command,
+    stdoutMaxBytes: rawOutputMaxBytes,
+    ...cwd !== undefined ? { workdir: cwd } : {},
+    ...exec.signal ? { signal: exec.signal } : {},
+  })
+  let result: BashRunResult
+  try {
+    result = await ctx.bash.run(spec)
+  } catch (error: unknown) {
+    // The seam contract: run() REJECTS only for infrastructure failures — a
+    // pre-aborted signal, an unusable workdir, a missing shell. Translate them
+    // so these failures stay machine-routable under the SEARCH_* taxonomy.
+    if (spec.signal?.aborted === true) {
+      throw new SearchError(`${toolName} was aborted before completion (tool timeout or caller cancellation)`, 'SEARCH_ABORTED', { cause: error })
+    }
+    throw new SearchError(`${toolName} could not start its search command (unusable working directory or missing shell)`, 'SEARCH_FAILED', { cause: error })
+  }
+  if (result.aborted) {
+    throw new SearchError(`${toolName} was aborted before completion (tool timeout or caller cancellation)`, 'SEARCH_ABORTED')
+  }
+  if (result.timedOut) {
+    throw new SearchError(`${toolName} timed out after ${result.timeoutMs}ms in the bash executor; narrow pattern, path, or include and retry`, 'SEARCH_ABORTED')
+  }
+  if (result.signal !== null || result.exitCode === null) {
+    throw new SearchError(`${toolName} search command was killed by signal ${result.signal ?? '(unknown)'}`, 'SEARCH_FAILED')
+  }
+  if (result.exitCode !== 0 && result.exitCode !== 1) {
+    throw classifyRunFailure(toolName, result)
+  }
+  const stdout = completeStdout(toolName, result, rawOutputMaxBytes)
+  return { stdout, noMatches: result.exitCode === 1, workdir: spec.workdir }
+}
+
+/**
+ * Map an `rg` output path to its display form: absolute paths inside the
+ * resolved bash workdir become workdir-relative; everything else (relative
+ * output, paths outside the workdir) passes through unchanged. Display-only —
+ * returned paths are follow-up-readable in co-located bash/filesystem
+ * deployments where both resolve the same workspace (the documented v1
+ * deployment requirement).
+ *
+ * @param path - one path as ripgrep printed it.
+ * @param workdir - the resolved bash workdir the command ran in.
+ * @returns the workdir-relative display path when possible, else `path` unchanged.
+ */
+export function toWorkdirRelative(path: string, workdir: string): string {
+  if (!isAbsolute(path)) return path
+  const rel = relative(workdir, path)
+  if (rel.length === 0) return '.'
+  if (rel === '..' || rel.startsWith(`..${sep}`)) return path
+  return rel
+}
+
+/**
+ * Best-effort save of one COMPLETE formatted search result through
+ * `ctx.spillStore.saveText()` — the model-facing recovery path for a capped
+ * result. `spillStore` is read with `ctx.get()` (not static inject) because
+ * formatted-result spill is optional; the spill owner is the calling agent's
+ * session header id and the source is the tool execution identity. A missing
+ * backend, a call with no session owner, or a `saveText()` rejection logs a
+ * warning and returns `undefined` — the caller keeps the inline result and
+ * reports that the complete result could not be saved; search success never
+ * turns into `isError` because spill storage is unavailable.
+ *
+ * @param ctx - the plugin context; `spillStore` is looked up opportunistically.
+ * @param exec - the tool-execution context; supplies the owning session, tool name, and call id.
+ * @param suggestedName - the backend-sanitized filename hint (e.g. `grep-results.txt`).
+ * @param content - the complete formatted result to persist.
+ * @returns the saved spill reference, or `undefined` when the result could not be saved.
+ */
+export async function trySaveFormattedResult(
+  ctx: Context,
+  exec: ToolExecution,
+  suggestedName: string,
+  content: string,
+): Promise<SpillRef | undefined> {
+  const sessionId = exec.agent?.session.header.id
+  if (sessionId === undefined) {
+    ctx.logger.warn(`tool-fs-search: no session owner for ${exec.name} result; complete result not saved`)
+    return undefined
+  }
+  const spillStore = ctx.get('spillStore')
+  if (!spillStore) {
+    ctx.logger.warn(`tool-fs-search: no ctx.spillStore backend loaded; complete ${exec.name} result not saved`)
+    return undefined
+  }
+  const save: SaveTextSpill = {
+    owner: { sessionId },
+    source: { toolName: exec.name, callId: exec.callId, label: 'result' },
+    suggestedName,
+    content,
+  }
+  try {
+    return await spillStore.saveText(save)
+  } catch (error: unknown) {
+    // Best-effort: a storage failure must never fail the search or hide the
+    // inline result — the footer reports the unsaved remainder instead.
+    ctx.logger.warn(`tool-fs-search: saveText failed for ${exec.name}: ${String(error)}; complete result not saved`)
+    return undefined
+  }
+}
diff --git a/packages/fs/tool-fs-search/src/shell-quote.ts b/packages/fs/tool-fs-search/src/shell-quote.ts
new file mode 100644
index 0000000000..9453b8e255
--- /dev/null
+++ b/packages/fs/tool-fs-search/src/shell-quote.ts
@@ -0,0 +1,27 @@
+/**
+ * The one shell-quoting helper both search tools MUST route every
+ * model-controlled value through before it enters an `rg` command string. The
+ * bash seam (`ctx.bash`) accepts a command STRING, not an argv vector, so this
+ * is the safety boundary that stops a `pattern`, `path`, or `include` from
+ * breaking out of its argument and injecting shell syntax.
+ *
+ * Command builders in `glob.ts` / `grep.ts` must never hand-roll quoting or
+ * concatenate an unquoted model value — they call {@link singleQuote}.
+ *
+ * @module @deepseek-ai/dsh-tool-fs-search/shell-quote
+ */
+
+/**
+ * POSIX single-quote a string for safe use as ONE shell word. Wraps the value
+ * in single quotes and rewrites every embedded single quote as `'\''` (close
+ * quote, an escaped literal quote, reopen quote). Inside single quotes the shell
+ * treats every other byte literally — spaces, newlines, `$`, backticks, `;`,
+ * `|`, `&`, glob metacharacters, and a leading `-` are all inert — so the result
+ * is a single, injection-safe argument regardless of the input.
+ *
+ * @param value - the raw, possibly model-controlled string to quote.
+ * @returns the value wrapped as one safe single-quoted shell word.
+ */
+export function singleQuote(value: string): string {
+  return `'${value.replaceAll("'", "'\\''")}'`
+}
diff --git a/packages/fs/tool-fs-search/tests/integration.spec.ts b/packages/fs/tool-fs-search/tests/integration.spec.ts
new file mode 100644
index 0000000000..36fb2c28e6
--- /dev/null
+++ b/packages/fs/tool-fs-search/tests/integration.spec.ts
@@ -0,0 +1,190 @@
+/**
+ * Integration tests: the REAL local bash executor (`dsh-bash-local`) plus a
+ * REAL ripgrep binary, exercised through `ctx.tools.execute()`. These verify
+ * the WORLD — actual files on disk are discovered and grepped, hostile
+ * patterns stay inert in a real shell, and real `rg` stderr classifies into
+ * the `SEARCH_*` vocabulary. The whole suite self-skips when `rg` is not on
+ * PATH (a CI accommodation mirroring the keyless e2e skip); the fake-executor
+ * suite (tools.spec.ts) carries the coverage gate.
+ */
+
+import { afterEach, beforeEach, describe, expect, it } from 'vitest'
+import { spawnSync } from 'node:child_process'
+import { mkdir, mkdtemp, rm, utimes, writeFile } from 'node:fs/promises'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { Context } from 'cordis'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import ToolRegistry from '@deepseek-ai/dsh-tools'
+import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
+import * as ToolFsSearch from '@deepseek-ai/dsh-tool-fs-search'
+
+const hasRg = spawnSync('rg', ['--version'], { encoding: 'utf8' }).status === 0
+
+let dir: string
+let ctx: Context
+
+let callCounter = 0
+function call(name: string, args: unknown, agentObj?: object) {
+  return ctx.tools.execute({
+    callId: CallId(`it-${++callCounter}`),
+    name,
+    arguments: args,
+    ...agentObj ? { agent: agentObj as never } : {},
+  })
+}
+
+function text(result: { content: { type: string; text?: string }[] }): string {
+  return result.content.filter(b => b.type === 'text').map(b => b.text).join('')
+}
+
+describe.skipIf(!hasRg)('search tools over the real bash executor + real rg', () => {
+  beforeEach(async () => {
+    dir = await mkdtemp(join(tmpdir(), 'dsh-search-int-'))
+    await mkdir(join(dir, 'src'), { recursive: true })
+    await mkdir(join(dir, '.git'), { recursive: true })
+    await mkdir(join(dir, 'spaced dir'), { recursive: true })
+    await writeFile(join(dir, 'src', 'alpha.ts'), 'export const alpha = 1\n// TODO: refit alpha\n')
+    await writeFile(join(dir, 'src', 'beta.ts'), 'export const beta = 2\n')
+    await writeFile(join(dir, 'notes.md'), 'alpha appears here too\n')
+    await writeFile(join(dir, '.hidden.ts'), 'export const hidden = 3\n')
+    await writeFile(join(dir, '.git', 'config.ts'), 'never listed\n')
+    await writeFile(join(dir, 'spaced dir', "wei'rd \"name\".ts"), 'const inside = true\n')
+    // Deterministic --sort=modified order: alpha oldest, beta newest.
+    await utimes(join(dir, 'src', 'alpha.ts'), new Date(2000, 0, 1), new Date(2000, 0, 1))
+    await utimes(join(dir, 'src', 'beta.ts'), new Date(2020, 0, 1), new Date(2020, 0, 1))
+
+    ctx = new Context()
+    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(LocalBashExecutor, { cwd: dir, timeoutMs: 20_000 })
+    await ctx.plugin(ToolFsSearch)
+  })
+
+  afterEach(async () => {
+    await rm(dir, { recursive: true, force: true })
+  })
+
+  describe('glob', () => {
+    it('discovers files by pattern, sorted by modification time, hidden included, .git excluded', async () => {
+      const result = await call('glob', { pattern: '**/*.ts' })
+      expect(result.isError).toBe(false)
+      const paths = text(result).split('\n')
+      expect(paths.indexOf('src/alpha.ts')).toBeLessThan(paths.indexOf('src/beta.ts'))
+      expect(paths).toContain('.hidden.ts')
+      expect(paths).toContain("spaced dir/wei'rd \"name\".ts")
+      expect(paths).not.toContain('.git/config.ts')
+      expect(paths).not.toContain('notes.md')
+    })
+
+    it('scopes to a directory search root (path arg)', async () => {
+      const result = await call('glob', { pattern: '*.ts', path: 'src' })
+      expect(text(result).split('\n').sort()).toEqual(['src/alpha.ts', 'src/beta.ts'])
+    })
+
+    it('reports zero discoveries as No files found', async () => {
+      expect(text(await call('glob', { pattern: '*.nomatch' }))).toBe('No files found')
+    })
+
+    it('excludes VCS internals even when the search root IS the VCS directory', async () => {
+      // The prune glob alone never matches root-prefixed paths when rg is
+      // rooted at .git; the paired contents glob keeps the exclusion airtight.
+      expect(text(await call('glob', { pattern: '*', path: '.git' }))).toBe('No files found')
+    })
+
+    it('classifies an invalid glob as SEARCH_INVALID_PATTERN', async () => {
+      const result = await call('glob', { pattern: '[' })
+      expect(result.isError).toBe(true)
+      expect(result.error).toMatchObject({ name: 'SearchError', code: 'SEARCH_INVALID_PATTERN' })
+    })
+  })
+
+  describe('grep', () => {
+    it('greps a directory tree with grouped, line-numbered output', async () => {
+      const result = await call('grep', { pattern: 'alpha' })
+      expect(result.isError).toBe(false)
+      const output = text(result)
+      expect(output).toContain('Found 3 matches')
+      expect(output).toContain('src/alpha.ts\nLine 1: export const alpha = 1\nLine 2: // TODO: refit alpha')
+      expect(output).toContain('notes.md\nLine 1: alpha appears here too')
+    })
+
+    it('greps a single FILE target', async () => {
+      const result = await call('grep', { pattern: 'alpha', path: 'notes.md' })
+      expect(text(result)).toBe('Found 1 match\n\nnotes.md\nLine 1: alpha appears here too')
+    })
+
+    it('greps a directory target with an include filter', async () => {
+      const result = await call('grep', { pattern: 'alpha', path: '.', include: '*.ts' })
+      const output = text(result)
+      expect(output).toContain('alpha.ts')
+      expect(output).not.toContain('notes.md')
+    })
+
+    it('a hostile pattern stays inert (no command substitution, the world untouched)', async () => {
+      const canary = join(dir, 'pwned')
+      const result = await call('grep', { pattern: `$(touch ${canary})` })
+      expect(result.isError).toBe(false) // exit 1: found nothing, executed nothing
+      expect(text(result)).toBe('No matches found')
+      expect(spawnSync('test', ['-e', canary]).status).not.toBe(0)
+    })
+
+    it('a leading-dash pattern is a pattern, not a flag', async () => {
+      await writeFile(join(dir, 'dashes.txt'), 'value --flag value\n')
+      const result = await call('grep', { pattern: '--flag', path: 'dashes.txt' })
+      expect(text(result)).toBe('Found 1 match\n\ndashes.txt\nLine 1: value --flag value')
+    })
+
+    it('classifies a real rg regex error as SEARCH_INVALID_PATTERN', async () => {
+      const result = await call('grep', { pattern: '(unclosed' })
+      expect(result.isError).toBe(true)
+      expect(result.error).toMatchObject({ code: 'SEARCH_INVALID_PATTERN' })
+    })
+
+    it('classifies a missing target as SEARCH_FAILED', async () => {
+      const result = await call('grep', { pattern: 'x', path: 'no-such-dir' })
+      expect(result.isError).toBe(true)
+      expect(result.error).toMatchObject({ code: 'SEARCH_FAILED' })
+    })
+  })
+
+  describe('per-session cwd', () => {
+    it('resolves the search in the SESSION workspace, not the executor config cwd', async () => {
+      const sessionDir = await mkdtemp(join(tmpdir(), 'dsh-search-session-'))
+      try {
+        await writeFile(join(sessionDir, 'only-here.ts'), 'const sessionFile = true\n')
+        const agentObj = { session: { header: { id: 'session-int', cwd: sessionDir } } }
+        const globbed = await call('glob', { pattern: '*.ts' }, agentObj)
+        expect(text(globbed)).toBe('only-here.ts')
+        const grepped = await call('grep', { pattern: 'sessionFile' }, agentObj)
+        expect(text(grepped)).toContain('only-here.ts\nLine 1: const sessionFile = true')
+      } finally {
+        await rm(sessionDir, { recursive: true, force: true })
+      }
+    })
+  })
+
+  describe('bash-start infrastructure failures stay in the SEARCH_* taxonomy', () => {
+    it('a pre-aborted exec.signal (real executor rejects before spawn) is SEARCH_ABORTED', async () => {
+      const controller = new AbortController()
+      controller.abort()
+      const result = await ctx.tools.execute({
+        callId: CallId(`it-${++callCounter}`),
+        name: 'grep',
+        arguments: { pattern: 'x' },
+        signal: controller.signal,
+      })
+      expect(result.isError).toBe(true)
+      expect(result.error).toMatchObject({ name: 'SearchError', code: 'SEARCH_ABORTED' })
+    })
+
+    it('an unusable session cwd (spawn failure) is SEARCH_FAILED', async () => {
+      const gone = join(dir, 'deleted-session-dir')
+      const result = await call('glob', { pattern: '*' }, { session: { header: { id: 'session-int', cwd: gone } } })
+      expect(result.isError).toBe(true)
+      expect(result.error).toMatchObject({ name: 'SearchError', code: 'SEARCH_FAILED' })
+      expect(text(result)).toContain('could not start')
+    })
+  })
+})
diff --git a/packages/fs/tool-fs-search/tests/load-path.spec.ts b/packages/fs/tool-fs-search/tests/load-path.spec.ts
new file mode 100644
index 0000000000..d3c28619a3
--- /dev/null
+++ b/packages/fs/tool-fs-search/tests/load-path.spec.ts
@@ -0,0 +1,50 @@
+/**
+ * Real-load-path guard for @deepseek-ai/dsh-tool-fs-search. `tool-fs-search` is
+ * a NAMESPACE plugin with `inject` — so a stray `export default apply` would
+ * make the cordis Loader's `unwrapExports` (`exports.default ?? exports`)
+ * collapse the module to the bare `apply` function, DROPPING `inject`. The
+ * plugin would then read `ctx.bash` without having injected it and throw
+ * `cannot get property … without inject` the moment it loads (postmortem 0001).
+ *
+ * A hand-built `ctx.plugin({ apply, inject })` mount CANNOT catch that — it
+ * bypasses `unwrapExports`. So this test unwraps the module through the REAL
+ * `Loader.prototype.unwrapExports` and mounts the result over a bash executor,
+ * exercising the exact path the Loader uses. Prove the guard bites: add
+ * `export default apply` to `src/index.ts`, watch this go red, revert.
+ */
+
+import { describe, expect, it } from 'vitest'
+import { Context } from 'cordis'
+import Loader from '@cordisjs/plugin-loader'
+import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import ToolRegistry from '@deepseek-ai/dsh-tools'
+import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
+import * as toolFsSearch from '@deepseek-ai/dsh-tool-fs-search'
+
+describe('dsh-tool-fs-search real-load-path guard', () => {
+  it('has no default export and keeps name/inject/Config through unwrapExports', () => {
+    expect('default' in toolFsSearch).toBe(false)
+
+    const loader = Object.create(Loader.prototype) as Loader
+    const unwrapped = loader.unwrapExports(toolFsSearch) as Record<string, unknown>
+    expect(unwrapped).toBe(toolFsSearch)
+    expect(unwrapped.name).toBe('tool-fs-search')
+    expect(unwrapped.inject).toEqual(['tools', 'systemPrompt', 'bash'])
+    expect(typeof unwrapped.Config).toBe('function')
+    expect(typeof unwrapped.apply).toBe('function')
+  })
+
+  it('boots over ctx.bash through the unwrapped module without an inject error', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(LocalBashExecutor, {})
+
+    const loader = Object.create(Loader.prototype) as Loader
+    const unwrapped = loader.unwrapExports(toolFsSearch) as Parameters<Context['plugin']>[0]
+    // A collapsed export shape (dropped inject) would throw "without inject" here.
+    const fiber = await ctx.plugin(unwrapped)
+    expect(ctx.tools.schemas().map(s => s.name)).toEqual(expect.arrayContaining(['glob', 'grep']))
+    await fiber.dispose()
+  })
+})
diff --git a/packages/fs/tool-fs-search/tests/shell-quote.spec.ts b/packages/fs/tool-fs-search/tests/shell-quote.spec.ts
new file mode 100644
index 0000000000..84c8506be1
--- /dev/null
+++ b/packages/fs/tool-fs-search/tests/shell-quote.spec.ts
@@ -0,0 +1,59 @@
+/**
+ * Unit tests for the shell-quoting safety boundary, plus a REAL round-trip:
+ * every adversarial value, quoted, must survive `bash -c "printf '%s' <quoted>"`
+ * byte-for-byte — proving the quoting is inert in an actual shell, not just
+ * against a mental model of one.
+ */
+
+import { describe, expect, it } from 'vitest'
+import { spawnSync } from 'node:child_process'
+import { singleQuote } from '@deepseek-ai/dsh-tool-fs-search'
+
+/** Adversarial values a model could pass as pattern / path / include. */
+const HOSTILE: readonly string[] = [
+  'plain',
+  'with spaces',
+  "it's got 'quotes'",
+  '"double quoted"',
+  '$(rm -rf /tmp/nope)',
+  '`touch /tmp/nope`',
+  '$HOME and ${PATH}',
+  'semi;colon && chain || pipe | bg &',
+  'newline\nin the middle',
+  '-leading-dash',
+  '--leading-double-dash',
+  '*?[a-z]{x,y}',
+  '!bang',
+  '\\backslash\\',
+  '~tilde',
+  '# not a comment',
+  '>redirect <input 2>&1',
+]
+
+describe('singleQuote', () => {
+  it('wraps a plain value in single quotes', () => {
+    expect(singleQuote('abc')).toBe("'abc'")
+  })
+
+  it("rewrites embedded single quotes as '\\''", () => {
+    expect(singleQuote("a'b")).toBe("'a'\\''b'")
+    expect(singleQuote("''")).toBe("''\\'''\\'''")
+  })
+
+  it.each(HOSTILE.map(value => [JSON.stringify(value), value] as const))(
+    'round-trips %s through a real bash -c unchanged',
+    (_label, value) => {
+      const result = spawnSync('bash', ['-c', `printf '%s' ${singleQuote(value)}`], { encoding: 'utf8' })
+      expect(result.status).toBe(0)
+      expect(result.stdout).toBe(value)
+    },
+  )
+
+  it('a quoted command substitution does not execute (the world stays untouched)', () => {
+    const canary = `/tmp/dsh-quote-canary-${process.pid}`
+    const result = spawnSync('bash', ['-c', `printf '%s' ${singleQuote(`$(touch ${canary})`)}`], { encoding: 'utf8' })
+    expect(result.stdout).toBe(`$(touch ${canary})`)
+    // The canary file must NOT exist — the substitution stayed literal.
+    expect(spawnSync('test', ['-e', canary]).status).not.toBe(0)
+  })
+})
diff --git a/packages/fs/tool-fs-search/tests/tools.spec.ts b/packages/fs/tool-fs-search/tests/tools.spec.ts
new file mode 100644
index 0000000000..5363344f07
--- /dev/null
+++ b/packages/fs/tool-fs-search/tests/tools.spec.ts
@@ -0,0 +1,632 @@
+/**
+ * Consumer-surface tests for the search tools over a FAKE bash executor and a
+ * FAKE spill backend, exercised through `ctx.tools.execute()` so nothing
+ * bypasses the tool registry. The fake executor makes every seam outcome
+ * scriptable — truncated stdout with/without a raw spill path, abort/timeout,
+ * signal kills, ripgrep exit codes — so these tests verify schemas, argument
+ * validation, shell-safe command construction, workdir derivation, signal
+ * forwarding, `SEARCH_*` error classification, retention, formatted-result
+ * spill handoff, and the no-background-task invariant. Real-`rg` behavior is
+ * pinned separately in integration.spec.ts.
+ */
+
+import { describe, expect, it } from 'vitest'
+import { Context } from 'cordis'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import SystemPrompt, { renderPrompt } from '@deepseek-ai/dsh-system-prompt'
+import ToolRegistry from '@deepseek-ai/dsh-tools'
+import { BashExecutor } from '@deepseek-ai/dsh-bash'
+import type { BashExecRequest, BashExecSpec, BashProcess, BashRunResult } from '@deepseek-ai/dsh-bash'
+import { SpillLocator, SpillStore } from '@deepseek-ai/dsh-spill'
+import type { SaveTextSpill, SpillRef } from '@deepseek-ai/dsh-spill'
+import * as ToolFsSearch from '@deepseek-ai/dsh-tool-fs-search'
+import {
+  buildGlobCommand,
+  buildGrepCommand,
+  formatGrepMatches,
+  parseGrepMatches,
+  presentGlobCall,
+  presentGrepCall,
+  previewLine,
+  toWorkdirRelative,
+} from '@deepseek-ai/dsh-tool-fs-search'
+
+/** A successful run result over the given stdout; overrides script the failure shapes. */
+function runResult(stdout: string, overrides?: Partial<BashRunResult>): BashRunResult {
+  return {
+    exitCode: 0,
+    signal: null,
+    timedOut: false,
+    aborted: false,
+    timeoutMs: 60_000,
+    stdout: { text: stdout, truncated: false },
+    stderr: { text: '', truncated: false },
+    ...overrides,
+  }
+}
+
+/**
+ * A scriptable fake executor: `resolve()` mirrors the real request→spec
+ * defaulting (workdir falls back to `/work`), `run()` returns whatever the
+ * test armed via `handler`, and `start()` throws — the search tools must NEVER
+ * create a background task.
+ */
+class FakeBash extends BashExecutor {
+  requests: BashExecRequest[] = []
+  specs: BashExecSpec[] = []
+  startCalls = 0
+  handler: (spec: BashExecSpec) => BashRunResult = () => runResult('')
+
+  override resolve(request: BashExecRequest): BashExecSpec {
+    this.requests.push(request)
+    return {
+      command: request.command,
+      workdir: request.workdir ?? '/work',
+      timeoutMs: request.timeoutMs ?? 60_000,
+      stdoutMaxBytes: request.stdoutMaxBytes ?? 64_000,
+      signal: request.signal,
+      sandboxMode: request.sandboxMode,
+    }
+  }
+  override run(spec: BashExecSpec): Promise<BashRunResult> {
+    this.specs.push(spec)
+    return Promise.resolve(this.handler(spec))
+  }
+  override start(): BashProcess {
+    this.startCalls++
+    throw new Error('search tools must never start a background task')
+  }
+}
+
+/** A recording spill backend; arm `failWith` to script a storage failure. */
+class FakeSpill extends SpillStore {
+  saves: SaveTextSpill[] = []
+  failWith?: Error
+
+  override saveText(input: SaveTextSpill): Promise<SpillRef> {
+    if (this.failWith) return Promise.reject(this.failWith)
+    this.saves.push(input)
+    return Promise.resolve({
+      locator: SpillLocator(`/spill/${input.suggestedName}`),
+      bytes: Buffer.byteLength(input.content, 'utf8'),
+      retrievalHint: 'Use the fake retrieval hint.',
+    })
+  }
+}
+
+interface SetupOptions {
+  config?: ToolFsSearch.Config
+  spill?: boolean
+}
+
+async function setup(options: SetupOptions = {}) {
+  const ctx = new Context()
+  await ctx.plugin(SystemPrompt)
+  await ctx.plugin(ToolRegistry)
+  await ctx.plugin(FakeBash)
+  if (options.spill === true) await ctx.plugin(FakeSpill)
+  const fiber = await ctx.plugin(ToolFsSearch, options.config)
+  const bash = ctx.bash as FakeBash
+  const spill = options.spill === true ? ctx.get('spillStore') as FakeSpill : undefined
+  return { ctx, bash, spill, fiber }
+}
+
+/** A stand-in agent whose session header carries the given cwd (and a stable id). */
+const agent = (cwd?: string) => ({ session: { header: { id: 'session-1', ...cwd !== undefined ? { cwd } : {} } } })
+
+let callCounter = 0
+function call(ctx: Context, name: string, args: unknown, options: { agent?: object; signal?: AbortSignal } = {}) {
+  return ctx.tools.execute({
+    callId: CallId(`call-${++callCounter}`),
+    name,
+    arguments: args,
+    ...options.agent ? { agent: options.agent as never } : {},
+    ...options.signal ? { signal: options.signal } : {},
+  })
+}
+
+function text(result: { content: { type: string; text?: string }[] }): string {
+  return result.content.filter(b => b.type === 'text').map(b => b.text).join('')
+}
+
+/** One rg --json match record line. */
+function matchLine(path: string, lineNumber: number, lineText: string): string {
+  return JSON.stringify({ type: 'match', data: { path: { text: path }, lines: { text: lineText }, line_number: lineNumber, absolute_offset: 0, submatches: [] } })
+}
+
+describe('registration', () => {
+  it('registers glob and grep with their prompt sections', async () => {
+    const { ctx } = await setup()
+    expect(ctx.tools.schemas().map(s => s.name).sort()).toEqual(['glob', 'grep'])
+    const prompt = renderPrompt(await ctx.systemPrompt.assemble())
+    expect(prompt).toContain('Use the glob tool')
+    expect(prompt).toContain('Use the grep tool')
+  })
+
+  it('stays pending until ctx.bash exists (inject)', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(ToolFsSearch) // no bash executor
+    expect(ctx.tools.schemas()).toHaveLength(0)
+  })
+
+  it('unregisters everything on fiber disposal (HMR safety)', async () => {
+    const { ctx, fiber } = await setup()
+    expect(ctx.tools.schemas()).toHaveLength(2)
+    await fiber.dispose()
+    expect(ctx.tools.schemas()).toHaveLength(0)
+    const sections = (await ctx.systemPrompt.assemble()).sections.map(s => s.name)
+    expect(sections).not.toContain('tool:glob')
+    expect(sections).not.toContain('tool:grep')
+  })
+
+  it('attaches the configured timeoutMs to both tool definitions', async () => {
+    const { ctx } = await setup({ config: { timeoutMs: 5000 } })
+    expect(ctx.tools.get('glob')?.timeoutMs).toBe(5000)
+    expect(ctx.tools.get('grep')?.timeoutMs).toBe(5000)
+  })
+
+  it('defaults the timeout budget to 30 seconds', async () => {
+    const { ctx } = await setup()
+    expect(ctx.tools.get('glob')?.timeoutMs).toBe(30_000)
+    expect(ctx.tools.get('grep')?.timeoutMs).toBe(30_000)
+  })
+})
+
+describe('config validation', () => {
+  it.each([
+    ['globMaxResults', { globMaxResults: 0 }],
+    ['grepMaxMatches', { grepMaxMatches: -1 }],
+    ['grepMaxLineBytes', { grepMaxLineBytes: 1.5 }],
+    ['rawOutputMaxBytes', { rawOutputMaxBytes: 0 }],
+    ['timeoutMs', { timeoutMs: -100 }],
+  ] as const)('rejects a non-positive or fractional %s at load', async (name, config) => {
+    const ctx = new Context()
+    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(FakeBash)
+    await expect(ctx.plugin(ToolFsSearch, config)).rejects.toThrow(new RegExp(`tool-fs-search: ${name} must be a positive integer`))
+  })
+})
+
+describe('command construction (shell-safe)', () => {
+  it('glob: fixed rg --files template with quoted pattern and paired VCS excludes', () => {
+    const command = buildGlobCommand({ pattern: '**/*.ts' })
+    expect(command).toBe(
+      "rg --files --glob='**/*.ts' --sort=modified --no-ignore --hidden "
+      + "--glob='!**/.git' --glob='!**/.git/**' --glob='!**/.svn' --glob='!**/.svn/**' "
+      + "--glob='!**/.hg' --glob='!**/.hg/**' --glob='!**/.bzr' --glob='!**/.bzr/**' "
+      + "--glob='!**/.jj' --glob='!**/.jj/**' --glob='!**/.sl' --glob='!**/.sl/**'",
+    )
+  })
+
+  it('glob: the search root rides behind -- and is quoted', () => {
+    const command = buildGlobCommand({ pattern: '*.md', path: 'docs dir' })
+    expect(command).toContain("-- 'docs dir'")
+  })
+
+  it('grep: fixed rg --json template with the pattern in --regexp= form', () => {
+    expect(buildGrepCommand({ pattern: 'foo.*bar' })).toBe("rg --json --regexp='foo.*bar'")
+  })
+
+  it('grep: include and path are quoted, include in --glob= form, path behind --', () => {
+    const command = buildGrepCommand({ pattern: 'x', path: '-leading-dash', include: '*.{ts,tsx}' })
+    expect(command).toBe("rg --json --regexp='x' --glob='*.{ts,tsx}' -- '-leading-dash'")
+  })
+
+  it.each([
+    ['a command-substitution pattern', '$(rm -rf /)', "'$(rm -rf /)'"],
+    ['a backtick pattern', '`touch pwned`', "'`touch pwned`'"],
+    ['a pattern with double quotes and spaces', 'say "hi there"', '\'say "hi there"\''],
+    ['a pattern with single quotes', "it's", '\'it\'\\\'\'s\''],
+    ['a pattern with newlines', 'a\nb', "'a\nb'"],
+    ['a leading-dash pattern', '--flag', "'--flag'"],
+    ['glob metacharacters', '*?[a-z]{x,y}', "'*?[a-z]{x,y}'"],
+  ])('quotes %s into one inert shell word', (_label, raw, quoted) => {
+    expect(buildGrepCommand({ pattern: raw })).toBe(`rg --json --regexp=${quoted}`)
+  })
+})
+
+describe('workdir derivation and signal forwarding', () => {
+  it('forwards the session cwd as the request workdir', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('a.ts\n')
+    await call(ctx, 'glob', { pattern: '*' }, { agent: agent('/sessions/s1') })
+    expect(bash.requests[0]?.workdir).toBe('/sessions/s1')
+    expect(bash.specs[0]?.workdir).toBe('/sessions/s1')
+  })
+
+  it('omits the request workdir without a session cwd so resolve() defaults apply', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('a.ts\n')
+    await call(ctx, 'glob', { pattern: '*' }, { agent: agent() })
+    expect(bash.requests[0]).not.toHaveProperty('workdir')
+    expect(bash.specs[0]?.workdir).toBe('/work')
+    // A non-agent caller takes the same default path.
+    await call(ctx, 'grep', { pattern: 'x' })
+    expect(bash.requests[1]).not.toHaveProperty('workdir')
+  })
+
+  it('forwards exec.signal into the bash spec (the abort reaches the backend)', async () => {
+    const { ctx, bash } = await setup()
+    const controller = new AbortController()
+    controller.abort()
+    bash.handler = spec => runResult('', { aborted: spec.signal?.aborted === true })
+    const result = await call(ctx, 'grep', { pattern: 'x' }, { signal: controller.signal })
+    expect(bash.specs[0]?.signal).toBe(controller.signal)
+    expect(result.isError).toBe(true)
+    expect(result.error).toMatchObject({ name: 'SearchError', code: 'SEARCH_ABORTED' })
+    expect(text(result)).toContain('aborted')
+  })
+
+  it('reports the bash executor timeout as SEARCH_ABORTED with the budget', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { timedOut: true, timeoutMs: 1234, exitCode: null, signal: 'SIGTERM' })
+    const result = await call(ctx, 'glob', { pattern: '*' })
+    expect(result.isError).toBe(true)
+    expect(result.error).toMatchObject({ code: 'SEARCH_ABORTED' })
+    expect(text(result)).toContain('timed out after 1234ms')
+  })
+
+  it('translates a run() rejection under a pre-aborted signal into SEARCH_ABORTED', async () => {
+    // The seam contract: run() REJECTS for a pre-aborted signal (it never
+    // spawns). The plain rejection must not escape the SEARCH_* taxonomy.
+    const { ctx, bash } = await setup()
+    const controller = new AbortController()
+    controller.abort()
+    bash.handler = () => { throw new Error('aborted before spawn') }
+    const result = await call(ctx, 'grep', { pattern: 'x' }, { signal: controller.signal })
+    expect(result.isError).toBe(true)
+    expect(result.error).toMatchObject({ name: 'SearchError', code: 'SEARCH_ABORTED' })
+  })
+
+  it('translates a run() rejection without an abort (unusable workdir) into SEARCH_FAILED', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => { throw new Error('spawn bash ENOENT') }
+    const result = await call(ctx, 'glob', { pattern: '*' })
+    expect(result.isError).toBe(true)
+    expect(result.error).toMatchObject({ name: 'SearchError', code: 'SEARCH_FAILED' })
+    expect(text(result)).toContain('could not start')
+  })
+})
+
+describe('exit semantics and failure classification', () => {
+  it('exit 1 is a successful empty search', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { exitCode: 1 })
+    const glob = await call(ctx, 'glob', { pattern: '*.nope' })
+    expect(glob.isError).toBe(false)
+    expect(text(glob)).toBe('No files found')
+    const grep = await call(ctx, 'grep', { pattern: 'nope' })
+    expect(grep.isError).toBe(false)
+    expect(text(grep)).toBe('No matches found')
+  })
+
+  it('a regex parse error classifies as SEARCH_INVALID_PATTERN', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { exitCode: 2, stderr: { text: 'rg: regex parse error:\n    (\nerror: unclosed group', truncated: false } })
+    const result = await call(ctx, 'grep', { pattern: '(' })
+    expect(result.error).toMatchObject({ code: 'SEARCH_INVALID_PATTERN' })
+    expect(text(result)).toContain('regex parse error')
+  })
+
+  it('a glob parse error classifies as SEARCH_INVALID_PATTERN', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { exitCode: 2, stderr: { text: 'rg: error parsing glob \'[\': unclosed character class', truncated: false } })
+    const result = await call(ctx, 'glob', { pattern: '[' })
+    expect(result.error).toMatchObject({ code: 'SEARCH_INVALID_PATTERN' })
+  })
+
+  it('a missing rg binary classifies as SEARCH_FAILED naming ripgrep', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { exitCode: 127, stderr: { text: 'bash: line 1: rg: command not found', truncated: false } })
+    const result = await call(ctx, 'glob', { pattern: '*' })
+    expect(result.error).toMatchObject({ code: 'SEARCH_FAILED' })
+    expect(text(result)).toContain('requires ripgrep (rg)')
+    // The same classification holds from either evidence alone: the 127 exit
+    // with silent stderr, or a shell's command-not-found text on another exit.
+    bash.handler = () => runResult('', { exitCode: 127 })
+    expect(text(await call(ctx, 'glob', { pattern: '*' }))).toContain('requires ripgrep (rg)')
+    bash.handler = () => runResult('', { exitCode: 2, stderr: { text: 'sh: rg: command not found', truncated: false } })
+    expect(text(await call(ctx, 'grep', { pattern: 'x' }))).toContain('requires ripgrep (rg)')
+  })
+
+  it('other nonzero exits are SEARCH_FAILED carrying the stderr excerpt', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { exitCode: 2, stderr: { text: 'rg: missing.dir: IO error: no such file or directory', truncated: false } })
+    const result = await call(ctx, 'grep', { pattern: 'x', path: 'missing.dir' })
+    expect(result.error).toMatchObject({ code: 'SEARCH_FAILED' })
+    expect(text(result)).toContain('IO error')
+  })
+
+  it('a nonzero exit with EMPTY stderr still reports the exit code', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { exitCode: 3 })
+    const result = await call(ctx, 'glob', { pattern: '*' })
+    expect(result.error).toMatchObject({ code: 'SEARCH_FAILED' })
+    expect(text(result)).toContain('exit 3')
+  })
+
+  it('truncated stderr gains a truncation note and stderr.spillPath is never read', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', {
+      exitCode: 2,
+      stderr: { text: 'tail of diagnostics', truncated: true, spillPath: '/does/not/exist-and-never-read' },
+    })
+    const result = await call(ctx, 'grep', { pattern: 'x' })
+    expect(text(result)).toContain('tail of diagnostics [stderr truncated]')
+  })
+
+  it('a signal kill (not timeout, not abort) is SEARCH_FAILED', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { exitCode: null, signal: 'SIGKILL' })
+    const result = await call(ctx, 'grep', { pattern: 'x' })
+    expect(result.error).toMatchObject({ code: 'SEARCH_FAILED' })
+    expect(text(result)).toContain('SIGKILL')
+  })
+
+  it('a null exit with no signal (defensive) is SEARCH_FAILED', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { exitCode: null, signal: null })
+    const result = await call(ctx, 'glob', { pattern: '*' })
+    expect(result.error).toMatchObject({ code: 'SEARCH_FAILED' })
+  })
+})
+
+describe('raw output acquisition', () => {
+  it('passes rawOutputMaxBytes to bash as the stdout capture budget', async () => {
+    const { ctx, bash } = await setup({ config: { rawOutputMaxBytes: 1234 } })
+    bash.handler = () => runResult('', { exitCode: 1 })
+    await call(ctx, 'glob', { pattern: '*.ts' })
+    await call(ctx, 'grep', { pattern: 'needle' })
+    expect(bash.requests.map(request => request.stdoutMaxBytes)).toEqual([1234, 1234])
+    expect(bash.specs.map(spec => spec.stdoutMaxBytes)).toEqual([1234, 1234])
+  })
+
+  it('fails with SEARCH_RAW_OUTPUT_OVERFLOW when truncated stdout has a raw spill path', async () => {
+    const { ctx, bash } = await setup({ config: { rawOutputMaxBytes: 16 } })
+    bash.handler = () => runResult('', { stdout: { text: 'x', truncated: true, spillPath: '/does/not/get-read' } })
+    const result = await call(ctx, 'glob', { pattern: '*' })
+    expect(result.error).toMatchObject({ code: 'SEARCH_RAW_OUTPUT_OVERFLOW' })
+    expect(text(result)).toContain('narrow pattern, path, or include')
+  })
+
+  it('fails with SEARCH_RAW_OUTPUT_OVERFLOW when UNTRUNCATED inline stdout exceeds the cap', async () => {
+    // An executor retaining more inline than this package's cap (or a
+    // deployment lowering rawOutputMaxBytes below the bash retention) must not
+    // smuggle an over-cap parse through the untruncated path.
+    const { ctx, bash } = await setup({ config: { rawOutputMaxBytes: 16 } })
+    bash.handler = () => runResult(`${'x'.repeat(64)}\n`)
+    const result = await call(ctx, 'grep', { pattern: 'x' })
+    expect(result.error).toMatchObject({ name: 'SearchError', code: 'SEARCH_RAW_OUTPUT_OVERFLOW' })
+    expect(text(result)).toContain('narrow pattern, path, or include')
+  })
+
+  it('fails with SEARCH_RAW_OUTPUT_OVERFLOW when truncated stdout has no spill path', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { stdout: { text: 'partial', truncated: true } })
+    const result = await call(ctx, 'grep', { pattern: 'x' })
+    expect(result.error).toMatchObject({ code: 'SEARCH_RAW_OUTPUT_OVERFLOW' })
+  })
+})
+
+describe('glob results', () => {
+  it('lists workdir-relative paths (absolute output under the workdir is relativized)', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('/sessions/s1/src/a.ts\n/elsewhere/b.ts\nrel/c.ts\n')
+    const result = await call(ctx, 'glob', { pattern: '*' }, { agent: agent('/sessions/s1') })
+    expect(text(result)).toBe('src/a.ts\n/elsewhere/b.ts\nrel/c.ts')
+  })
+
+  it('validates arguments (blank pattern, blank path)', async () => {
+    const { ctx } = await setup()
+    expect(text(await call(ctx, 'glob', { pattern: '  ' }))).toContain('pattern must be a non-empty string')
+    expect(text(await call(ctx, 'glob', { pattern: '*', path: ' ' }))).toContain('path must be a non-empty string')
+  })
+
+  it('threads a valid path through to the command as the quoted search root', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('sub/a.ts\n')
+    const result = await call(ctx, 'glob', { pattern: '*.ts', path: 'sub' })
+    expect(result.isError).toBe(false)
+    expect(bash.specs[0]?.command).toContain("-- 'sub'")
+  })
+
+  it('caps at globMaxResults and saves the FULL sorted list through spillStore', async () => {
+    const { ctx, bash, spill } = await setup({ config: { globMaxResults: 2 }, spill: true })
+    bash.handler = () => runResult('a.ts\nb.ts\nc.ts\nd.ts\n')
+    const result = await call(ctx, 'glob', { pattern: '*.ts' }, { agent: agent('/w') })
+    expect(result.isError).toBe(false)
+    expect(text(result)).toBe('a.ts\nb.ts\n\n(Showing 2 of 4 paths. Full sorted result stored at: /spill/glob-results.txt. Use the fake retrieval hint.)')
+    expect(spill?.saves).toHaveLength(1)
+    expect(spill?.saves[0]).toMatchObject({
+      owner: { sessionId: 'session-1' },
+      source: { toolName: 'glob', label: 'result' },
+      suggestedName: 'glob-results.txt',
+      content: 'a.ts\nb.ts\nc.ts\nd.ts',
+    })
+    expect(spill?.saves[0]?.source.callId).toBeDefined()
+  })
+
+  it('does not create a spill file when the result fits inline', async () => {
+    const { ctx, bash, spill } = await setup({ spill: true })
+    bash.handler = () => runResult('a.ts\nb.ts\n')
+    const result = await call(ctx, 'glob', { pattern: '*' }, { agent: agent('/w') })
+    expect(text(result)).toBe('a.ts\nb.ts')
+    expect(spill?.saves).toHaveLength(0)
+  })
+
+  it.each([
+    ['no spill backend loaded', { fail: false, spill: false, ownerless: false }],
+    ['saveText fails', { fail: true, spill: true, ownerless: false }],
+    ['no session owner', { fail: false, spill: true, ownerless: true }],
+  ])('keeps the inline page and reports the unsaved remainder when %s', async (_label, mode) => {
+    const { ctx, bash, spill } = await setup({ config: { globMaxResults: 1 }, spill: mode.spill })
+    if (mode.fail && spill) spill.failWith = new Error('disk full')
+    bash.handler = () => runResult('a.ts\nb.ts\n')
+    const result = await call(ctx, 'glob', { pattern: '*' }, mode.ownerless ? {} : { agent: agent('/w') })
+    expect(result.isError).toBe(false) // spill unavailability never fails the search
+    expect(text(result)).toBe('a.ts\n\n(Showing 1 of 2 paths. The complete result could not be saved; narrow pattern or path to see more.)')
+  })
+})
+
+describe('grep results', () => {
+  it('groups matches by file with line numbers', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult([
+      JSON.stringify({ type: 'begin', data: { path: { text: 'a.ts' } } }),
+      matchLine('a.ts', 3, 'const x = 1\n'),
+      matchLine('a.ts', 9, 'const y = 2\n'),
+      JSON.stringify({ type: 'end', data: { path: { text: 'a.ts' } } }),
+      matchLine('b.ts', 1, 'const z = 3'),
+      JSON.stringify({ type: 'summary', data: {} }),
+      '',
+    ].join('\n'))
+    const result = await call(ctx, 'grep', { pattern: 'const' })
+    expect(result.isError).toBe(false)
+    expect(text(result)).toBe('Found 3 matches\n\na.ts\nLine 3: const x = 1\nLine 9: const y = 2\n\nb.ts\nLine 1: const z = 3')
+  })
+
+  it('reports a single match in the singular', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult(`${matchLine('a.ts', 1, 'hit')}\n`)
+    expect(text(await call(ctx, 'grep', { pattern: 'hit' }))).toBe('Found 1 match\n\na.ts\nLine 1: hit')
+  })
+
+  it('relativizes absolute match paths against the resolved workdir', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult(`${matchLine('/sessions/s1/deep/a.ts', 2, 'hit')}\n`)
+    const result = await call(ctx, 'grep', { pattern: 'hit', path: '/sessions/s1' }, { agent: agent('/sessions/s1') })
+    expect(text(result)).toContain('deep/a.ts\nLine 2: hit')
+  })
+
+  it('previews a long matched line at grepMaxLineBytes preserving UTF-8', async () => {
+    const { ctx, bash } = await setup({ config: { grepMaxLineBytes: 7 } })
+    // 'héllo wörld' cut at 7 bytes lands mid-'é'? h(1)é(2)l(1)l(1)o(1)=6, space=7 → clean cut at 7.
+    // Use a multibyte straddle instead: 'aé' repeated — cut at 7 bytes: a(1)é(2)a(1)é(2)=6 +a(1)=7 → next é straddles: trimmed.
+    bash.handler = () => runResult(`${matchLine('a.txt', 1, 'aéaéaéaé')}\n`)
+    const result = await call(ctx, 'grep', { pattern: 'a' })
+    expect(text(result)).toContain('Line 1: aéaéa (line truncated)')
+  })
+
+  it('renders a non-UTF-8 line (rg bytes form) as a placeholder instead of failing', async () => {
+    const { ctx, bash } = await setup()
+    const record = JSON.stringify({ type: 'match', data: { path: { text: 'bin.dat' }, lines: { bytes: 'AAECww==' }, line_number: 4 } })
+    bash.handler = () => runResult(`${record}\n`)
+    expect(text(await call(ctx, 'grep', { pattern: 'x' }))).toContain('Line 4: (line is not valid UTF-8)')
+  })
+
+  it('strips a CRLF terminator from the matched line text', () => {
+    const matches = parseGrepMatches(`${matchLine('a.txt', 1, 'windows line\r\n')}\n`)
+    expect(matches[0]?.line).toBe('windows line')
+  })
+
+  it('caps at grepMaxMatches and spills the full formatted match list', async () => {
+    const { ctx, bash, spill } = await setup({ config: { grepMaxMatches: 2 }, spill: true })
+    bash.handler = () => runResult([
+      matchLine('a.ts', 1, 'one'),
+      matchLine('a.ts', 2, 'two'),
+      matchLine('b.ts', 3, 'three'),
+      '',
+    ].join('\n'))
+    const result = await call(ctx, 'grep', { pattern: 'e' }, { agent: agent('/w') })
+    expect(text(result)).toBe('Found 2 of 3 matches\n\na.ts\nLine 1: one\nLine 2: two\n\n(Full grep result stored at: /spill/grep-results.txt. Use the fake retrieval hint.)')
+    expect(spill?.saves[0]).toMatchObject({
+      source: { toolName: 'grep', label: 'result' },
+      suggestedName: 'grep-results.txt',
+      content: 'Found 3 matches\n\na.ts\nLine 1: one\nLine 2: two\n\nb.ts\nLine 3: three',
+    })
+  })
+
+  it('reports the unsaved remainder when capped with no spill backend', async () => {
+    const { ctx, bash } = await setup({ config: { grepMaxMatches: 1 } })
+    bash.handler = () => runResult(`${matchLine('a.ts', 1, 'one')}\n${matchLine('a.ts', 2, 'two')}\n`)
+    const result = await call(ctx, 'grep', { pattern: 'o' }, { agent: agent('/w') })
+    expect(result.isError).toBe(false)
+    expect(text(result)).toBe('Found 1 of 2 matches\n\na.ts\nLine 1: one\n\n(The complete result could not be saved; narrow pattern, path, or include to see more.)')
+  })
+
+  it('validates arguments (empty pattern, blank path, bad include)', async () => {
+    const { ctx } = await setup()
+    expect(text(await call(ctx, 'grep', { pattern: '' }))).toContain('pattern must be a non-empty string')
+    expect(text(await call(ctx, 'grep', { pattern: 'x', path: '  ' }))).toContain('path must be a non-empty string')
+    expect(text(await call(ctx, 'grep', { pattern: 'x', include: '  ' }))).toContain('include must be a non-empty glob')
+    expect(text(await call(ctx, 'grep', { pattern: 'x', include: '!*.ts' }))).toContain('negated patterns')
+    expect(text(await call(ctx, 'grep', { pattern: 'x', include: '*.ts,*.js' }))).toContain('comma-separated list')
+  })
+
+  it('accepts a whitespace-only pattern (a legitimate regex) and brace alternation in include', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('', { exitCode: 1 })
+    const result = await call(ctx, 'grep', { pattern: '  ', include: '*.{ts,tsx}' })
+    expect(result.isError).toBe(false)
+  })
+})
+
+describe('rg --json transport failures (SEARCH_FAILED)', () => {
+  it.each([
+    ['a non-JSON line', 'not json at all'],
+    ['a non-object record', '42'],
+    ['a match record with no data', JSON.stringify({ type: 'match' })],
+    ['a match record with no path text', JSON.stringify({ type: 'match', data: { path: {}, lines: { text: 'x' }, line_number: 1 } })],
+    ['a match record with a non-object path', JSON.stringify({ type: 'match', data: { path: 'a.ts', lines: { text: 'x' }, line_number: 1 } })],
+    ['a match record with no line number', JSON.stringify({ type: 'match', data: { path: { text: 'a.ts' }, lines: { text: 'x' } } })],
+    ['a match record with no line content', JSON.stringify({ type: 'match', data: { path: { text: 'a.ts' }, line_number: 1 } })],
+    ['a match record with neither text nor bytes', JSON.stringify({ type: 'match', data: { path: { text: 'a.ts' }, lines: {}, line_number: 1 } })],
+  ])('%s fails the search', async (_label, line) => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult(`${line}\n`)
+    const result = await call(ctx, 'grep', { pattern: 'x' })
+    expect(result.isError).toBe(true)
+    expect(result.error).toMatchObject({ name: 'SearchError', code: 'SEARCH_FAILED' })
+  })
+})
+
+describe('the no-background-task invariant', () => {
+  it('never calls ctx.bash.start() across successful and failed searches', async () => {
+    const { ctx, bash } = await setup()
+    bash.handler = () => runResult('a.ts\n')
+    await call(ctx, 'glob', { pattern: '*' })
+    bash.handler = () => runResult('', { exitCode: 2, stderr: { text: 'boom', truncated: false } })
+    await call(ctx, 'grep', { pattern: 'x' })
+    expect(bash.startCalls).toBe(0)
+  })
+})
+
+describe('presentation', () => {
+  it('glob titles carry the pattern and optional root', () => {
+    expect(presentGlobCall({ pattern: '**/*.ts' })).toMatchObject({ card: 'generic', title: 'Glob **/*.ts', kind: 'search' })
+    expect(presentGlobCall({ pattern: '*.md', path: 'docs' }).title).toBe('Glob *.md in docs')
+  })
+
+  it('grep titles carry the pattern, target, and include filter', () => {
+    expect(presentGrepCall({ pattern: 'todo' })).toMatchObject({ card: 'generic', title: 'Grep todo', kind: 'search' })
+    expect(presentGrepCall({ pattern: 'todo', path: 'src', include: '*.ts' }).title).toBe('Grep todo in src (*.ts)')
+  })
+})
+
+describe('helpers', () => {
+  it('toWorkdirRelative maps inside-workdir absolutes and passes everything else through', () => {
+    expect(toWorkdirRelative('/w/a/b.ts', '/w')).toBe('a/b.ts')
+    expect(toWorkdirRelative('/w', '/w')).toBe('.')
+    expect(toWorkdirRelative('/other/b.ts', '/w')).toBe('/other/b.ts')
+    expect(toWorkdirRelative('/w-sibling/b.ts', '/w')).toBe('/w-sibling/b.ts')
+    expect(toWorkdirRelative('rel/b.ts', '/w')).toBe('rel/b.ts')
+    // Normalization makes this land OUTSIDE the workdir → original path kept.
+    expect(toWorkdirRelative('/w/../up.ts', '/w')).toBe('/w/../up.ts')
+  })
+
+  it('previewLine keeps a within-budget line untouched', () => {
+    expect(previewLine('short', 100)).toBe('short')
+  })
+
+  it('formatGrepMatches groups by first-seen file order', () => {
+    const grouped = formatGrepMatches([
+      { path: 'b.ts', lineNumber: 2, line: 'x' },
+      { path: 'a.ts', lineNumber: 1, line: 'y' },
+      { path: 'b.ts', lineNumber: 5, line: 'z' },
+    ])
+    expect(grouped).toBe('b.ts\nLine 2: x\nLine 5: z\n\na.ts\nLine 1: y')
+  })
+})
diff --git a/packages/fs/tool-fs-search/tsconfig.json b/packages/fs/tool-fs-search/tsconfig.json
new file mode 100644
index 0000000000..9241aca15b
--- /dev/null
+++ b/packages/fs/tool-fs-search/tsconfig.json
@@ -0,0 +1,20 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": ["src"],
+  "references": [
+    { "path": "../../../vendor/cosmokit" },
+    { "path": "../../../vendor/cordis" },
+    { "path": "../../../vendor/schemastery" },
+    { "path": "../../util/retention" },
+    { "path": "../../llm/llm" },
+    { "path": "../../core/session" },
+    { "path": "../../core/tools" },
+    { "path": "../../core/system-prompt" },
+    { "path": "../../bash/bash" },
+    { "path": "../../spill/spill" }
+  ]
+}
diff --git a/packages/fs/tool-fs/README.md b/packages/fs/tool-fs/README.md
index bf3633a253..2f116a2b71 100644
--- a/packages/fs/tool-fs/README.md
+++ b/packages/fs/tool-fs/README.md
@@ -34,7 +34,7 @@ Field names are snake_case to match Claude Code and existing harness tool schema
 
 ## The tool is the executor; policy is an event gate
 
-The tools do **not** inject a policy service or inspect any cache. Each tool resolves the path via `ctx.fs.resolve(path, { cwd })` — passing the calling agent's session cwd (`exec.agent.session.header.cwd`) so a relative path resolves against the session's workspace, matching `dsh-tool-bash` (see [the per-session cwd RFC](../../../docs/rfc/implemented/architecture/2026-07-02-fs-per-session-cwd.md)) — then:
+The tools do **not** inject a policy service or inspect any cache. Each tool resolves the path via `ctx.fs.resolve(path, { cwd, signal })` — passing the calling agent's session cwd (`exec.agent.session.header.cwd`) so a relative path resolves against the session's workspace, matching `dsh-tool-bash`, and forwarding tool cancellation through resolution (see [the per-session cwd RFC](../../../docs/rfc/implemented/architecture/2026-07-02-fs-per-session-cwd.md)) — then:
 
 - **read** — one `ctx.fs.stat` (type + size routing + version), then `readText`/`streamText`, then builds the line window, then emits `fs/observed` with a plain `ctx.emit`. (1 stat.)
 - **write** — `ctx.waterfall('fs/write-intent', target, exec, () => undefined)` for the optional guard, then `ctx.fs.writeText(target, content, intent)`, then `fs/observed`. (0 stat.)
@@ -46,6 +46,8 @@ The tool passes `exec` (the tool-execution context) as the opaque `actor` on eve
 
 `fs/observed` fires AFTER the read/write/edit already succeeded, via a plain `ctx.emit`. A listener is contractually a synchronous, side-effect-only recorder (`@deepseek-ai/dsh-fs-policy`'s is a `WeakMap.set`); the tool does not guard the emit, so a listener that throws would surface as the tool's `isError` result — async or fallible observation does not belong on this event.
 
+`read` opts into concurrent scheduling because its only mutation is the synchronous version recorder. Recorder races fail closed when a later `write` or `edit` re-checks the version under its target lock; both mutation tools remain exclusive. See the [parallel tool-call RFC](../../../docs/rfc/implemented/feature/2026-07-10-parallel-tool-call-execution.md).
+
 The package root exports only the Cordis plugin contract (`name`, `inject`, `Config`, and `apply`). Read rendering (line windowing + output formatting) lives in `src/read-render.ts` (Cordis-free, independently unit-tested); `src/read.ts`/`write.ts`/`edit.ts` are the tool executors and `src/index.ts` composes them.
 
 ## Model Experience
@@ -100,6 +102,6 @@ Use the edit tool for targeted changes to existing UTF-8 text files. It replaces
 
 ## Known Limitations and Deferred Work
 
-- **No directory-listing, glob, grep, or search tools ship** — a deferral of [the tool-schemas RFC](../../../docs/rfc/implemented/feature/2026-06-17-filesystem-tool-schemas.md); `ctx.fs.listDir` serves provider code such as skill discovery but still has no model-facing consumer, so models fall back to `bash`.
+- **No model-facing directory listing ships** — `ctx.fs.listDir` serves provider code such as skill discovery, while the sibling [`dsh-tool-fs-search`](../tool-fs-search/) package supplies bash-backed `glob` and `grep` rather than extending the filesystem seam.
 - **`read` handles UTF-8 text files only** — binary-safe reads and PDF/image/multimodal content are deferred; a directory target is `FS_NOT_REGULAR_FILE`.
 - **No timeout surface** — `read`/`write`/`edit` take no timeout argument and declare no `timeout-policy` budget; cancellation rides `exec.signal` only (the deliberate [fs-family stance](../README.md)).
diff --git a/packages/fs/tool-fs/package.json b/packages/fs/tool-fs/package.json
index c21c806e98..f9ef3136bf 100644
--- a/packages/fs/tool-fs/package.json
+++ b/packages/fs/tool-fs/package.json
@@ -36,6 +36,7 @@
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-fs-policy": "workspace:^",
     "@deepseek-ai/dsh-fs": "workspace:^",
     "@deepseek-ai/dsh-fs-local": "workspace:^",
diff --git a/packages/fs/tool-fs/src/edit.ts b/packages/fs/tool-fs/src/edit.ts
index c347d7d958..bc6f7a8fb8 100644
--- a/packages/fs/tool-fs/src/edit.ts
+++ b/packages/fs/tool-fs/src/edit.ts
@@ -12,7 +12,7 @@ import type { ContentBlock } from '@deepseek-ai/dsh-llm'
 import type {} from '@deepseek-ai/dsh-fs'
 import type {} from '@deepseek-ai/dsh-system-prompt'
 import { computeHunkDiffs, diffsFromMeta, type FsDiffMeta } from './diff.ts'
-import { sessionCwd } from './session-cwd.ts'
+import { sessionResolveOptions } from './session-cwd.ts'
 
 /** Validated `edit` arguments after defaulting. */
 interface EditInput {
@@ -75,8 +75,7 @@ export function applyEditTool(ctx: Context): void {
     },
     async execute(args, exec): Promise<{ content: ContentBlock[]; meta?: FsDiffMeta }> {
       const input = parseEditArgs(args)
-      const cwd = sessionCwd(exec)
-      const target = await ctx.fs.resolve(input.filePath, cwd !== undefined ? { cwd } : undefined)
+      const target = await ctx.fs.resolve(input.filePath, sessionResolveOptions(exec))
       // Single-slot decision: the policy plugin returns { version: vObserved } or
       // throws FS_NOT_OBSERVED; the bare default is undefined (unconditional edit).
       // No stat — the bare default never manufactures a version basis.
diff --git a/packages/fs/tool-fs/src/read.ts b/packages/fs/tool-fs/src/read.ts
index 77bb76eb20..a19b073514 100644
--- a/packages/fs/tool-fs/src/read.ts
+++ b/packages/fs/tool-fs/src/read.ts
@@ -13,7 +13,7 @@ import type {} from '@deepseek-ai/dsh-fs'
 import type {} from '@deepseek-ai/dsh-system-prompt'
 import { buildWindow, formatReadOutput } from './read-render.ts'
 import type { FileReadOutcome } from './read-render.ts'
-import { sessionCwd } from './session-cwd.ts'
+import { sessionResolveOptions } from './session-cwd.ts'
 
 /** Default and maximum number of lines returned by one `read` call (the `readLimit` config). */
 export const READ_LIMIT = 2000
@@ -84,10 +84,11 @@ export function applyReadTool(ctx: Context, caps: ReadToolCaps): void {
       offset: { type: 'number', description: '1-based first line to return. Defaults to 1.' },
       limit: { type: 'number', description: `Maximum number of lines to return. Defaults to ${caps.limit}.` },
     },
+    // Observation races fail closed because guarded mutations re-check the version in-lock.
+    isConcurrencySafe: () => true,
     async execute(args, exec): Promise<ContentBlock[]> {
       const input = parseReadArgs(args, caps.limit)
-      const cwd = sessionCwd(exec)
-      const target = await ctx.fs.resolve(input.filePath, cwd !== undefined ? { cwd } : undefined)
+      const target = await ctx.fs.resolve(input.filePath, sessionResolveOptions(exec))
 
       // One stat: type check + size routing + the version recorded as observed.
       // A concurrent write can only make a later guarded mutation fail stale and require reread.
diff --git a/packages/fs/tool-fs/src/session-cwd.ts b/packages/fs/tool-fs/src/session-cwd.ts
index 2f53d630ce..4f98a41a94 100644
--- a/packages/fs/tool-fs/src/session-cwd.ts
+++ b/packages/fs/tool-fs/src/session-cwd.ts
@@ -18,3 +18,16 @@ import type { ToolExecution } from '@deepseek-ai/dsh-tools'
 export function sessionCwd(exec: ToolExecution): string | undefined {
   return exec.agent?.session.header.cwd
 }
+
+/**
+ * Resolution options shared by all model-facing filesystem tools.
+ * @param exec - the tool-execution context supplying session cwd and cancellation.
+ * @returns provider resolution options for the current tool call.
+ */
+export function sessionResolveOptions(exec: ToolExecution): { cwd?: string; signal?: AbortSignal } {
+  const cwd = sessionCwd(exec)
+  return {
+    ...cwd !== undefined ? { cwd } : {},
+    ...exec.signal !== undefined ? { signal: exec.signal } : {},
+  }
+}
diff --git a/packages/fs/tool-fs/src/write.ts b/packages/fs/tool-fs/src/write.ts
index 86fe186e3b..8ab3ce0f62 100644
--- a/packages/fs/tool-fs/src/write.ts
+++ b/packages/fs/tool-fs/src/write.ts
@@ -13,7 +13,7 @@ import type { FsWriteOutcome } from '@deepseek-ai/dsh-fs'
 import type {} from '@deepseek-ai/dsh-fs'
 import type {} from '@deepseek-ai/dsh-system-prompt'
 import { computeHunkDiffs, diffsFromMeta, type FsDiffMeta } from './diff.ts'
-import { sessionCwd } from './session-cwd.ts'
+import { sessionResolveOptions } from './session-cwd.ts'
 
 /**
  * Validate value constraints the schema DSL can't express: only a non-blank
@@ -61,8 +61,7 @@ export function applyWriteTool(ctx: Context): void {
     },
     async execute(args, exec): Promise<{ content: ContentBlock[]; meta?: FsDiffMeta }> {
       const input = parseWriteArgs(args)
-      const cwd = sessionCwd(exec)
-      const target = await ctx.fs.resolve(input.filePath, cwd !== undefined ? { cwd } : undefined)
+      const target = await ctx.fs.resolve(input.filePath, sessionResolveOptions(exec))
       // Single-slot decision: the policy plugin produces createIfAbsent/
       // replaceIfVersion; the bare default is undefined (unconditional). No stat.
       const intent = await ctx.waterfall('fs/write-intent', target, exec, () => undefined)
diff --git a/packages/fs/tool-fs/tests/fs-tools.e2e.ts b/packages/fs/tool-fs/tests/fs-tools.e2e.ts
index a270e7c6d8..ca93bad63d 100644
--- a/packages/fs/tool-fs/tests/fs-tools.e2e.ts
+++ b/packages/fs/tool-fs/tests/fs-tools.e2e.ts
@@ -35,7 +35,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('fs tools with-key smoke', () =>
     ctx = await fsHarness(workdir, SYSTEM)
     // agentLoop.create prepares a session with no cwd, so the provider default
     // (config.cwd = workdir) is the workspace.
-    const agent = ctx.agentLoop.create(AgentId('fs-e2e'), { model: 'deepseek-v4-flash' })
+    const agent = ctx.agentLoop.create(AgentId('fs-e2e'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     agent.send([{ type: 'text', text:
       'Create a file named note.txt containing exactly the line: status: draft. '
@@ -68,7 +68,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('fs tools with-key smoke', () =>
         agentId: AgentId('fs-e2e-cwd'),
         sessionId: SessionId(`fs-e2e-cwd-${Date.now()}`),
         meta: { cwd: sessionDir },
-        agentOptions: { model: 'deepseek-v4-flash' },
+        agentOptions: { provider: 'deepseek', model: 'deepseek-v4-flash' },
       })
       handle.agent.send([{ type: 'text', text:
         'Use the write tool to create a file named where.txt containing exactly the line: here. Tell me when done.' }])
diff --git a/packages/fs/tool-fs/tests/harness.ts b/packages/fs/tool-fs/tests/harness.ts
index 0487962922..eff2588ec2 100644
--- a/packages/fs/tool-fs/tests/harness.ts
+++ b/packages/fs/tool-fs/tests/harness.ts
@@ -1,10 +1,7 @@
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { type Agent } from '@deepseek-ai/dsh-agent'
+import type { Agent } from '@deepseek-ai/dsh-agent'
 import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import LocalFileSystem from '@deepseek-ai/dsh-fs-local'
 import * as FsPolicy from '@deepseek-ai/dsh-fs-policy'
 import * as ToolFs from '@deepseek-ai/dsh-tool-fs'
@@ -17,13 +14,9 @@ import * as LlmDeepSeek from '@deepseek-ai/dsh-llm-deepseek'
  */
 export async function fsHarness(fsCwd: string, persona = ''): Promise<Context> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt, { persona })
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx, { systemPrompt: { persona } })
   await ctx.plugin(AgentLoop, { agents: [] })
-  await ctx.plugin(LlmDeepSeek, { models: ['deepseek-v4-flash'] })
+  await ctx.plugin(LlmDeepSeek)
   await ctx.plugin(LocalFileSystem, { cwd: fsCwd })
   await ctx.plugin(FsPolicy)
   await ctx.plugin(ToolFs)
diff --git a/packages/fs/tool-fs/tests/integration.spec.ts b/packages/fs/tool-fs/tests/integration.spec.ts
index 26a81343ff..6a9e6f3568 100644
--- a/packages/fs/tool-fs/tests/integration.spec.ts
+++ b/packages/fs/tool-fs/tests/integration.spec.ts
@@ -384,6 +384,33 @@ describe('signal, concurrency, and the fs/observed contract', () => {
     expect(onDisk === 'ONE value here' || onDisk === 'base TWO here').toBe(true)
   })
 
+  it('a stale observed version from an older read fails closed at edit CAS', async () => {
+    await writeFile(join(dir, 'a.txt'), 'older content\n')
+    const target = await ctx.fs.resolve('a.txt')
+    const firstInfo = await ctx.fs.stat(target)
+    if (!firstInfo) throw new Error('expected first stat')
+
+    expect((await callOwned('read', { file_path: 'a.txt' })).isError).toBe(false)
+
+    await writeFile(join(dir, 'a.txt'), 'newer current content\n')
+    const secondInfo = await ctx.fs.stat(target)
+    if (!secondInfo) throw new Error('expected second stat')
+    expect(secondInfo.version).not.toBe(firstInfo.version)
+    expect((await callOwned('read', { file_path: 'a.txt' })).isError).toBe(false)
+
+    // Reproduce an older concurrent read winning the observation race.
+    ctx.emit('fs/observed', target, firstInfo.version, { agent: { session } })
+
+    const edit = await callOwned('edit', {
+      file_path: 'a.txt',
+      old_string: 'newer',
+      new_string: 'edited',
+    })
+    expect(edit.isError).toBe(true)
+    expect(edit.error).toMatchObject({ code: 'FS_STALE_VERSION' })
+    expect(await readFile(join(dir, 'a.txt'), 'utf8')).toBe('newer current content\n')
+  })
+
   it('a throwing fs/observed listener surfaces as isError, but the mutation already hit disk', async () => {
     // fs/observed is a plain ctx.emit after the write succeeded; a throwing listener cannot
     // roll the write back — it only turns the tool result into isError.
diff --git a/packages/fs/tool-fs/tests/tools.spec.ts b/packages/fs/tool-fs/tests/tools.spec.ts
index e4367b8702..f08db928af 100644
--- a/packages/fs/tool-fs/tests/tools.spec.ts
+++ b/packages/fs/tool-fs/tests/tools.spec.ts
@@ -14,6 +14,7 @@ import type {
   FsEditOutcome,
   FsEditRequest,
   FsInfo,
+  FsPathInfo,
   FsTarget,
   FsWriteIntent,
   FsWriteOutcome,
@@ -44,6 +45,11 @@ class FakeFs extends FileSystem {
     if (content === undefined) return undefined
     return { version: FsVersion('v1'), type: 'file', size: content.length }
   }
+  override async lstat(path: string): Promise<FsPathInfo | undefined> {
+    const content = this.files.get(`key:${path}`)
+    if (content === undefined) return undefined
+    return { version: FsVersion('v1'), type: 'file', size: content.length }
+  }
   override async readText(target: FsTarget): Promise<string> {
     return this.files.get(target.targetKey) ?? ''
   }
@@ -102,6 +108,16 @@ describe('registration', () => {
     expect(ctx.tools.schemas().map(s => s.name).sort()).toEqual(['edit', 'read', 'write'])
   })
 
+  it('declares read parallel-safe while write/edit remain exclusive', async () => {
+    const { ctx } = await setup()
+    expect(ctx.tools.executionMode({ callId: CallId('read-safe'), name: 'read', arguments: { file_path: 'a.txt' } }))
+      .toEqual({ kind: 'parallel' })
+    expect(ctx.tools.executionMode({ callId: CallId('write-exclusive'), name: 'write', arguments: { file_path: 'a.txt', content: 'x' } }))
+      .toEqual({ kind: 'exclusive' })
+    expect(ctx.tools.executionMode({ callId: CallId('edit-exclusive'), name: 'edit', arguments: { file_path: 'a.txt', old_string: 'x', new_string: 'y' } }))
+      .toEqual({ kind: 'exclusive' })
+  })
+
   it('registers prompt sections for each tool', async () => {
     const { ctx } = await setup()
     const prompt = renderPrompt(await ctx.systemPrompt.assemble())
diff --git a/packages/guard/README.md b/packages/guard/README.md
index 9198698b20..05c9625cb0 100644
--- a/packages/guard/README.md
+++ b/packages/guard/README.md
@@ -6,4 +6,4 @@ Behavioral guard plugins that watch the agent loop for unproductive patterns and
 |---|---|---|
 | `repeat-tool-guard/` | Advisory reminders when an agent loops on identical tool calls | (listens on `ctx.tools`' waterfalls) |
 
-Reminders travel as `additionalContext` on the `tools/post-execute` decision; the agent loop appends them as logged `context/message` events after the step's tool results (see [the tools package](../core/tools)), so everything a guard says to the model is reconstructable from the session log.
+Reminders travel as `additionalContexts` on the `tools/post-execute` decision; the agent loop appends them as logged `context/message` events after the step's tool results (see [the tools package](../core/tools)), so everything a guard says to the model is reconstructable from the session log.
diff --git a/packages/guard/repeat-tool-guard/README.md b/packages/guard/repeat-tool-guard/README.md
index 843611c2a7..5a9fed247f 100644
--- a/packages/guard/repeat-tool-guard/README.md
+++ b/packages/guard/repeat-tool-guard/README.md
@@ -30,7 +30,7 @@ The chain key is `(tool name, canonical arguments)` — canonicalization is a de
 
 ## Reminder delivery
 
-Reminders use source-attributed `additionalContext`, preserving the tool's original result. The loop records them after the step's results as reconstructable `context/message` events. The guard always delegates and folds its reminder onto downstream context, including blocked calls.
+Reminders ride the post-execute decision's `additionalContexts` (source `{kind: 'plugin', plugin: 'repeat-tool-guard'}`), never a `content` replacement: the `tool/result` event stays the tool's own output for audit. The loop buffers the context and appends it as a `context/message` after the step's tool results, which the session renders as the tagged synthetic-user envelope — so the reminder is model-visible, source-attributed, and reconstructable from the session log with no new session event. The guard always delegates via `next()` and prepends its reminder to the downstream decision's context array (both variants — a blocked call still gets the nudge); every entry retains its own source, envelope, and metadata.
 
 ## Testing
 
diff --git a/packages/guard/repeat-tool-guard/package.json b/packages/guard/repeat-tool-guard/package.json
index 0cc99b6976..92d49c548f 100644
--- a/packages/guard/repeat-tool-guard/package.json
+++ b/packages/guard/repeat-tool-guard/package.json
@@ -32,9 +32,9 @@
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
-    "@deepseek-ai/dsh-system-prompt": "workspace:^",
     "@deepseek-ai/dsh-tools": "workspace:^",
     "cordis": "^4.0.0-rc.7"
   }
diff --git a/packages/guard/repeat-tool-guard/src/index.ts b/packages/guard/repeat-tool-guard/src/index.ts
index ca4e6b5c0e..9df8dd399f 100644
--- a/packages/guard/repeat-tool-guard/src/index.ts
+++ b/packages/guard/repeat-tool-guard/src/index.ts
@@ -140,16 +140,11 @@ function validateThresholds(values: number[]): number[] {
 }
 
 /**
- * Concatenate the guard's reminder context with a downstream listener's
- * optional one so folding drops neither. The merged block carries the guard's
- * `source` — a `HookContext` holds one `MessageSource` and the seam cannot
- * represent mixed provenance; the rendered `context/message` only
- * distinguishes by `source.kind`, so a downstream plugin's text is still
- * correctly framed as plugin context.
+ * Prepend the guard's reminder while preserving every downstream context's
+ * source, envelope, and metadata.
  */
-function concatContext(ours: HookContext, theirs: HookContext | undefined): HookContext {
-  if (!theirs) return ours
-  return { content: [...ours.content, ...theirs.content], source: ours.source }
+function prependContext(ours: HookContext, theirs: HookContext[] | undefined): HookContext[] {
+  return [ours, ...theirs ?? []]
 }
 
 /** One agent's consecutive-repeat chain: the last tracked call's identity key and its run length. */
@@ -211,19 +206,19 @@ export function apply(ctx: Context, config: Config): void {
 
   // Observe-and-enrich, never veto: count first (state advances regardless of
   // the downstream outcome), DELEGATE so a later listener can still block or
-  // replace, then fold the reminder onto whatever came back — additionalContext
+  // replace, then fold the reminder onto whatever came back — additionalContexts
   // rides both decision variants, so a blocked call still gets the nudge.
   ctx.on('tools/post-execute', async (exec, _result, next): Promise<PostToolDecision> => {
     const reminder = observe(exec)
     const downstream = await next()
     if (!reminder) return downstream
     if (downstream.kind === 'block') {
-      return { kind: 'block', feedback: downstream.feedback, additionalContext: concatContext(reminder, downstream.additionalContext) }
+      return { kind: 'block', feedback: downstream.feedback, additionalContexts: prependContext(reminder, downstream.additionalContexts) }
     }
     return {
       kind: 'accept',
       ...downstream.content !== undefined ? { content: downstream.content } : {},
-      additionalContext: concatContext(reminder, downstream.additionalContext),
+      additionalContexts: prependContext(reminder, downstream.additionalContexts),
     }
   })
 
diff --git a/packages/guard/repeat-tool-guard/tests/repeat-tool-guard.spec.ts b/packages/guard/repeat-tool-guard/tests/repeat-tool-guard.spec.ts
index 565f1076b5..3c5c9051bf 100644
--- a/packages/guard/repeat-tool-guard/tests/repeat-tool-guard.spec.ts
+++ b/packages/guard/repeat-tool-guard/tests/repeat-tool-guard.spec.ts
@@ -1,11 +1,11 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import LlmService, { CallId } from '@deepseek-ai/dsh-llm'
-import SessionStore, { type SessionEvent } from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry, { defineTool } from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import type { SessionEvent } from '@deepseek-ai/dsh-session'
+import { defineTool } from '@deepseek-ai/dsh-tools'
+import { AgentId } from '@deepseek-ai/dsh-agent'
 import AgentLoop, { type ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as RepeatToolGuard from '@deepseek-ai/dsh-repeat-tool-guard'
 import type { Config } from '@deepseek-ai/dsh-repeat-tool-guard'
 import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
@@ -21,11 +21,7 @@ import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent
 /** Boot the core spine + the guard; the caller registers adapters and extra listeners. */
 async function harness(config: Config = {}): Promise<Context> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(RepeatToolGuard, config)
   ctx.tools.register(defineTool({ name: 'probe', description: 'p', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
@@ -57,7 +53,7 @@ describe('threshold escalation', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -78,7 +74,7 @@ describe('threshold escalation', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -100,7 +96,7 @@ describe('chain semantics', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -124,7 +120,7 @@ describe('chain semantics', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -142,7 +138,7 @@ describe('chain semantics', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -163,7 +159,7 @@ describe('chain semantics', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -179,7 +175,7 @@ describe('chain semantics', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -195,7 +191,7 @@ describe('chain semantics', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -215,8 +211,8 @@ describe('chain semantics', () => {
       toolCallResponse('b3', 'probe', { q: 1 }),
       textResponse('done'),
     ]))
-    const agentA = ctx.agentLoop.create(AgentId('a'), { model: 'mock-a' })
-    const agentB = ctx.agentLoop.create(AgentId('b'), { model: 'mock-b' })
+    const agentA = ctx.agentLoop.create(AgentId('a'), { provider: 'mock-a', model: 'model-a' })
+    const agentB = ctx.agentLoop.create(AgentId('b'), { provider: 'mock-b', model: 'model-b' })
     agentA.send([{ type: 'text', text: 'go' }])
     agentB.send([{ type: 'text', text: 'go' }])
     await Promise.all([waitForIdle(ctx, agentA), waitForIdle(ctx, agentB)])
@@ -235,7 +231,7 @@ describe('chain semantics', () => {
       textResponse('turn two done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
     agent.send([{ type: 'text', text: 'again' }])
@@ -256,14 +252,14 @@ describe('chain semantics', () => {
     // (the loop.spec pattern): a child plugin fiber owns `first`.
     let first!: ReactLoopAgent
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
-      first = inner.agentLoop.create(AgentId('reused'), { model: 'mock' })
+      first = inner.agentLoop.create(AgentId('reused'), { provider: 'mock', model: 'mock' })
     }, { inject: ['agentLoop'] }))
     first.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, first)
     await fiber.dispose()
     await first.done
 
-    const second = ctx.agentLoop.create(AgentId('reused'), { model: 'mock' })
+    const second = ctx.agentLoop.create(AgentId('reused'), { provider: 'mock', model: 'mock' })
     second.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, second)
 
@@ -279,7 +275,7 @@ describe('chain semantics', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -295,7 +291,7 @@ describe('chain semantics', () => {
       toolCallResponse('c1', 'probe', { q: 1 }), // if the direct call had counted, this would be #2
       textResponse('done'),
     ]))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -309,7 +305,7 @@ describe('fold onto the downstream decision', () => {
     ctx.on('tools/post-execute', async () => ({
       kind: 'block' as const,
       feedback: [{ type: 'text' as const, text: 'nope' }],
-      additionalContext: { content: [{ type: 'text' as const, text: 'downstream-ctx' }], source: { kind: 'plugin' as const, plugin: 'test' } },
+      additionalContexts: [{ content: [{ type: 'text' as const, text: 'downstream-ctx' }], source: { kind: 'plugin' as const, plugin: 'test' } }],
     }))
     const adapter = new MockAdapter([
       toolCallResponse('c1', 'probe', { q: 1 }),
@@ -317,19 +313,19 @@ describe('fold onto the downstream decision', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
     const found = reminders(agent)
-    expect(found).toHaveLength(2)
+    expect(found).toHaveLength(3)
     // Call 1: below threshold — the downstream context passes through untouched.
     expect(found[0]!.text).toBe('downstream-ctx')
     expect(found[0]!.source).toEqual({ kind: 'plugin', plugin: 'test' })
-    // Call 2: reminder folded in front, single merged context, the guard's source.
+    // Call 2: reminder and downstream context retain separate provenance.
     expect(found[1]!.text).toContain('repeating the exact same tool call')
-    expect(found[1]!.text).toContain('|downstream-ctx')
     expect(found[1]!.source).toEqual(GUARD_SOURCE)
+    expect(found[2]).toEqual({ text: 'downstream-ctx', source: { kind: 'plugin', plugin: 'test' } })
     // The block's feedback reached the tool result unchanged.
     const results = [...agent.session.events].filter((e): e is SessionEvent<'tool/result'> => e.type === 'tool/result')
     expect(results.every(r => r.data.isError)).toBe(true)
@@ -348,7 +344,7 @@ describe('fold onto the downstream decision', () => {
       textResponse('done'),
     ])
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -363,11 +359,7 @@ describe('fold onto the downstream decision', () => {
 describe('config validation fails loud', () => {
   async function spine(): Promise<Context> {
     const ctx = new Context()
-    await ctx.plugin(LlmService)
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
+    await mountAgentLoopTestDependencies(ctx)
     await ctx.plugin(AgentLoop, { agents: [] })
     return ctx
   }
diff --git a/packages/hooks/hook-protocol/tests/runner.spec.ts b/packages/hooks/hook-protocol/tests/runner.spec.ts
index 6f52198a93..c2990e10be 100644
--- a/packages/hooks/hook-protocol/tests/runner.spec.ts
+++ b/packages/hooks/hook-protocol/tests/runner.spec.ts
@@ -22,10 +22,10 @@ function recordingBash(run: (spec: BashExecSpec) => Promise<BashRunResult>): {
         command: request.command,
         workdir: request.workdir ?? '/stub',
         timeoutMs: request.timeoutMs ?? 0,
+        stdoutMaxBytes: request.stdoutMaxBytes ?? 64_000,
         ...request.signal ? { signal: request.signal } : {},
         ...request.stdin !== undefined ? { stdin: request.stdin } : {},
         ...request.env !== undefined ? { env: request.env } : {},
-        owner: request.owner,
         sandboxMode: request.sandboxMode,
       }
     },
diff --git a/packages/hooks/hooks-claude/README.md b/packages/hooks/hooks-claude/README.md
index efd5df183c..4430f5511a 100644
--- a/packages/hooks/hooks-claude/README.md
+++ b/packages/hooks/hooks-claude/README.md
@@ -35,9 +35,9 @@ The hooks **themselves** run in the agent's session workspace: for the agent-sco
 | CC hook | Harness seam | Mapping |
 |---|---|---|
 | `SessionStart` | `agent/session-start` (emit) | additionalContext → `agent.inject()` into the new session (cannot block) |
-| `UserPromptSubmit` | `agent/prompt-submit` (waterfall) | `deny` → `PromptDecision.block`; additionalContext-only → delegate via `next()` then fold context onto the downstream decision (a later listener can still block/rewrite) |
+| `UserPromptSubmit` | `agent/prompt-submit` (waterfall) | `deny` → `PromptDecision.block`; additionalContext-only → delegate via `next()` then prepend a separately sourced context to downstream `additionalContexts` (a later listener can still block/rewrite) |
 | `PreToolUse` | `tools/pre-execute` (waterfall) | `deny` → `PreToolDecision.deny`; `ask` → `PreToolDecision.ask` |
-| `PostToolUse` | `tools/post-execute` (waterfall) | `deny` → `block` with feedback; additionalContext-only → delegate via `next()` then fold context onto the downstream decision (a Code Mode sub-call’s context is dropped by the run_code bridge — see [the pipeline doc](../../../docs/tool-execution-pipeline.md)) |
+| `PostToolUse` | `tools/post-execute` (waterfall) | `deny` → `block` with feedback; additionalContext-only → delegate via `next()` then prepend a separately sourced context to the downstream decision; Code Mode defers sub-call contexts until the outer `run_code` result |
 | `Stop` | `agent/turn-continuation` (waterfall) | a blocking Stop hook forces `continue`, feeding its reason as next-step steering |
 | `SubagentStart` | `subagent/start` (emit) | additionalContext → `agent.inject()` into a live in-process child; a remote child has no local injection target |
 | `SubagentStop` | `subagent/end` (emit) | observe-only |
@@ -46,6 +46,8 @@ The three emit points run detached — no seam awaits a `SessionStart`/`Subagent
 
 The matcher subject is the tool name (`PreToolUse`/`PostToolUse`), the session source (`SessionStart`), or a constant `agent_type` of `general-purpose` (`SubagentStart`/`SubagentStop` — the harness subagent seam carries no per-kind label, so the bridge reports Claude Code's own Task-tool default; a default/`*`/empty `agent_type` matcher fires, a specific-kind matcher does not); `UserPromptSubmit`/`Stop` ignore matchers. Multiple file-configured hooks on one point run **serially, in config order**, and fold most-restrictively (`deny > ask > allow`, see `dsh-hook-protocol`); serial keeps each hook's `hook/invoked`/`hook/result` pair adjacent in the log, and the fold is order-independent for the decision (see the RFC's "run serially, not concurrently" note).
 
+Every agent-scoped stdin payload carries `session_id` and string-shaped `transcript_path`. The bridge resolves the latter through `ctx.sessionPersistence.locate(session.header)` when available and otherwise sends `''`. Lookup does not create or flush the artifact, so a path can be absent before the first turn-end checkpoint or omit the current open turn.
+
 ## Context source
 
 Injected context carries an explicit `{ kind: 'plugin', plugin: 'hooks-claude' }` source. `agent.inject()` defaults a missing source to `{ kind: 'user' }`, which would mislabel plugin context as a user prompt — so the bridge always names itself.
diff --git a/packages/hooks/hooks-claude/package.json b/packages/hooks/hooks-claude/package.json
index 21f08965d8..c6870a6471 100644
--- a/packages/hooks/hooks-claude/package.json
+++ b/packages/hooks/hooks-claude/package.json
@@ -29,6 +29,7 @@
     "@deepseek-ai/dsh-hook-protocol": "^0.0.1",
     "@deepseek-ai/dsh-llm": "^0.0.1",
     "@deepseek-ai/dsh-session": "^0.0.1",
+    "@deepseek-ai/dsh-session-persistence": "^0.0.1",
     "@deepseek-ai/dsh-subagent": "^0.0.1",
     "@deepseek-ai/dsh-tools": "^0.0.1",
     "cordis": "^4.0.0-rc.7"
@@ -36,13 +37,15 @@
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-bash": "workspace:^",
     "@deepseek-ai/dsh-bash-local": "workspace:^",
     "@deepseek-ai/dsh-hook-protocol": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
+    "@deepseek-ai/dsh-session-persistence": "workspace:^",
+    "@deepseek-ai/dsh-session-persistence-jsonl": "workspace:^",
     "@deepseek-ai/dsh-subagent": "workspace:^",
-    "@deepseek-ai/dsh-system-prompt": "workspace:^",
     "@deepseek-ai/dsh-tools": "workspace:^",
     "cordis": "^4.0.0-rc.7"
   }
diff --git a/packages/hooks/hooks-claude/src/index.ts b/packages/hooks/hooks-claude/src/index.ts
index 08a2d26c9d..121e5cf75a 100644
--- a/packages/hooks/hooks-claude/src/index.ts
+++ b/packages/hooks/hooks-claude/src/index.ts
@@ -14,6 +14,7 @@ import type { Context } from 'cordis'
 import z from 'schemastery'
 import type { Agent, ContinuationDecision, HookContext, PromptDecision } from '@deepseek-ai/dsh-agent'
 import type { ContentBlock, MessageSource } from '@deepseek-ai/dsh-llm'
+import type {} from '@deepseek-ai/dsh-session-persistence'
 import type { PostToolDecision, PreToolDecision, ToolExecution, ToolExecutionResult } from '@deepseek-ai/dsh-tools'
 import {
   appendHookInvoked,
@@ -188,17 +189,16 @@ export function apply(ctx: Context, config: Config): void {
     return { content, source: PLUGIN_SOURCE }
   }
 
-  /** Merge hook context while retaining this bridge's plugin-level source. */
-  function concatContext(ours: HookContext, theirs: HookContext | undefined): HookContext {
-    if (!theirs) return ours
-    return { content: [...ours.content, ...theirs.content], source: ours.source }
+  /** Prepend one context without flattening downstream provenance or metadata. */
+  function prependContext(ours: HookContext, theirs: HookContext[] | undefined): HookContext[] {
+    return [ours, ...theirs ?? []]
   }
 
   // SessionStart injects context when its detached hook resolves; a slow hook
   // may miss the first request.
   // TODO(session-start-gating): add a startup gate before promising first-turn delivery.
   ctx.on('agent/session-start', (agent, source) => {
-    detached.track(runPoint('SessionStart', source, sessionStartPayload(agent, source), { agent, signal: detached.signal })
+    detached.track(runPoint('SessionStart', source, sessionStartPayload(ctx, agent, source), { agent, signal: detached.signal })
       .then((merged) => {
         const context = contextFrom(merged)
         if (context) agent.inject(context.content, { source: context.source })
@@ -212,7 +212,7 @@ export function apply(ctx: Context, config: Config): void {
   // matcher subject (CC ignores matchers for this event). ---
   ctx.on('agent/prompt-submit', async (agent, content, _source, next): Promise<PromptDecision> => {
     const turn = lastTurn(agent)
-    const merged = await runPoint('UserPromptSubmit', '', promptPayload(agent, content), { agent, turn })
+    const merged = await runPoint('UserPromptSubmit', '', promptPayload(ctx, agent, content), { agent, turn })
     if (merged.decision === 'deny') {
       return { kind: 'block', reason: merged.reason ?? 'blocked by UserPromptSubmit hook' }
     }
@@ -224,14 +224,14 @@ export function apply(ctx: Context, config: Config): void {
     return {
       kind: 'allow',
       ...downstream.content !== undefined ? { content: downstream.content } : {},
-      additionalContext: concatContext(ours, downstream.additionalContext),
+      additionalContexts: prependContext(ours, downstream.additionalContexts),
     }
   })
 
   // --- PreToolUse → PreToolDecision. Matcher subject is the tool name. ---
   ctx.on('tools/pre-execute', async (exec, next): Promise<PreToolDecision> => {
     const turn = lastTurn(exec.agent)
-    const merged = await runPoint('PreToolUse', exec.name, preToolPayload(exec), { ...exec.agent ? { agent: exec.agent } : {}, turn, ...exec.signal ? { signal: exec.signal } : {} })
+    const merged = await runPoint('PreToolUse', exec.name, preToolPayload(ctx, exec), { ...exec.agent ? { agent: exec.agent } : {}, turn, ...exec.signal ? { signal: exec.signal } : {} })
     if (merged.decision === 'deny') return { kind: 'deny', reason: merged.reason ?? 'blocked by PreToolUse hook' }
     if (merged.decision === 'ask') return { kind: 'ask', ...merged.reason !== undefined ? { reason: merged.reason } : {} }
     return next()
@@ -240,29 +240,29 @@ export function apply(ctx: Context, config: Config): void {
   // --- PostToolUse → PostToolDecision. Matcher subject is the tool name. ---
   ctx.on('tools/post-execute', async (exec, result, next): Promise<PostToolDecision> => {
     const turn = lastTurn(exec.agent)
-    const merged = await runPoint('PostToolUse', exec.name, postToolPayload(exec, result), { ...exec.agent ? { agent: exec.agent } : {}, turn, ...exec.signal ? { signal: exec.signal } : {} })
+    const merged = await runPoint('PostToolUse', exec.name, postToolPayload(ctx, exec, result), { ...exec.agent ? { agent: exec.agent } : {}, turn, ...exec.signal ? { signal: exec.signal } : {} })
     const context = contextFrom(merged)
     if (merged.decision === 'deny') {
-      return { kind: 'block', feedback: [{ type: 'text', text: merged.reason ?? 'blocked by PostToolUse hook' }], ...context ? { additionalContext: context } : {} }
+      return { kind: 'block', feedback: [{ type: 'text', text: merged.reason ?? 'blocked by PostToolUse hook' }], ...context ? { additionalContexts: [context] } : {} }
     }
     // Our hooks did not block. DELEGATE so a later listener can still block/replace,
     // then fold our context onto its decision (a downstream block carries it too).
     const downstream = await next()
     if (!context) return downstream
     if (downstream.kind === 'block') {
-      return { ...downstream, additionalContext: concatContext(context, downstream.additionalContext) }
+      return { ...downstream, additionalContexts: prependContext(context, downstream.additionalContexts) }
     }
     return {
       kind: 'accept',
       ...downstream.content !== undefined ? { content: downstream.content } : {},
-      additionalContext: concatContext(context, downstream.additionalContext),
+      additionalContexts: prependContext(context, downstream.additionalContexts),
     }
   })
 
   // A blocking Stop hook forces continuation with its reason.
   // TODO(stop-loop-guard): cap consecutive forced continuations; hooks must self-limit meanwhile.
   ctx.on('agent/turn-continuation', async (agent, turn, _default, next): Promise<ContinuationDecision> => {
-    const merged = await runPoint('Stop', '', stopPayload(agent), { agent, turn })
+    const merged = await runPoint('Stop', '', stopPayload(ctx, agent), { agent, turn })
     if (merged.decision === 'deny') {
       // A blocking Stop hook forces continuation.
       const text = merged.reason ?? 'continue: blocked by Stop hook'
@@ -275,7 +275,7 @@ export function apply(ctx: Context, config: Config): void {
   // use the live child's workspace and the generic agent-type matcher subject.
   ctx.on('subagent/start', (info) => {
     const child = ctx.get('agents')?.get(info.id)
-    detached.track(runPoint('SubagentStart', SUBAGENT_TYPE, subagentPayload('SubagentStart', info, child), { ...child ? { agent: child } : {}, signal: detached.signal })
+    detached.track(runPoint('SubagentStart', SUBAGENT_TYPE, subagentPayload(ctx, 'SubagentStart', info, child), { ...child ? { agent: child } : {}, signal: detached.signal })
       .then((merged) => {
         const context = contextFrom(merged)
         if (context && child) child.inject(context.content, { source: context.source })
@@ -287,7 +287,7 @@ export function apply(ctx: Context, config: Config): void {
     // `.then` before the tool caller's `await run.result` disposes it) so the hook runs in the
     // child's cwd, not the server default.
     const child = ctx.get('agents')?.get(info.id)
-    detached.track(runPoint('SubagentStop', SUBAGENT_TYPE, subagentPayload('SubagentStop', info, child), { ...child ? { agent: child } : {}, signal: detached.signal }))
+    detached.track(runPoint('SubagentStop', SUBAGENT_TYPE, subagentPayload(ctx, 'SubagentStop', info, child), { ...child ? { agent: child } : {}, signal: detached.signal }))
   })
 }
 
@@ -317,28 +317,31 @@ function blocksToText(content: ContentBlock[]): string {
   return content.filter((b): b is Extract<ContentBlock, { type: 'text' }> => b.type === 'text').map(b => b.text).join('')
 }
 
-function base(agent: Agent | undefined, event: string): Record<string, unknown> {
+function base(ctx: Context, agent: Agent | undefined, event: string): Record<string, unknown> {
   return {
     session_id: agent?.session.header.id ?? '',
+    transcript_path: agent === undefined
+      ? ''
+      : ctx.get('sessionPersistence')?.locate(agent.session.header)?.path ?? '',
     cwd: agent?.session.header.cwd ?? process.cwd(),
     hook_event_name: event,
   }
 }
 
-function sessionStartPayload(agent: Agent, source: string): Record<string, unknown> {
-  return { ...base(agent, 'SessionStart'), source }
+function sessionStartPayload(ctx: Context, agent: Agent, source: string): Record<string, unknown> {
+  return { ...base(ctx, agent, 'SessionStart'), source }
 }
-function promptPayload(agent: Agent, content: ContentBlock[]): Record<string, unknown> {
-  return { ...base(agent, 'UserPromptSubmit'), prompt: blocksToText(content) }
+function promptPayload(ctx: Context, agent: Agent, content: ContentBlock[]): Record<string, unknown> {
+  return { ...base(ctx, agent, 'UserPromptSubmit'), prompt: blocksToText(content) }
 }
-function preToolPayload(exec: ToolExecution): Record<string, unknown> {
-  return { ...base(exec.agent, 'PreToolUse'), tool_name: exec.name, tool_input: exec.arguments, tool_use_id: exec.callId }
+function preToolPayload(ctx: Context, exec: ToolExecution): Record<string, unknown> {
+  return { ...base(ctx, exec.agent, 'PreToolUse'), tool_name: exec.name, tool_input: exec.arguments, tool_use_id: exec.callId }
 }
-function postToolPayload(exec: ToolExecution, result: ToolExecutionResult): Record<string, unknown> {
-  return { ...base(exec.agent, 'PostToolUse'), tool_name: exec.name, tool_input: exec.arguments, tool_use_id: exec.callId, tool_response: blocksToText(result.content) }
+function postToolPayload(ctx: Context, exec: ToolExecution, result: ToolExecutionResult): Record<string, unknown> {
+  return { ...base(ctx, exec.agent, 'PostToolUse'), tool_name: exec.name, tool_input: exec.arguments, tool_use_id: exec.callId, tool_response: blocksToText(result.content) }
 }
-function stopPayload(agent: Agent): Record<string, unknown> {
-  return { ...base(agent, 'Stop'), stop_hook_active: false }
+function stopPayload(ctx: Context, agent: Agent): Record<string, unknown> {
+  return { ...base(ctx, agent, 'Stop'), stop_hook_active: false }
 }
 /**
  * Build a SubagentStart/SubagentStop payload from the CC base (the child's
@@ -346,9 +349,9 @@ function stopPayload(agent: Agent): Record<string, unknown> {
  * fields. `agent_type` is the CC-default {@link SUBAGENT_TYPE}; `stop_hook_active`
  * is present on SubagentStop only (the loop-guard flag, always false this cut).
  */
-function subagentPayload(event: 'SubagentStart' | 'SubagentStop', info: { id: string }, child: Agent | undefined): Record<string, unknown> {
+function subagentPayload(ctx: Context, event: 'SubagentStart' | 'SubagentStop', info: { id: string }, child: Agent | undefined): Record<string, unknown> {
   return {
-    ...base(child, event),
+    ...base(ctx, child, event),
     agent_id: info.id,
     agent_type: SUBAGENT_TYPE,
     ...event === 'SubagentStop' ? { stop_hook_active: false } : {},
diff --git a/packages/hooks/hooks-claude/tests/bridge.spec.ts b/packages/hooks/hooks-claude/tests/bridge.spec.ts
index 506d613c73..672a89188a 100644
--- a/packages/hooks/hooks-claude/tests/bridge.spec.ts
+++ b/packages/hooks/hooks-claude/tests/bridge.spec.ts
@@ -4,12 +4,11 @@ import { tmpdir } from 'node:os'
 import { join } from 'node:path'
 import { Context, type Fiber } from 'cordis'
 import Loader from '@cordisjs/plugin-loader'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore, { type SessionEvent } from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry, { defineTool } from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import type { SessionEvent } from '@deepseek-ai/dsh-session'
+import { defineTool } from '@deepseek-ai/dsh-tools'
+import { AgentId } from '@deepseek-ai/dsh-agent'
 import AgentLoop, { type ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
 import * as HooksClaude from '@deepseek-ai/dsh-hooks-claude'
 import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
@@ -42,11 +41,7 @@ async function harness(configDir: string, adapter: MockAdapter): Promise<Context
 /** {@link harness}, also exposing the bridge's fiber for tests that dispose it. */
 async function harnessWithFiber(configDir: string, adapter: MockAdapter): Promise<{ ctx: Context; hooks: Fiber }> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
   const hooks = await ctx.plugin(HooksClaude, { configPath: join(configDir, 'hooks.json') })
@@ -92,7 +87,7 @@ describe('hooks-claude bridge — UserPromptSubmit', () => {
 
     const adapter = new MockAdapter([textResponse('should not run')])
     const ctx = await harness(dir, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'do something' }])
     await waitForIdle(ctx, agent)
 
@@ -115,7 +110,7 @@ describe('hooks-claude bridge — UserPromptSubmit', () => {
 
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(dir, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -140,7 +135,7 @@ describe('hooks-claude bridge — PreToolUse', () => {
     const ctx = await harness(dir, adapter)
     let ran = false
     ctx.tools.register(defineTool({ name: 'danger', description: 'd', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'should not run' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'use danger' }])
     await waitForIdle(ctx, agent)
 
@@ -163,7 +158,7 @@ describe('hooks-claude bridge — PreToolUse', () => {
     const ctx = await harness(dir, adapter)
     let ran = false
     ctx.tools.register(defineTool({ name: 'safe', description: 's', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'ran ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'use safe' }])
     await waitForIdle(ctx, agent)
 
@@ -185,7 +180,7 @@ describe('hooks-claude bridge — PostToolUse', () => {
     const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
     const ctx = await harness(dir, adapter)
     ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'raw output' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -205,7 +200,7 @@ describe('hooks-claude bridge — PostToolUse', () => {
     const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
     const ctx = await harness(dir, adapter)
     ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -229,7 +224,7 @@ describe('hooks-claude bridge — PostToolUse', () => {
     const ctx = await harness(dir, adapter)
     let ran = false
     ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'x' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -253,7 +248,7 @@ describe('hooks-claude bridge — SessionStart', () => {
 
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(dir, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     // session-start fires async (detached .then → agent.inject); wait for the
     // injected context/message to actually land before sending, rather than a
     // fixed sleep that flakes under load.
@@ -337,16 +332,12 @@ describe('hooks-claude bridge — load resilience', () => {
   it('a missing config file registers no hooks and does not crash the loop', async () => {
     const adapter = new MockAdapter([textResponse('fine')])
     const ctx = new Context()
-    await ctx.plugin(LlmService)
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
+    await mountAgentLoopTestDependencies(ctx)
     await ctx.plugin(AgentLoop, { agents: [] })
     await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
     await ctx.plugin(HooksClaude, { configPath: '/nonexistent/hooks.json' })
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
     // The turn ran normally — no hooks, no crash.
@@ -359,17 +350,13 @@ describe('hooks-claude bridge — load resilience', () => {
     const dir = writeConfig({ UserPromptSubmit: [{ hooks: [{ type: 'command', command: 'exit 2' }] }] })
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = new Context()
-    await ctx.plugin(LlmService)
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
+    await mountAgentLoopTestDependencies(ctx)
     await ctx.plugin(AgentLoop, { agents: [] })
     await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
     const fiber = await ctx.plugin(HooksClaude, { configPath: join(dir, 'hooks.json') })
     await fiber.dispose()
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
     expect(adapter.requests).toHaveLength(1) // not blocked → the listener is gone
diff --git a/packages/hooks/hooks-claude/tests/coverage-cases.ts b/packages/hooks/hooks-claude/tests/coverage-cases.ts
new file mode 100644
index 0000000000..774e2a8451
--- /dev/null
+++ b/packages/hooks/hooks-claude/tests/coverage-cases.ts
@@ -0,0 +1,741 @@
+import { afterEach, describe, expect, it, vi } from 'vitest'
+import { mkdtempSync, rmSync, writeFileSync, chmodSync, existsSync, readFileSync } from 'node:fs'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { Context } from 'cordis'
+import type { SessionEvent } from '@deepseek-ai/dsh-session'
+import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
+import { defineTool } from '@deepseek-ai/dsh-tools'
+import { AgentId } from '@deepseek-ai/dsh-agent'
+import AgentLoop, { type ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
+import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
+import * as HooksClaude from '@deepseek-ai/dsh-hooks-claude'
+import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
+
+/** Targeted branch coverage for the CC bridge: option arms, warn paths, no-agent
+ * fallbacks, contextFrom-empty, and the detached-listener catch handlers. */
+
+const dirs: string[] = []
+afterEach(() => { for (const d of dirs.splice(0)) rmSync(d, { recursive: true, force: true }) })
+
+function dir(): string { const d = mkdtempSync(join(tmpdir(), 'dsh-hc-cov-')); dirs.push(d); return d }
+function sh(d: string, name: string, body: string): string {
+  const p = join(d, name); writeFileSync(p, body); chmodSync(p, 0o755); return p
+}
+function hooks(d: string, h: unknown): string {
+  writeFileSync(join(d, 'hooks.json'), JSON.stringify({ hooks: h })); return join(d, 'hooks.json')
+}
+
+type HarnessOpts = { pluginRoot?: string; projectDir?: string; stderrSummaryMaxChars?: number; sessionRoot?: string }
+async function harness(configPath: string, adapter: MockAdapter, opts: HarnessOpts = {}): Promise<Context> {
+  const ctx = new Context()
+  await mountAgentLoopTestDependencies(ctx)
+  if (opts.sessionRoot !== undefined) await ctx.plugin(SessionPersistenceJsonl, { root: opts.sessionRoot })
+  await ctx.plugin(AgentLoop, { agents: [] })
+  await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
+  await ctx.plugin(HooksClaude, { configPath, ...opts })
+  ctx.llm.registerAdapter(['mock'], adapter)
+  return ctx
+}
+function waitForIdle(ctx: Context, agent: ReactLoopAgent): Promise<void> {
+  return new Promise((resolve) => { const d = ctx.on('agent/status', (s, st) => { if (s === agent && st === 'idle') { d(); resolve() } }) })
+}
+function events(agent: ReactLoopAgent): SessionEvent[] { return [...agent.session.events] }
+/** Poll until `predicate` holds or the deadline passes — robust to detached
+ * emit-listener hooks firing on a `.then` (a fixed sleep flakes under load). */
+async function waitFor(predicate: () => boolean, timeout = 5000, interval = 10): Promise<void> {
+  const deadline = Date.now() + timeout
+  while (!predicate()) {
+    if (Date.now() > deadline) throw new Error('waitFor: condition not met before deadline')
+    await new Promise(r => setTimeout(r, interval))
+  }
+}
+
+export type CoverageGroup = 'config' | 'stop' | 'context' | 'edge-paths'
+
+/** Register independently schedulable slices of the hooks-claude coverage matrix. */
+export function defineCoverageCases(group: CoverageGroup): void {
+  if (group === 'config') describe('hooks-claude coverage — config option arms + substitution + skip warning', () => {
+    it('uses the persistence locator for transcript_path and an empty string without one', async () => {
+      async function capture(sessionRoot?: string): Promise<{ payload: { transcript_path: string }; expected: string | undefined }> {
+        const d = dir()
+        const cap = join(d, 'payload')
+        const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'capture.sh', `#!/usr/bin/env bash\ncat > "${cap}"\n`) }] }] })
+        const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+        const ctx = await harness(path, adapter, { ...sessionRoot !== undefined ? { sessionRoot } : {} })
+        ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+        const agent = ctx.agentLoop.create(AgentId('transcript'), { provider: 'mock', model: 'mock' })
+        agent.send([{ type: 'text', text: 'go' }])
+        await waitForIdle(ctx, agent)
+        return {
+          payload: JSON.parse(readFileSync(cap, 'utf8')) as { transcript_path: string },
+          expected: ctx.get('sessionPersistence')?.locate(agent.session.header)?.path,
+        }
+      }
+
+      const located = await capture(dir())
+      expect(located.payload.transcript_path).toBe(located.expected)
+      expect((await capture()).payload.transcript_path).toBe('')
+    }, 15_000) // Two real agent/hook subprocess loops need loaded pre-push runner headroom.
+
+    it('honors pluginRoot + projectDir substitution and warns on a skipped non-command hook', async () => {
+      const d = dir()
+      // ${CLAUDE_PLUGIN_ROOT} resolves to d; the script writes its own cwd-independent marker.
+      const marker = join(d, 'ran')
+      sh(d, 'h.sh', `#!/usr/bin/env bash\ntouch "${marker}"\n`)
+      const path = hooks(d, {
+        PreToolUse: [{ hooks: [
+          { type: 'prompt', prompt: 'skipme' }, // skipped → warn loop
+          { type: 'command', command: '${CLAUDE_PLUGIN_ROOT}/h.sh' }, // substituted
+        ] }],
+      })
+      const warn = vi.fn()
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter, { pluginRoot: d, projectDir: d })
+      ctx.logger.warn = warn as never
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      expect(existsSync(marker)).toBe(true) // substituted command ran
+    })
+
+    it('warns and honors updatedInput as a no-op (input rewrite deferred)', async () => {
+      const d = dir()
+      const s = sh(d, 'u.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PreToolUse","permissionDecision":"allow","updatedInput":{"command":"rewritten"}}}\'\n')
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const warn = vi.fn()
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', { command: 'original' }), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      ctx.logger.warn = warn as never
+      let sawArgs: unknown
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: { command: { type: 'string' } }, async execute(args) { sawArgs = args; return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      // updatedInput is NOT honored — the tool ran with the ORIGINAL args.
+      expect((sawArgs as { command?: string }).command).toBe('original')
+      expect(warn).toHaveBeenCalledWith(expect.stringContaining('updatedInput'))
+    })
+  })
+
+  if (group === 'config') describe('hooks-claude coverage — empty/no-op outcomes and no-agent paths', () => {
+    it('a clean exit-0 hook with no output is a no-op (contextFrom empty → next())', async () => {
+      const d = dir()
+      const s = sh(d, 'noop.sh', '#!/usr/bin/env bash\nexit 0\n')
+      const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('ran')])
+      const ctx = await harness(path, adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      // The prompt proceeded unchanged; no context/message injected.
+      expect(adapter.requests).toHaveLength(1)
+      expect(events(agent).some(e => e.type === 'context/message')).toBe(false)
+    })
+
+    it('a PreToolUse hook fires for a no-agent direct tool call (no session/turn to record into)', async () => {
+      const d = dir()
+      const s = sh(d, 'deny.sh', '#!/usr/bin/env bash\necho "no" >&2\nexit 2\n')
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const ctx = await harness(path, new MockAdapter([]))
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'x' }] } }))
+      // Call execute() directly with NO agent — the bridge's no-agent/no-turn path.
+      const { CallId } = await import('@deepseek-ai/dsh-llm')
+      const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'echo', arguments: {} })
+      expect(ran).toBe(false)
+      expect(result.isError).toBe(true)
+    })
+
+    it('a long stderr is truncated in the hook/result summary', async () => {
+      const d = dir()
+      // Emit >500 chars of stderr then exit 2.
+      const s = sh(d, 'long.sh', '#!/usr/bin/env bash\nprintf "x%.0s" {1..600} >&2\nexit 2\n')
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && res.data.stderrSummary?.endsWith('…')).toBe(true)
+      expect(res?.type === 'hook/result' && res.data.stderrSummary?.length).toBe(501) // default 500-char cap + ellipsis
+    })
+
+    it('rejects a non-positive or fractional stderrSummaryMaxChars at load', async () => {
+      const d = dir()
+      const path = hooks(d, {})
+      for (const bad of [0, -5, 1.5, Number.NaN]) {
+        const adapter = new MockAdapter([])
+        await expect(harness(path, adapter, { stderrSummaryMaxChars: bad }))
+          .rejects.toThrow(/hooks-claude: stderrSummaryMaxChars must be a positive integer/)
+      }
+    })
+
+    it('the stderr summary cap is plugin config (stderrSummaryMaxChars)', async () => {
+      const d = dir()
+      const s = sh(d, 'long.sh', '#!/usr/bin/env bash\nprintf "x%.0s" {1..600} >&2\nexit 2\n')
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter, { stderrSummaryMaxChars: 40 })
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && res.data.stderrSummary).toBe('x'.repeat(40) + '…')
+    })
+  })
+
+  if (group === 'stop') describe('hooks-claude coverage — Stop continuation + subagent inject/catch', () => {
+    it('a Stop hook that blocks (exit 2) forces the turn to continue (CC dialect)', async () => {
+      const d = dir()
+      const marker = join(d, 'fired')
+      const s = sh(d, 'stop.sh', `#!/usr/bin/env bash\nif [ -e "${marker}" ]; then exit 0; fi\ntouch "${marker}"\necho "continue please" >&2\nexit 2\n`)
+      const path = hooks(d, { Stop: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
+      const ctx = await harness(path, adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      expect(adapter.requests).toHaveLength(2)
+      expect(JSON.stringify(adapter.requests[1]!.messages)).toContain('continue please')
+    })
+
+    it('a Stop hook that blocks with EMPTY stderr still forces continuation (no reason required)', async () => {
+    // A blocking Stop hook with no stderr yields `deny` without a reason. The block still forces
+    // continuation; the script self-limits to one block to avoid a loop.
+      const d = dir()
+      const marker = join(d, 'fired')
+      const s = sh(d, 'stop.sh', `#!/usr/bin/env bash\nif [ -e "${marker}" ]; then exit 0; fi\ntouch "${marker}"\nexit 2\n`)
+      const path = hooks(d, { Stop: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
+      const ctx = await harness(path, adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      // A second model request ran → the empty-reason block forced continuation.
+      expect(adapter.requests).toHaveLength(2)
+      // The steering carried the fallback reason (no stderr to use).
+      expect(JSON.stringify(adapter.requests[1]!.messages)).toContain('blocked by Stop hook')
+    })
+
+    it('SubagentStart additionalContext is injected into a REGISTERED live child', async () => {
+      const d = dir()
+      const s = sh(d, 'sa.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SubagentStart","additionalContext":"child guidance"}}\'\n')
+      const path = hooks(d, { SubagentStart: [{ hooks: [{ type: 'command', command: s }] }] })
+      const ctx = await harness(path, new MockAdapter([]))
+      // Register a fake child agent under the id the event carries.
+      const injected: string[] = []
+      const child = { id: AgentId('child-x'), inject: (content: { type: string; text?: string }[]) => { injected.push(content.map(b => b.text ?? '').join('')) }, session: { header: { id: 'child-x' } } } as unknown as Parameters<typeof ctx.agents.register>[0]
+      ctx.agents.register(child)
+      ctx.emit('subagent/start', { provider: 'p', id: AgentId('child-x') })
+      await waitFor(() => injected.includes('child guidance'))
+      expect(injected).toContain('child guidance')
+    })
+
+    it('a throwing SubagentStart/SubagentStop hook run is contained (logged)', async () => {
+      const d = dir()
+      // A hook command that does not exist makes runHook resolve a non-blocking
+      // error (not a throw), so to hit the .catch we make the .then throw: register
+      // a child whose inject throws for SubagentStart.
+      const s = sh(d, 'sa.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SubagentStart","additionalContext":"x"}}\'\n')
+      const path = hooks(d, { SubagentStart: [{ hooks: [{ type: 'command', command: s }] }] })
+      const ctx = await harness(path, new MockAdapter([]))
+      const warn = vi.fn(); ctx.logger.warn = warn as never
+      const child = { id: AgentId('child-y'), inject: () => { throw new Error('inject boom') }, session: { header: { id: 'child-y' } } } as unknown as Parameters<typeof ctx.agents.register>[0]
+      ctx.agents.register(child)
+      ctx.emit('subagent/start', { provider: 'p', id: AgentId('child-y') })
+      await waitFor(() => warn.mock.calls.some(c => String(c[0]).includes('SubagentStart hook failed')))
+      expect(warn).toHaveBeenCalledWith(expect.stringContaining('SubagentStart hook failed'))
+    })
+  })
+
+  if (group === 'stop') describe('hooks-claude coverage — default reasons + sparse payloads', () => {
+    it('PreToolUse deny with EMPTY stderr uses the default reason', async () => {
+      const d = dir()
+      const s = sh(d, 'deny.sh', '#!/usr/bin/env bash\nexit 2\n') // exit 2, no stderr
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'x' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const result = events(agent).find(e => e.type === 'tool/result')
+      expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text.includes('blocked by PreToolUse hook'))).toBe(true)
+    })
+
+    it('PostToolUse deny with EMPTY stderr + no context uses the default feedback', async () => {
+      const d = dir()
+      const s = sh(d, 'block.sh', '#!/usr/bin/env bash\nexit 2\n')
+      const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const result = events(agent).find(e => e.type === 'tool/result')
+      expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text.includes('blocked by PostToolUse hook'))).toBe(true)
+    })
+
+    it('SubagentStop with no registered child runs the hook cleanly (fire-and-forget)', async () => {
+      const d = dir()
+      // The agents registry has no entry for the id, so the child lookup yields
+      // undefined and the payload falls back to base(undefined) — assert the
+      // observe-only SubagentStop run still executes the hook without crashing.
+      const marker = join(d, 'stopran')
+      const s = sh(d, 'stop.sh', `#!/usr/bin/env bash\ntouch "${marker}"\n`)
+      const path = hooks(d, { SubagentStop: [{ hooks: [{ type: 'command', command: s }] }] })
+      const ctx = await harness(path, new MockAdapter([]))
+      ctx.emit('subagent/end', { provider: 'p', id: AgentId('child-z'), stopReason: 'completed' })
+      await waitFor(() => existsSync(marker))
+      expect(existsSync(marker)).toBe(true)
+    })
+  })
+
+  if (group === 'edge-paths') describe('hooks-claude coverage — more default/sparse arms', () => {
+    it('UserPromptSubmit deny with EMPTY stderr uses the default block reason', async () => {
+      const d = dir()
+      const s = sh(d, 'block.sh', '#!/usr/bin/env bash\nexit 2\n')
+      const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('no')])
+      const ctx = await harness(path, adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const turnEnd = events(agent).findLast(e => e.type === 'turn/end')
+      expect(turnEnd?.type === 'turn/end' && turnEnd.data.reason.kind === 'rejected' && turnEnd.data.reason.reason).toContain('blocked by UserPromptSubmit hook')
+    })
+
+    it('a PreToolUse ask with NO reason omits the reason (false arm)', async () => {
+      const d = dir()
+      const s = sh(d, 'ask.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PreToolUse","permissionDecision":"ask"}}\'\n')
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'x' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      // ask (no reason) → degrades to deny with the registry's generic message.
+      expect(ran).toBe(false)
+      expect(events(agent).some(e => e.type === 'tool/result' && e.data.isError)).toBe(true)
+    })
+
+    it('a recorded clean exit-0 hook with no stderr omits exitCode-extra/stderrSummary fields', async () => {
+      const d = dir()
+      const s = sh(d, 'noop.sh', '#!/usr/bin/env bash\nexit 0\n')
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && res.data.exitCode).toBe(0)
+      expect(res?.type === 'hook/result' && 'stderrSummary' in res.data).toBe(false)
+    })
+  })
+
+  if (group === 'edge-paths') describe('hooks-claude coverage — schema-bypass apply + unspawnable hook', () => {
+    it('a direct apply() (schema bypass) with only configPath runs', async () => {
+      const d = dir()
+      const marker = join(d, 'ran')
+      const s = sh(d, 'h.sh', `#!/usr/bin/env bash\ntouch "${marker}"\n`)
+      hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = new Context()
+      await mountAgentLoopTestDependencies(ctx)
+      await ctx.plugin(AgentLoop, { agents: [] })
+      await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
+      // Direct apply with only configPath — bypasses schemastery's defaults, so
+      // the bridge must run on the raw minimal config (the per-hook timeout is
+      // the protocol lib's reference default, not a config knob).
+      HooksClaude.apply(ctx, { configPath: join(d, 'hooks.json') })
+      ctx.llm.registerAdapter(['mock'], adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      expect(existsSync(marker)).toBe(true)
+    })
+
+    it('a non-zero non-2 hook exit (e.g. a command-not-found 127) is a non-blocking error; the tool still runs', async () => {
+      const d = dir()
+      // `bash -c` of a missing program exits 127 — a non-blocking error (not 0, not
+      // 2 → no decision), so the tool proceeds; the hook/result records exit 127.
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: '/nonexistent/definitely/not/a/command' }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      expect(ran).toBe(true)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && res.data.exitCode).toBe(127)
+    })
+
+    it('a PostToolUse deny with empty stderr + no context uses the default feedback (no context arm)', async () => {
+      const d = dir()
+      const s = sh(d, 'block.sh', '#!/usr/bin/env bash\nexit 2\n')
+      const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const result = events(agent).find(e => e.type === 'tool/result')
+      expect(result?.type === 'tool/result' && result.data.isError).toBe(true)
+    })
+  })
+
+  if (group === 'context') describe('hooks-claude coverage — continue:false, context arm, no-cwd', () => {
+    it('a {"continue":false} hook is RECORDED as decision "stop" but does not halt the run (TODO(hook-continue-false))', async () => {
+    // The seams cannot yet honor `continue:false` as a hard halt. The log must still record the
+    // stop decision while execution and the turn continue normally.
+      const d = dir()
+      const s = sh(d, 'stop.sh', '#!/usr/bin/env bash\necho \'{"continue":false,"stopReason":"halt"}\'\n')
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && res.data.decision).toBe('stop') // recorded
+      expect(ran).toBe(true) // NOT honored: the tool still ran (halt is deferred)
+      const turnEnd = events(agent).findLast(e => e.type === 'turn/end')
+      expect(turnEnd?.type === 'turn/end' && turnEnd.data.reason.kind).toBe('completed') // ran to completion
+    })
+
+    it('a PostToolUse hook that BOTH blocks AND attaches additionalContext', async () => {
+      const d = dir()
+      const s = sh(d, 'b.sh', '#!/usr/bin/env bash\necho \'{"decision":"block","reason":"bad","hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"context too"}}\'\n')
+      const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const result = events(agent).find(e => e.type === 'tool/result')
+      expect(result?.type === 'tool/result' && result.data.isError).toBe(true)
+      expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text.includes('bad'))).toBe(true)
+      // additionalContext also injected (the block + context arm).
+      expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('context too')))).toBe(true)
+    })
+
+    it('a PreToolUse hook whose hookSpecificOutput names a DIFFERENT event does NOT deny the tool', async () => {
+    // The block's hookEventName (UserPromptSubmit) mismatches the firing event
+    // (PreToolUse), so its permissionDecision:"deny" is discarded — the tool runs.
+      const d = dir()
+      const s = sh(d, 'x.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","permissionDecision":"deny"}}\'\n')
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      expect(ran).toBe(true) // the mismatched deny was discarded → the tool ran
+    })
+
+    it('defaults CLAUDE_PROJECT_DIR to the session workspace when no projectDir is configured', async () => {
+    // The default ACP wiring sets no projectDir. A stock CC hook that references
+    // $CLAUDE_PROJECT_DIR (shell expansion) must still get the session workspace,
+    // not an empty string. The hook echoes the var as additionalContext.
+      const d = dir()
+      const workspace = dir()
+      const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\nprintf \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"dir=%s"}}\' "$CLAUDE_PROJECT_DIR"\n')
+      const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('ran')])
+      const ctx = await harness(path, adapter) // NB: no projectDir
+      // The factory create() path honors meta.cwd (the plain agentLoop.create() does not).
+      const { SessionId } = await import('@deepseek-ai/dsh-session')
+      const handle = await ctx.agents.create({ agentId: AgentId('a1'), sessionId: SessionId('s1'), meta: { cwd: workspace }, agentOptions: { provider: 'mock', model: 'mock' } })
+      handle.agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, handle.agent as ReactLoopAgent)
+      expect(events(handle.agent as ReactLoopAgent).some(e => e.type === 'context/message'
+      && e.data.content.some(b => b.type === 'text' && b.text.includes(`dir=${workspace}`)))).toBe(true)
+      await handle.dispose()
+    })
+
+    it('a context-only UserPromptSubmit hook DELEGATES so a later listener can still block', async () => {
+    // A context-only hook delegates with `next()` and folds its context, so a downstream policy
+    // listener can still veto the prompt.
+      const d = dir()
+      const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"bridge ctx"}}\'\n')
+      const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('should not run')])
+      const ctx = await harness(path, adapter)
+      // A later listener that blocks every prompt (registered AFTER the bridge).
+      const { AgentId: AId } = await import('@deepseek-ai/dsh-agent')
+      ctx.on('agent/prompt-submit', async () => ({ kind: 'block' as const, reason: 'policy veto' }))
+      const agent = ctx.agentLoop.create(AId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      // the downstream block won: the model was never called, no user/message was
+      // recorded, and the (sole, fully-blocked) prompt closed the turn `rejected`
+      expect(adapter.requests).toHaveLength(0)
+      expect(events(agent).some(e => e.type === 'user/message')).toBe(false)
+      const turnEnd = events(agent).findLast(e => e.type === 'turn/end')
+      expect(turnEnd?.type === 'turn/end' && turnEnd.data.reason).toMatchObject({ kind: 'rejected', reason: 'policy veto' })
+    })
+
+    it('preserves separate bridge and downstream prompt contexts with framing and metadata', async () => {
+    // Both the bridge hook and a later prompt-submit listener attach context; the
+    // request must see both as separately sourced durable events.
+      const d = dir()
+      const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"from-bridge"}}\'\n')
+      const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(path, adapter)
+      ctx.on('agent/prompt-submit', async () => ({
+        kind: 'allow' as const,
+        content: [{ type: 'text' as const, text: 'rewritten-prompt' }],
+        additionalContexts: [{
+          content: [{ type: 'text' as const, text: 'from-downstream' }],
+          source: { kind: 'plugin' as const, plugin: 'policy' },
+          envelope: 'raw' as const,
+          meta: { owner: 'policy' },
+        }],
+      }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const req = JSON.stringify(adapter.requests[0]!.messages)
+      expect(req).toContain('from-bridge')
+      expect(req).toContain('from-downstream')
+      expect(req).toContain('rewritten-prompt') // downstream content rewrite preserved
+      // the original prompt was replaced by the downstream rewrite
+      const userMsg = events(agent).find(e => e.type === 'user/message')
+      expect(userMsg?.type === 'user/message' && userMsg.data.content.some(b => b.type === 'text' && b.text === 'rewritten-prompt')).toBe(true)
+      const contexts = events(agent).filter(event => event.type === 'context/message')
+      expect(contexts.map(event => event.type === 'context/message' && event.data.source)).toEqual([
+        { kind: 'plugin', plugin: 'hooks-claude' },
+        { kind: 'plugin', plugin: 'policy' },
+      ])
+      expect(contexts[1]?.type === 'context/message' && contexts[1].data.envelope).toBe('raw')
+      expect(contexts[1]?.type === 'context/message' && contexts[1].data.meta).toEqual({ owner: 'policy' })
+    })
+
+    it('folds the bridge PostToolUse context onto a downstream ACCEPT that replaces content', async () => {
+    // The bridge hook adds context; a later post-execute listener accepts with a
+    // content rewrite. Both the rewrite and the bridge context survive.
+      const d = dir()
+      const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"bridge-note"}}\'\n')
+      const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      ctx.on('tools/post-execute', async () => ({ kind: 'accept' as const, content: [{ type: 'text' as const, text: 'rewritten-result' }] }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const result = events(agent).find(e => e.type === 'tool/result')
+      expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text === 'rewritten-result')).toBe(true)
+      expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('bridge-note')))).toBe(true)
+    })
+
+    it('keeps bridge and downstream PostToolUse contexts as separate sourced events', async () => {
+      const d = dir()
+      const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"bridge-note"}}\'\n')
+      const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      ctx.on('tools/post-execute', async () => ({
+        kind: 'accept' as const,
+        additionalContexts: [{
+          content: [{ type: 'text' as const, text: 'downstream-note' }],
+          source: { kind: 'plugin' as const, plugin: 'policy' },
+          envelope: 'raw' as const,
+          meta: { owner: 'policy' },
+        }],
+      }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+
+      const contexts = events(agent).filter(event => event.type === 'context/message')
+      expect(contexts.map(event => event.type === 'context/message' && event.data.source)).toEqual([
+        { kind: 'plugin', plugin: 'hooks-claude' },
+        { kind: 'plugin', plugin: 'policy' },
+      ])
+      expect(contexts[1]?.type === 'context/message' && contexts[1].data.envelope).toBe('raw')
+      expect(contexts[1]?.type === 'context/message' && contexts[1].data.meta).toEqual({ owner: 'policy' })
+    })
+
+    it('folds the bridge PostToolUse context onto a downstream listener BLOCK', async () => {
+    // The bridge hook only adds context; a later post-execute listener blocks the
+    // result. The block wins AND carries the bridge context (concatContext on the
+    // block arm).
+      const d = dir()
+      const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"bridge-note"}}\'\n')
+      const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      ctx.on('tools/post-execute', async () => ({ kind: 'block' as const, feedback: [{ type: 'text' as const, text: 'downstream-block' }] }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const result = events(agent).find(e => e.type === 'tool/result')
+      expect(result?.type === 'tool/result' && result.data.isError).toBe(true)
+      expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text.includes('downstream-block'))).toBe(true)
+      // the bridge's context still landed (folded onto the block)
+      expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('bridge-note')))).toBe(true)
+    })
+
+  })
+
+  if (group === 'edge-paths') describe('hooks-claude coverage — executor reject + no-open-turn', () => {
+    it('when the bash executor REJECTS a hook run, the hook/result omits exitCode (non-blocking)', async () => {
+      const d = dir()
+      const s = sh(d, 'h.sh', '#!/usr/bin/env bash\nexit 0\n')
+      const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(path, adapter)
+      // Force the executor to reject (an infrastructure fault) so runHook's catch
+      // yields a HookOutput with exitCode undefined → the `exitCode` spread false arm.
+      const bash = ctx.bash
+      bash.run = (() => Promise.reject(new Error('executor down')))
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && 'exitCode' in res.data).toBe(false)
+    })
+
+  })
+
+  if (group === 'edge-paths') describe('hooks-claude coverage — detached-listener catch handlers', () => {
+    it('a throwing SessionStart inject is contained (logged, agent still runs)', async () => {
+      const d = dir()
+      const s = sh(d, 'start.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SessionStart","additionalContext":"x"}}\'\n')
+      const path = hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(path, adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      // Make inject throw, forcing the SessionStart .catch path.
+      const original = agent.inject.bind(agent)
+      let threw = false
+      agent.inject = (() => { threw = true; throw new Error('inject boom') })
+      await waitFor(() => threw)
+      expect(threw).toBe(true)
+      agent.inject = original
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      expect(adapter.requests).toHaveLength(1) // loop survived the thrown inject
+    })
+  })
+
+  if (group === 'stop') describe('hooks-claude coverage — hook runs in the session cwd, not the server cwd', () => {
+    it('runs an agent-scoped hook in the session workspace even when the executor default differs', async () => {
+    // The server launch directory and session cwd deliberately differ. The marker proves the
+    // bridge passes `session/new.cwd` instead of falling back to the executor default.
+      const serverDir = dir()
+      const sessionDir = dir()
+      const marker = join(sessionDir, 'where')
+      // The hook is invoked with cwd = session dir, so a relative marker path lands there.
+      hooks(serverDir, { PreToolUse: [{ hooks: [{ type: 'command', command: 'pwd > where' }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = new Context()
+      await mountAgentLoopTestDependencies(ctx)
+      await ctx.plugin(AgentLoop, { agents: [] })
+      // Executor default cwd = serverDir (deliberately NOT the session cwd).
+      await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000, cwd: serverDir })
+      await ctx.plugin(HooksClaude, { configPath: join(serverDir, 'hooks.json') })
+      ctx.llm.registerAdapter(['mock'], adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+
+      const { SessionId } = await import('@deepseek-ai/dsh-session')
+      const handle = await ctx.agents.create({ agentId: AgentId('a1'), sessionId: SessionId('s1'), meta: { cwd: sessionDir }, agentOptions: { provider: 'mock', model: 'mock' } })
+      handle.agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, handle.agent as ReactLoopAgent)
+
+      expect(existsSync(marker)).toBe(true) // the marker landed in the SESSION dir
+      const { readFileSync } = await import('node:fs')
+      const where = readFileSync(marker, 'utf8').trim()
+      // `pwd` may resolve symlinks (/var → /private/var etc.), so compare basenames.
+      expect(where.endsWith(sessionDir.split('/').pop()!)).toBe(true)
+      await handle.dispose()
+    })
+
+    it('runs a SubagentStop hook in the CHILD session workspace, not the server cwd', async () => {
+    // `SubagentStop` recovers the child at `subagent/end`; a relative marker proves `runPoint`
+    // receives that agent and runs in the child's cwd rather than the executor default.
+      const serverDir = dir()
+      const childDir = dir()
+      const marker = join(childDir, 'stopwhere')
+      hooks(serverDir, { SubagentStop: [{ hooks: [{ type: 'command', command: 'pwd > stopwhere' }] }] })
+      const ctx = new Context()
+      await mountAgentLoopTestDependencies(ctx)
+      await ctx.plugin(AgentLoop, { agents: [] })
+      // Executor default cwd = serverDir (deliberately NOT the child session cwd).
+      await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000, cwd: serverDir })
+      await ctx.plugin(HooksClaude, { configPath: join(serverDir, 'hooks.json') })
+      ctx.llm.registerAdapter(['mock'], new MockAdapter([]))
+
+      // Register a live child on its own session cwd; emit subagent/end with its id.
+      const { SessionId } = await import('@deepseek-ai/dsh-session')
+      const childHandle = await ctx.agents.create({ agentId: AgentId('child-stop'), sessionId: SessionId('child-stop-session'), meta: { cwd: childDir }, agentOptions: { provider: 'mock', model: 'mock' } })
+      ctx.emit('subagent/end', { provider: 'inproc', id: childHandle.agent.id, stopReason: 'completed' })
+
+      await waitFor(() => existsSync(marker))
+      expect(existsSync(marker)).toBe(true) // the marker landed in the CHILD dir
+      const { readFileSync } = await import('node:fs')
+      const where = readFileSync(marker, 'utf8').trim()
+      // `pwd` may resolve symlinks (/var → /private/var etc.), so compare basenames.
+      expect(where.endsWith(childDir.split('/').pop()!)).toBe(true)
+      await childHandle.dispose()
+    })
+  })
+
+  if (group === 'config') describe('hooks-claude coverage — systemMessage is warned, not surfaced', () => {
+    it('a hook emitting a systemMessage is logged as not-yet-surfaced', async () => {
+      const d = dir()
+      const s = sh(d, 'sm.sh', '#!/usr/bin/env bash\necho \'{"systemMessage":"heads up"}\'\n')
+      const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(path, adapter)
+      const warn = vi.fn(); ctx.logger.warn = warn as never
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      expect(warn).toHaveBeenCalledWith(expect.stringContaining('systemMessage'))
+      // Not surfaced: the systemMessage text never reaches the model request.
+      expect(JSON.stringify(adapter.requests[0]!.messages)).not.toContain('heads up')
+    })
+  })
+
+  if (group === 'edge-paths') describe('hooks-claude coverage — SessionStart timing is best-effort (no-wait)', () => {
+    it('does NOT crash or block when the prompt is sent immediately (context is best-effort, may miss the first request)', async () => {
+    // Session-start injection is detached, so an immediate prompt need not observe it. Assert only
+    // the guaranteed behavior—no crash and a completed turn—without pre-waiting away the race.
+      const d = dir()
+      const s = sh(d, 'start.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SessionStart","additionalContext":"late ctx"}}\'\n')
+      const path = hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: s }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(path, adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      // Send immediately — do NOT wait for the session-start inject.
+      agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, agent)
+      expect(adapter.requests).toHaveLength(1) // the turn ran regardless of hook timing
+    })
+  })
+}
diff --git a/packages/hooks/hooks-claude/tests/coverage-config.spec.ts b/packages/hooks/hooks-claude/tests/coverage-config.spec.ts
new file mode 100644
index 0000000000..1afa18c4ff
--- /dev/null
+++ b/packages/hooks/hooks-claude/tests/coverage-config.spec.ts
@@ -0,0 +1,3 @@
+import { defineCoverageCases } from './coverage-cases.ts'
+
+defineCoverageCases('config')
diff --git a/packages/hooks/hooks-claude/tests/coverage-context.spec.ts b/packages/hooks/hooks-claude/tests/coverage-context.spec.ts
new file mode 100644
index 0000000000..e0f3fb0ef8
--- /dev/null
+++ b/packages/hooks/hooks-claude/tests/coverage-context.spec.ts
@@ -0,0 +1,3 @@
+import { defineCoverageCases } from './coverage-cases.ts'
+
+defineCoverageCases('context')
diff --git a/packages/hooks/hooks-claude/tests/coverage-edge-paths.spec.ts b/packages/hooks/hooks-claude/tests/coverage-edge-paths.spec.ts
new file mode 100644
index 0000000000..0bbcb53b03
--- /dev/null
+++ b/packages/hooks/hooks-claude/tests/coverage-edge-paths.spec.ts
@@ -0,0 +1,3 @@
+import { defineCoverageCases } from './coverage-cases.ts'
+
+defineCoverageCases('edge-paths')
diff --git a/packages/hooks/hooks-claude/tests/coverage-stop.spec.ts b/packages/hooks/hooks-claude/tests/coverage-stop.spec.ts
new file mode 100644
index 0000000000..651cb1f6f0
--- /dev/null
+++ b/packages/hooks/hooks-claude/tests/coverage-stop.spec.ts
@@ -0,0 +1,3 @@
+import { defineCoverageCases } from './coverage-cases.ts'
+
+defineCoverageCases('stop')
diff --git a/packages/hooks/hooks-claude/tests/coverage.spec.ts b/packages/hooks/hooks-claude/tests/coverage.spec.ts
deleted file mode 100644
index f376708688..0000000000
--- a/packages/hooks/hooks-claude/tests/coverage.spec.ts
+++ /dev/null
@@ -1,688 +0,0 @@
-import { afterEach, describe, expect, it, vi } from 'vitest'
-import { mkdtempSync, rmSync, writeFileSync, chmodSync, existsSync } from 'node:fs'
-import { tmpdir } from 'node:os'
-import { join } from 'node:path'
-import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore, { type SessionEvent } from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry, { defineTool } from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
-import AgentLoop, { type ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
-import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
-import * as HooksClaude from '@deepseek-ai/dsh-hooks-claude'
-import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
-
-/** Targeted branch coverage for the CC bridge: option arms, warn paths, no-agent
- * fallbacks, contextFrom-empty, and the detached-listener catch handlers. */
-
-const dirs: string[] = []
-afterEach(() => { for (const d of dirs.splice(0)) rmSync(d, { recursive: true, force: true }) })
-
-function dir(): string { const d = mkdtempSync(join(tmpdir(), 'dsh-hc-cov-')); dirs.push(d); return d }
-function sh(d: string, name: string, body: string): string {
-  const p = join(d, name); writeFileSync(p, body); chmodSync(p, 0o755); return p
-}
-function hooks(d: string, h: unknown): string {
-  writeFileSync(join(d, 'hooks.json'), JSON.stringify({ hooks: h })); return join(d, 'hooks.json')
-}
-
-type HarnessOpts = { pluginRoot?: string; projectDir?: string; stderrSummaryMaxChars?: number }
-async function harness(configPath: string, adapter: MockAdapter, opts: HarnessOpts = {}): Promise<Context> {
-  const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
-  await ctx.plugin(AgentLoop, { agents: [] })
-  await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
-  await ctx.plugin(HooksClaude, { configPath, ...opts })
-  ctx.llm.registerAdapter(['mock'], adapter)
-  return ctx
-}
-function waitForIdle(ctx: Context, agent: ReactLoopAgent): Promise<void> {
-  return new Promise((resolve) => { const d = ctx.on('agent/status', (s, st) => { if (s === agent && st === 'idle') { d(); resolve() } }) })
-}
-function events(agent: ReactLoopAgent): SessionEvent[] { return [...agent.session.events] }
-/** Poll until `predicate` holds or the deadline passes — robust to detached
- * emit-listener hooks firing on a `.then` (a fixed sleep flakes under load). */
-async function waitFor(predicate: () => boolean, timeout = 5000, interval = 10): Promise<void> {
-  const deadline = Date.now() + timeout
-  while (!predicate()) {
-    if (Date.now() > deadline) throw new Error('waitFor: condition not met before deadline')
-    await new Promise(r => setTimeout(r, interval))
-  }
-}
-
-describe('hooks-claude coverage — config option arms + substitution + skip warning', () => {
-  it('honors pluginRoot + projectDir substitution and warns on a skipped non-command hook', async () => {
-    const d = dir()
-    // ${CLAUDE_PLUGIN_ROOT} resolves to d; the script writes its own cwd-independent marker.
-    const marker = join(d, 'ran')
-    sh(d, 'h.sh', `#!/usr/bin/env bash\ntouch "${marker}"\n`)
-    const path = hooks(d, {
-      PreToolUse: [{ hooks: [
-        { type: 'prompt', prompt: 'skipme' }, // skipped → warn loop
-        { type: 'command', command: '${CLAUDE_PLUGIN_ROOT}/h.sh' }, // substituted
-      ] }],
-    })
-    const warn = vi.fn()
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter, { pluginRoot: d, projectDir: d })
-    ctx.logger.warn = warn as never
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    expect(existsSync(marker)).toBe(true) // substituted command ran
-  })
-
-  it('warns and honors updatedInput as a no-op (input rewrite deferred)', async () => {
-    const d = dir()
-    const s = sh(d, 'u.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PreToolUse","permissionDecision":"allow","updatedInput":{"command":"rewritten"}}}\'\n')
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const warn = vi.fn()
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', { command: 'original' }), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    ctx.logger.warn = warn as never
-    let sawArgs: unknown
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: { command: { type: 'string' } }, async execute(args) { sawArgs = args; return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    // updatedInput is NOT honored — the tool ran with the ORIGINAL args.
-    expect((sawArgs as { command?: string }).command).toBe('original')
-    expect(warn).toHaveBeenCalledWith(expect.stringContaining('updatedInput'))
-  })
-})
-
-describe('hooks-claude coverage — empty/no-op outcomes and no-agent paths', () => {
-  it('a clean exit-0 hook with no output is a no-op (contextFrom empty → next())', async () => {
-    const d = dir()
-    const s = sh(d, 'noop.sh', '#!/usr/bin/env bash\nexit 0\n')
-    const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('ran')])
-    const ctx = await harness(path, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    // The prompt proceeded unchanged; no context/message injected.
-    expect(adapter.requests).toHaveLength(1)
-    expect(events(agent).some(e => e.type === 'context/message')).toBe(false)
-  })
-
-  it('a PreToolUse hook fires for a no-agent direct tool call (no session/turn to record into)', async () => {
-    const d = dir()
-    const s = sh(d, 'deny.sh', '#!/usr/bin/env bash\necho "no" >&2\nexit 2\n')
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const ctx = await harness(path, new MockAdapter([]))
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'x' }] } }))
-    // Call execute() directly with NO agent — the bridge's no-agent/no-turn path.
-    const { CallId } = await import('@deepseek-ai/dsh-llm')
-    const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'echo', arguments: {} })
-    expect(ran).toBe(false)
-    expect(result.isError).toBe(true)
-  })
-
-  it('a long stderr is truncated in the hook/result summary', async () => {
-    const d = dir()
-    // Emit >500 chars of stderr then exit 2.
-    const s = sh(d, 'long.sh', '#!/usr/bin/env bash\nprintf "x%.0s" {1..600} >&2\nexit 2\n')
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && res.data.stderrSummary?.endsWith('…')).toBe(true)
-    expect(res?.type === 'hook/result' && res.data.stderrSummary?.length).toBe(501) // default 500-char cap + ellipsis
-  })
-
-  it('rejects a non-positive or fractional stderrSummaryMaxChars at load', async () => {
-    const d = dir()
-    const path = hooks(d, {})
-    for (const bad of [0, -5, 1.5, Number.NaN]) {
-      const adapter = new MockAdapter([])
-      await expect(harness(path, adapter, { stderrSummaryMaxChars: bad }))
-        .rejects.toThrow(/hooks-claude: stderrSummaryMaxChars must be a positive integer/)
-    }
-  })
-
-  it('the stderr summary cap is plugin config (stderrSummaryMaxChars)', async () => {
-    const d = dir()
-    const s = sh(d, 'long.sh', '#!/usr/bin/env bash\nprintf "x%.0s" {1..600} >&2\nexit 2\n')
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter, { stderrSummaryMaxChars: 40 })
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && res.data.stderrSummary).toBe('x'.repeat(40) + '…')
-  })
-})
-
-describe('hooks-claude coverage — Stop continuation + subagent inject/catch', () => {
-  it('a Stop hook that blocks (exit 2) forces the turn to continue (CC dialect)', async () => {
-    const d = dir()
-    const marker = join(d, 'fired')
-    const s = sh(d, 'stop.sh', `#!/usr/bin/env bash\nif [ -e "${marker}" ]; then exit 0; fi\ntouch "${marker}"\necho "continue please" >&2\nexit 2\n`)
-    const path = hooks(d, { Stop: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
-    const ctx = await harness(path, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    expect(adapter.requests).toHaveLength(2)
-    expect(JSON.stringify(adapter.requests[1]!.messages)).toContain('continue please')
-  })
-
-  it('a Stop hook that blocks with EMPTY stderr still forces continuation (no reason required)', async () => {
-    // A blocking Stop hook with no stderr yields `deny` without a reason. The block still forces
-    // continuation; the script self-limits to one block to avoid a loop.
-    const d = dir()
-    const marker = join(d, 'fired')
-    const s = sh(d, 'stop.sh', `#!/usr/bin/env bash\nif [ -e "${marker}" ]; then exit 0; fi\ntouch "${marker}"\nexit 2\n`)
-    const path = hooks(d, { Stop: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
-    const ctx = await harness(path, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    // A second model request ran → the empty-reason block forced continuation.
-    expect(adapter.requests).toHaveLength(2)
-    // The steering carried the fallback reason (no stderr to use).
-    expect(JSON.stringify(adapter.requests[1]!.messages)).toContain('blocked by Stop hook')
-  })
-
-  it('SubagentStart additionalContext is injected into a REGISTERED live child', async () => {
-    const d = dir()
-    const s = sh(d, 'sa.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SubagentStart","additionalContext":"child guidance"}}\'\n')
-    const path = hooks(d, { SubagentStart: [{ hooks: [{ type: 'command', command: s }] }] })
-    const ctx = await harness(path, new MockAdapter([]))
-    // Register a fake child agent under the id the event carries.
-    const injected: string[] = []
-    const child = { id: AgentId('child-x'), inject: (content: { type: string; text?: string }[]) => { injected.push(content.map(b => b.text ?? '').join('')) }, session: { header: { id: 'child-x' } } } as unknown as Parameters<typeof ctx.agents.register>[0]
-    ctx.agents.register(child)
-    ctx.emit('subagent/start', { provider: 'p', id: AgentId('child-x') })
-    await waitFor(() => injected.includes('child guidance'))
-    expect(injected).toContain('child guidance')
-  })
-
-  it('a throwing SubagentStart/SubagentStop hook run is contained (logged)', async () => {
-    const d = dir()
-    // A hook command that does not exist makes runHook resolve a non-blocking
-    // error (not a throw), so to hit the .catch we make the .then throw: register
-    // a child whose inject throws for SubagentStart.
-    const s = sh(d, 'sa.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SubagentStart","additionalContext":"x"}}\'\n')
-    const path = hooks(d, { SubagentStart: [{ hooks: [{ type: 'command', command: s }] }] })
-    const ctx = await harness(path, new MockAdapter([]))
-    const warn = vi.fn(); ctx.logger.warn = warn as never
-    const child = { id: AgentId('child-y'), inject: () => { throw new Error('inject boom') }, session: { header: { id: 'child-y' } } } as unknown as Parameters<typeof ctx.agents.register>[0]
-    ctx.agents.register(child)
-    ctx.emit('subagent/start', { provider: 'p', id: AgentId('child-y') })
-    await waitFor(() => warn.mock.calls.some(c => String(c[0]).includes('SubagentStart hook failed')))
-    expect(warn).toHaveBeenCalledWith(expect.stringContaining('SubagentStart hook failed'))
-  })
-})
-
-describe('hooks-claude coverage — default reasons + sparse payloads', () => {
-  it('PreToolUse deny with EMPTY stderr uses the default reason', async () => {
-    const d = dir()
-    const s = sh(d, 'deny.sh', '#!/usr/bin/env bash\nexit 2\n') // exit 2, no stderr
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'x' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const result = events(agent).find(e => e.type === 'tool/result')
-    expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text.includes('blocked by PreToolUse hook'))).toBe(true)
-  })
-
-  it('PostToolUse deny with EMPTY stderr + no context uses the default feedback', async () => {
-    const d = dir()
-    const s = sh(d, 'block.sh', '#!/usr/bin/env bash\nexit 2\n')
-    const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const result = events(agent).find(e => e.type === 'tool/result')
-    expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text.includes('blocked by PostToolUse hook'))).toBe(true)
-  })
-
-  it('SubagentStop with no registered child runs the hook cleanly (fire-and-forget)', async () => {
-    const d = dir()
-    // The agents registry has no entry for the id, so the child lookup yields
-    // undefined and the payload falls back to base(undefined) — assert the
-    // observe-only SubagentStop run still executes the hook without crashing.
-    const marker = join(d, 'stopran')
-    const s = sh(d, 'stop.sh', `#!/usr/bin/env bash\ntouch "${marker}"\n`)
-    const path = hooks(d, { SubagentStop: [{ hooks: [{ type: 'command', command: s }] }] })
-    const ctx = await harness(path, new MockAdapter([]))
-    ctx.emit('subagent/end', { provider: 'p', id: AgentId('child-z'), stopReason: 'completed' })
-    await waitFor(() => existsSync(marker))
-    expect(existsSync(marker)).toBe(true)
-  })
-})
-
-describe('hooks-claude coverage — more default/sparse arms', () => {
-  it('UserPromptSubmit deny with EMPTY stderr uses the default block reason', async () => {
-    const d = dir()
-    const s = sh(d, 'block.sh', '#!/usr/bin/env bash\nexit 2\n')
-    const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('no')])
-    const ctx = await harness(path, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const turnEnd = events(agent).findLast(e => e.type === 'turn/end')
-    expect(turnEnd?.type === 'turn/end' && turnEnd.data.reason.kind === 'rejected' && turnEnd.data.reason.reason).toContain('blocked by UserPromptSubmit hook')
-  })
-
-  it('a PreToolUse ask with NO reason omits the reason (false arm)', async () => {
-    const d = dir()
-    const s = sh(d, 'ask.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PreToolUse","permissionDecision":"ask"}}\'\n')
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'x' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    // ask (no reason) → degrades to deny with the registry's generic message.
-    expect(ran).toBe(false)
-    expect(events(agent).some(e => e.type === 'tool/result' && e.data.isError)).toBe(true)
-  })
-
-  it('a recorded clean exit-0 hook with no stderr omits exitCode-extra/stderrSummary fields', async () => {
-    const d = dir()
-    const s = sh(d, 'noop.sh', '#!/usr/bin/env bash\nexit 0\n')
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && res.data.exitCode).toBe(0)
-    expect(res?.type === 'hook/result' && 'stderrSummary' in res.data).toBe(false)
-  })
-})
-
-describe('hooks-claude coverage — schema-bypass apply + unspawnable hook', () => {
-  it('a direct apply() (schema bypass) with only configPath runs', async () => {
-    const d = dir()
-    const marker = join(d, 'ran')
-    const s = sh(d, 'h.sh', `#!/usr/bin/env bash\ntouch "${marker}"\n`)
-    hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = new Context()
-    await ctx.plugin(LlmService)
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
-    await ctx.plugin(AgentLoop, { agents: [] })
-    await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
-    // Direct apply with only configPath — bypasses schemastery's defaults, so
-    // the bridge must run on the raw minimal config (the per-hook timeout is
-    // the protocol lib's reference default, not a config knob).
-    HooksClaude.apply(ctx, { configPath: join(d, 'hooks.json') })
-    ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    expect(existsSync(marker)).toBe(true)
-  })
-
-  it('a non-zero non-2 hook exit (e.g. a command-not-found 127) is a non-blocking error; the tool still runs', async () => {
-    const d = dir()
-    // `bash -c` of a missing program exits 127 — a non-blocking error (not 0, not
-    // 2 → no decision), so the tool proceeds; the hook/result records exit 127.
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: '/nonexistent/definitely/not/a/command' }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    expect(ran).toBe(true)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && res.data.exitCode).toBe(127)
-  })
-
-  it('a PostToolUse deny with empty stderr + no context uses the default feedback (no context arm)', async () => {
-    const d = dir()
-    const s = sh(d, 'block.sh', '#!/usr/bin/env bash\nexit 2\n')
-    const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const result = events(agent).find(e => e.type === 'tool/result')
-    expect(result?.type === 'tool/result' && result.data.isError).toBe(true)
-  })
-})
-
-describe('hooks-claude coverage — continue:false, context arm, no-cwd', () => {
-  it('a {"continue":false} hook is RECORDED as decision "stop" but does not halt the run (TODO(hook-continue-false))', async () => {
-    // The seams cannot yet honor `continue:false` as a hard halt. The log must still record the
-    // stop decision while execution and the turn continue normally.
-    const d = dir()
-    const s = sh(d, 'stop.sh', '#!/usr/bin/env bash\necho \'{"continue":false,"stopReason":"halt"}\'\n')
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && res.data.decision).toBe('stop') // recorded
-    expect(ran).toBe(true) // NOT honored: the tool still ran (halt is deferred)
-    const turnEnd = events(agent).findLast(e => e.type === 'turn/end')
-    expect(turnEnd?.type === 'turn/end' && turnEnd.data.reason.kind).toBe('completed') // ran to completion
-  })
-
-  it('a PostToolUse hook that BOTH blocks AND attaches additionalContext', async () => {
-    const d = dir()
-    const s = sh(d, 'b.sh', '#!/usr/bin/env bash\necho \'{"decision":"block","reason":"bad","hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"context too"}}\'\n')
-    const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const result = events(agent).find(e => e.type === 'tool/result')
-    expect(result?.type === 'tool/result' && result.data.isError).toBe(true)
-    expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text.includes('bad'))).toBe(true)
-    // additionalContext also injected (the block + context arm).
-    expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('context too')))).toBe(true)
-  })
-
-  it('a PreToolUse hook whose hookSpecificOutput names a DIFFERENT event does NOT deny the tool', async () => {
-    // The block's hookEventName (UserPromptSubmit) mismatches the firing event
-    // (PreToolUse), so its permissionDecision:"deny" is discarded — the tool runs.
-    const d = dir()
-    const s = sh(d, 'x.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","permissionDecision":"deny"}}\'\n')
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    expect(ran).toBe(true) // the mismatched deny was discarded → the tool ran
-  })
-
-  it('defaults CLAUDE_PROJECT_DIR to the session workspace when no projectDir is configured', async () => {
-    // The default ACP wiring sets no projectDir. A stock CC hook that references
-    // $CLAUDE_PROJECT_DIR (shell expansion) must still get the session workspace,
-    // not an empty string. The hook echoes the var as additionalContext.
-    const d = dir()
-    const workspace = dir()
-    const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\nprintf \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"dir=%s"}}\' "$CLAUDE_PROJECT_DIR"\n')
-    const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('ran')])
-    const ctx = await harness(path, adapter) // NB: no projectDir
-    // The factory create() path honors meta.cwd (the plain agentLoop.create() does not).
-    const { SessionId } = await import('@deepseek-ai/dsh-session')
-    const handle = await ctx.agents.create({ agentId: AgentId('a1'), sessionId: SessionId('s1'), meta: { cwd: workspace }, agentOptions: { model: 'mock' } })
-    handle.agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, handle.agent as ReactLoopAgent)
-    expect(events(handle.agent as ReactLoopAgent).some(e => e.type === 'context/message'
-      && e.data.content.some(b => b.type === 'text' && b.text.includes(`dir=${workspace}`)))).toBe(true)
-    await handle.dispose()
-  })
-
-  it('a context-only UserPromptSubmit hook DELEGATES so a later listener can still block', async () => {
-    // A context-only hook delegates with `next()` and folds its context, so a downstream policy
-    // listener can still veto the prompt.
-    const d = dir()
-    const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"bridge ctx"}}\'\n')
-    const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('should not run')])
-    const ctx = await harness(path, adapter)
-    // A later listener that blocks every prompt (registered AFTER the bridge).
-    const { AgentId: AId } = await import('@deepseek-ai/dsh-agent')
-    ctx.on('agent/prompt-submit', async () => ({ kind: 'block' as const, reason: 'policy veto' }))
-    const agent = ctx.agentLoop.create(AId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    // the downstream block won: the model was never called, no user/message was
-    // recorded, and the (sole, fully-blocked) prompt closed the turn `rejected`
-    expect(adapter.requests).toHaveLength(0)
-    expect(events(agent).some(e => e.type === 'user/message')).toBe(false)
-    const turnEnd = events(agent).findLast(e => e.type === 'turn/end')
-    expect(turnEnd?.type === 'turn/end' && turnEnd.data.reason).toMatchObject({ kind: 'rejected', reason: 'policy veto' })
-  })
-
-  it('folds the bridge additionalContext WITH a downstream listener that also adds context', async () => {
-    // Both the bridge hook and a later prompt-submit listener attach context; the
-    // request must see BOTH (concatContext keeps the downstream one too).
-    const d = dir()
-    const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"from-bridge"}}\'\n')
-    const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(path, adapter)
-    ctx.on('agent/prompt-submit', async () => ({
-      kind: 'allow' as const,
-      content: [{ type: 'text' as const, text: 'rewritten-prompt' }],
-      additionalContext: { content: [{ type: 'text' as const, text: 'from-downstream' }], source: { kind: 'plugin' as const, plugin: 'policy' } },
-    }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const req = JSON.stringify(adapter.requests[0]!.messages)
-    expect(req).toContain('from-bridge')
-    expect(req).toContain('from-downstream')
-    expect(req).toContain('rewritten-prompt') // downstream content rewrite preserved
-    // the original prompt was replaced by the downstream rewrite
-    const userMsg = events(agent).find(e => e.type === 'user/message')
-    expect(userMsg?.type === 'user/message' && userMsg.data.content.some(b => b.type === 'text' && b.text === 'rewritten-prompt')).toBe(true)
-  })
-
-  it('folds the bridge PostToolUse context onto a downstream ACCEPT that replaces content', async () => {
-    // The bridge hook adds context; a later post-execute listener accepts with a
-    // content rewrite. Both the rewrite and the bridge context survive.
-    const d = dir()
-    const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"bridge-note"}}\'\n')
-    const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    ctx.on('tools/post-execute', async () => ({ kind: 'accept' as const, content: [{ type: 'text' as const, text: 'rewritten-result' }] }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const result = events(agent).find(e => e.type === 'tool/result')
-    expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text === 'rewritten-result')).toBe(true)
-    expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('bridge-note')))).toBe(true)
-  })
-
-  it('folds the bridge PostToolUse context onto a downstream listener BLOCK', async () => {
-    // The bridge hook only adds context; a later post-execute listener blocks the
-    // result. The block wins AND carries the bridge context (concatContext on the
-    // block arm).
-    const d = dir()
-    const s = sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"bridge-note"}}\'\n')
-    const path = hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    ctx.on('tools/post-execute', async () => ({ kind: 'block' as const, feedback: [{ type: 'text' as const, text: 'downstream-block' }] }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const result = events(agent).find(e => e.type === 'tool/result')
-    expect(result?.type === 'tool/result' && result.data.isError).toBe(true)
-    expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text.includes('downstream-block'))).toBe(true)
-    // the bridge's context still landed (folded onto the block)
-    expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('bridge-note')))).toBe(true)
-  })
-
-})
-
-describe('hooks-claude coverage — executor reject + no-open-turn', () => {
-  it('when the bash executor REJECTS a hook run, the hook/result omits exitCode (non-blocking)', async () => {
-    const d = dir()
-    const s = sh(d, 'h.sh', '#!/usr/bin/env bash\nexit 0\n')
-    const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(path, adapter)
-    // Force the executor to reject (an infrastructure fault) so runHook's catch
-    // yields a HookOutput with exitCode undefined → the `exitCode` spread false arm.
-    const bash = ctx.bash
-    bash.run = (() => Promise.reject(new Error('executor down')))
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && 'exitCode' in res.data).toBe(false)
-  })
-
-})
-
-describe('hooks-claude coverage — detached-listener catch handlers', () => {
-  it('a throwing SessionStart inject is contained (logged, agent still runs)', async () => {
-    const d = dir()
-    const s = sh(d, 'start.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SessionStart","additionalContext":"x"}}\'\n')
-    const path = hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(path, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    // Make inject throw, forcing the SessionStart .catch path.
-    const original = agent.inject.bind(agent)
-    let threw = false
-    agent.inject = (() => { threw = true; throw new Error('inject boom') })
-    await waitFor(() => threw)
-    expect(threw).toBe(true)
-    agent.inject = original
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    expect(adapter.requests).toHaveLength(1) // loop survived the thrown inject
-  })
-})
-
-describe('hooks-claude coverage — hook runs in the session cwd, not the server cwd', () => {
-  it('runs an agent-scoped hook in the session workspace even when the executor default differs', async () => {
-    // The server launch directory and session cwd deliberately differ. The marker proves the
-    // bridge passes `session/new.cwd` instead of falling back to the executor default.
-    const serverDir = dir()
-    const sessionDir = dir()
-    const marker = join(sessionDir, 'where')
-    // The hook is invoked with cwd = session dir, so a relative marker path lands there.
-    hooks(serverDir, { PreToolUse: [{ hooks: [{ type: 'command', command: 'pwd > where' }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = new Context()
-    await ctx.plugin(LlmService)
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
-    await ctx.plugin(AgentLoop, { agents: [] })
-    // Executor default cwd = serverDir (deliberately NOT the session cwd).
-    await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000, cwd: serverDir })
-    await ctx.plugin(HooksClaude, { configPath: join(serverDir, 'hooks.json') })
-    ctx.llm.registerAdapter(['mock'], adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-
-    const { SessionId } = await import('@deepseek-ai/dsh-session')
-    const handle = await ctx.agents.create({ agentId: AgentId('a1'), sessionId: SessionId('s1'), meta: { cwd: sessionDir }, agentOptions: { model: 'mock' } })
-    handle.agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, handle.agent as ReactLoopAgent)
-
-    expect(existsSync(marker)).toBe(true) // the marker landed in the SESSION dir
-    const { readFileSync } = await import('node:fs')
-    const where = readFileSync(marker, 'utf8').trim()
-    // `pwd` may resolve symlinks (/var → /private/var etc.), so compare basenames.
-    expect(where.endsWith(sessionDir.split('/').pop()!)).toBe(true)
-    await handle.dispose()
-  })
-
-  it('runs a SubagentStop hook in the CHILD session workspace, not the server cwd', async () => {
-    // `SubagentStop` recovers the child at `subagent/end`; a relative marker proves `runPoint`
-    // receives that agent and runs in the child's cwd rather than the executor default.
-    const serverDir = dir()
-    const childDir = dir()
-    const marker = join(childDir, 'stopwhere')
-    hooks(serverDir, { SubagentStop: [{ hooks: [{ type: 'command', command: 'pwd > stopwhere' }] }] })
-    const ctx = new Context()
-    await ctx.plugin(LlmService)
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
-    await ctx.plugin(AgentLoop, { agents: [] })
-    // Executor default cwd = serverDir (deliberately NOT the child session cwd).
-    await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000, cwd: serverDir })
-    await ctx.plugin(HooksClaude, { configPath: join(serverDir, 'hooks.json') })
-    ctx.llm.registerAdapter(['mock'], new MockAdapter([]))
-
-    // Register a live child on its own session cwd; emit subagent/end with its id.
-    const { SessionId } = await import('@deepseek-ai/dsh-session')
-    const childHandle = await ctx.agents.create({ agentId: AgentId('child-stop'), sessionId: SessionId('child-stop-session'), meta: { cwd: childDir }, agentOptions: { model: 'mock' } })
-    ctx.emit('subagent/end', { provider: 'inproc', id: childHandle.agent.id, stopReason: 'completed' })
-
-    await waitFor(() => existsSync(marker))
-    expect(existsSync(marker)).toBe(true) // the marker landed in the CHILD dir
-    const { readFileSync } = await import('node:fs')
-    const where = readFileSync(marker, 'utf8').trim()
-    // `pwd` may resolve symlinks (/var → /private/var etc.), so compare basenames.
-    expect(where.endsWith(childDir.split('/').pop()!)).toBe(true)
-    await childHandle.dispose()
-  })
-})
-
-describe('hooks-claude coverage — systemMessage is warned, not surfaced', () => {
-  it('a hook emitting a systemMessage is logged as not-yet-surfaced', async () => {
-    const d = dir()
-    const s = sh(d, 'sm.sh', '#!/usr/bin/env bash\necho \'{"systemMessage":"heads up"}\'\n')
-    const path = hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(path, adapter)
-    const warn = vi.fn(); ctx.logger.warn = warn as never
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    expect(warn).toHaveBeenCalledWith(expect.stringContaining('systemMessage'))
-    // Not surfaced: the systemMessage text never reaches the model request.
-    expect(JSON.stringify(adapter.requests[0]!.messages)).not.toContain('heads up')
-  })
-})
-
-describe('hooks-claude coverage — SessionStart timing is best-effort (no-wait)', () => {
-  it('does NOT crash or block when the prompt is sent immediately (context is best-effort, may miss the first request)', async () => {
-    // Session-start injection is detached, so an immediate prompt need not observe it. Assert only
-    // the guaranteed behavior—no crash and a completed turn—without pre-waiting away the race.
-    const d = dir()
-    const s = sh(d, 'start.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SessionStart","additionalContext":"late ctx"}}\'\n')
-    const path = hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: s }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(path, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    // Send immediately — do NOT wait for the session-start inject.
-    agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, agent)
-    expect(adapter.requests).toHaveLength(1) // the turn ran regardless of hook timing
-  })
-})
diff --git a/packages/hooks/hooks-claude/tsconfig.json b/packages/hooks/hooks-claude/tsconfig.json
index 909db9b5c3..07c88610f9 100644
--- a/packages/hooks/hooks-claude/tsconfig.json
+++ b/packages/hooks/hooks-claude/tsconfig.json
@@ -29,6 +29,9 @@
     {
       "path": "../../core/session"
     },
+    {
+      "path": "../../session-persistence/session-persistence"
+    },
     {
       "path": "../../subagent/subagent"
     },
diff --git a/packages/hooks/hooks-codex/README.md b/packages/hooks/hooks-codex/README.md
index 06cafe1297..0784857286 100644
--- a/packages/hooks/hooks-codex/README.md
+++ b/packages/hooks/hooks-codex/README.md
@@ -41,13 +41,15 @@ The hooks themselves run in the agent's session workspace: for the agent-scoped
 | Codex hook | Harness seam | Mapping |
 |---|---|---|
 | `SessionStart` | `agent/session-start` (emit) | a plain-stdout hook's output → additionalContext → `agent.inject()` |
-| `UserPromptSubmit` | `agent/prompt-submit` (waterfall) | `block` (exit 2) → `PromptDecision.block`; additionalContext-only → delegate via `next()` then fold context onto the downstream decision |
+| `UserPromptSubmit` | `agent/prompt-submit` (waterfall) | `block` (exit 2) → `PromptDecision.block`; additionalContext-only → delegate via `next()` then prepend a separately sourced context to downstream `additionalContexts` |
 | `PreToolUse` | `tools/pre-execute` (waterfall) | `block` → `PreToolDecision.deny` (no `allow`/`ask`) |
-| `PostToolUse` | `tools/post-execute` (waterfall) | `block` → `block` with feedback; additionalContext-only → delegate via `next()` then fold context onto the downstream decision (a Code Mode sub-call’s context is dropped by the run_code bridge — see [the pipeline doc](../../../docs/tool-execution-pipeline.md)) |
+| `PostToolUse` | `tools/post-execute` (waterfall) | `block` → `block` with feedback; additionalContext-only → delegate via `next()` then prepend a separately sourced context to the downstream decision; Code Mode defers sub-call contexts until the outer `run_code` result |
 | `Stop` | `agent/turn-continuation` (waterfall) | a blocking Stop hook forces `continue` with the reason as next-step steering |
 
 A tool call's payload carries the real `tool_name` (the same value the matcher tests) and Codex's `tool_input: { command }` shape (the `command` arg when present, else `''`). The matcher subject is the tool name (`PreToolUse`/`PostToolUse`) or the session source (`SessionStart`); `UserPromptSubmit`/`Stop` ignore matchers.
 
+Every agent-scoped stdin payload carries `session_id` and `transcript_path`. The bridge resolves the latter through `ctx.sessionPersistence.locate(session.header)` when available and otherwise sends `null`, preserving the Codex `string | null` shape. Lookup does not create or flush the artifact, so a path can be absent before the first turn-end checkpoint or omit the current open turn.
+
 `SessionStart` — the one emit point — runs detached; each run chain is tracked, and disposing the bridge aborts a still-running hook process, then drains the continuation before the dispose resolves (`createDetachedRuns` in `dsh-hook-protocol`).
 
 ## Context source
diff --git a/packages/hooks/hooks-codex/package.json b/packages/hooks/hooks-codex/package.json
index fe667b0302..8c8686c539 100644
--- a/packages/hooks/hooks-codex/package.json
+++ b/packages/hooks/hooks-codex/package.json
@@ -29,18 +29,21 @@
     "@deepseek-ai/dsh-hook-protocol": "^0.0.1",
     "@deepseek-ai/dsh-llm": "^0.0.1",
     "@deepseek-ai/dsh-session": "^0.0.1",
+    "@deepseek-ai/dsh-session-persistence": "^0.0.1",
     "@deepseek-ai/dsh-tools": "^0.0.1",
     "cordis": "^4.0.0-rc.7"
   },
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-bash": "workspace:^",
     "@deepseek-ai/dsh-bash-local": "workspace:^",
     "@deepseek-ai/dsh-hook-protocol": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
-    "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-session-persistence": "workspace:^",
+    "@deepseek-ai/dsh-session-persistence-jsonl": "workspace:^",
     "@deepseek-ai/dsh-tools": "workspace:^",
     "cordis": "^4.0.0-rc.7"
   }
diff --git a/packages/hooks/hooks-codex/src/index.ts b/packages/hooks/hooks-codex/src/index.ts
index 924b181151..8fa89a469d 100644
--- a/packages/hooks/hooks-codex/src/index.ts
+++ b/packages/hooks/hooks-codex/src/index.ts
@@ -17,6 +17,7 @@ import type { Context } from 'cordis'
 import z from 'schemastery'
 import type { Agent, ContinuationDecision, HookContext, PromptDecision } from '@deepseek-ai/dsh-agent'
 import type { ContentBlock, MessageSource } from '@deepseek-ai/dsh-llm'
+import type {} from '@deepseek-ai/dsh-session-persistence'
 import type { PostToolDecision, PreToolDecision, ToolExecution, ToolExecutionResult } from '@deepseek-ai/dsh-tools'
 import {
   appendHookInvoked,
@@ -163,17 +164,16 @@ export function apply(ctx: Context, config: Config): void {
     return { content, source: PLUGIN_SOURCE }
   }
 
-  /** Merge hook context while retaining this bridge's plugin-level source. */
-  function concatContext(ours: HookContext, theirs: HookContext | undefined): HookContext {
-    if (!theirs) return ours
-    return { content: [...ours.content, ...theirs.content], source: ours.source }
+  /** Prepend one context without flattening downstream provenance or metadata. */
+  function prependContext(ours: HookContext, theirs: HookContext[] | undefined): HookContext[] {
+    return [ours, ...theirs ?? []]
   }
 
   // SessionStart injects plain stdout when its detached hook resolves; a slow
   // hook may miss the first request.
   // TODO(session-start-gating): add a startup gate before promising first-turn delivery.
   ctx.on('agent/session-start', (agent, source) => {
-    detached.track(runPoint('SessionStart', source, { ...base(agent, 'SessionStart', model), source }, { agent, plainStdoutAsContext: true, signal: detached.signal })
+    detached.track(runPoint('SessionStart', source, { ...base(ctx, agent, 'SessionStart', model), source }, { agent, plainStdoutAsContext: true, signal: detached.signal })
       .then((merged) => {
         const context = contextFrom(merged)
         if (context) agent.inject(context.content, { source: context.source })
@@ -185,7 +185,7 @@ export function apply(ctx: Context, config: Config): void {
   // UserPromptSubmit → PromptDecision. Codex supports block, not allow or ask.
   ctx.on('agent/prompt-submit', async (agent, content, _source, next): Promise<PromptDecision> => {
     const turn = lastTurn(agent)
-    const merged = await runPoint('UserPromptSubmit', '', { ...turnBase(agent, 'UserPromptSubmit', model), prompt: blocksToText(content) }, { agent, turn, plainStdoutAsContext: true })
+    const merged = await runPoint('UserPromptSubmit', '', { ...turnBase(ctx, agent, 'UserPromptSubmit', model), prompt: blocksToText(content) }, { agent, turn, plainStdoutAsContext: true })
     /* jscpd:ignore-start */
     if (merged.decision === 'deny') return { kind: 'block', reason: merged.reason ?? 'blocked by UserPromptSubmit hook' }
     // Context alone is not a veto: DELEGATE so a later prompt-submit listener can
@@ -196,14 +196,14 @@ export function apply(ctx: Context, config: Config): void {
     return {
       kind: 'allow',
       ...downstream.content !== undefined ? { content: downstream.content } : {},
-      additionalContext: concatContext(ours, downstream.additionalContext),
+      additionalContexts: prependContext(ours, downstream.additionalContexts),
     }
   })
 
   // PreToolUse → PreToolDecision. Codex blocks only (no allow/ask honored).
   ctx.on('tools/pre-execute', async (exec, next): Promise<PreToolDecision> => {
     const turn = lastTurn(exec.agent)
-    const merged = await runPoint('PreToolUse', exec.name, preToolPayload(exec, model), { ...exec.agent ? { agent: exec.agent } : {}, turn, ...exec.signal ? { signal: exec.signal } : {} })
+    const merged = await runPoint('PreToolUse', exec.name, preToolPayload(ctx, exec, model), { ...exec.agent ? { agent: exec.agent } : {}, turn, ...exec.signal ? { signal: exec.signal } : {} })
     /* jscpd:ignore-end */
     if (merged.decision === 'deny') return { kind: 'deny', reason: merged.reason ?? 'blocked by PreToolUse hook' }
     return next()
@@ -213,22 +213,22 @@ export function apply(ctx: Context, config: Config): void {
   ctx.on('tools/post-execute', async (exec, result, next): Promise<PostToolDecision> => {
     const turn = lastTurn(exec.agent)
     /* jscpd:ignore-start */
-    const merged = await runPoint('PostToolUse', exec.name, postToolPayload(exec, result, model), { ...exec.agent ? { agent: exec.agent } : {}, turn, ...exec.signal ? { signal: exec.signal } : {} })
+    const merged = await runPoint('PostToolUse', exec.name, postToolPayload(ctx, exec, result, model), { ...exec.agent ? { agent: exec.agent } : {}, turn, ...exec.signal ? { signal: exec.signal } : {} })
     const context = contextFrom(merged)
     if (merged.decision === 'deny') {
-      return { kind: 'block', feedback: [{ type: 'text', text: merged.reason ?? 'blocked by PostToolUse hook' }], ...context ? { additionalContext: context } : {} }
+      return { kind: 'block', feedback: [{ type: 'text', text: merged.reason ?? 'blocked by PostToolUse hook' }], ...context ? { additionalContexts: [context] } : {} }
     }
     // Context alone is not a veto: DELEGATE, then fold our context onto the
     // downstream decision (a downstream block carries it too).
     const downstream = await next()
     if (!context) return downstream
     if (downstream.kind === 'block') {
-      return { ...downstream, additionalContext: concatContext(context, downstream.additionalContext) }
+      return { ...downstream, additionalContexts: prependContext(context, downstream.additionalContexts) }
     }
     return {
       kind: 'accept',
       ...downstream.content !== undefined ? { content: downstream.content } : {},
-      additionalContext: concatContext(context, downstream.additionalContext),
+      additionalContexts: prependContext(context, downstream.additionalContexts),
     }
   })
 
@@ -237,7 +237,7 @@ export function apply(ctx: Context, config: Config): void {
   // avoid continuing the same turn indefinitely. It is always false here, so an
   // unconditionally blocking hook force-continues every step until it self-limits.
   ctx.on('agent/turn-continuation', async (agent, turn, _default, next): Promise<ContinuationDecision> => {
-    const merged = await runPoint('Stop', '', { ...turnBase(agent, 'Stop', model), stop_hook_active: false, last_assistant_message: null }, { agent, turn })
+    const merged = await runPoint('Stop', '', { ...turnBase(ctx, agent, 'Stop', model), stop_hook_active: false, last_assistant_message: null }, { agent, turn })
     /* jscpd:ignore-end */
     if (merged.decision === 'deny') {
       // A blocking Stop hook forces continuation; a block with no reason (exit 2,
@@ -271,10 +271,12 @@ function blocksToText(content: ContentBlock[]): string {
 /* jscpd:ignore-end */
 
 /** Base fields on every Codex payload (no turn_id). */
-function base(agent: Agent | undefined, event: string, model: string): Record<string, unknown> {
+function base(ctx: Context, agent: Agent | undefined, event: string, model: string): Record<string, unknown> {
   return {
     session_id: agent?.session.header.id ?? '',
-    transcript_path: null,
+    transcript_path: agent === undefined
+      ? null
+      : ctx.get('sessionPersistence')?.locate(agent.session.header)?.path ?? null,
     cwd: agent?.session.header.cwd ?? process.cwd(),
     hook_event_name: event,
     model,
@@ -283,8 +285,8 @@ function base(agent: Agent | undefined, event: string, model: string): Record<st
 }
 
 /** Base + turn_id, for the turn-scoped events (PreToolUse/PostToolUse/UserPromptSubmit/Stop). */
-function turnBase(agent: Agent | undefined, event: string, model: string): Record<string, unknown> {
-  return { ...base(agent, event, model), turn_id: String(lastTurn(agent)) }
+function turnBase(ctx: Context, agent: Agent | undefined, event: string, model: string): Record<string, unknown> {
+  return { ...base(ctx, agent, event, model), turn_id: String(lastTurn(agent)) }
 }
 
 /** Extract a `command` string from a tool call's parsed arguments, else ''. */
@@ -296,14 +298,14 @@ function commandOf(args: unknown): string {
   return ''
 }
 
-function preToolPayload(exec: ToolExecution, model: string): Record<string, unknown> {
+function preToolPayload(ctx: Context, exec: ToolExecution, model: string): Record<string, unknown> {
   // `tool_name` is the REAL tool name (matching the `exec.name` matcher subject);
   // a hardcoded constant would disagree with what the matcher tests and make a
   // config's tool matcher never fire. `tool_input` keeps Codex's `{ command }`
   // shape (its shell payload), derived from the call's `command` arg when present.
-  return { ...turnBase(exec.agent, 'PreToolUse', model), tool_name: exec.name, tool_input: { command: commandOf(exec.arguments) }, tool_use_id: exec.callId }
+  return { ...turnBase(ctx, exec.agent, 'PreToolUse', model), tool_name: exec.name, tool_input: { command: commandOf(exec.arguments) }, tool_use_id: exec.callId }
 }
 
-function postToolPayload(exec: ToolExecution, result: ToolExecutionResult, model: string): Record<string, unknown> {
-  return { ...turnBase(exec.agent, 'PostToolUse', model), tool_name: exec.name, tool_input: { command: commandOf(exec.arguments) }, tool_use_id: exec.callId, tool_response: blocksToText(result.content) }
+function postToolPayload(ctx: Context, exec: ToolExecution, result: ToolExecutionResult, model: string): Record<string, unknown> {
+  return { ...turnBase(ctx, exec.agent, 'PostToolUse', model), tool_name: exec.name, tool_input: { command: commandOf(exec.arguments) }, tool_use_id: exec.callId, tool_response: blocksToText(result.content) }
 }
diff --git a/packages/hooks/hooks-codex/tests/bridge.spec.ts b/packages/hooks/hooks-codex/tests/bridge.spec.ts
index d4a5797b96..e686b565f2 100644
--- a/packages/hooks/hooks-codex/tests/bridge.spec.ts
+++ b/packages/hooks/hooks-codex/tests/bridge.spec.ts
@@ -4,12 +4,11 @@ import { tmpdir } from 'node:os'
 import { join } from 'node:path'
 import { Context } from 'cordis'
 import Loader from '@cordisjs/plugin-loader'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore, { type SessionEvent } from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry, { defineTool } from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import type { SessionEvent } from '@deepseek-ai/dsh-session'
+import { defineTool } from '@deepseek-ai/dsh-tools'
+import { AgentId } from '@deepseek-ai/dsh-agent'
 import AgentLoop, { type ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
 import * as HooksCodex from '@deepseek-ai/dsh-hooks-codex'
 import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
@@ -40,11 +39,7 @@ function writeHooks(dir: string, hooks: unknown): void {
 
 async function harness(dir: string, adapter: MockAdapter): Promise<Context> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
   await ctx.plugin(HooksCodex, { configPath: join(dir, 'hooks.json'), model: 'test-model' })
@@ -81,7 +76,7 @@ describe('hooks-codex bridge', () => {
     const ctx = await harness(dir, adapter)
     let ran = false
     ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'no' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'run ls' }])
     await waitForIdle(ctx, agent)
 
@@ -102,7 +97,7 @@ describe('hooks-codex bridge', () => {
 
     const adapter = new MockAdapter([textResponse('first answer'), textResponse('second answer after goal')])
     const ctx = await harness(dir, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
 
@@ -117,7 +112,7 @@ describe('hooks-codex bridge', () => {
 
     const adapter = new MockAdapter([textResponse('fine')])
     const ctx = await harness(dir, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
     expect(adapter.requests).toHaveLength(1)
@@ -127,7 +122,7 @@ describe('hooks-codex bridge', () => {
     const dir = configDir() // no hooks.json written
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = await harness(dir, adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
     expect(adapter.requests).toHaveLength(1)
@@ -141,17 +136,13 @@ describe('hooks-codex bridge', () => {
     writeHooks(dir, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: deny }] }] })
     const adapter = new MockAdapter([textResponse('ok')])
     const ctx = new Context()
-    await ctx.plugin(LlmService)
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
+    await mountAgentLoopTestDependencies(ctx)
     await ctx.plugin(AgentLoop, { agents: [] })
     await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
     const fiber = await ctx.plugin(HooksCodex, { configPath: join(dir, 'hooks.json'), model: 'm' })
     await fiber.dispose()
     ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
     agent.send([{ type: 'text', text: 'go' }])
     await waitForIdle(ctx, agent)
     expect(adapter.requests).toHaveLength(1) // not blocked → the listener is gone
@@ -167,18 +158,14 @@ describe('hooks-codex bridge', () => {
     const slow = script(dir, 'slow.sh', `#!/usr/bin/env bash\necho $$ > "${pidFile}"\ntouch "${marker}"\nsleep 30\n`)
     writeHooks(dir, { SessionStart: [{ hooks: [{ type: 'command', command: slow }] }] })
     const ctx = new Context()
-    await ctx.plugin(LlmService)
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
+    await mountAgentLoopTestDependencies(ctx)
     await ctx.plugin(AgentLoop, { agents: [] })
     await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
     const fiber = await ctx.plugin(HooksCodex, { configPath: join(dir, 'hooks.json'), model: 'm' })
     ctx.llm.registerAdapter(['mock'], new MockAdapter([]))
     const warn = vi.fn()
     ctx.logger.warn = warn as never
-    ctx.agentLoop.create(AgentId('a1'), { model: 'mock' }) // fires agent/session-start
+    ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' }) // fires agent/session-start
     await waitFor(() => existsSync(marker))
     const pid = Number(readFileSync(pidFile, 'utf8').trim())
     await fiber.dispose()
diff --git a/packages/hooks/hooks-codex/tests/coverage-cases.ts b/packages/hooks/hooks-codex/tests/coverage-cases.ts
new file mode 100644
index 0000000000..d11c74951d
--- /dev/null
+++ b/packages/hooks/hooks-codex/tests/coverage-cases.ts
@@ -0,0 +1,637 @@
+import { afterEach, describe, expect, it, vi } from 'vitest'
+import { mkdtempSync, rmSync, writeFileSync, chmodSync, existsSync, readFileSync } from 'node:fs'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { Context } from 'cordis'
+import type { SessionEvent } from '@deepseek-ai/dsh-session'
+import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
+import { defineTool } from '@deepseek-ai/dsh-tools'
+import { AgentId } from '@deepseek-ai/dsh-agent'
+import AgentLoop, { type ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
+import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
+import * as HooksCodex from '@deepseek-ai/dsh-hooks-codex'
+import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
+
+const dirs: string[] = []
+afterEach(() => { for (const d of dirs.splice(0)) rmSync(d, { recursive: true, force: true }) })
+function dir(): string { const d = mkdtempSync(join(tmpdir(), 'dsh-hx-cov-')); dirs.push(d); return d }
+function sh(d: string, name: string, body: string): string {
+  const p = join(d, name); writeFileSync(p, body); chmodSync(p, 0o755); return p
+}
+function hooks(d: string, h: unknown): string {
+  writeFileSync(join(d, 'hooks.json'), JSON.stringify({ hooks: h })); return join(d, 'hooks.json')
+}
+
+type HarnessOpts = { stderrSummaryMaxChars?: number; sessionRoot?: string }
+async function harness(configPath: string, adapter: MockAdapter, opts: HarnessOpts = {}): Promise<Context> {
+  const ctx = new Context()
+  await mountAgentLoopTestDependencies(ctx)
+  if (opts.sessionRoot !== undefined) await ctx.plugin(SessionPersistenceJsonl, { root: opts.sessionRoot })
+  await ctx.plugin(AgentLoop, { agents: [] })
+  await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
+  await ctx.plugin(HooksCodex, { configPath, model: 'm', ...opts })
+  ctx.llm.registerAdapter(['mock'], adapter)
+  return ctx
+}
+function waitForIdle(ctx: Context, agent: ReactLoopAgent): Promise<void> {
+  return new Promise((resolve) => { const d = ctx.on('agent/status', (s, st) => { if (s === agent && st === 'idle') { d(); resolve() } }) })
+}
+function events(agent: ReactLoopAgent): SessionEvent[] { return [...agent.session.events] }
+/** Poll until `predicate` holds or the deadline passes — robust to detached
+ * emit-listener hooks firing on a `.then` (a fixed sleep flakes under load). */
+async function waitFor(predicate: () => boolean, timeout = 5000, interval = 10): Promise<void> {
+  const deadline = Date.now() + timeout
+  while (!predicate()) {
+    if (Date.now() > deadline) throw new Error('waitFor: condition not met before deadline')
+    await new Promise(r => setTimeout(r, interval))
+  }
+}
+
+export type CoverageGroup = 'prompt' | 'post-tool' | 'result-shape' | 'edge-paths' | 'payload'
+
+/** Register independently schedulable slices of the hooks-codex coverage matrix. */
+export function defineCoverageCases(groups: CoverageGroup | readonly CoverageGroup[]): void {
+  const selected = new Set(typeof groups === 'string' ? [groups] : groups)
+  if (selected.has('prompt')) describe('hooks-codex coverage — prompt decision mapping', () => {
+    it('uses the persistence locator for transcript_path and null without one', async () => {
+      async function capture(sessionRoot?: string): Promise<{ payload: { transcript_path: string | null }; expected: string | undefined }> {
+        const d = dir()
+        const cap = join(d, 'payload')
+        const path = hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'capture.sh', `#!/usr/bin/env bash\ncat > "${cap}"\n`) }] }] })
+        const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+        const ctx = await harness(path, adapter, { ...sessionRoot !== undefined ? { sessionRoot } : {} })
+        ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+        const agent = ctx.agentLoop.create(AgentId('transcript'), { provider: 'mock', model: 'mock' })
+        agent.send([{ type: 'text', text: 'go' }])
+        await waitForIdle(ctx, agent)
+        return {
+          payload: JSON.parse(readFileSync(cap, 'utf8')) as { transcript_path: string | null },
+          expected: ctx.get('sessionPersistence')?.locate(agent.session.header)?.path,
+        }
+      }
+
+      const located = await capture(dir())
+      expect(located.payload.transcript_path).toBe(located.expected)
+      expect((await capture()).payload.transcript_path).toBeNull()
+    }, 15_000) // Two real agent/hook subprocess loops need loaded pre-push runner headroom.
+
+    it('UserPromptSubmit block (exit 2) → rejected turn; default reason on empty stderr', async () => {
+      const d = dir()
+      hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'b.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
+      const adapter = new MockAdapter([textResponse('no')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(adapter.requests).toHaveLength(0)
+      const te = events(agent).findLast(e => e.type === 'turn/end')
+      expect(te?.type === 'turn/end' && te.data.reason.kind).toBe('rejected')
+    })
+
+    it('UserPromptSubmit additionalContext is injected; a no-op hook proceeds', async () => {
+      const d = dir()
+      hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'c.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"ctx-x"}}\'\n') }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(JSON.stringify(adapter.requests[0]!.messages)).toContain('ctx-x')
+    })
+
+    it('a context-only UserPromptSubmit hook DELEGATES so a later listener can still block', async () => {
+    // Context alone is not a veto: the bridge delegates with `next()` and folds its context, so a
+    // downstream policy listener can still block.
+      const d = dir()
+      hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'c.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"bridge ctx"}}\'\n') }] }] })
+      const adapter = new MockAdapter([textResponse('should not run')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.on('agent/prompt-submit', async () => ({ kind: 'block' as const, reason: 'policy veto' }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(adapter.requests).toHaveLength(0)
+      expect(events(agent).some(e => e.type === 'user/message')).toBe(false)
+      const te = events(agent).findLast(e => e.type === 'turn/end')
+      expect(te?.type === 'turn/end' && te.data.reason).toMatchObject({ kind: 'rejected', reason: 'policy veto' })
+    })
+
+    it('preserves separate bridge and downstream prompt contexts with framing and metadata', async () => {
+      const d = dir()
+      hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'c.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"from-bridge"}}\'\n') }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.on('agent/prompt-submit', async () => ({
+        kind: 'allow' as const,
+        content: [{ type: 'text' as const, text: 'rewritten-prompt' }],
+        additionalContexts: [{
+          content: [{ type: 'text' as const, text: 'from-downstream' }],
+          source: { kind: 'plugin' as const, plugin: 'policy' },
+          envelope: 'raw' as const,
+          meta: { owner: 'policy' },
+        }],
+      }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const req = JSON.stringify(adapter.requests[0]!.messages)
+      expect(req).toContain('from-bridge')
+      expect(req).toContain('from-downstream')
+      expect(req).toContain('rewritten-prompt')
+      const contexts = events(agent).filter(event => event.type === 'context/message')
+      expect(contexts.map(event => event.type === 'context/message' && event.data.source)).toEqual([
+        { kind: 'plugin', plugin: 'hooks-codex' },
+        { kind: 'plugin', plugin: 'policy' },
+      ])
+      expect(contexts[1]?.type === 'context/message' && contexts[1].data.envelope).toBe('raw')
+      expect(contexts[1]?.type === 'context/message' && contexts[1].data.meta).toEqual({ owner: 'policy' })
+    })
+  })
+
+  if (selected.has('post-tool')) describe('hooks-codex coverage — post-tool and session context mapping', () => {
+    it('folds the bridge PostToolUse context onto a downstream ACCEPT that replaces content', async () => {
+      const d = dir()
+      hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pc.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"bridge-note"}}\'\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      ctx.on('tools/post-execute', async () => ({ kind: 'accept' as const, content: [{ type: 'text' as const, text: 'rewritten-result' }] }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const result = events(agent).find(e => e.type === 'tool/result')
+      expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text === 'rewritten-result')).toBe(true)
+      expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('bridge-note')))).toBe(true)
+    })
+
+    it('keeps bridge and downstream PostToolUse contexts as separate sourced events', async () => {
+      const d = dir()
+      hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pc.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"bridge-note"}}\'\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      ctx.on('tools/post-execute', async () => ({
+        kind: 'accept' as const,
+        additionalContexts: [{
+          content: [{ type: 'text' as const, text: 'downstream-note' }],
+          source: { kind: 'plugin' as const, plugin: 'policy' },
+          envelope: 'raw' as const,
+          meta: { owner: 'policy' },
+        }],
+      }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+
+      const contexts = events(agent).filter(event => event.type === 'context/message')
+      expect(contexts.map(event => event.type === 'context/message' && event.data.source)).toEqual([
+        { kind: 'plugin', plugin: 'hooks-codex' },
+        { kind: 'plugin', plugin: 'policy' },
+      ])
+      expect(contexts[1]?.type === 'context/message' && contexts[1].data.envelope).toBe('raw')
+      expect(contexts[1]?.type === 'context/message' && contexts[1].data.meta).toEqual({ owner: 'policy' })
+    })
+
+    it('folds the bridge PostToolUse context onto a downstream listener BLOCK', async () => {
+      const d = dir()
+      hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pc.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"bridge-note"}}\'\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      ctx.on('tools/post-execute', async () => ({ kind: 'block' as const, feedback: [{ type: 'text' as const, text: 'downstream-block' }] }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const result = events(agent).find(e => e.type === 'tool/result')
+      expect(result?.type === 'tool/result' && result.data.isError).toBe(true)
+      expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text.includes('downstream-block'))).toBe(true)
+      expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('bridge-note')))).toBe(true)
+    })
+
+    it('SessionStart additionalContext is injected for the first request', async () => {
+      const d = dir()
+      hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: sh(d, 's.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SessionStart","additionalContext":"start-ctx"}}\'\n') }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      await waitFor(() => events(agent).some(e => e.type === 'context/message'
+      && e.data.content.some(b => b.type === 'text' && b.text.includes('start-ctx'))))
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(JSON.stringify(adapter.requests[0]!.messages)).toContain('start-ctx')
+    })
+
+    it('PostToolUse block (exit 2) → isError feedback; default reason', async () => {
+      const d = dir()
+      hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'p.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'ls' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const r = events(agent).find(e => e.type === 'tool/result')
+      expect(r?.type === 'tool/result' && r.data.isError).toBe(true)
+      expect(r?.type === 'tool/result' && r.data.content.some(b => b.type === 'text' && b.text.includes('blocked by PostToolUse hook'))).toBe(true)
+    })
+
+    it('PostToolUse additionalContext (clean exit) is attached after the result', async () => {
+      const d = dir()
+      hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pc.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"post-ctx"}}\'\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'ls' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('post-ctx')))).toBe(true)
+    })
+  })
+
+  if (selected.has('result-shape')) describe('hooks-codex coverage — hook result shape and configuration', () => {
+    it('PreToolUse for a tool call WITHOUT a command arg passes an empty command (commandOf non-object/missing arm)', async () => {
+      const d = dir()
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pre.sh', '#!/usr/bin/env bash\ncat >/dev/null\nexit 0\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', {}), textResponse('done')]) // no command arg
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(ran).toBe(true) // clean-exit hook allows; commandOf returned ''
+    })
+
+    it('a clean exit-0 hook records exitCode 0 and omits stderrSummary', async () => {
+      const d = dir()
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'n.sh', '#!/usr/bin/env bash\nexit 0\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && res.data.exitCode).toBe(0)
+      expect(res?.type === 'hook/result' && 'stderrSummary' in res.data).toBe(false)
+    })
+
+    it('a long stderr is truncated in the hook/result summary', async () => {
+      const d = dir()
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'l.sh', '#!/usr/bin/env bash\nprintf "x%.0s" {1..600} >&2\nexit 2\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && res.data.stderrSummary?.endsWith('…')).toBe(true)
+      expect(res?.type === 'hook/result' && res.data.stderrSummary?.length).toBe(501) // default 500-char cap + ellipsis
+    })
+
+    it('rejects a non-positive or fractional stderrSummaryMaxChars at load', async () => {
+      const d = dir()
+      hooks(d, {})
+      for (const bad of [0, -5, 1.5, Number.NaN]) {
+        const adapter = new MockAdapter([])
+        await expect(harness(join(d, 'hooks.json'), adapter, { stderrSummaryMaxChars: bad }))
+          .rejects.toThrow(/hooks-codex: stderrSummaryMaxChars must be a positive integer/)
+      }
+    })
+
+    it('the stderr summary cap is plugin config (stderrSummaryMaxChars)', async () => {
+      const d = dir()
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'l.sh', '#!/usr/bin/env bash\nprintf "x%.0s" {1..600} >&2\nexit 2\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter, { stderrSummaryMaxChars: 40 })
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && res.data.stderrSummary).toBe('x'.repeat(40) + '…')
+    })
+
+    it('warns on a skipped async hook and a direct apply() (schema bypass) runs', async () => {
+      const d = dir()
+      const marker = join(d, 'ran')
+      hooks(d, { UserPromptSubmit: [{ hooks: [
+        { type: 'command', command: 'bg.sh', async: true }, // skipped → warn
+        { type: 'command', command: sh(d, 'h.sh', `#!/usr/bin/env bash\ntouch "${marker}"\n`) },
+      ] }] })
+      const warn = vi.fn()
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = new Context()
+      await mountAgentLoopTestDependencies(ctx)
+      await ctx.plugin(AgentLoop, { agents: [] })
+      await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
+      ctx.logger.warn = warn as never
+      // Direct apply (schema bypass) → the `model ?? ''` fallback is exercised.
+      HooksCodex.apply(ctx, { configPath: join(d, 'hooks.json') })
+      ctx.llm.registerAdapter(['mock'], adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(existsSync(marker)).toBe(true)
+      expect(warn).toHaveBeenCalledWith(expect.stringContaining('async hook'))
+    })
+
+    it('a no-op clean hook proceeds (contextFrom empty → next)', async () => {
+      const d = dir()
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'n.sh', '#!/usr/bin/env bash\nexit 0\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(ran).toBe(true)
+    })
+
+    it('SessionStart with no additionalContext is a no-op (contextFrom empty)', async () => {
+      const d = dir()
+      // The hook touches a marker so we can wait for it to ACTUALLY FINISH before
+      // asserting absence — a completed turn alone would not prove the detached
+      // session-start hook ran, making the absence check a false pass.
+      const marker = join(d, 'ss-ran')
+      hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: sh(d, 's.sh', `#!/usr/bin/env bash\ntouch "${marker}"\nexit 0\n`) }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      await waitFor(() => existsSync(marker)) // the clean no-output hook has finished
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(events(agent).some(e => e.type === 'context/message')).toBe(false)
+    })
+
+    it('a throwing SessionStart inject is contained (logged)', async () => {
+      const d = dir()
+      hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: sh(d, 's.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SessionStart","additionalContext":"x"}}\'\n') }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const warn = vi.fn(); ctx.logger.warn = warn as never
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.inject = (() => { throw new Error('inject boom') })
+      await waitFor(() => warn.mock.calls.some(c => String(c[0]).includes('SessionStart hook failed')))
+      expect(warn).toHaveBeenCalledWith(expect.stringContaining('SessionStart hook failed'))
+    })
+  })
+
+  if (selected.has('edge-paths')) describe('hooks-codex coverage — matching and no-agent edge paths', () => {
+    it('a clean PreToolUse with no decision allows the tool (no deny)', async () => {
+      const d = dir()
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'ok.sh', '#!/usr/bin/env bash\nexit 0\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(ran).toBe(true)
+    })
+
+    it('a non-matching regex matcher skips the hook (matchesMatcher false → continue)', async () => {
+      const d = dir()
+      // /^Edit$/ does not match the tool name "Bash" → the group is skipped.
+      hooks(d, { PreToolUse: [{ matcher: '^Edit$', hooks: [{ type: 'command', command: sh(d, 'deny.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(ran).toBe(true) // matcher didn't match → no hook ran → tool proceeded
+      expect(events(agent).some(e => e.type === 'hook/invoked')).toBe(false)
+    })
+
+    it('a {"continue":false} hook is RECORDED as "stop" but does not halt the run (TODO(hook-continue-false))', async () => {
+    // Honoring `continue:false` is deferred — the seams have no hard-halt
+    // primitive. Assert the LOG records the halt request AND that the run is not
+    // actually halted (the tool still runs, the turn completes).
+      const d = dir()
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 's.sh', '#!/usr/bin/env bash\necho \'{"continue":false,"stopReason":"halt"}\'\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && res.data.decision).toBe('stop') // recorded
+      expect(ran).toBe(true) // NOT honored: the tool still ran (halt is deferred)
+    })
+
+    it('PreToolUse deny with EMPTY stderr uses the default reason (?? right arm)', async () => {
+      const d = dir()
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'd.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const r = events(agent).find(e => e.type === 'tool/result')
+      expect(r?.type === 'tool/result' && r.data.content.some(b => b.type === 'text' && b.text.includes('blocked by PreToolUse hook'))).toBe(true)
+    })
+
+    it('PostToolUse block AND additionalContext are surfaced together', async () => {
+      const d = dir()
+      hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'bc.sh', '#!/usr/bin/env bash\necho \'{"decision":"block","reason":"bad","hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"ctx too"}}\'\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const r = events(agent).find(e => e.type === 'tool/result')
+      expect(r?.type === 'tool/result' && r.data.isError).toBe(true)
+      expect(r?.type === 'tool/result' && r.data.content.some(b => b.type === 'text' && b.text.includes('bad'))).toBe(true)
+      expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('ctx too')))).toBe(true)
+    })
+
+    it('commandOf reads a non-string command arg as an empty command', async () => {
+      const d = dir()
+      // The tool-call arguments carry `command` as a NUMBER → commandOf's
+      // `typeof command === 'string'` false arm → '' (the payload's tool_input.command).
+      const cap = join(d, 'payload')
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'cap.sh', `#!/usr/bin/env bash\ncat > "${cap}"\nexit 0\n`) }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 7 }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'number' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const payload = JSON.parse(readFileSync(cap, 'utf8')) as { tool_input: { command: string } }
+      expect(payload.tool_input.command).toBe('')
+    })
+
+    it('a no-agent direct PreToolUse run uses process.cwd() and turn 0 (no session to record)', async () => {
+      const d = dir()
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'd.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
+      const ctx = await harness(join(d, 'hooks.json'), new MockAdapter([]))
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'x' }] } }))
+      const { CallId } = await import('@deepseek-ai/dsh-llm')
+      const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'Bash', arguments: { command: 'x' } })
+      expect(ran).toBe(false) // denied
+      expect(result.isError).toBe(true)
+    })
+
+    it('a no-agent direct PostToolUse run attaches context with no session to record', async () => {
+      const d = dir()
+      hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pc.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"x"}}\'\n') }] }] })
+      const ctx = await harness(join(d, 'hooks.json'), new MockAdapter([]))
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const { CallId } = await import('@deepseek-ai/dsh-llm')
+      const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'Bash', arguments: { command: 'x' } })
+      expect(result.isError).toBeFalsy()
+      expect(result.additionalContexts?.[0]?.content.some(b => b.type === 'text' && b.text === 'x')).toBe(true)
+    })
+
+    it('when the bash executor REJECTS, the hook/result omits exitCode (non-blocking)', async () => {
+      const d = dir()
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'h.sh', '#!/usr/bin/env bash\nexit 0\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.bash.run = (() => Promise.reject(new Error('executor down')))
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const res = events(agent).find(e => e.type === 'hook/result')
+      expect(res?.type === 'hook/result' && 'exitCode' in res.data).toBe(false)
+    })
+  })
+
+  if (selected.has('payload')) describe('hooks-codex coverage — continuation, payload, and cwd mapping', () => {
+    it('a blocking Stop hook with EMPTY stderr still forces continuation (no reason required)', async () => {
+    // Regression: an exit-2 Stop hook with no stderr yields decision 'deny' +
+    // reason undefined; the turn must STILL force-continue, not silently stop.
+      const d = dir()
+      const marker = join(d, 'fired')
+      hooks(d, { Stop: [{ hooks: [{ type: 'command', command: sh(d, 's.sh', `#!/usr/bin/env bash\nif [ -e "${marker}" ]; then exit 0; fi\ntouch "${marker}"\nexit 2\n`) }] }] })
+      const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(adapter.requests).toHaveLength(2) // empty-reason block forced continuation
+      expect(JSON.stringify(adapter.requests[1]!.messages)).toContain('blocked by Stop hook')
+    })
+
+    it('a clean UserPromptSubmit hook that prints PLAIN stdout injects it as context', async () => {
+    // Codex feeds a SessionStart/UserPromptSubmit hook's PLAIN (non-JSON) stdout
+    // as additionalContext (unlike CC, which needs a JSON hookSpecificOutput).
+      const d = dir()
+      hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho "extra guidance from a plain hook"\nexit 0\n') }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(JSON.stringify(adapter.requests[0]!.messages)).toContain('extra guidance from a plain hook')
+    })
+
+    it('a NON-clean SessionStart hook (exit 2) does NOT inject its stdout as context', async () => {
+    // SessionStart cannot block, but non-clean stdout still must not become context. The marker
+    // waits for detached completion; `echo stale; exit 2` then proves the exit-code gate matches
+    // the codec's structured-stdout rule.
+      const d = dir()
+      const marker = join(d, 'ran')
+      hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: sh(d, 'b.sh', `#!/usr/bin/env bash\ntouch "${marker}"\necho "stale"\nexit 2\n`) }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      await waitFor(() => existsSync(marker)) // the exit-2 hook has finished
+      expect(events(agent).some(e => e.type === 'context/message'
+      && e.data.content.some(b => b.type === 'text' && b.text.includes('stale')))).toBe(false)
+    })
+
+    it('a UserPromptSubmit hook with a non-blocking error exit (1) + stdout does NOT inject it', async () => {
+    // Exit 1 is a non-blocking error (no decision), so the prompt is NOT blocked
+    // and the handler falls through to the context path — the gate must still
+    // suppress the error hook's stdout ("stale" never reaches the model).
+      const d = dir()
+      hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'e.sh', '#!/usr/bin/env bash\necho "stale"\nexit 1\n') }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(adapter.requests).toHaveLength(1) // exit 1 is non-blocking → the turn ran
+      expect(JSON.stringify(adapter.requests[0]!.messages)).not.toContain('stale')
+    })
+
+    it('a clean SessionStart hook that prints PLAIN stdout injects it (not JSON)', async () => {
+      const d = dir()
+      hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: sh(d, 'ss.sh', '#!/usr/bin/env bash\necho "session preamble"\nexit 0\n') }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      await waitFor(() => events(agent).some(e => e.type === 'context/message'
+      && e.data.content.some(b => b.type === 'text' && b.text.includes('session preamble'))))
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(JSON.stringify(adapter.requests[0]!.messages)).toContain('session preamble')
+    })
+
+    it('a clean hook that prints JSON is NOT injected as prose (plain-stdout gate)', async () => {
+    // A structured (JSON) stdout must go through the hookSpecificOutput path, not
+    // be dumped verbatim as context — the `!startsWith('{')` gate guards this.
+      const d = dir()
+      hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'j.sh', '#!/usr/bin/env bash\necho \'{"unrelated":"json"}\'\nexit 0\n') }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(JSON.stringify(adapter.requests[0]!.messages)).not.toContain('unrelated')
+    })
+
+    it('the PreToolUse payload carries the REAL tool name (matches the matcher subject)', async () => {
+    // Regression: the payload once hardcoded tool_name "Bash", disagreeing with
+    // the exec.name matcher subject — a config matcher on the real name would
+    // then never fire. Capture the payload and assert tool_name === the real name.
+      const d = dir()
+      const cap = join(d, 'payload')
+      hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'cap.sh', `#!/usr/bin/env bash\ncat > "${cap}"\nexit 0\n`) }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'shell', { command: 'ls' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      ctx.tools.register(defineTool({ name: 'shell', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      const payload = JSON.parse(readFileSync(cap, 'utf8')) as { tool_name: string; tool_input: { command: string } }
+      expect(payload.tool_name).toBe('shell')
+      expect(payload.tool_input.command).toBe('ls')
+    })
+
+    it('a Codex matcher on the REAL tool name fires (matcher subject === payload tool_name)', async () => {
+    // A regex matcher matching the real tool name must select the hook — proving
+    // the matcher subject and the payload tool_name agree.
+      const d = dir()
+      hooks(d, { PreToolUse: [{ matcher: 'shell', hooks: [{ type: 'command', command: sh(d, 'd.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'shell', { command: 'ls' }), textResponse('done')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      let ran = false
+      ctx.tools.register(defineTool({ name: 'shell', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(ran).toBe(false) // the matcher fired → the hook denied the tool
+      expect(events(agent).some(e => e.type === 'hook/invoked' && e.data.point === 'PreToolUse')).toBe(true)
+    })
+
+    it('a hook emitting a systemMessage is warned as not-yet-surfaced', async () => {
+      const d = dir()
+      hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'sm.sh', '#!/usr/bin/env bash\necho \'{"systemMessage":"heads up"}\'\n') }] }] })
+      const adapter = new MockAdapter([textResponse('ok')])
+      const ctx = await harness(join(d, 'hooks.json'), adapter)
+      const warn = vi.fn(); ctx.logger.warn = warn as never
+      const agent = ctx.agentLoop.create(AgentId('a1'), { provider: 'mock', model: 'mock' })
+      agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
+      expect(warn).toHaveBeenCalledWith(expect.stringContaining('systemMessage'))
+      expect(JSON.stringify(adapter.requests[0]!.messages)).not.toContain('heads up')
+    })
+
+    it('runs an agent-scoped hook in the session cwd, not the executor default', async () => {
+    // Same regression as the CC bridge: the Codex bridge must thread the session
+    // cwd as the hook workdir. Executor default = serverDir; session cwd =
+    // sessionDir; the PreToolUse hook's `pwd` marker must land in sessionDir.
+      const serverDir = dir()
+      const sessionDir = dir()
+      const marker = join(sessionDir, 'where')
+      hooks(serverDir, { PreToolUse: [{ hooks: [{ type: 'command', command: 'pwd > where' }] }] })
+      const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
+      const ctx = new Context()
+      await mountAgentLoopTestDependencies(ctx)
+      await ctx.plugin(AgentLoop, { agents: [] })
+      await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000, cwd: serverDir })
+      await ctx.plugin(HooksCodex, { configPath: join(serverDir, 'hooks.json'), model: 'm' })
+      ctx.llm.registerAdapter(['mock'], adapter)
+      ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
+      const { SessionId } = await import('@deepseek-ai/dsh-session')
+      const handle = await ctx.agents.create({ agentId: AgentId('a1'), sessionId: SessionId('s1'), meta: { cwd: sessionDir }, agentOptions: { provider: 'mock', model: 'mock' } })
+      handle.agent.send([{ type: 'text', text: 'go' }])
+      await waitForIdle(ctx, handle.agent as ReactLoopAgent)
+      expect(existsSync(marker)).toBe(true)
+      expect(readFileSync(marker, 'utf8').trim().endsWith(sessionDir.split('/').pop()!)).toBe(true)
+      await handle.dispose()
+    })
+  })
+}
diff --git a/packages/hooks/hooks-codex/tests/coverage-post-tool.spec.ts b/packages/hooks/hooks-codex/tests/coverage-post-tool.spec.ts
new file mode 100644
index 0000000000..0cd39dbe20
--- /dev/null
+++ b/packages/hooks/hooks-codex/tests/coverage-post-tool.spec.ts
@@ -0,0 +1,3 @@
+import { defineCoverageCases } from './coverage-cases.ts'
+
+defineCoverageCases(['post-tool', 'payload'])
diff --git a/packages/hooks/hooks-codex/tests/coverage-prompt.spec.ts b/packages/hooks/hooks-codex/tests/coverage-prompt.spec.ts
new file mode 100644
index 0000000000..be18c719f6
--- /dev/null
+++ b/packages/hooks/hooks-codex/tests/coverage-prompt.spec.ts
@@ -0,0 +1,3 @@
+import { defineCoverageCases } from './coverage-cases.ts'
+
+defineCoverageCases(['prompt', 'edge-paths'])
diff --git a/packages/hooks/hooks-codex/tests/coverage-result-shape.spec.ts b/packages/hooks/hooks-codex/tests/coverage-result-shape.spec.ts
new file mode 100644
index 0000000000..9546872e3c
--- /dev/null
+++ b/packages/hooks/hooks-codex/tests/coverage-result-shape.spec.ts
@@ -0,0 +1,3 @@
+import { defineCoverageCases } from './coverage-cases.ts'
+
+defineCoverageCases('result-shape')
diff --git a/packages/hooks/hooks-codex/tests/coverage.spec.ts b/packages/hooks/hooks-codex/tests/coverage.spec.ts
deleted file mode 100644
index c287d86b23..0000000000
--- a/packages/hooks/hooks-codex/tests/coverage.spec.ts
+++ /dev/null
@@ -1,560 +0,0 @@
-import { afterEach, describe, expect, it, vi } from 'vitest'
-import { mkdtempSync, rmSync, writeFileSync, chmodSync, existsSync, readFileSync } from 'node:fs'
-import { tmpdir } from 'node:os'
-import { join } from 'node:path'
-import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore, { type SessionEvent } from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry, { defineTool } from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
-import AgentLoop, { type ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
-import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
-import * as HooksCodex from '@deepseek-ai/dsh-hooks-codex'
-import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
-
-const dirs: string[] = []
-afterEach(() => { for (const d of dirs.splice(0)) rmSync(d, { recursive: true, force: true }) })
-function dir(): string { const d = mkdtempSync(join(tmpdir(), 'dsh-hx-cov-')); dirs.push(d); return d }
-function sh(d: string, name: string, body: string): string {
-  const p = join(d, name); writeFileSync(p, body); chmodSync(p, 0o755); return p
-}
-function hooks(d: string, h: unknown): string {
-  writeFileSync(join(d, 'hooks.json'), JSON.stringify({ hooks: h })); return join(d, 'hooks.json')
-}
-
-async function harness(configPath: string, adapter: MockAdapter, opts: { stderrSummaryMaxChars?: number } = {}): Promise<Context> {
-  const ctx = new Context()
-  await ctx.plugin(LlmService); await ctx.plugin(SessionStore); await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry); await ctx.plugin(AgentRegistry); await ctx.plugin(AgentLoop, { agents: [] })
-  await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
-  await ctx.plugin(HooksCodex, { configPath, model: 'm', ...opts })
-  ctx.llm.registerAdapter(['mock'], adapter)
-  return ctx
-}
-function waitForIdle(ctx: Context, agent: ReactLoopAgent): Promise<void> {
-  return new Promise((resolve) => { const d = ctx.on('agent/status', (s, st) => { if (s === agent && st === 'idle') { d(); resolve() } }) })
-}
-function events(agent: ReactLoopAgent): SessionEvent[] { return [...agent.session.events] }
-/** Poll until `predicate` holds or the deadline passes — robust to detached
- * emit-listener hooks firing on a `.then` (a fixed sleep flakes under load). */
-async function waitFor(predicate: () => boolean, timeout = 5000, interval = 10): Promise<void> {
-  const deadline = Date.now() + timeout
-  while (!predicate()) {
-    if (Date.now() > deadline) throw new Error('waitFor: condition not met before deadline')
-    await new Promise(r => setTimeout(r, interval))
-  }
-}
-
-describe('hooks-codex coverage — decision mapping paths', () => {
-  it('UserPromptSubmit block (exit 2) → rejected turn; default reason on empty stderr', async () => {
-    const d = dir()
-    hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'b.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
-    const adapter = new MockAdapter([textResponse('no')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(adapter.requests).toHaveLength(0)
-    const te = events(agent).findLast(e => e.type === 'turn/end')
-    expect(te?.type === 'turn/end' && te.data.reason.kind).toBe('rejected')
-  })
-
-  it('UserPromptSubmit additionalContext is injected; a no-op hook proceeds', async () => {
-    const d = dir()
-    hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'c.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"ctx-x"}}\'\n') }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(JSON.stringify(adapter.requests[0]!.messages)).toContain('ctx-x')
-  })
-
-  it('a context-only UserPromptSubmit hook DELEGATES so a later listener can still block', async () => {
-    // Context alone is not a veto: the bridge delegates with `next()` and folds its context, so a
-    // downstream policy listener can still block.
-    const d = dir()
-    hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'c.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"bridge ctx"}}\'\n') }] }] })
-    const adapter = new MockAdapter([textResponse('should not run')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.on('agent/prompt-submit', async () => ({ kind: 'block' as const, reason: 'policy veto' }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(adapter.requests).toHaveLength(0)
-    expect(events(agent).some(e => e.type === 'user/message')).toBe(false)
-    const te = events(agent).findLast(e => e.type === 'turn/end')
-    expect(te?.type === 'turn/end' && te.data.reason).toMatchObject({ kind: 'rejected', reason: 'policy veto' })
-  })
-
-  it('folds the bridge additionalContext WITH a downstream listener that also adds context', async () => {
-    const d = dir()
-    hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'c.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"UserPromptSubmit","additionalContext":"from-bridge"}}\'\n') }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.on('agent/prompt-submit', async () => ({
-      kind: 'allow' as const,
-      content: [{ type: 'text' as const, text: 'rewritten-prompt' }],
-      additionalContext: { content: [{ type: 'text' as const, text: 'from-downstream' }], source: { kind: 'plugin' as const, plugin: 'policy' } },
-    }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const req = JSON.stringify(adapter.requests[0]!.messages)
-    expect(req).toContain('from-bridge')
-    expect(req).toContain('from-downstream')
-    expect(req).toContain('rewritten-prompt')
-  })
-
-  it('folds the bridge PostToolUse context onto a downstream ACCEPT that replaces content', async () => {
-    const d = dir()
-    hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pc.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"bridge-note"}}\'\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    ctx.on('tools/post-execute', async () => ({ kind: 'accept' as const, content: [{ type: 'text' as const, text: 'rewritten-result' }] }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const result = events(agent).find(e => e.type === 'tool/result')
-    expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text === 'rewritten-result')).toBe(true)
-    expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('bridge-note')))).toBe(true)
-  })
-
-  it('folds the bridge PostToolUse context onto a downstream listener BLOCK', async () => {
-    const d = dir()
-    hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pc.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"bridge-note"}}\'\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'echo', {}), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.tools.register(defineTool({ name: 'echo', description: 'e', parameters: {}, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    ctx.on('tools/post-execute', async () => ({ kind: 'block' as const, feedback: [{ type: 'text' as const, text: 'downstream-block' }] }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const result = events(agent).find(e => e.type === 'tool/result')
-    expect(result?.type === 'tool/result' && result.data.isError).toBe(true)
-    expect(result?.type === 'tool/result' && result.data.content.some(b => b.type === 'text' && b.text.includes('downstream-block'))).toBe(true)
-    expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('bridge-note')))).toBe(true)
-  })
-
-  it('SessionStart additionalContext is injected for the first request', async () => {
-    const d = dir()
-    hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: sh(d, 's.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SessionStart","additionalContext":"start-ctx"}}\'\n') }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    await waitFor(() => events(agent).some(e => e.type === 'context/message'
-      && e.data.content.some(b => b.type === 'text' && b.text.includes('start-ctx'))))
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(JSON.stringify(adapter.requests[0]!.messages)).toContain('start-ctx')
-  })
-
-  it('PostToolUse block (exit 2) → isError feedback; default reason', async () => {
-    const d = dir()
-    hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'p.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'ls' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const r = events(agent).find(e => e.type === 'tool/result')
-    expect(r?.type === 'tool/result' && r.data.isError).toBe(true)
-    expect(r?.type === 'tool/result' && r.data.content.some(b => b.type === 'text' && b.text.includes('blocked by PostToolUse hook'))).toBe(true)
-  })
-
-  it('PostToolUse additionalContext (clean exit) is attached after the result', async () => {
-    const d = dir()
-    hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pc.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"post-ctx"}}\'\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'ls' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('post-ctx')))).toBe(true)
-  })
-
-  it('PreToolUse for a tool call WITHOUT a command arg passes an empty command (commandOf non-object/missing arm)', async () => {
-    const d = dir()
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pre.sh', '#!/usr/bin/env bash\ncat >/dev/null\nexit 0\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', {}), textResponse('done')]) // no command arg
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: {}, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(ran).toBe(true) // clean-exit hook allows; commandOf returned ''
-  })
-
-  it('a clean exit-0 hook records exitCode 0 and omits stderrSummary', async () => {
-    const d = dir()
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'n.sh', '#!/usr/bin/env bash\nexit 0\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && res.data.exitCode).toBe(0)
-    expect(res?.type === 'hook/result' && 'stderrSummary' in res.data).toBe(false)
-  })
-
-  it('a long stderr is truncated in the hook/result summary', async () => {
-    const d = dir()
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'l.sh', '#!/usr/bin/env bash\nprintf "x%.0s" {1..600} >&2\nexit 2\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && res.data.stderrSummary?.endsWith('…')).toBe(true)
-    expect(res?.type === 'hook/result' && res.data.stderrSummary?.length).toBe(501) // default 500-char cap + ellipsis
-  })
-
-  it('rejects a non-positive or fractional stderrSummaryMaxChars at load', async () => {
-    const d = dir()
-    hooks(d, {})
-    for (const bad of [0, -5, 1.5, Number.NaN]) {
-      const adapter = new MockAdapter([])
-      await expect(harness(join(d, 'hooks.json'), adapter, { stderrSummaryMaxChars: bad }))
-        .rejects.toThrow(/hooks-codex: stderrSummaryMaxChars must be a positive integer/)
-    }
-  })
-
-  it('the stderr summary cap is plugin config (stderrSummaryMaxChars)', async () => {
-    const d = dir()
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'l.sh', '#!/usr/bin/env bash\nprintf "x%.0s" {1..600} >&2\nexit 2\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter, { stderrSummaryMaxChars: 40 })
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && res.data.stderrSummary).toBe('x'.repeat(40) + '…')
-  })
-
-  it('warns on a skipped async hook and a direct apply() (schema bypass) runs', async () => {
-    const d = dir()
-    const marker = join(d, 'ran')
-    hooks(d, { UserPromptSubmit: [{ hooks: [
-      { type: 'command', command: 'bg.sh', async: true }, // skipped → warn
-      { type: 'command', command: sh(d, 'h.sh', `#!/usr/bin/env bash\ntouch "${marker}"\n`) },
-    ] }] })
-    const warn = vi.fn()
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = new Context()
-    await ctx.plugin(LlmService); await ctx.plugin(SessionStore); await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry); await ctx.plugin(AgentRegistry); await ctx.plugin(AgentLoop, { agents: [] })
-    await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000 })
-    ctx.logger.warn = warn as never
-    // Direct apply (schema bypass) → the `model ?? ''` fallback is exercised.
-    HooksCodex.apply(ctx, { configPath: join(d, 'hooks.json') })
-    ctx.llm.registerAdapter(['mock'], adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(existsSync(marker)).toBe(true)
-    expect(warn).toHaveBeenCalledWith(expect.stringContaining('async hook'))
-  })
-
-  it('a no-op clean hook proceeds (contextFrom empty → next)', async () => {
-    const d = dir()
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'n.sh', '#!/usr/bin/env bash\nexit 0\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(ran).toBe(true)
-  })
-
-  it('SessionStart with no additionalContext is a no-op (contextFrom empty)', async () => {
-    const d = dir()
-    // The hook touches a marker so we can wait for it to ACTUALLY FINISH before
-    // asserting absence — a completed turn alone would not prove the detached
-    // session-start hook ran, making the absence check a false pass.
-    const marker = join(d, 'ss-ran')
-    hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: sh(d, 's.sh', `#!/usr/bin/env bash\ntouch "${marker}"\nexit 0\n`) }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    await waitFor(() => existsSync(marker)) // the clean no-output hook has finished
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(events(agent).some(e => e.type === 'context/message')).toBe(false)
-  })
-
-  it('a throwing SessionStart inject is contained (logged)', async () => {
-    const d = dir()
-    hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: sh(d, 's.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"SessionStart","additionalContext":"x"}}\'\n') }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const warn = vi.fn(); ctx.logger.warn = warn as never
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.inject = (() => { throw new Error('inject boom') })
-    await waitFor(() => warn.mock.calls.some(c => String(c[0]).includes('SessionStart hook failed')))
-    expect(warn).toHaveBeenCalledWith(expect.stringContaining('SessionStart hook failed'))
-  })
-
-  it('a clean PreToolUse with no decision allows the tool (no deny)', async () => {
-    const d = dir()
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'ok.sh', '#!/usr/bin/env bash\nexit 0\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(ran).toBe(true)
-  })
-
-  it('a non-matching regex matcher skips the hook (matchesMatcher false → continue)', async () => {
-    const d = dir()
-    // /^Edit$/ does not match the tool name "Bash" → the group is skipped.
-    hooks(d, { PreToolUse: [{ matcher: '^Edit$', hooks: [{ type: 'command', command: sh(d, 'deny.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(ran).toBe(true) // matcher didn't match → no hook ran → tool proceeded
-    expect(events(agent).some(e => e.type === 'hook/invoked')).toBe(false)
-  })
-
-  it('a {"continue":false} hook is RECORDED as "stop" but does not halt the run (TODO(hook-continue-false))', async () => {
-    // Honoring `continue:false` is deferred — the seams have no hard-halt
-    // primitive. Assert the LOG records the halt request AND that the run is not
-    // actually halted (the tool still runs, the turn completes).
-    const d = dir()
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 's.sh', '#!/usr/bin/env bash\necho \'{"continue":false,"stopReason":"halt"}\'\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && res.data.decision).toBe('stop') // recorded
-    expect(ran).toBe(true) // NOT honored: the tool still ran (halt is deferred)
-  })
-
-  it('PreToolUse deny with EMPTY stderr uses the default reason (?? right arm)', async () => {
-    const d = dir()
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'd.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const r = events(agent).find(e => e.type === 'tool/result')
-    expect(r?.type === 'tool/result' && r.data.content.some(b => b.type === 'text' && b.text.includes('blocked by PreToolUse hook'))).toBe(true)
-  })
-
-  it('PostToolUse block AND additionalContext are surfaced together', async () => {
-    const d = dir()
-    hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'bc.sh', '#!/usr/bin/env bash\necho \'{"decision":"block","reason":"bad","hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"ctx too"}}\'\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const r = events(agent).find(e => e.type === 'tool/result')
-    expect(r?.type === 'tool/result' && r.data.isError).toBe(true)
-    expect(r?.type === 'tool/result' && r.data.content.some(b => b.type === 'text' && b.text.includes('bad'))).toBe(true)
-    expect(events(agent).some(e => e.type === 'context/message' && e.data.content.some(b => b.type === 'text' && b.text.includes('ctx too')))).toBe(true)
-  })
-
-  it('commandOf reads a non-string command arg as an empty command', async () => {
-    const d = dir()
-    // The tool-call arguments carry `command` as a NUMBER → commandOf's
-    // `typeof command === 'string'` false arm → '' (the payload's tool_input.command).
-    const cap = join(d, 'payload')
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'cap.sh', `#!/usr/bin/env bash\ncat > "${cap}"\nexit 0\n`) }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 7 }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'number' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const payload = JSON.parse(readFileSync(cap, 'utf8')) as { tool_input: { command: string } }
-    expect(payload.tool_input.command).toBe('')
-  })
-
-  it('a no-agent direct PreToolUse run uses process.cwd() and turn 0 (no session to record)', async () => {
-    const d = dir()
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'd.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
-    const ctx = await harness(join(d, 'hooks.json'), new MockAdapter([]))
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'x' }] } }))
-    const { CallId } = await import('@deepseek-ai/dsh-llm')
-    const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'Bash', arguments: { command: 'x' } })
-    expect(ran).toBe(false) // denied
-    expect(result.isError).toBe(true)
-  })
-
-  it('a no-agent direct PostToolUse run attaches context with no session to record', async () => {
-    const d = dir()
-    hooks(d, { PostToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'pc.sh', '#!/usr/bin/env bash\necho \'{"hookSpecificOutput":{"hookEventName":"PostToolUse","additionalContext":"x"}}\'\n') }] }] })
-    const ctx = await harness(join(d, 'hooks.json'), new MockAdapter([]))
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const { CallId } = await import('@deepseek-ai/dsh-llm')
-    const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'Bash', arguments: { command: 'x' } })
-    expect(result.isError).toBeFalsy()
-    expect(result.additionalContext?.content.some(b => b.type === 'text' && b.text === 'x')).toBe(true)
-  })
-
-  it('when the bash executor REJECTS, the hook/result omits exitCode (non-blocking)', async () => {
-    const d = dir()
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'h.sh', '#!/usr/bin/env bash\nexit 0\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.bash.run = (() => Promise.reject(new Error('executor down')))
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const res = events(agent).find(e => e.type === 'hook/result')
-    expect(res?.type === 'hook/result' && 'exitCode' in res.data).toBe(false)
-  })
-
-  it('a blocking Stop hook with EMPTY stderr still forces continuation (no reason required)', async () => {
-    // Regression: an exit-2 Stop hook with no stderr yields decision 'deny' +
-    // reason undefined; the turn must STILL force-continue, not silently stop.
-    const d = dir()
-    const marker = join(d, 'fired')
-    hooks(d, { Stop: [{ hooks: [{ type: 'command', command: sh(d, 's.sh', `#!/usr/bin/env bash\nif [ -e "${marker}" ]; then exit 0; fi\ntouch "${marker}"\nexit 2\n`) }] }] })
-    const adapter = new MockAdapter([textResponse('one'), textResponse('two')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(adapter.requests).toHaveLength(2) // empty-reason block forced continuation
-    expect(JSON.stringify(adapter.requests[1]!.messages)).toContain('blocked by Stop hook')
-  })
-
-  it('a clean UserPromptSubmit hook that prints PLAIN stdout injects it as context', async () => {
-    // Codex feeds a SessionStart/UserPromptSubmit hook's PLAIN (non-JSON) stdout
-    // as additionalContext (unlike CC, which needs a JSON hookSpecificOutput).
-    const d = dir()
-    hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'ctx.sh', '#!/usr/bin/env bash\necho "extra guidance from a plain hook"\nexit 0\n') }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(JSON.stringify(adapter.requests[0]!.messages)).toContain('extra guidance from a plain hook')
-  })
-
-  it('a NON-clean SessionStart hook (exit 2) does NOT inject its stdout as context', async () => {
-    // SessionStart cannot block, but non-clean stdout still must not become context. The marker
-    // waits for detached completion; `echo stale; exit 2` then proves the exit-code gate matches
-    // the codec's structured-stdout rule.
-    const d = dir()
-    const marker = join(d, 'ran')
-    hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: sh(d, 'b.sh', `#!/usr/bin/env bash\ntouch "${marker}"\necho "stale"\nexit 2\n`) }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    await waitFor(() => existsSync(marker)) // the exit-2 hook has finished
-    expect(events(agent).some(e => e.type === 'context/message'
-      && e.data.content.some(b => b.type === 'text' && b.text.includes('stale')))).toBe(false)
-  })
-
-  it('a UserPromptSubmit hook with a non-blocking error exit (1) + stdout does NOT inject it', async () => {
-    // Exit 1 is a non-blocking error (no decision), so the prompt is NOT blocked
-    // and the handler falls through to the context path — the gate must still
-    // suppress the error hook's stdout ("stale" never reaches the model).
-    const d = dir()
-    hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'e.sh', '#!/usr/bin/env bash\necho "stale"\nexit 1\n') }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(adapter.requests).toHaveLength(1) // exit 1 is non-blocking → the turn ran
-    expect(JSON.stringify(adapter.requests[0]!.messages)).not.toContain('stale')
-  })
-
-  it('a clean SessionStart hook that prints PLAIN stdout injects it (not JSON)', async () => {
-    const d = dir()
-    hooks(d, { SessionStart: [{ hooks: [{ type: 'command', command: sh(d, 'ss.sh', '#!/usr/bin/env bash\necho "session preamble"\nexit 0\n') }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    await waitFor(() => events(agent).some(e => e.type === 'context/message'
-      && e.data.content.some(b => b.type === 'text' && b.text.includes('session preamble'))))
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(JSON.stringify(adapter.requests[0]!.messages)).toContain('session preamble')
-  })
-
-  it('a clean hook that prints JSON is NOT injected as prose (plain-stdout gate)', async () => {
-    // A structured (JSON) stdout must go through the hookSpecificOutput path, not
-    // be dumped verbatim as context — the `!startsWith('{')` gate guards this.
-    const d = dir()
-    hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'j.sh', '#!/usr/bin/env bash\necho \'{"unrelated":"json"}\'\nexit 0\n') }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(JSON.stringify(adapter.requests[0]!.messages)).not.toContain('unrelated')
-  })
-
-  it('the PreToolUse payload carries the REAL tool name (matches the matcher subject)', async () => {
-    // Regression: the payload once hardcoded tool_name "Bash", disagreeing with
-    // the exec.name matcher subject — a config matcher on the real name would
-    // then never fire. Capture the payload and assert tool_name === the real name.
-    const d = dir()
-    const cap = join(d, 'payload')
-    hooks(d, { PreToolUse: [{ hooks: [{ type: 'command', command: sh(d, 'cap.sh', `#!/usr/bin/env bash\ncat > "${cap}"\nexit 0\n`) }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'shell', { command: 'ls' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    ctx.tools.register(defineTool({ name: 'shell', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    const payload = JSON.parse(readFileSync(cap, 'utf8')) as { tool_name: string; tool_input: { command: string } }
-    expect(payload.tool_name).toBe('shell')
-    expect(payload.tool_input.command).toBe('ls')
-  })
-
-  it('a Codex matcher on the REAL tool name fires (matcher subject === payload tool_name)', async () => {
-    // A regex matcher matching the real tool name must select the hook — proving
-    // the matcher subject and the payload tool_name agree.
-    const d = dir()
-    hooks(d, { PreToolUse: [{ matcher: 'shell', hooks: [{ type: 'command', command: sh(d, 'd.sh', '#!/usr/bin/env bash\nexit 2\n') }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'shell', { command: 'ls' }), textResponse('done')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    let ran = false
-    ctx.tools.register(defineTool({ name: 'shell', description: 'b', parameters: { command: { type: 'string' } }, async execute() { ran = true; return [{ type: 'text', text: 'ok' }] } }))
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(ran).toBe(false) // the matcher fired → the hook denied the tool
-    expect(events(agent).some(e => e.type === 'hook/invoked' && e.data.point === 'PreToolUse')).toBe(true)
-  })
-
-  it('a hook emitting a systemMessage is warned as not-yet-surfaced', async () => {
-    const d = dir()
-    hooks(d, { UserPromptSubmit: [{ hooks: [{ type: 'command', command: sh(d, 'sm.sh', '#!/usr/bin/env bash\necho \'{"systemMessage":"heads up"}\'\n') }] }] })
-    const adapter = new MockAdapter([textResponse('ok')])
-    const ctx = await harness(join(d, 'hooks.json'), adapter)
-    const warn = vi.fn(); ctx.logger.warn = warn as never
-    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
-    agent.send([{ type: 'text', text: 'go' }]); await waitForIdle(ctx, agent)
-    expect(warn).toHaveBeenCalledWith(expect.stringContaining('systemMessage'))
-    expect(JSON.stringify(adapter.requests[0]!.messages)).not.toContain('heads up')
-  })
-
-  it('runs an agent-scoped hook in the session cwd, not the executor default', async () => {
-    // Same regression as the CC bridge: the Codex bridge must thread the session
-    // cwd as the hook workdir. Executor default = serverDir; session cwd =
-    // sessionDir; the PreToolUse hook's `pwd` marker must land in sessionDir.
-    const serverDir = dir()
-    const sessionDir = dir()
-    const marker = join(sessionDir, 'where')
-    hooks(serverDir, { PreToolUse: [{ hooks: [{ type: 'command', command: 'pwd > where' }] }] })
-    const adapter = new MockAdapter([toolCallResponse('c1', 'Bash', { command: 'x' }), textResponse('done')])
-    const ctx = new Context()
-    await ctx.plugin(LlmService); await ctx.plugin(SessionStore); await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry); await ctx.plugin(AgentRegistry); await ctx.plugin(AgentLoop, { agents: [] })
-    await ctx.plugin(LocalBashExecutor, { timeoutMs: 10_000, cwd: serverDir })
-    await ctx.plugin(HooksCodex, { configPath: join(serverDir, 'hooks.json'), model: 'm' })
-    ctx.llm.registerAdapter(['mock'], adapter)
-    ctx.tools.register(defineTool({ name: 'Bash', description: 'b', parameters: { command: { type: 'string' } }, async execute() { return [{ type: 'text', text: 'ok' }] } }))
-    const { SessionId } = await import('@deepseek-ai/dsh-session')
-    const handle = await ctx.agents.create({ agentId: AgentId('a1'), sessionId: SessionId('s1'), meta: { cwd: sessionDir }, agentOptions: { model: 'mock' } })
-    handle.agent.send([{ type: 'text', text: 'go' }])
-    await waitForIdle(ctx, handle.agent as ReactLoopAgent)
-    expect(existsSync(marker)).toBe(true)
-    expect(readFileSync(marker, 'utf8').trim().endsWith(sessionDir.split('/').pop()!)).toBe(true)
-    await handle.dispose()
-  })
-})
diff --git a/packages/hooks/hooks-codex/tsconfig.json b/packages/hooks/hooks-codex/tsconfig.json
index f936b500aa..ae3c91e9dd 100644
--- a/packages/hooks/hooks-codex/tsconfig.json
+++ b/packages/hooks/hooks-codex/tsconfig.json
@@ -29,6 +29,9 @@
     {
       "path": "../../core/session"
     },
+    {
+      "path": "../../session-persistence/session-persistence"
+    },
     {
       "path": "../../llm/llm"
     },
diff --git a/packages/llm/README.md b/packages/llm/README.md
index 3fc5c9cf9e..4a5e6769e2 100644
--- a/packages/llm/README.md
+++ b/packages/llm/README.md
@@ -5,7 +5,8 @@ The LLM seam and its provider adapters. The interface package (`llm`) owns the a
 | Package | Role | ctx key |
 |---|---|---|
 | `llm/` | Abstract LLM service + content-block vocabulary + chunk assembler | `ctx.llm` |
+| `token-meter/` | Replay-aware request and surface token measurement | `ctx.tokenMeter` |
 | `llm-deepseek/` | DeepSeek API adapter (hand-rolled fetch/SSE) | (registers on `ctx.llm`) |
-| `llm-pi-ai/` | DeepSeek adapter via `@earendil-works/pi-ai` (design twin) | (registers on `ctx.llm`) |
+| `llm-pi-ai/` | Multi-provider adapter via `@earendil-works/pi-ai` | (registers on `ctx.llm`) |
 
-The interface lives at `llm/llm/`; adapters are flat siblings under the group. A new provider adapter joins here and registers on `ctx.llm` without touching the interface. See [twin LLM adapters](../../docs/rfc/implemented/architecture/2026-06-13-twin-llm-adapters.md) for why two adapters exist.
+The interface lives at `llm/llm/`; adapters and the reusable token meter are flat siblings under the group. Requests route by `provider`, while `model` is passed through to the selected adapter. A new provider adapter joins here and registers one or more provider routes on `ctx.llm` without touching the interface. See [twin LLM adapters](../../docs/rfc/implemented/architecture/2026-06-13-twin-llm-adapters.md) for the contract-validation origin of the two shipping implementations and the [replay token meter RFC](../../docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.md) for measurement ownership.
diff --git a/packages/llm/llm-deepseek/README.md b/packages/llm/llm-deepseek/README.md
index 1270ad40a5..7da98cdebf 100644
--- a/packages/llm/llm-deepseek/README.md
+++ b/packages/llm/llm-deepseek/README.md
@@ -2,7 +2,7 @@
 
 DeepSeek chat-completions adapter for the harness LLM seam: hand-rolled `fetch` + SSE translation from the official wire format (source of truth: the API docs — guides/thinking_mode, guides/tool_calls, api/create-chat-completion) into the `StreamChunk` protocol.
 
-A second, independent implementation of the same seam exists in `@deepseek-ai/dsh-llm-pi-ai` (library-backed). Same Config shape — pick one per context (registering both for the same model names throws by design).
+A second, library-backed implementation of the same seam exists in `@deepseek-ai/dsh-llm-pi-ai`. This package always owns the `deepseek` provider route; mounting a pi-ai profile with `provider: deepseek` in the same context throws `LlmError('DUPLICATE_ADAPTER')` by design.
 
 ## Config
 
@@ -12,12 +12,16 @@ A second, independent implementation of the same seam exists in `@deepseek-ai/ds
   config:
     apiKey: !!js process.env.DEEPSEEK_API_KEY    # or rely on the env fallback
     baseURL: !!js process.env.DEEPSEEK_BASE_URL  # default: https://api.deepseek.com
-    models: [deepseek-v4-flash, deepseek-v4-pro] # one adapter, registered for each name
     thinking: enabled        # optional; provider default is enabled
     reasoningEffort: high    # optional; high | max — omitted ⇒ not sent
+    models:                  # optional; defaults to V4 Flash and V4 Pro
+      - id: deepseek-v4-flash
+        name: DeepSeek V4 Flash
+      - id: private-reasoner
+        description: Company-hosted reasoning model
 ```
 
-`models` lists every model name this one adapter instance serves: the adapter registers itself for each (the harness model name IS the wire `model` string), so a `generate`/`stream` call routes to it whenever `options.model` is any of them. Registering a second adapter for a name already taken throws `LlmError('DUPLICATE_ADAPTER')` (the LLM service enforces one adapter per model, all-or-nothing).
+The plugin registers the single provider route `deepseek`. A request selects it with `provider: deepseek`; its `model` is passed through as the wire `model` string, so changing DeepSeek models does not require lifecycle-time registration. Omitting `models` advertises `deepseek-v4-flash` and `deepseek-v4-pro`; an explicit list replaces those defaults, while `models: []` advertises none. Catalog entries are exposed through `ctx.llm.listModels('deepseek')` for clients such as ACP editors, but remain advisory: unlisted model ids still pass through unchanged. An omitted entry name defaults to its id. Registering another adapter for `deepseek` throws `LlmError('DUPLICATE_ADAPTER')`.
 
 `reasoningEffort` is **omitted by default** — when unset, the `reasoning_effort` wire field is not sent and the server applies its own default for the model. The only accepted values are `high` and `max` (DeepSeek's official effort levels). It is meaningful only with thinking enabled (the provider default).
 
diff --git a/packages/llm/llm-deepseek/src/adapter.ts b/packages/llm/llm-deepseek/src/adapter.ts
index 30760a8fbc..ece053c6b3 100644
--- a/packages/llm/llm-deepseek/src/adapter.ts
+++ b/packages/llm/llm-deepseek/src/adapter.ts
@@ -6,13 +6,23 @@
  */
 
 import { attributionHeaders, LlmAdapter, LlmError } from '@deepseek-ai/dsh-llm'
-import type { GenerateOptions, StreamChunk } from '@deepseek-ai/dsh-llm'
+import type { GenerateOptions, LlmModelInfo, LlmProviderInfo, StreamChunk } from '@deepseek-ai/dsh-llm'
 import { serializeRequest } from './serialize.ts'
 import type { RequestDefaults } from './serialize.ts'
 import { parseSse } from './sse.ts'
 import { translate } from './translate.ts'
 import type { WireError } from './types.ts'
 
+/** One optional model entry advertised by the hand-written adapter. */
+export interface DeepSeekCatalogModel {
+  /** Wire model id accepted by the configured endpoint. */
+  id: string
+  /** Selector label; defaults to {@link id}. */
+  name?: string
+  /** Optional selector detail for deployments with similar model variants. */
+  description?: string
+}
+
 /** Constructor options for {@link DeepSeekAdapter}; the plugin's `apply` resolves them from Config + environment. */
 export interface DeepSeekAdapterOptions {
   /** Bearer token sent in the `authorization` header on every request. */
@@ -21,6 +31,8 @@ export interface DeepSeekAdapterOptions {
   baseURL: string
   /** Request defaults applied to every call (thinking mode, effort). */
   defaults?: RequestDefaults
+  /** Advisory models exposed to discovery consumers; requests remain unrestricted. */
+  models?: readonly DeepSeekCatalogModel[]
 }
 
 /**
@@ -49,6 +61,19 @@ export class DeepSeekAdapter extends LlmAdapter {
     super()
   }
 
+  override providerInfo(provider: string): LlmProviderInfo {
+    return { id: provider, name: 'DeepSeek' }
+  }
+
+  override listModels(provider: string): Promise<readonly LlmModelInfo[]> {
+    return Promise.resolve((this.options.models ?? []).map(model => ({
+      provider,
+      id: model.id,
+      name: model.name ?? model.id,
+      ...model.description === undefined ? {} : { description: model.description },
+    })))
+  }
+
   async * stream(options: GenerateOptions): AsyncIterable<StreamChunk> {
     const body = serializeRequest(options, this.options.defaults ?? {})
 
diff --git a/packages/llm/llm-deepseek/src/index.ts b/packages/llm/llm-deepseek/src/index.ts
index b816e2e7cf..695831b1bd 100644
--- a/packages/llm/llm-deepseek/src/index.ts
+++ b/packages/llm/llm-deepseek/src/index.ts
@@ -1,5 +1,5 @@
 /**
- * Register a {@link DeepSeekAdapter} for configured model names on `ctx.llm`. Configuration uses
+ * Register a {@link DeepSeekAdapter} for the `deepseek` provider route on `ctx.llm`. Configuration uses
  * Cordis schemastery; pass secrets from environment variables through `cordis.yml` with `!!js`,
  * as shown in the package README, rather than reading ad hoc files.
  * @module @deepseek-ai/dsh-llm-deepseek
@@ -9,9 +9,10 @@ import type { Context } from 'cordis'
 import z from 'schemastery'
 import type {} from '@deepseek-ai/dsh-llm'
 import { DeepSeekAdapter } from './adapter.ts'
+import type { DeepSeekCatalogModel } from './adapter.ts'
 
 export { DeepSeekAdapter, httpErrorCode } from './adapter.ts'
-export type { DeepSeekAdapterOptions } from './adapter.ts'
+export type { DeepSeekAdapterOptions, DeepSeekCatalogModel } from './adapter.ts'
 export { serializeMessages, serializeRequest } from './serialize.ts'
 export type { RequestDefaults } from './serialize.ts'
 export { DONE, parseSse } from './sse.ts'
@@ -21,6 +22,11 @@ export type * from './types.ts'
 export const name = 'llm-deepseek'
 export const inject = ['llm']
 
+const DEFAULT_MODELS: DeepSeekCatalogModel[] = [
+  { id: 'deepseek-v4-flash' },
+  { id: 'deepseek-v4-pro' },
+]
+
 /**
  * Plugin config, validated by the same-named schemastery schema. Every field
  * is optional in yml: credentials/endpoint fall back to the environment (a
@@ -32,40 +38,62 @@ export interface Config {
   apiKey?: string
   /** Endpoint base; falls back to $DEEPSEEK_BASE_URL, then the public API. */
   baseURL?: string
-  /** Model names to register (sent verbatim on the wire). */
-  models?: string[]
   /** Thinking-mode default for every request (provider default: enabled). */
   thinking?: 'enabled' | 'disabled'
   /** Thinking effort (only meaningful with thinking enabled). */
   reasoningEffort?: 'high' | 'max'
+  /** Advisory models shown by discovery consumers; defaults to V4 Flash and V4 Pro. */
+  models?: DeepSeekCatalogModel[]
 }
 
+const catalogModel: z<DeepSeekCatalogModel> = z.object({
+  id: z.string().required(),
+  name: z.string(),
+  description: z.string(),
+})
+
 export const Config: z<Config> = z.object({
   apiKey: z.string(),
   baseURL: z.string(),
-  models: z.array(z.string()).default(['deepseek-v4-flash', 'deepseek-v4-pro']),
   thinking: z.union(['enabled', 'disabled']),
   reasoningEffort: z.union(['high', 'max']),
+  models: z.array(catalogModel).default(DEFAULT_MODELS),
 })
 
 /** Public API default; the internal endpoint comes from $DEEPSEEK_BASE_URL. */
 export const PUBLIC_BASE_URL = 'https://api.deepseek.com'
 
+/** Resolve, validate, and detach the advisory model catalog. */
+function resolveModels(models: readonly DeepSeekCatalogModel[] | undefined): DeepSeekCatalogModel[] {
+  const seen = new Set<string>()
+  return (models ?? DEFAULT_MODELS).map((model) => {
+    if (model.id.length === 0) throw new Error('llm-deepseek: catalog model ids must be non-empty')
+    if (model.name !== undefined && model.name.length === 0) {
+      throw new Error(`llm-deepseek: catalog model "${model.id}" has an empty name`)
+    }
+    if (seen.has(model.id)) throw new Error(`llm-deepseek: duplicate catalog model "${model.id}"`)
+    seen.add(model.id)
+    return {
+      id: model.id,
+      ...model.name === undefined ? {} : { name: model.name },
+      ...model.description === undefined ? {} : { description: model.description },
+    }
+  })
+}
+
 export function apply(ctx: Context, config: Config): void {
   const apiKey = config.apiKey ?? process.env.DEEPSEEK_API_KEY
   if (apiKey === undefined || apiKey.length === 0) {
     throw new Error('llm-deepseek: an API key is required (Config.apiKey or $DEEPSEEK_API_KEY)')
   }
   const baseURL = config.baseURL ?? process.env.DEEPSEEK_BASE_URL ?? PUBLIC_BASE_URL
-  // schemastery's .default() guarantees models is set after validation.
-  const models = config.models as string[]
-
-  ctx.llm.registerAdapter(models, new DeepSeekAdapter({
+  ctx.llm.registerAdapter(['deepseek'], new DeepSeekAdapter({
     apiKey,
     baseURL,
     defaults: {
       thinking: config.thinking,
       reasoningEffort: config.reasoningEffort,
     },
+    models: resolveModels(config.models),
   }))
 }
diff --git a/packages/llm/llm-deepseek/tests/adapter.e2e.ts b/packages/llm/llm-deepseek/tests/adapter.e2e.ts
index b01b498dff..e02476eaef 100644
--- a/packages/llm/llm-deepseek/tests/adapter.e2e.ts
+++ b/packages/llm/llm-deepseek/tests/adapter.e2e.ts
@@ -16,11 +16,11 @@ const FLASH = 'deepseek-v4-flash'
 const PRO = 'deepseek-v4-pro'
 const contexts: Context[] = []
 
-async function harness(model: string, config: Partial<Config> = {}) {
+async function harness(_model: string, config: Partial<Config> = {}) {
   const ctx = new Context()
   contexts.push(ctx)
   await ctx.plugin(LlmService)
-  await ctx.plugin(LlmDeepSeek, { models: [model], ...config })
+  await ctx.plugin(LlmDeepSeek, config)
   return ctx
 }
 
@@ -134,6 +134,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('llm-deepseek e2e (real API)', ()
     const ctx = await harness(FLASH, { thinking: 'disabled' })
     const kinds: string[] = []
     for await (const chunk of ctx.llm.stream({
+      provider: 'deepseek',
       model: FLASH,
       messages: ask('Count from 1 to 5, digits only.'),
       maxTokens: 50,
diff --git a/packages/llm/llm-deepseek/tests/adapter.spec.ts b/packages/llm/llm-deepseek/tests/adapter.spec.ts
index 46f123a1c7..c0a2f8f482 100644
--- a/packages/llm/llm-deepseek/tests/adapter.spec.ts
+++ b/packages/llm/llm-deepseek/tests/adapter.spec.ts
@@ -86,7 +86,7 @@ const textEvents = [
 async function harness(baseURL: string, config: object = {}) {
   const ctx = new Context()
   await ctx.plugin(LlmService)
-  await ctx.plugin(LlmDeepSeek, { apiKey: 'test-key', baseURL, models: ['deepseek-v4-flash'], ...config })
+  await ctx.plugin(LlmDeepSeek, { apiKey: 'test-key', baseURL, ...config })
   return ctx
 }
 
@@ -123,6 +123,7 @@ describe('DeepSeekAdapter against a mock server', () => {
 
     const kinds: string[] = []
     for await (const chunk of ctx.llm.stream({
+      provider: 'deepseek',
       model: 'deepseek-v4-flash',
       messages: [{ role: 'user', content: [{ type: 'text', text: 'hi' }] }],
     })) {
@@ -198,7 +199,7 @@ describe('DeepSeekAdapter against a mock server', () => {
     )
     try {
       const iterate = async (): Promise<void> => {
-        for await (const _chunk of adapter.stream({ model: 'm', messages: [] })) { /* drain */ }
+        for await (const _chunk of adapter.stream({ provider: 'deepseek', model: 'm', messages: [] })) { /* drain */ }
       }
       await expect(iterate()).rejects.toThrow(/no response body/)
     } finally {
@@ -224,6 +225,7 @@ describe('DeepSeekAdapter against a mock server', () => {
     const pending = (async () => {
       const chunks = []
       for await (const chunk of ctx.llm.stream({
+        provider: 'deepseek',
         model: 'deepseek-v4-flash',
         messages: [],
         signal: controller.signal,
@@ -239,25 +241,81 @@ describe('DeepSeekAdapter against a mock server', () => {
 })
 
 describe('plugin registration and config', () => {
-  it('registers the configured models and unregisters on dispose (HMR safety)', async () => {
+  it('registers the deepseek provider and unregisters on dispose (HMR safety)', async () => {
     const server = await mockServer([])
     const ctx = new Context()
     await ctx.plugin(LlmService)
     const fiber = await ctx.plugin(LlmDeepSeek, {
       apiKey: 'k',
       baseURL: server.url,
-      models: ['deepseek-v4-flash', 'deepseek-v4-pro'],
     })
-    expect(ctx.llm.models().sort()).toEqual(['deepseek-v4-flash', 'deepseek-v4-pro'])
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'deepseek', name: 'DeepSeek' }])
     await fiber.dispose()
-    expect(ctx.llm.models()).toEqual([])
+    expect(ctx.llm.listProviders()).toEqual([])
   })
 
-  it('defaults the model list', async () => {
+  it('owns the deepseek provider and advertises the default models', async () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
     await ctx.plugin(LlmDeepSeek, { apiKey: 'k', baseURL: 'http://127.0.0.1:1' })
-    expect(ctx.llm.models().sort()).toEqual(['deepseek-v4-flash', 'deepseek-v4-pro'])
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'deepseek', name: 'DeepSeek' }])
+    await expect(ctx.llm.listModels('deepseek')).resolves.toEqual([
+      { provider: 'deepseek', id: 'deepseek-v4-flash', name: 'deepseek-v4-flash' },
+      { provider: 'deepseek', id: 'deepseek-v4-pro', name: 'deepseek-v4-pro' },
+    ])
+  })
+
+  it('uses the default model catalog when apply is called directly', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    LlmDeepSeek.apply(ctx, { apiKey: 'k', baseURL: 'http://127.0.0.1:1' })
+    await expect(ctx.llm.listModels('deepseek')).resolves.toEqual([
+      { provider: 'deepseek', id: 'deepseek-v4-flash', name: 'deepseek-v4-flash' },
+      { provider: 'deepseek', id: 'deepseek-v4-pro', name: 'deepseek-v4-pro' },
+    ])
+  })
+
+  it('advertises configured models without restricting arbitrary request ids', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    await ctx.plugin(LlmDeepSeek, {
+      apiKey: 'k',
+      baseURL: 'http://127.0.0.1:1',
+      models: [
+        { id: 'private-fast' },
+        { id: 'private-reasoner', name: 'Private Reasoner', description: 'Higher reasoning budget' },
+      ],
+    })
+    await expect(ctx.llm.listModels('deepseek')).resolves.toEqual([
+      { provider: 'deepseek', id: 'private-fast', name: 'private-fast' },
+      { provider: 'deepseek', id: 'private-reasoner', name: 'Private Reasoner', description: 'Higher reasoning budget' },
+    ])
+  })
+
+  it('allows an explicit empty model catalog', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    await ctx.plugin(LlmDeepSeek, {
+      apiKey: 'k',
+      baseURL: 'http://127.0.0.1:1',
+      models: [],
+    })
+    await expect(ctx.llm.listModels('deepseek')).resolves.toEqual([])
+  })
+
+  it.each([
+    [[{ id: '' }], /ids must be non-empty/],
+    [[{ id: 'm', name: '' }], /empty name/],
+    [[{ id: 'm' }, { id: 'm' }], /duplicate catalog model/],
+  ] as const)('rejects invalid advisory model config', async (models, message) => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    await expect(ctx.plugin(LlmDeepSeek, {
+      apiKey: 'k',
+      baseURL: 'http://127.0.0.1:1',
+      models: [...models],
+    })).rejects.toThrow(message)
+    expect(ctx.llm.listProviders()).toEqual([])
   })
 
   it('falls back to DEEPSEEK_API_KEY and DEEPSEEK_BASE_URL env vars', async () => {
@@ -266,7 +324,7 @@ describe('plugin registration and config', () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
     await ctx.plugin(LlmDeepSeek, {})
-    expect(ctx.llm.models().length).toBeGreaterThan(0)
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'deepseek', name: 'DeepSeek' }])
   })
 
   it('throws a clear error when no API key is available', async () => {
@@ -275,7 +333,7 @@ describe('plugin registration and config', () => {
     await ctx.plugin(LlmService)
     await expect(ctx.plugin(LlmDeepSeek, {}))
       .rejects.toThrow(/an API key is required/)
-    expect(ctx.llm.models()).toEqual([])
+    expect(ctx.llm.listProviders()).toEqual([])
   })
 
   it('prefers explicit config over env for key and base URL', async () => {
@@ -292,7 +350,7 @@ describe('plugin registration and config', () => {
     vi.stubEnv('DEEPSEEK_BASE_URL', server.url)
     const ctx = new Context()
     await ctx.plugin(LlmService)
-    await ctx.plugin(LlmDeepSeek, { apiKey: 'k', models: ['deepseek-v4-flash'] })
+    await ctx.plugin(LlmDeepSeek, { apiKey: 'k' })
     await assemble(ctx,{ model: 'deepseek-v4-flash', messages: [] })
     expect(server.requests).toHaveLength(1)
   })
@@ -304,11 +362,12 @@ describe('plugin registration and config', () => {
     await ctx.plugin(LlmService)
     // Registration succeeds; no call is made (would hit api.deepseek.com).
     await ctx.plugin(LlmDeepSeek, {})
-    expect(ctx.llm.models().length).toBeGreaterThan(0)
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'deepseek', name: 'DeepSeek' }])
   })
 
-  it('adapter is constructible directly for embedding', () => {
+  it('adapter is constructible directly for embedding', async () => {
     const adapter = new DeepSeekAdapter({ apiKey: 'k', baseURL: 'http://127.0.0.1:1' })
     expect(adapter).toBeInstanceOf(DeepSeekAdapter)
+    await expect(adapter.listModels('deepseek')).resolves.toEqual([])
   })
 })
diff --git a/packages/llm/llm-deepseek/tests/assemble.ts b/packages/llm/llm-deepseek/tests/assemble.ts
index b0182615e0..494eeac494 100644
--- a/packages/llm/llm-deepseek/tests/assemble.ts
+++ b/packages/llm/llm-deepseek/tests/assemble.ts
@@ -15,11 +15,19 @@ export interface AssembledResult {
   finish: FinishReason
 }
 
-export async function assemble(ctx: Context, options: GenerateOptions): Promise<AssembledResult> {
+export async function assemble(ctx: Context, options: Omit<GenerateOptions, 'provider'> & { provider?: string }): Promise<AssembledResult> {
   const assembler = new BlockAssembler()
-  for await (const chunk of ctx.llm.stream(options)) assembler.push(chunk)
+  const request = { provider: 'deepseek', ...options }
+  for await (const chunk of ctx.llm.stream(request)) assembler.push(chunk)
   return {
-    message: assembler.message(),
+    message: {
+      ...assembler.message(),
+      provenance: {
+        provider: request.provider,
+        model: request.model,
+        ...assembler.replayState === undefined ? {} : { replayState: assembler.replayState },
+      },
+    },
     ...assembler.usage !== undefined ? { usage: assembler.usage } : {},
     finish: assembler.finish,
   }
diff --git a/packages/llm/llm-deepseek/tests/serialize.spec.ts b/packages/llm/llm-deepseek/tests/serialize.spec.ts
index 5944f8d30a..0a5d10e23b 100644
--- a/packages/llm/llm-deepseek/tests/serialize.spec.ts
+++ b/packages/llm/llm-deepseek/tests/serialize.spec.ts
@@ -4,7 +4,7 @@ import type { ContentBlock, GenerateOptions, Message } from '@deepseek-ai/dsh-ll
 import { serializeMessages, serializeRequest } from '@deepseek-ai/dsh-llm-deepseek'
 
 function request(overrides: Partial<GenerateOptions> = {}): GenerateOptions {
-  return { model: 'deepseek-v4-flash', messages: [], ...overrides }
+  return { provider: 'deepseek', model: 'deepseek-v4-flash', messages: [], ...overrides }
 }
 
 describe('serializeMessages', () => {
diff --git a/packages/llm/llm-pi-ai/README.md b/packages/llm/llm-pi-ai/README.md
index dd1cb5b48c..bb63d82010 100644
--- a/packages/llm/llm-pi-ai/README.md
+++ b/packages/llm/llm-pi-ai/README.md
@@ -1,60 +1,79 @@
 # @deepseek-ai/dsh-llm-pi-ai
 
-DeepSeek adapter for the harness LLM seam backed by [`@earendil-works/pi-ai`](https://www.npmjs.com/package/@earendil-works/pi-ai) (the LLM library behind the pi agent).
-
-## Why a second adapter exists
-
-`@deepseek-ai/dsh-llm-deepseek` already talks to the same endpoint. This package is its **design-verification twin**: same models, same wire protocol, completely different internals — a unified LLM library with its own event vocabulary versus hand-rolled fetch/SSE. Anything the harness `StreamChunk` protocol cannot express for BOTH implementations is a core-vocabulary bug. The differences it exercised on purpose:
-
-- pi-ai hands tool-call `arguments` around as **parsed objects**; the harness keeps raw JSON strings. The adapter patches replay payloads back to the original raw strings before sending them, and re-stringifies parsed output tool calls at `block-end`.
-- pi-ai reports failures as **in-stream error events** (it never throws mid-stream); these map to `finish {kind:'error'|'aborted'}` chunks — the protocol's other sanctioned error path besides throwing (which llm-deepseek uses).
-- pi-ai folds reasoning tokens into `usage.output`; there is no separate reasoning count to map.
-- pi-ai's options omit some DeepSeek/OpenAI-compatible details; the adapter uses its `onPayload` hook to preserve the harness contract (`stop`, scrubbing pi-ai's own per-tool `strict` default — the hand-rolled twin sends no such field — omitted reasoning effort, raw replayed tool arguments).
+Generic multi-provider adapter for the harness LLM seam backed by [`@earendil-works/pi-ai`](https://www.npmjs.com/package/@earendil-works/pi-ai). One plugin instance owns an explicit list of provider profiles; every request selects a profile with `GenerateOptions.provider` and resolves `GenerateOptions.model` dynamically from pi-ai's installed catalog.
 
 ## Config
 
-Same shape as llm-deepseek (one-line swap in cordis.yml), with pi-ai's thinking-level vocabulary:
+Configure credentials and deployment-specific transport settings per provider. Omitting `apiKey` delegates authentication to pi-ai's provider-native ambient discovery. `baseURL` overrides only the endpoint of the selected catalog model, preserving its API family and compatibility metadata, so private proxies such as `https://proxy.example.com:8443` remain supported.
 
 ```yaml
 - id: llm
   name: '@deepseek-ai/dsh-llm-pi-ai'
   config:
-    apiKey: !!js process.env.DEEPSEEK_API_KEY
-    baseURL: !!js process.env.DEEPSEEK_BASE_URL
-    models: [deepseek-v4-flash, deepseek-v4-pro]
-    reasoning: high   # off | high | xhigh (xhigh → wire 'max')
+    providers:
+      - provider: openai
+        apiKey: !!js process.env.OPENAI_API_KEY
+        baseURL: https://proxy.example.com:8443
+        reasoning: high
+      - provider: anthropic
+        apiKey: !!js process.env.ANTHROPIC_API_KEY
+        maxRetries: 2
+      - provider: openrouter
+        apiKey: !!js process.env.OPENROUTER_API_KEY
+        headers:
+          X-Deployment: production
 ```
 
+Each provider name must exist in pi-ai's installed catalog and may appear only once in this plugin instance. Registration with `ctx.llm` is atomic: a collision with any provider route already owned by another adapter fails plugin loading without registering the remaining routes. Model ids are not lifecycle config; an unknown model fails before any provider request with `LlmError('UNKNOWN_MODEL')`.
+
+The adapter exposes each configured provider's installed pi-ai models through `ctx.llm.listModels(provider)`. This is provider-neutral selector metadata derived from `getModels(provider)`; request-time resolution still performs the authoritative catalog lookup, so discovery does not create a second model registry.
+
+Supported profile fields are `provider`, `apiKey`, `baseURL`, `headers`, `reasoning`, `thinkingBudgets`, `cacheRetention`, `transport`, `timeoutMs`, `websocketConnectTimeoutMs`, `maxRetries`, and `maxRetryDelayMs`. They map to pi-ai's common stream options. Harness app attribution wins a conflicting configured header name.
+
+## Provider/model routing and replay
+
+The selected pi-ai catalog descriptor supplies the protocol implementation. This includes native API differences such as OpenAI models whose descriptor uses the Responses API rather than Chat Completions; the harness adapter does not hardcode endpoint selection by model name.
+
+Successful assistant responses store a versioned, lossless-JSON replay state beside their durable provider/model provenance. At request time, `LlmService` passes replay state only when the historical provider route and target provider route are currently owned by this same `PiAiAdapter` instance. The adapter validates the state and restores pi-ai response ids and provider signatures even when the target provider or model changes; pi-ai then decides which metadata its target API can reuse. History without replay state is translated as foreign provider-neutral content and never impersonates a native pi-ai response.
+
+If a listener rewrites assembled assistant content, the loop drops replay state before logging the message because its provider metadata no longer describes the content. Invalid versions, malformed metadata, provenance provider/model mismatches, and content/block mismatches fail explicitly with `LlmError('INVALID_REPLAY_STATE')`.
+
+## Vocabulary differences
+
+- pi-ai tool-call arguments are parsed objects; the harness stores raw JSON strings. The adapter parses input and re-stringifies output.
+- pi-ai reports failures as in-stream error events; these map to `finish {kind:'error'|'aborted'}` chunks.
+- pi-ai folds reasoning tokens into output usage; there is no separate reasoning count to map.
+- `GenerateOptions.stop` is rejected with `UNSUPPORTED_OPTION` because pi-ai's common streaming surface cannot guarantee it across providers.
+
 ## App attribution
 
-Every request carries the shared attribution header from dsh-llm's `attributionHeaders()`, passed through pi-ai's `headers` stream option (pi-ai merges caller headers last, so it always reaches the wire - the unit suite asserts arrival on the mock server, same as llm-deepseek). OpenRouter-specific app attribution headers are intentionally not sent by this adapter contract; they are deferred to a future explicit OpenRouter adapter or mode. See [dsh-llm § App attribution](../llm/README.md#app-attribution-attributionts).
+Every request carries the shared attribution header from dsh-llm's `attributionHeaders()`, merged through pi-ai's `headers` stream option. Provider-specific app-attribution headers are not synthesized. See [dsh-llm § App attribution](../llm/README.md#app-attribution-attributionts).
 
 ## Dependency weight
 
-pi-ai declares the openai/anthropic/google/mistral/AWS SDKs as install-time dependencies. They are lazy-loaded — only the openai SDK actually loads for this adapter — but they do land in `node_modules`. Accepted for a package whose purpose is design verification.
+pi-ai installs several provider SDKs and lazy-loads the one selected by the catalog model. The dependency weight is isolated to this opt-in adapter package.
 
 ## Testing
 
-Unit suites run against a local `node:http` mock SSE server (pi-ai's openai SDK happily talks to any base URL). Real-API coverage in `tests/adapter.e2e.ts` (`pnpm run test:e2e`, key-gated): V4 Flash + V4 Pro across all exposed reasoning levels (off/high/xhigh), the thinking+tools round trip, and a cross-adapter structural-equivalence check against llm-deepseek.
+Unit tests use pi-ai catalog models redirected to local mock servers and cover provider/profile routing, native API selection, endpoint overrides, attribution, conversion, replay-state validation, and cross-provider/model replay within one adapter instance. Real-API coverage remains key-gated under `pnpm run test:e2e`.
 
 ## Model Experience
 
-### DeepSeek request through pi-ai
+### Provider request through pi-ai
 
-**What the model sees**: The selected model receives the same logical system prompt, history, tools, stop sequences, and raw replayed tool arguments as the hand-written adapter. This package adds no prompt prose and removes pi-ai's own per-tool `strict` default to preserve that contract.
+**What the model sees**: The selected catalog model receives `GenerateOptions.system`, history, tools, and sampling fields supported by pi-ai's common streaming API. This package adds no prompt prose. Provider-native replay metadata is restored only when the adapter validates it for the historical content.
 
-**Token effect**: Provider tokenization governs exact input. Reasoning level changes generated and passback content; pi-ai reports reasoning inside output usage rather than as a separate count.
+**Token effect**: Provider tokenization governs exact input. Conversion adds no model-visible text; replay metadata may let a native API reuse provider-side state.
 
-### DeepSeek response
+### Provider response
 
-**What the model sees**: pi-ai events become harness reasoning, text, tool-call, usage, and finish chunks; parsed tool arguments are restored to raw JSON strings at the harness boundary.
+**What the model sees**: pi-ai events become harness reasoning, text, tool-call, usage, and finish chunks. Parsed tool arguments cross the harness boundary as raw JSON strings.
 
-**Token effect**: Generated content affects later inputs only after the loop records it; adapter conversion adds no model-visible text.
+**Token effect**: Generated content affects later inputs only after the loop records it. pi-ai folds reasoning tokens into output usage when the provider does not report them separately.
 
 ## Known Limitations and Deferred Work
 
-- **`tool_choice` is not mapped** — same MVP contract as llm-deepseek.
-- **In-history `system`-role messages fold into `user`-role wire messages** — pi-ai exposes a single `systemPrompt` slot, diverging from the hand-rolled twin's `role: 'system'` passthrough.
-- **`LlmError.status` is never set** — pi-ai reports failures as in-stream events with no HTTP status, so error codes are regex-classified from the error text.
-- **`buildModel` hardcodes descriptor metadata** — `contextWindow: 128000`, `maxTokens: 64000`, zero cost, identically for every registered model name; not configurable.
-- **pi-ai's built-in retries are disabled (`maxRetries: 0`)** — failures surface immediately; retry policy belongs to `llm/stream` listeners.
+- **Catalog membership is required** — custom model ids that are absent from the installed pi-ai catalog fail with `UNKNOWN_MODEL`, even when a provider profile supplies a custom endpoint.
+- **`GenerateOptions.stop` is unsupported** — pi-ai's common stream options cannot guarantee stop-sequence behavior across providers, so the adapter rejects the field.
+- **In-history `system` messages use pi-ai's common context conversion** — provider-specific placement follows pi-ai rather than a harness-owned wire override.
+- **`LlmError.status` is unavailable for in-stream failures** — pi-ai error events do not expose a stable HTTP status across providers.
diff --git a/packages/llm/llm-pi-ai/src/adapter.ts b/packages/llm/llm-pi-ai/src/adapter.ts
index 4b93b5e87d..26e28ecd1f 100644
--- a/packages/llm/llm-pi-ai/src/adapter.ts
+++ b/packages/llm/llm-pi-ai/src/adapter.ts
@@ -1,147 +1,101 @@
 /**
- * Pi-ai-backed DeepSeek adapter and design twin of the hand-rolled adapter.
- * Both implementations must fit the same provider-neutral stream vocabulary.
+ * Generic pi-ai-backed implementation of the Harness LLM seam.
+ *
  * @module dsh-llm-pi-ai/adapter
  */
 
-import { stream as piStream } from '@earendil-works/pi-ai'
-import type { Model } from '@earendil-works/pi-ai'
-import { attributionHeaders, LlmAdapter } from '@deepseek-ai/dsh-llm'
-import { CallId } from '@deepseek-ai/dsh-llm'
-import type { GenerateOptions, StreamChunk } from '@deepseek-ai/dsh-llm'
-import { toPiContext, toStreamChunks } from './convert.ts'
+import {
+  getModels,
+  streamSimple,
+} from '@earendil-works/pi-ai'
+import type {
+  Api,
+  KnownProvider,
+  Model,
+  SimpleStreamOptions,
+} from '@earendil-works/pi-ai'
+import { attributionHeaders, LlmAdapter, LlmError } from '@deepseek-ai/dsh-llm'
+import type { GenerateOptions, LlmModelInfo, StreamChunk } from '@deepseek-ai/dsh-llm'
+import type { PiAiProviderProfile } from './config.ts'
+import { toPiContext } from './context.ts'
+import { toStreamChunks } from './stream.ts'
 
-/** Reasoning levels surfaced by this adapter (DeepSeek wire: high|max). */
-export type PiAiReasoning = 'off' | 'high' | 'xhigh'
-
-/** Constructor options for {@link PiAiAdapter}; the plugin's `apply` resolves them from Config + environment. */
+/** Constructor options for {@link PiAiAdapter}. */
 export interface PiAiAdapterOptions {
-  /** Bearer token pi-ai sends on every request. */
-  apiKey: string
-  /** Endpoint base; `/chat/completions` is appended. */
-  baseURL: string
-  /** Thinking level applied to every request ('off' disables thinking). */
-  reasoning?: PiAiReasoning | undefined
+  /** Validated provider profiles this adapter instance owns. */
+  profiles: readonly PiAiProviderProfile[]
 }
 
 /**
- * Build the inline pi-ai model descriptor for one DeepSeek model name.
- * @param modelId - harness model name; sent verbatim on the wire.
- * @param options - adapter options; only `baseURL` is read here (key and reasoning apply per request, not per descriptor).
- * @returns a descriptor with every DeepSeek compat flag explicit — pi-ai's URL-based auto-detection is never relied on.
+ * Resolve a catalog model dynamically and apply only the configured endpoint
+ * override, preserving the catalog's API/capability/compatibility metadata.
  */
-export function buildModel(modelId: string, options: PiAiAdapterOptions): Model<'openai-completions'> {
+function resolveModel(profile: PiAiProviderProfile, modelId: string): Model<Api> {
+  const model = getModels(profile.provider as KnownProvider).find(candidate => candidate.id === modelId) as Model<Api> | undefined
+  if (model === undefined) {
+    throw new LlmError(`pi-ai provider "${profile.provider}" has no catalog model "${modelId}"`, 'UNKNOWN_MODEL')
+  }
+  return profile.baseURL === undefined ? model : { ...model, baseUrl: profile.baseURL }
+}
+
+/** Copy profile stream knobs into pi-ai's common option vocabulary. */
+function profileOptions(profile: PiAiProviderProfile): SimpleStreamOptions {
   return {
-    id: modelId,
-    name: modelId,
-    api: 'openai-completions',
-    provider: 'deepseek',
-    baseUrl: options.baseURL,
-    // Keep reasoning support enabled so `off` can send DeepSeek's explicit
-    // disabled marker rather than falling back to the provider's enabled default.
-    reasoning: true,
-    // DeepSeek's official effort levels: high|max (xhigh maps to max).
-    thinkingLevelMap: { minimal: null, low: null, medium: null, high: 'high', xhigh: 'max' },
-    input: ['text'],
-    cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
-    contextWindow: 128_000,
-    maxTokens: 64_000,
-    compat: {
-      // Auto-detection only fires for *.deepseek.com base URLs; the internal
-      // endpoint (and test mocks) need these set explicitly.
-      thinkingFormat: 'deepseek',
-      requiresReasoningContentOnAssistantMessages: true,
-      supportsReasoningEffort: true,
-      // DeepSeek documents max_tokens (not OpenAI's max_completion_tokens).
-      maxTokensField: 'max_tokens',
-    },
+    ...profile.apiKey === undefined ? {} : { apiKey: profile.apiKey },
+    ...profile.reasoning === undefined ? {} : { reasoning: profile.reasoning },
+    ...profile.thinkingBudgets === undefined ? {} : { thinkingBudgets: profile.thinkingBudgets },
+    ...profile.cacheRetention === undefined ? {} : { cacheRetention: profile.cacheRetention },
+    ...profile.transport === undefined ? {} : { transport: profile.transport },
+    ...profile.timeoutMs === undefined ? {} : { timeoutMs: profile.timeoutMs },
+    ...profile.websocketConnectTimeoutMs === undefined ? {} : { websocketConnectTimeoutMs: profile.websocketConnectTimeoutMs },
+    ...profile.maxRetries === undefined ? {} : { maxRetries: profile.maxRetries },
+    ...profile.maxRetryDelayMs === undefined ? {} : { maxRetryDelayMs: profile.maxRetryDelayMs },
   }
 }
 
-type Payload = {
-  tools?: { function?: { strict?: unknown } }[]
-  messages?: {
-    role?: unknown
-    tool_calls?: { id?: unknown; function?: { arguments?: unknown } }[]
-  }[]
-  reasoning_effort?: unknown
-  stop?: unknown
-}
-
-function rawToolArguments(options: GenerateOptions): Map<CallId, string> {
-  const raw = new Map<CallId, string>()
-  for (const message of options.messages) {
-    if (message.role !== 'assistant') continue
-    for (const block of message.content) {
-      if (block.type === 'tool-call') raw.set(block.id, block.arguments)
-    }
+/** Merge deployment headers while removing case-insensitive attribution collisions. */
+function requestHeaders(headers: Readonly<Record<string, string>> | undefined): Record<string, string> {
+  const attribution = attributionHeaders()
+  const reserved = new Set(Object.keys(attribution).map(name => name.toLowerCase()))
+  return {
+    ...Object.fromEntries(Object.entries(headers ?? {}).filter(([name]) => !reserved.has(name.toLowerCase()))),
+    ...attribution,
   }
-  return raw
-}
-
-function patchPayload(payload: unknown, options: GenerateOptions, reasoning: PiAiReasoning | undefined): unknown {
-  /* v8 ignore next -- pi-ai onPayload always receives an object; tolerate unusual future hooks defensively */
-  if (typeof payload !== 'object' || payload === null) return payload
-  const body = payload as Payload
-
-  if (reasoning === undefined) {
-    delete body.reasoning_effort
-  }
-  if (options.stop !== undefined) {
-    body.stop = options.stop
-  }
-
-  // pi-ai stamps its own `strict` default on every serialized tool; the
-  // harness tool contract has no strict field and the hand-rolled twin sends
-  // none, so scrub it for wire parity.
-  for (const tool of body.tools ?? []) {
-    /* v8 ignore next -- malformed pi-ai payload guard: real tool entries always carry function */
-    if (tool.function === undefined) continue
-    delete tool.function.strict
-  }
-
-  const rawById = rawToolArguments(options)
-  /* v8 ignore next -- defensive for non-chat payloads; OpenAI chat payloads always carry messages */
-  for (const message of body.messages ?? []) {
-    if (message.role !== 'assistant') continue
-    /* v8 ignore next -- assistant messages without tool_calls need no raw-argument patch */
-    for (const call of message.tool_calls ?? []) {
-      /* v8 ignore next -- malformed pi-ai payload guard: real tool calls always carry a string id */
-      if (typeof call.id !== 'string') continue
-      const raw = rawById.get(CallId(call.id))
-      /* v8 ignore next -- pi-ai always emits a function object for assistant tool_calls; guard malformed payloads defensively */
-      if (raw !== undefined && call.function !== undefined) call.function.arguments = raw
-    }
-  }
-
-  return body
 }
 
 /**
- * pi-ai-backed adapter. One instance serves every registered model name.
- *
- * Implementation notes:
- * - `onPayload` patches provider payload details pi-ai cannot express directly:
- *   stop sequences, scrubbing pi-ai's own per-tool `strict` default (the
- *   hand-rolled twin sends no such field), omitted reasoning effort, and raw
- *   replayed tool-call arguments.
- * - pi-ai reports request failures as in-stream error events; convert.ts
- *   maps them to `finish {kind:'error'|'aborted'}` chunks rather than
- *   throwing — both are sanctioned StreamChunk error paths.
+ * pi-ai-backed multi-provider adapter. Model descriptors are resolved for each
+ * request, so models need not be registered during the Cordis lifecycle.
  */
 export class PiAiAdapter extends LlmAdapter {
-  constructor(private readonly options: PiAiAdapterOptions) {
+  private readonly profiles: ReadonlyMap<string, PiAiProviderProfile>
+
+  constructor(options: PiAiAdapterOptions) {
     super()
+    this.profiles = new Map(options.profiles.map(profile => [profile.provider, profile]))
+  }
+
+  override listModels(provider: string): Promise<readonly LlmModelInfo[]> {
+    const profile = this.profiles.get(provider)
+    if (profile === undefined) {
+      return Promise.reject(new LlmError(`pi-ai adapter does not own provider "${provider}"`, 'NO_ADAPTER'))
+    }
+    return Promise.resolve(getModels(profile.provider as KnownProvider).map(model => ({
+      provider,
+      id: model.id,
+      name: model.name,
+    })))
   }
 
   async * stream(options: GenerateOptions): AsyncIterable<StreamChunk> {
-    const model = buildModel(options.model, this.options)
-    // Undefined config means "provider default" (DeepSeek: thinking ENABLED),
-    // matching llm-deepseek's omission semantics. pi-ai derives the wire
-    // thinking toggle from whether reasoningEffort is passed, so undefined maps
-    // internally to 'high' to get `thinking: enabled`; patchPayload then removes
-    // `reasoning_effort` so the provider chooses its default effort.
-    const reasoning = this.options.reasoning ?? 'high'
+    if (options.stop !== undefined) {
+      throw new LlmError('llm-pi-ai does not support GenerateOptions.stop', 'UNSUPPORTED_OPTION')
+    }
+    const profile = this.profiles.get(options.provider)
+    if (profile === undefined) {
+      throw new LlmError(`pi-ai adapter does not own provider "${options.provider}"`, 'NO_ADAPTER')
+    }
+    const model = resolveModel(profile, options.model)
 
     // Pi-ai has no iterator-return cancellation hook. Chain an internal signal
     // and abort it when this generator exits so early consumers stop the HTTP stream.
@@ -151,19 +105,16 @@ export class PiAiAdapter extends LlmAdapter {
     else options.signal?.addEventListener('abort', onCallerAbort, { once: true })
 
     try {
-      const events = piStream(model, toPiContext(options), {
-        apiKey: this.options.apiKey,
-        // pi-ai merges caller headers last over its provider defaults, so the
-        // harness attribution always reaches the wire.
-        headers: attributionHeaders(),
-        ...options.temperature !== undefined ? { temperature: options.temperature } : {},
-        ...options.maxTokens !== undefined ? { maxTokens: options.maxTokens } : {},
+      const events = streamSimple(model, toPiContext(options), {
+        ...profileOptions(profile),
+        ...options.temperature === undefined ? {} : { temperature: options.temperature },
+        ...options.maxTokens === undefined ? {} : { maxTokens: options.maxTokens },
+        ...options.sessionId === undefined ? {} : { sessionId: String(options.sessionId) },
         signal: controller.signal,
-        ...reasoning !== 'off' ? { reasoningEffort: reasoning } : {},
-        onPayload: payload => patchPayload(payload, options, this.options.reasoning),
-        maxRetries: 0,
+        // Profile headers are deployment-owned; attribution names are
+        // Harness-owned and therefore win collisions.
+        headers: requestHeaders(profile.headers),
       })
-
       yield* toStreamChunks(events)
     } finally {
       options.signal?.removeEventListener('abort', onCallerAbort)
diff --git a/packages/llm/llm-pi-ai/src/config.ts b/packages/llm/llm-pi-ai/src/config.ts
new file mode 100644
index 0000000000..f7570aff64
--- /dev/null
+++ b/packages/llm/llm-pi-ai/src/config.ts
@@ -0,0 +1,99 @@
+/**
+ * Configuration schema and provider-profile validation for the pi-ai adapter.
+ *
+ * @module dsh-llm-pi-ai/config
+ */
+
+import { getProviders } from '@earendil-works/pi-ai'
+import type { CacheRetention, ThinkingBudgets, ThinkingLevel, Transport } from '@earendil-works/pi-ai'
+import z from 'schemastery'
+
+/** Configuration for one pi-ai provider route. */
+export interface PiAiProviderProfile {
+  /** pi-ai provider catalog name and Harness route key. */
+  provider: string
+  /** Provider credential; when absent pi-ai uses its provider-native ambient discovery. */
+  apiKey?: string
+  /** Override the selected catalog model's endpoint without changing its protocol metadata. */
+  baseURL?: string
+  /** Provider request headers; Harness attribution wins reserved names. */
+  headers?: Record<string, string>
+  /** Provider-neutral pi-ai reasoning level. */
+  reasoning?: ThinkingLevel
+  /** Token budgets used by reasoning providers that support them. */
+  thinkingBudgets?: ThinkingBudgets
+  /** Prompt-cache retention preference. */
+  cacheRetention?: CacheRetention
+  /** Streaming transport preference. */
+  transport?: Transport
+  /** HTTP/provider SDK timeout in milliseconds. */
+  timeoutMs?: number
+  /** WebSocket connection timeout in milliseconds. */
+  websocketConnectTimeoutMs?: number
+  /** Provider SDK retry count. */
+  maxRetries?: number
+  /** Maximum provider-requested retry delay in milliseconds. */
+  maxRetryDelayMs?: number
+}
+
+/** Plugin configuration: the non-empty provider profiles this instance owns. */
+export interface Config {
+  /** Non-empty set of pi-ai provider routes this adapter instance owns. */
+  providers: PiAiProviderProfile[]
+}
+
+const thinkingBudgets = z.object({
+  minimal: z.number(),
+  low: z.number(),
+  medium: z.number(),
+  high: z.number(),
+})
+
+const profile = z.object({
+  provider: z.string().required(),
+  apiKey: z.string(),
+  baseURL: z.string(),
+  headers: z.dict(z.string()),
+  reasoning: z.union(['minimal', 'low', 'medium', 'high', 'xhigh']),
+  thinkingBudgets,
+  cacheRetention: z.union(['none', 'short', 'long']),
+  transport: z.union(['sse', 'websocket', 'websocket-cached', 'auto']),
+  timeoutMs: z.natural(),
+  websocketConnectTimeoutMs: z.natural(),
+  maxRetries: z.natural(),
+  maxRetryDelayMs: z.natural(),
+})
+
+/** Runtime schema for {@link Config}. */
+export const Config: z<Config> = z.object({
+  providers: z.array(profile).required(),
+})
+
+/**
+ * Validate profiles against the installed pi-ai catalog and return a detached
+ * shallow copy suitable for adapter construction.
+ * @param profiles - configured provider profiles.
+ * @returns validated profiles in configuration order.
+ */
+export function resolveProfiles(profiles: readonly PiAiProviderProfile[]): PiAiProviderProfile[] {
+  if (profiles.length === 0) throw new Error('llm-pi-ai: providers must contain at least one profile')
+  const supported = new Set<string>(getProviders())
+  const seen = new Set<string>()
+  return profiles.map((source) => {
+    if (source.provider.length === 0) throw new Error('llm-pi-ai: provider names must be non-empty')
+    if (!supported.has(source.provider)) throw new Error(`llm-pi-ai: unknown pi-ai provider "${source.provider}"`)
+    if (seen.has(source.provider)) throw new Error(`llm-pi-ai: duplicate provider profile "${source.provider}"`)
+    if (source.apiKey !== undefined && source.apiKey.trim().length === 0) {
+      throw new Error(`llm-pi-ai: provider "${source.provider}" has an empty apiKey; omit it to use ambient authentication`)
+    }
+    if (source.baseURL !== undefined && source.baseURL.length === 0) {
+      throw new Error(`llm-pi-ai: provider "${source.provider}" has an empty baseURL`)
+    }
+    seen.add(source.provider)
+    return {
+      ...source,
+      ...source.headers === undefined ? {} : { headers: { ...source.headers } },
+      ...source.thinkingBudgets === undefined ? {} : { thinkingBudgets: { ...source.thinkingBudgets } },
+    }
+  })
+}
diff --git a/packages/llm/llm-pi-ai/src/context.ts b/packages/llm/llm-pi-ai/src/context.ts
new file mode 100644
index 0000000000..ddb8284448
--- /dev/null
+++ b/packages/llm/llm-pi-ai/src/context.ts
@@ -0,0 +1,85 @@
+/**
+ * Harness request-history conversion into pi-ai's Context vocabulary.
+ *
+ * @module dsh-llm-pi-ai/context
+ */
+
+import { CallId } from '@deepseek-ai/dsh-llm'
+import type { GenerateOptions, Message } from '@deepseek-ai/dsh-llm'
+import type { Context as PiContext, Message as PiMessage, Tool as PiTool } from '@earendil-works/pi-ai'
+import { toPiAssistant } from './replay.ts'
+
+/** Join the text blocks of a harness message. */
+function flattenText(message: Message): string {
+  return message.content
+    .filter(block => block.type === 'text')
+    .map(block => block.text)
+    .join('')
+}
+
+/**
+ * Convert harness history to a pi-ai Context. Tool results need the tool
+ * NAME (pi-ai's `toolName`), which the harness doesn't carry on the result
+ * block — it is recovered from the preceding assistant tool-call with the
+ * same id.
+ * @param options - the harness request; `options.system` maps to pi-ai's single `systemPrompt` slot.
+ * @returns the pi-ai context; `tools` is omitted entirely when the request declares none.
+ */
+export function toPiContext(options: GenerateOptions): PiContext {
+  const toolNames = new Map<CallId, string>()
+  const messages: PiMessage[] = []
+
+  for (const message of options.messages) {
+    if (message.role === 'system') {
+      // pi-ai has a single systemPrompt slot; in-history system messages are
+      // folded into user messages to preserve order (rare in practice — the
+      // harness sends the system prompt via options.system).
+      messages.push({ role: 'user', content: flattenText(message), timestamp: 0 })
+      continue
+    }
+    if (message.role === 'assistant') {
+      const assistant = toPiAssistant(message)
+      for (const block of assistant.content) {
+        if (block.type === 'toolCall') toolNames.set(CallId(block.id), block.name)
+      }
+      messages.push(assistant)
+      continue
+    }
+    // user role: text + tool results (each result becomes its own message).
+    const text = flattenText(message)
+    const results = message.content.filter(block => block.type === 'tool-result')
+    if (text.length > 0 || results.length === 0) {
+      messages.push({ role: 'user', content: text, timestamp: 0 })
+    }
+    for (const result of results) {
+      messages.push({
+        role: 'toolResult',
+        toolCallId: result.toolCallId,
+        toolName: toolNames.get(result.toolCallId) ?? 'unknown',
+        content: [{
+          type: 'text',
+          text: result.content
+            .filter(block => block.type === 'text')
+            .map(block => block.text)
+            .join('') || '(no output)',
+        }],
+        isError: result.isError ?? false,
+        timestamp: 0,
+      })
+    }
+  }
+
+  const tools: PiTool[] | undefined = options.tools?.map(tool => ({
+    name: tool.name,
+    description: tool.description,
+    // ToolSchema.parameters is a JSON Schema object; pi-ai's TSchema
+    // (TypeBox) is structurally JSON Schema, so it assigns directly.
+    parameters: tool.parameters,
+  }))
+
+  return {
+    ...options.system !== undefined ? { systemPrompt: options.system } : {},
+    messages,
+    ...tools !== undefined && tools.length > 0 ? { tools } : {},
+  }
+}
diff --git a/packages/llm/llm-pi-ai/src/index.ts b/packages/llm/llm-pi-ai/src/index.ts
index 43468ba507..cbd0dcd435 100644
--- a/packages/llm/llm-pi-ai/src/index.ts
+++ b/packages/llm/llm-pi-ai/src/index.ts
@@ -1,76 +1,45 @@
 /**
- * pi-ai-backed DeepSeek adapter plugin. Same Config shape as
- * `@deepseek-ai/dsh-llm-deepseek` (one-line swap in cordis.yml), different
- * implementation underneath — see `./adapter.ts` for why both exist.
+ * Generic pi-ai-backed LLM adapter plugin. One plugin instance registers an
+ * explicit set of provider profiles; requests select a profile by provider and
+ * resolve the model dynamically from pi-ai's installed catalog.
  *
  * ```yaml
  * - id: llm
  *   name: '@deepseek-ai/dsh-llm-pi-ai'
  *   config:
- *     apiKey: !!js process.env.DEEPSEEK_API_KEY
- *     baseURL: !!js process.env.DEEPSEEK_BASE_URL
- *     models: [deepseek-v4-flash, deepseek-v4-pro]
- *     reasoning: high
+ *     providers:
+ *       - provider: openai
+ *         apiKey: !!js process.env.OPENAI_API_KEY
+ *       - provider: anthropic
+ *         apiKey: !!js process.env.ANTHROPIC_API_KEY
+ *       - provider: openrouter
+ *         apiKey: !!js process.env.OPENROUTER_API_KEY
+ *         baseURL: https://proxy.example.com/v1
  * ```
  *
  * @module @deepseek-ai/dsh-llm-pi-ai
  */
 
 import type { Context } from 'cordis'
-import z from 'schemastery'
 import type {} from '@deepseek-ai/dsh-llm'
 import { PiAiAdapter } from './adapter.ts'
-import type { PiAiReasoning } from './adapter.ts'
+import { Config, resolveProfiles } from './config.ts'
 
-export { buildModel, PiAiAdapter } from './adapter.ts'
-export type { PiAiAdapterOptions, PiAiReasoning } from './adapter.ts'
-export { mapStopReason, mapUsage, toPiContext, toStreamChunks } from './convert.ts'
+export { PiAiAdapter } from './adapter.ts'
+export type { PiAiAdapterOptions } from './adapter.ts'
+export { Config, resolveProfiles } from './config.ts'
+export type { PiAiProviderProfile } from './config.ts'
+export { toPiContext } from './context.ts'
+export { toPiReplayState } from './replay.ts'
+export type { PiAiReplayState } from './replay.ts'
+export { mapStopReason, mapUsage, toStreamChunks } from './stream.ts'
 
 export const name = 'llm-pi-ai'
 export const inject = ['llm']
 
-/**
- * Plugin config, validated by the same-named schemastery schema. Every field
- * is optional in yml: credentials/endpoint fall back to the environment (a
- * missing API key fails plugin load, not the first call).
- */
-export interface Config {
-  /** API key; falls back to $DEEPSEEK_API_KEY. Required one way or the other. */
-  apiKey?: string
-  /** Endpoint base; falls back to $DEEPSEEK_BASE_URL, then the public API. */
-  baseURL?: string
-  /** Model names to register (sent verbatim on the wire). */
-  models?: string[]
-  /**
-   * Thinking level for every request: 'off' disables thinking mode; 'high'
-   * and 'xhigh' (wire 'max') set the effort. Omitted = provider default
-   * (thinking enabled), matching llm-deepseek's omission semantics.
-   */
-  reasoning?: PiAiReasoning
-}
-
-export const Config: z<Config> = z.object({
-  apiKey: z.string(),
-  baseURL: z.string(),
-  models: z.array(z.string()).default(['deepseek-v4-flash', 'deepseek-v4-pro']),
-  reasoning: z.union(['off', 'high', 'xhigh']),
-})
-
-/** Public API default; the internal endpoint comes from $DEEPSEEK_BASE_URL. */
-export const PUBLIC_BASE_URL = 'https://api.deepseek.com'
-
+/** Register one generic pi-ai adapter for all configured provider routes. */
 export function apply(ctx: Context, config: Config): void {
-  const apiKey = config.apiKey ?? process.env.DEEPSEEK_API_KEY
-  if (apiKey === undefined || apiKey.length === 0) {
-    throw new Error('llm-pi-ai: an API key is required (Config.apiKey or $DEEPSEEK_API_KEY)')
-  }
-  const baseURL = config.baseURL ?? process.env.DEEPSEEK_BASE_URL ?? PUBLIC_BASE_URL
-  // schemastery's .default() guarantees models is set after validation.
-  const models = config.models as string[]
-
-  ctx.llm.registerAdapter(models, new PiAiAdapter({
-    apiKey,
-    baseURL,
-    reasoning: config.reasoning,
-  }))
+  const profiles = resolveProfiles(config.providers)
+  const adapter = new PiAiAdapter({ profiles })
+  ctx.llm.registerAdapter(profiles.map(entry => entry.provider), adapter)
 }
diff --git a/packages/llm/llm-pi-ai/src/replay.ts b/packages/llm/llm-pi-ai/src/replay.ts
new file mode 100644
index 0000000000..4e4fe679a5
--- /dev/null
+++ b/packages/llm/llm-pi-ai/src/replay.ts
@@ -0,0 +1,211 @@
+/**
+ * Durable pi-ai replay metadata and assistant-history reconstruction.
+ *
+ * Harness content remains the durable source for text and tool calls. This
+ * module stores only the provider-native metadata needed to reconstruct a
+ * pi-ai assistant message on a later request.
+ *
+ * @module dsh-llm-pi-ai/replay
+ */
+
+import { LlmError } from '@deepseek-ai/dsh-llm'
+import type { Message } from '@deepseek-ai/dsh-llm'
+import type { Api, AssistantMessage, Usage as PiUsage } from '@earendil-works/pi-ai'
+
+type PiAiReplayBlock =
+  | { type: 'text'; textSignature?: string }
+  | { type: 'reasoning'; thinkingSignature?: string; redacted?: boolean }
+  | { type: 'tool-call'; thoughtSignature?: string }
+
+/** Versioned adapter-private projection required to replay a pi-ai response. */
+export interface PiAiReplayState {
+  kind: 'pi-ai'
+  version: 1
+  api: Api
+  provider: string
+  model: string
+  responseModel?: string
+  responseId?: string
+  stopReason: AssistantMessage['stopReason']
+  blocks: PiAiReplayBlock[]
+}
+
+/** Parse tool-call argument JSON; tolerate model malformations with {}. */
+function parseArguments(raw: string): Record<string, unknown> {
+  try {
+    const parsed: unknown = JSON.parse(raw)
+    if (typeof parsed === 'object' && parsed !== null && !Array.isArray(parsed)) {
+      return parsed as Record<string, unknown>
+    }
+  } catch {
+    // fall through
+  }
+  return {}
+}
+
+/** Construct the zero usage value required by historical pi-ai messages. */
+function emptyPiUsage(): PiUsage {
+  return {
+    input: 0,
+    output: 0,
+    cacheRead: 0,
+    cacheWrite: 0,
+    totalTokens: 0,
+    cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0, total: 0 },
+  }
+}
+
+/**
+ * Project a successful pi-ai response into the minimal durable replay state.
+ * @param message - completed native pi-ai assistant response.
+ * @returns the versioned lossless-JSON replay projection.
+ */
+export function toPiReplayState(message: AssistantMessage): PiAiReplayState {
+  return {
+    kind: 'pi-ai',
+    version: 1,
+    api: message.api,
+    provider: message.provider,
+    model: message.model,
+    ...message.responseModel === undefined ? {} : { responseModel: message.responseModel },
+    ...message.responseId === undefined ? {} : { responseId: message.responseId },
+    stopReason: message.stopReason,
+    blocks: message.content.map((block): PiAiReplayBlock => {
+      switch (block.type) {
+        case 'text': return {
+          type: 'text',
+          ...block.textSignature === undefined ? {} : { textSignature: block.textSignature },
+        }
+        case 'thinking': return {
+          type: 'reasoning',
+          ...block.thinkingSignature === undefined ? {} : { thinkingSignature: block.thinkingSignature },
+          ...block.redacted === undefined ? {} : { redacted: block.redacted },
+        }
+        case 'toolCall': return {
+          type: 'tool-call',
+          ...block.thoughtSignature === undefined ? {} : { thoughtSignature: block.thoughtSignature },
+        }
+      }
+    }),
+  }
+}
+
+function invalidReplay(message: string): never {
+  throw new LlmError(`invalid pi-ai replay state: ${message}`, 'INVALID_REPLAY_STATE')
+}
+
+/** Validate the adapter-private state before it reaches pi-ai. */
+function readReplayState(value: unknown): PiAiReplayState {
+  if (typeof value !== 'object' || value === null || Array.isArray(value)) return invalidReplay('expected an object')
+  const state = value as Record<string, unknown>
+  if (state['kind'] !== 'pi-ai') return invalidReplay('unknown state kind')
+  if (state['version'] !== 1) return invalidReplay(`unsupported version ${String(state['version'])}`)
+  for (const key of ['api', 'provider', 'model'] as const) {
+    if (typeof state[key] !== 'string' || state[key].length === 0) return invalidReplay(`${key} must be a non-empty string`)
+  }
+  if (!['stop', 'length', 'toolUse', 'error', 'aborted'].includes(String(state['stopReason']))) {
+    return invalidReplay('unknown stopReason')
+  }
+  if (state['responseModel'] !== undefined && typeof state['responseModel'] !== 'string') return invalidReplay('responseModel must be a string')
+  if (state['responseId'] !== undefined && typeof state['responseId'] !== 'string') return invalidReplay('responseId must be a string')
+  if (!Array.isArray(state['blocks'])) return invalidReplay('blocks must be an array')
+  for (const [index, value] of state['blocks'].entries()) {
+    if (typeof value !== 'object' || value === null || Array.isArray(value)) return invalidReplay(`block ${index} must be an object`)
+    const block = value as Record<string, unknown>
+    if (!['text', 'reasoning', 'tool-call'].includes(String(block['type']))) return invalidReplay(`block ${index} has an unknown type`)
+    for (const signature of ['textSignature', 'thinkingSignature', 'thoughtSignature'] as const) {
+      if (block[signature] !== undefined && typeof block[signature] !== 'string') return invalidReplay(`block ${index} ${signature} must be a string`)
+    }
+    if (block['redacted'] !== undefined && typeof block['redacted'] !== 'boolean') return invalidReplay(`block ${index} redacted must be boolean`)
+  }
+  return state as unknown as PiAiReplayState
+}
+
+/** Convert provider-neutral blocks without trusting them as same-model replay. */
+function foreignAssistant(message: Message): AssistantMessage {
+  const content: AssistantMessage['content'] = []
+  for (const block of message.content) {
+    switch (block.type) {
+      case 'text': content.push({ type: 'text', text: block.text }); break
+      case 'reasoning': content.push({ type: 'thinking', thinking: block.text }); break
+      case 'tool-call': content.push({
+        type: 'toolCall',
+        id: block.id,
+        name: block.name,
+        arguments: parseArguments(block.arguments),
+      }); break
+      default:
+        // plugin-added block types are not representable in pi-ai.
+        break
+    }
+  }
+  return {
+    role: 'assistant',
+    content,
+    // Deliberately never equals a catalog API: absent replay state is foreign
+    // even if provenance names the same provider/model as this request.
+    api: 'dsh-foreign',
+    provider: message.provenance?.provider ?? 'dsh-foreign',
+    model: message.provenance?.model ?? 'dsh-foreign',
+    usage: emptyPiUsage(),
+    stopReason: content.some(piece => piece.type === 'toolCall') ? 'toolUse' : 'stop',
+    timestamp: 0,
+  }
+}
+
+/** Recombine durable Harness content with validated pi-ai replay metadata. */
+function replayedAssistant(message: Message, rawState: unknown): AssistantMessage {
+  const state = readReplayState(rawState)
+  const provenance = message.provenance
+  if (state.provider !== provenance?.provider) return invalidReplay('provider does not match assistant provenance')
+  if (state.model !== provenance.model) return invalidReplay('model does not match assistant provenance')
+  if (state.blocks.length !== message.content.length) return invalidReplay('block count does not match assistant content')
+  const content: AssistantMessage['content'] = message.content.map((block, index) => {
+    const replay = state.blocks[index]
+    if (replay === undefined || replay.type !== block.type) return invalidReplay(`block ${index} does not match assistant content`)
+    switch (block.type) {
+      case 'text': return {
+        type: 'text',
+        text: block.text,
+        ...replay.type === 'text' && replay.textSignature !== undefined ? { textSignature: replay.textSignature } : {},
+      }
+      case 'reasoning': return {
+        type: 'thinking',
+        thinking: block.text,
+        ...replay.type === 'reasoning' && replay.thinkingSignature !== undefined ? { thinkingSignature: replay.thinkingSignature } : {},
+        ...replay.type === 'reasoning' && replay.redacted !== undefined ? { redacted: replay.redacted } : {},
+      }
+      case 'tool-call': return {
+        type: 'toolCall',
+        id: block.id,
+        name: block.name,
+        arguments: parseArguments(block.arguments),
+        ...replay.type === 'tool-call' && replay.thoughtSignature !== undefined ? { thoughtSignature: replay.thoughtSignature } : {},
+      }
+      /* v8 ignore next -- readReplayState rejects unknown replay tags, so an equal plugin-added Harness tag cannot reach this switch */
+      default: return invalidReplay(`block ${index} has an unsupported Harness type`)
+    }
+  })
+  return {
+    role: 'assistant',
+    content,
+    api: state.api,
+    provider: state.provider,
+    model: state.model,
+    ...state.responseModel === undefined ? {} : { responseModel: state.responseModel },
+    ...state.responseId === undefined ? {} : { responseId: state.responseId },
+    usage: emptyPiUsage(),
+    stopReason: state.stopReason,
+    timestamp: 0,
+  }
+}
+
+/**
+ * Convert one durable Harness assistant message into pi-ai history.
+ * @param message - assistant content with optional adapter-owned replay metadata.
+ * @returns a native pi-ai assistant message reconstructed from durable content.
+ */
+export function toPiAssistant(message: Message): AssistantMessage {
+  const replayState = message.provenance?.replayState
+  return replayState === undefined ? foreignAssistant(message) : replayedAssistant(message, replayState)
+}
diff --git a/packages/llm/llm-pi-ai/src/convert.ts b/packages/llm/llm-pi-ai/src/stream.ts
similarity index 50%
rename from packages/llm/llm-pi-ai/src/convert.ts
rename to packages/llm/llm-pi-ai/src/stream.ts
index 098b93edb3..5bdce3c528 100644
--- a/packages/llm/llm-pi-ai/src/convert.ts
+++ b/packages/llm/llm-pi-ai/src/stream.ts
@@ -1,152 +1,17 @@
 /**
- * Bidirectional mapping between the harness vocabulary and pi-ai's:
- * Convert harness requests to pi-ai context and pi-ai assistant events to harness stream chunks.
- * pi-ai parses tool arguments while the harness preserves raw JSON, so conversion parses inbound
- * arguments and re-stringifies outbound values while the adapter restores provider payloads.
- * In-stream pi-ai errors become harness error/aborted finishes, and its reasoning tokens remain
- * folded into output usage because it reports no separate count.
- * @module dsh-llm-pi-ai/convert
+ * pi-ai assistant event translation into the Harness streaming protocol.
+ *
+ * pi-ai tool-call arguments are parsed objects while the Harness keeps their
+ * raw JSON representation. pi-ai also reports failures as terminal stream
+ * events, which this module maps into Harness finish chunks.
+ *
+ * @module dsh-llm-pi-ai/stream
  */
 
 import { CallId, LlmError } from '@deepseek-ai/dsh-llm'
-import type { FinishReason, GenerateOptions, Message, StreamChunk, TokenUsage } from '@deepseek-ai/dsh-llm'
-import type {
-  AssistantMessage,
-  AssistantMessageEvent,
-  Context as PiContext,
-  Message as PiMessage,
-  Tool as PiTool,
-  Usage as PiUsage,
-} from '@earendil-works/pi-ai'
-
-/** Join the text blocks of a harness message. */
-function flattenText(message: Message): string {
-  return message.content
-    .filter(block => block.type === 'text')
-    .map(block => block.text)
-    .join('')
-}
-
-/** Parse tool-call argument JSON; tolerate model malformations with {}. */
-function parseArguments(raw: string): Record<string, unknown> {
-  try {
-    const parsed: unknown = JSON.parse(raw)
-    if (typeof parsed === 'object' && parsed !== null && !Array.isArray(parsed)) {
-      return parsed as Record<string, unknown>
-    }
-  } catch {
-    // fall through
-  }
-  return {}
-}
-
-/**
- * Convert harness history to a pi-ai Context. Tool results need the tool
- * NAME (pi-ai's `toolName`), which the harness doesn't carry on the result
- * block — it is recovered from the preceding assistant tool-call with the
- * same id.
- * @param options - the harness request; `options.system` maps to pi-ai's single `systemPrompt` slot.
- * @returns the pi-ai context; `tools` is omitted entirely when the request declares none.
- */
-export function toPiContext(options: GenerateOptions): PiContext {
-  const toolNames = new Map<CallId, string>()
-  const messages: PiMessage[] = []
-
-  for (const message of options.messages) {
-    if (message.role === 'system') {
-      // pi-ai has a single systemPrompt slot; in-history system messages are
-      // folded into user messages to preserve order (rare in practice — the
-      // harness sends the system prompt via options.system).
-      messages.push({ role: 'user', content: flattenText(message), timestamp: 0 })
-      continue
-    }
-    if (message.role === 'assistant') {
-      const content: AssistantMessage['content'] = []
-      for (const block of message.content) {
-        switch (block.type) {
-          case 'text':
-            content.push({ type: 'text', text: block.text })
-            break
-          case 'reasoning':
-            // Without this wire-field name, pi-ai replays an empty `reasoning_content`, violating
-            // DeepSeek's thinking-mode passback rule on tool-call turns.
-            content.push({ type: 'thinking', thinking: block.text, thinkingSignature: 'reasoning_content' })
-            break
-          case 'tool-call':
-            toolNames.set(block.id, block.name)
-            content.push({
-              type: 'toolCall',
-              id: block.id,
-              name: block.name,
-              arguments: parseArguments(block.arguments),
-            })
-            break
-          default:
-            // plugin-added block types: not representable here.
-            break
-        }
-      }
-      messages.push({
-        role: 'assistant',
-        content,
-        api: 'openai-completions',
-        provider: 'deepseek',
-        model: options.model,
-        usage: emptyPiUsage(),
-        stopReason: content.some(piece => piece.type === 'toolCall') ? 'toolUse' : 'stop',
-        timestamp: 0,
-      })
-      continue
-    }
-    // user role: text + tool results (each result becomes its own message).
-    const text = flattenText(message)
-    const results = message.content.filter(block => block.type === 'tool-result')
-    if (text.length > 0 || results.length === 0) {
-      messages.push({ role: 'user', content: text, timestamp: 0 })
-    }
-    for (const result of results) {
-      messages.push({
-        role: 'toolResult',
-        toolCallId: result.toolCallId,
-        toolName: toolNames.get(result.toolCallId) ?? 'unknown',
-        content: [{
-          type: 'text',
-          text: result.content
-            .filter(block => block.type === 'text')
-            .map(block => block.text)
-            .join('') || '(no output)',
-        }],
-        isError: result.isError ?? false,
-        timestamp: 0,
-      })
-    }
-  }
-
-  const tools: PiTool[] | undefined = options.tools?.map(tool => ({
-    name: tool.name,
-    description: tool.description,
-    // ToolSchema.parameters is a JSON Schema object; pi-ai's TSchema
-    // (TypeBox) is structurally JSON Schema, so it assigns directly.
-    parameters: tool.parameters,
-  }))
-
-  return {
-    ...options.system !== undefined ? { systemPrompt: options.system } : {},
-    messages,
-    ...tools !== undefined && tools.length > 0 ? { tools } : {},
-  }
-}
-
-function emptyPiUsage(): PiUsage {
-  return {
-    input: 0,
-    output: 0,
-    cacheRead: 0,
-    cacheWrite: 0,
-    totalTokens: 0,
-    cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0, total: 0 },
-  }
-}
+import type { FinishReason, StreamChunk, TokenUsage } from '@deepseek-ai/dsh-llm'
+import type { AssistantMessage, AssistantMessageEvent, Usage as PiUsage } from '@earendil-works/pi-ai'
+import { toPiReplayState } from './replay.ts'
 
 /**
  * Map pi-ai usage (reasoning folded into output by pi-ai).
@@ -259,7 +124,7 @@ export async function* toStreamChunks(events: AsyncIterable<AssistantMessageEven
         break
       case 'done':
         yield { type: 'usage', usage: mapUsage(event.message.usage) }
-        yield { type: 'finish', reason: mapStopReason(event.message) }
+        yield { type: 'finish', reason: mapStopReason(event.message), replayState: toPiReplayState(event.message) }
         return
       case 'error':
         // In-stream error delivery (pi-ai's style) → error finish chunk
diff --git a/packages/llm/llm-pi-ai/tests/adapter.e2e.ts b/packages/llm/llm-pi-ai/tests/adapter.e2e.ts
index fa30226ddf..77d2cc81dd 100644
--- a/packages/llm/llm-pi-ai/tests/adapter.e2e.ts
+++ b/packages/llm/llm-pi-ai/tests/adapter.e2e.ts
@@ -3,26 +3,33 @@ import { Context } from 'cordis'
 import LlmService, { CallId } from '@deepseek-ai/dsh-llm'
 import type { Message, ToolSchema } from '@deepseek-ai/dsh-llm'
 import * as LlmPiAi from '@deepseek-ai/dsh-llm-pi-ai'
-import type { Config } from '@deepseek-ai/dsh-llm-pi-ai'
+import type { PiAiProviderProfile } from '@deepseek-ai/dsh-llm-pi-ai'
 import * as LlmDeepSeek from '@deepseek-ai/dsh-llm-deepseek'
 import { assemble, type AssembledResult } from './assemble.ts'
 
 /**
- * Real-API e2e for the pi-ai-backed adapter: V4 Flash + V4 Pro across all
- * reasoning levels the adapter exposes (off / high / xhigh→wire 'max').
- * Mirrors the llm-deepseek matrix so the two independent implementations
- * verify the same StreamChunk contract. Key-gated.
+ * Real-API e2e for the pi-ai-backed adapter: V4 Flash + V4 Pro with provider
+ * defaults and representative high/xhigh reasoning. Mirrors the native
+ * adapter's StreamChunk contract and exercises a replayed tool follow-up.
+ * Key-gated.
  */
 
 const FLASH = 'deepseek-v4-flash'
 const PRO = 'deepseek-v4-pro'
 const contexts: Context[] = []
 
-async function harness(model: string, config: Partial<Config> = {}) {
+async function harness(_model: string, config: Partial<PiAiProviderProfile> = {}) {
   const ctx = new Context()
   contexts.push(ctx)
   await ctx.plugin(LlmService)
-  await ctx.plugin(LlmPiAi, { models: [model], ...config })
+  await ctx.plugin(LlmPiAi, {
+    providers: [{
+      provider: 'deepseek',
+      ...process.env.DEEPSEEK_API_KEY === undefined ? {} : { apiKey: process.env.DEEPSEEK_API_KEY },
+      ...process.env.DEEPSEEK_BASE_URL === undefined ? {} : { baseURL: process.env.DEEPSEEK_BASE_URL },
+      ...config,
+    }],
+  })
   return ctx
 }
 
@@ -56,8 +63,8 @@ const weatherTool: ToolSchema = {
 }
 
 describe.skipIf(!process.env.DEEPSEEK_API_KEY)('llm-pi-ai e2e (real API)', () => {
-  it.each([FLASH, PRO])('%s + reasoning off: plain text generation', async (model) => {
-    const ctx = await harness(model, { reasoning: 'off' })
+  it.each([FLASH, PRO])('%s + provider-default reasoning: plain text generation', async (model) => {
+    const ctx = await harness(model)
     const result = await assemble(ctx,{
       model,
       messages: ask('Reply with exactly the word: pong'),
@@ -65,7 +72,6 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('llm-pi-ai e2e (real API)', () =>
     })
     expect(result.finish.kind).toBe('stop')
     expect(textOf(result).toLowerCase()).toContain('pong')
-    expect(result.message.content.some(block => block.type === 'reasoning')).toBe(false)
   })
 
   it.each([FLASH, PRO])('%s + reasoning high: reasoning blocks present', async (model) => {
@@ -99,7 +105,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('llm-pi-ai e2e (real API)', () =>
       model: PRO,
       messages: [
         ...ask('What is the weather in Paris right now? Use the get_weather tool.'),
-        { role: 'assistant', content: first.message.content },
+        first.message,
         {
           role: 'user',
           content: [{
@@ -123,9 +129,9 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('llm-pi-ai e2e (real API)', () =>
     const deepseekCtx = new Context()
     contexts.push(deepseekCtx)
     await deepseekCtx.plugin(LlmService)
-    await deepseekCtx.plugin(LlmDeepSeek, { models: [FLASH], thinking: 'disabled' })
+    await deepseekCtx.plugin(LlmDeepSeek, { thinking: 'disabled' })
 
-    const piCtx = await harness(FLASH, { reasoning: 'off' })
+    const piCtx = await harness(FLASH)
 
     const prompt = ask('Reply with exactly the word: pong')
     const [fromDeepSeek, fromPiAi] = await Promise.all([
diff --git a/packages/llm/llm-pi-ai/tests/adapter.spec.ts b/packages/llm/llm-pi-ai/tests/adapter.spec.ts
index f3f12666a3..1767447d7d 100644
--- a/packages/llm/llm-pi-ai/tests/adapter.spec.ts
+++ b/packages/llm/llm-pi-ai/tests/adapter.spec.ts
@@ -2,34 +2,35 @@ import { createServer } from 'node:http'
 import type { IncomingMessage, Server, ServerResponse } from 'node:http'
 import { afterEach, describe, expect, it, vi } from 'vitest'
 import { Context } from 'cordis'
-import LlmService, { CallId, userAgent } from '@deepseek-ai/dsh-llm'
+import LlmService, { LlmError, userAgent } from '@deepseek-ai/dsh-llm'
 import * as LlmPiAi from '@deepseek-ai/dsh-llm-pi-ai'
-import { buildModel, PiAiAdapter } from '@deepseek-ai/dsh-llm-pi-ai'
+import { PiAiAdapter, resolveProfiles } from '@deepseek-ai/dsh-llm-pi-ai'
 import { assemble } from './assemble.ts'
 
-/** Scripted SSE responses, one per request (OpenAI chat-completions shape). */
 interface MockServer {
   url: string
+  paths: string[]
   requests: unknown[]
-  /** Header bags of received requests, in order (parallel to `requests`). */
   headers: IncomingMessage['headers'][]
-  close(): Promise<void>
 }
 
 const servers: Server[] = []
 
 afterEach(async () => {
+  vi.unstubAllEnvs()
   await Promise.all(servers.splice(0).map(server => new Promise(resolve => server.close(resolve))))
 })
 
-async function mockServer(script: { status?: number; events?: string[]; body?: string }[]): Promise<MockServer> {
+async function mockServer(script: { status?: number; events?: string[]; body?: string; delayMs?: number }[]): Promise<MockServer> {
+  const paths: string[] = []
   const requests: unknown[] = []
   const headers: IncomingMessage['headers'][] = []
   const server = createServer((request: IncomingMessage, response: ServerResponse) => {
     let body = ''
     request.on('data', (chunk: Buffer) => { body += chunk.toString('utf8') })
     request.on('end', () => {
-      requests.push(JSON.parse(body))
+      paths.push(request.url ?? '')
+      requests.push(body.length === 0 ? undefined : JSON.parse(body))
       headers.push(request.headers)
       const behavior = script.shift() ?? { status: 500, body: 'script exhausted' }
       if (behavior.status !== undefined && behavior.status !== 200) {
@@ -38,20 +39,22 @@ async function mockServer(script: { status?: number; events?: string[]; body?: s
         return
       }
       response.writeHead(200, { 'content-type': 'text/event-stream' })
-      for (const event of behavior.events ?? []) response.write(`data: ${event}\n\n`)
-      response.end()
+      let index = 0
+      const writeNext = (): void => {
+        const event = behavior.events?.[index++]
+        if (event === undefined) { response.end(); return }
+        response.write(`data: ${event}\n\n`)
+        if (behavior.delayMs === undefined) writeNext()
+        else setTimeout(writeNext, behavior.delayMs)
+      }
+      writeNext()
     })
   })
   servers.push(server)
   await new Promise<void>(resolve => server.listen(0, '127.0.0.1', resolve))
   const address = server.address()
   if (address === null || typeof address === 'string') throw new Error('no port')
-  return {
-    url: `http://127.0.0.1:${address.port}`,
-    requests,
-    headers,
-    close: () => new Promise(resolve => server.close(() => { resolve() })),
-  }
+  return { url: `http://127.0.0.1:${address.port}`, paths, requests, headers }
 }
 
 const textEvents = [
@@ -61,347 +64,231 @@ const textEvents = [
   '[DONE]',
 ]
 
-const toolEvents = [
-  '{"choices":[{"delta":{"role":"assistant","content":null},"index":0,"finish_reason":null}]}',
-  '{"choices":[{"delta":{"tool_calls":[{"index":0,"id":"call-1","type":"function","function":{"name":"get_weather","arguments":""}}]},"index":0,"finish_reason":null}]}',
-  '{"choices":[{"delta":{"tool_calls":[{"index":0,"function":{"arguments":"{\\"city\\":\\"Paris\\"}"}}]},"index":0,"finish_reason":null}]}',
-  '{"choices":[{"delta":{},"index":0,"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":20,"completion_tokens":6}}',
-  '[DONE]',
-]
-
-const thinkingEvents = [
-  '{"choices":[{"delta":{"role":"assistant","content":null,"reasoning_content":""},"index":0,"finish_reason":null}]}',
-  '{"choices":[{"delta":{"reasoning_content":"pondering"},"index":0,"finish_reason":null}]}',
-  '{"choices":[{"delta":{"content":"answer","reasoning_content":null},"index":0,"finish_reason":null}]}',
-  '{"choices":[{"delta":{},"index":0,"finish_reason":"stop"}],"usage":{"prompt_tokens":5,"completion_tokens":9}}',
-  '[DONE]',
-]
-
-async function harness(baseURL: string, config: object = {}) {
+async function harness(baseURL: string, overrides: Record<string, unknown> = {}): Promise<Context> {
   const ctx = new Context()
   await ctx.plugin(LlmService)
-  await ctx.plugin(LlmPiAi, { apiKey: 'test-key', baseURL, models: ['deepseek-v4-flash'], ...config })
+  await ctx.plugin(LlmPiAi, {
+    providers: [{ provider: 'deepseek', apiKey: 'test-key', baseURL, ...overrides }],
+  })
   return ctx
 }
 
-describe('PiAiAdapter against a mock server', () => {
-  it('streams a text generation through the assembler', async () => {
+describe('PiAiAdapter provider routing', () => {
+  it('resolves a catalog model dynamically and uses a private endpoint', async () => {
     const server = await mockServer([{ events: textEvents }])
     const ctx = await harness(server.url)
-
     const result = await assemble(ctx, {
       model: 'deepseek-v4-flash',
       messages: [{ role: 'user', content: [{ type: 'text', text: 'hi' }] }],
     })
     expect(result.message.content).toEqual([{ type: 'text', text: 'hello' }])
     expect(result.finish).toEqual({ kind: 'stop' })
-    expect(result.usage).toMatchObject({ inputTokens: 3, outputTokens: 1 })
+    expect(result.usage).toEqual({ inputTokens: 3, outputTokens: 1 })
+    expect(server.paths).toEqual(['/chat/completions'])
+  })
 
-    // Attribution reaches the wire through pi-ai's headers hook: the exact
-    // shared User-Agent, and no provider-specific headers under the
-    // User-Agent-only contract.
+  it('merges profile headers with Harness attribution winning', async () => {
+    const server = await mockServer([{ events: textEvents }])
+    const ctx = await harness(server.url, {
+      headers: { 'x-company': 'private', 'User-Agent': 'wrong' },
+    })
+    await assemble(ctx, { model: 'deepseek-v4-flash', messages: [] })
+    expect(server.headers[0]?.['x-company']).toBe('private')
     expect(server.headers[0]?.['user-agent']).toBe(userAgent())
-    expect(server.headers[0]).not.toHaveProperty('http-referer')
-    expect(server.headers[0]).not.toHaveProperty('x-openrouter-title')
-    expect(server.headers[0]).not.toHaveProperty('x-openrouter-categories')
   })
 
-  it('streams tool calls with re-stringified arguments', async () => {
-    const server = await mockServer([{ events: toolEvents }])
-    const ctx = await harness(server.url)
-
-    const result = await assemble(ctx,{
-      model: 'deepseek-v4-flash',
-      messages: [{ role: 'user', content: [{ type: 'text', text: 'weather?' }] }],
-      tools: [{
-        name: 'get_weather',
-        description: 'Get weather',
-        parameters: { type: 'object', properties: { city: { type: 'string' } } },
-      }],
+  it('forwards common stream options and profile reasoning', async () => {
+    const server = await mockServer([{ events: textEvents }])
+    const ctx = await harness(server.url, {
+      reasoning: 'xhigh',
+      cacheRetention: 'none',
+      transport: 'sse',
+      timeoutMs: 5000,
+      websocketConnectTimeoutMs: 3000,
+      maxRetries: 0,
+      maxRetryDelayMs: 10,
+      thinkingBudgets: { high: 2048 },
     })
-    expect(result.finish).toEqual({ kind: 'tool-calls' })
-    const call = result.message.content.find(block => block.type === 'tool-call')
-    expect(call).toMatchObject({ name: 'get_weather', arguments: '{"city":"Paris"}' })
-  })
-
-  it('maps reasoning_content streams to reasoning blocks', async () => {
-    const server = await mockServer([{ events: thinkingEvents }])
-    const ctx = await harness(server.url, { reasoning: 'high' })
-
-    const result = await assemble(ctx,{
-      model: 'deepseek-v4-flash',
-      messages: [{ role: 'user', content: [{ type: 'text', text: 'think' }] }],
-    })
-    expect(result.message.content).toEqual([
-      { type: 'reasoning', text: 'pondering' },
-      { type: 'text', text: 'answer' },
-    ])
-  })
-
-  it('sends DeepSeek thinking fields when reasoning is configured', async () => {
-    const server = await mockServer([{ events: textEvents }])
-    const ctx = await harness(server.url, { reasoning: 'xhigh' })
-    await assemble(ctx,{ model: 'deepseek-v4-flash', messages: [] })
-    expect(server.requests[0]).toMatchObject({
-      thinking: { type: 'enabled' },
-      reasoning_effort: 'max', // xhigh maps to max via thinkingLevelMap
-    })
-  })
-
-  it('disables thinking for reasoning: off', async () => {
-    const server = await mockServer([{ events: textEvents }])
-    const ctx = await harness(server.url, { reasoning: 'off' })
-    await assemble(ctx,{ model: 'deepseek-v4-flash', messages: [] })
-    expect(server.requests[0]).toMatchObject({ thinking: { type: 'disabled' } })
-  })
-
-  it('injects stop sequences through onPayload', async () => {
-    const server = await mockServer([{ events: textEvents }])
-    const ctx = await harness(server.url)
-    await assemble(ctx,{ model: 'deepseek-v4-flash', messages: [], stop: ['END'] })
-    expect(server.requests[0]).toMatchObject({ stop: ['END'] })
-  })
-
-  it('scrubs pi-ai\'s own per-tool strict default through onPayload', async () => {
-    const server = await mockServer([{ events: textEvents }])
-    const ctx = await harness(server.url)
-    await assemble(ctx,{
+    await assemble(ctx, {
       model: 'deepseek-v4-flash',
       messages: [],
-      tools: [
-        { name: 'alpha', description: 'a', parameters: {} },
-        { name: 'beta', description: 'b', parameters: {} },
-      ],
+      temperature: 0.2,
+      maxTokens: 77,
+      sessionId: 'session-for-pi' as never,
     })
-
-    // pi-ai stamps `strict` on every serialized tool function; the harness
-    // contract has none and the hand-rolled twin sends no such field, so the
-    // payload fixup must have deleted it from every tool.
-    const request = server.requests[0] as { tools: { function: { name: string; strict?: boolean } }[] }
-    expect(request.tools.map(tool => tool.function.name)).toEqual(['alpha', 'beta'])
-    for (const tool of request.tools) {
-      expect('strict' in tool.function).toBe(false)
-    }
-  })
-
-  it('preserves raw replayed tool-call arguments in the provider payload', async () => {
-    const server = await mockServer([{ events: textEvents }])
-    const ctx = await harness(server.url)
-    await assemble(ctx,{
+    expect(server.requests[0]).toMatchObject({
       model: 'deepseek-v4-flash',
-      messages: [{
-        role: 'assistant',
-        content: [{ type: 'tool-call', id: CallId('broken'), name: 'f', arguments: '{broken' }],
-      }],
+      temperature: 0.2,
+      max_completion_tokens: 77,
+      thinking: { type: 'enabled' },
+      reasoning_effort: 'max',
     })
-
-    const request = server.requests[0] as { messages: { role: string; tool_calls?: { id: string; function: { arguments: string } }[] }[] }
-    const assistant = request.messages.find(message => message.role === 'assistant')
-    expect(assistant?.tool_calls?.[0]?.function.arguments).toBe('{broken')
   })
 
-  it('maps HTTP errors to error finish chunks (pi-ai in-stream style)', async () => {
-    const server = await mockServer([{
-      status: 401,
-      body: JSON.stringify({ error: { message: 'bad key' } }),
-    }])
+  it('preserves omitted profile options when constructing the adapter directly', async () => {
+    const server = await mockServer([{ events: textEvents }])
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    ctx.llm.registerAdapter(['deepseek'], new PiAiAdapter({
+      profiles: [{ provider: 'deepseek', apiKey: 'test-key', baseURL: server.url }],
+    }))
+
+    const result = await assemble(ctx, { model: 'deepseek-v4-flash', messages: [] })
+
+    expect(result.message.content).toEqual([{ type: 'text', text: 'hello' }])
+  })
+
+  it('rejects stop sequences rather than silently ignoring them', async () => {
+    const server = await mockServer([])
     const ctx = await harness(server.url)
-    const result = await assemble(ctx,{ model: 'deepseek-v4-flash', messages: [] })
-    expect(result.finish).toMatchObject({ kind: 'error', code: 'AUTH' })
-    expect((result.finish as { message: string }).message).toMatch(/bad key|401/)
+    await expect(assemble(ctx, { model: 'deepseek-v4-flash', messages: [], stop: ['END'] }))
+      .rejects.toMatchObject({ code: 'UNSUPPORTED_OPTION' })
+    expect(server.requests).toEqual([])
+  })
+
+  it('rejects unknown catalog models before network I/O', async () => {
+    const server = await mockServer([])
+    const ctx = await harness(server.url)
+    await expect(assemble(ctx, { model: 'not-in-the-catalog', messages: [] }))
+      .rejects.toMatchObject({ code: 'UNKNOWN_MODEL' })
+    expect(server.requests).toEqual([])
+  })
+
+  it('uses the catalog API implementation, including OpenAI Responses', async () => {
+    const server = await mockServer([{ status: 401, body: JSON.stringify({ error: { message: 'expected mock failure' } }) }])
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    await ctx.plugin(LlmPiAi, {
+      providers: [{ provider: 'openai', apiKey: 'test-key', baseURL: `${server.url}/v1`, maxRetries: 0 }],
+    })
+    const result = await assemble(ctx, { provider: 'openai', model: 'gpt-4.1', messages: [] })
+    expect(result.finish.kind).toBe('error')
+    expect(server.paths).toEqual(['/v1/responses'])
   })
 
   it.each([
+    [401, 'AUTH'],
     [400, 'INVALID_REQUEST'],
     [429, 'RATE_LIMIT'],
     [500, 'SERVER'],
-  ] as const)('maps HTTP %s to stable error code %s', async (status, code) => {
+  ] as const)('maps HTTP %s failures to %s', async (status, code) => {
     const server = await mockServer([{ status, body: JSON.stringify({ error: { message: `provider ${status}` } }) }])
-    const ctx = await harness(server.url)
-    const result = await assemble(ctx,{ model: 'deepseek-v4-flash', messages: [] })
+    const ctx = await harness(server.url, { maxRetries: 0 })
+    const result = await assemble(ctx, { model: 'deepseek-v4-flash', messages: [] })
     expect(result.finish).toMatchObject({ kind: 'error', code })
   })
+})
 
-  it('registers/unregisters models on the llm service (HMR safety)', async () => {
+describe('provider profile lifecycle', () => {
+  it('registers every profile atomically and unregisters on dispose', async () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
-    const fiber = await ctx.plugin(LlmPiAi, { apiKey: 'k', baseURL: 'http://127.0.0.1:1' })
-    expect(ctx.llm.models().sort()).toEqual(['deepseek-v4-flash', 'deepseek-v4-pro'])
+    const fiber = await ctx.plugin(LlmPiAi, {
+      providers: [{ provider: 'openai' }, { provider: 'anthropic' }],
+    })
+    expect(ctx.llm.listProviders()).toEqual([
+      { id: 'openai', name: 'openai' },
+      { id: 'anthropic', name: 'anthropic' },
+    ])
     await fiber.dispose()
-    expect(ctx.llm.models()).toEqual([])
+    expect(ctx.llm.listProviders()).toEqual([])
   })
 
-  it('throws a clear error when no API key is available', async () => {
-    const previous = process.env.DEEPSEEK_API_KEY
-    delete process.env.DEEPSEEK_API_KEY
-    try {
-      const ctx = new Context()
-      await ctx.plugin(LlmService)
-      await expect(ctx.plugin(LlmPiAi, {})).rejects.toThrow(/an API key is required/)
-    } finally {
-      if (previous !== undefined) process.env.DEEPSEEK_API_KEY = previous
-    }
-  })
-})
-
-describe('option spreads and env fallbacks', () => {
-  it('forwards temperature, maxTokens, and signal', async () => {
-    const server = await mockServer([{ events: textEvents }])
-    const ctx = await harness(server.url)
-    const controller = new AbortController()
-    await assemble(ctx,{
-      model: 'deepseek-v4-flash',
-      messages: [],
-      temperature: 0.5,
-      maxTokens: 40,
-      signal: controller.signal,
+  it('exposes the installed pi-ai model catalog through provider-neutral metadata', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    await ctx.plugin(LlmPiAi, { providers: [{ provider: 'openai' }] })
+    const models = await ctx.llm.listModels('openai')
+    expect(models.find(model => model.id === 'gpt-4.1')).toEqual({
+      provider: 'openai', id: 'gpt-4.1', name: 'GPT-4.1',
     })
-    expect(server.requests[0]).toMatchObject({ temperature: 0.5, max_tokens: 40 })
+    expect(models.every(model => model.provider === 'openai')).toBe(true)
   })
 
-  it('falls back to DEEPSEEK_API_KEY / DEEPSEEK_BASE_URL env vars', async () => {
+  it('accepts absent credentials for pi-ai ambient authentication', async () => {
+    vi.stubEnv('DEEPSEEK_API_KEY', 'ambient-key')
     const server = await mockServer([{ events: textEvents }])
-    vi.stubEnv('DEEPSEEK_API_KEY', 'env-key')
-    vi.stubEnv('DEEPSEEK_BASE_URL', server.url)
-    try {
-      const ctx = new Context()
-      await ctx.plugin(LlmService)
-      await ctx.plugin(LlmPiAi, { models: ['deepseek-v4-flash'] })
-      await assemble(ctx,{ model: 'deepseek-v4-flash', messages: [] })
-      expect(server.requests).toHaveLength(1)
-    } finally {
-      vi.unstubAllEnvs()
+    const ctx = await harness(server.url, { apiKey: undefined })
+    await assemble(ctx, { model: 'deepseek-v4-flash', messages: [] })
+    expect(server.headers[0]?.authorization).toBe('Bearer ambient-key')
+  })
+
+  it('validates empty, duplicate, unknown, and explicitly blank profiles', () => {
+    expect(() => resolveProfiles([])).toThrow(/at least one/)
+    expect(() => resolveProfiles([{ provider: '' }])).toThrow(/non-empty/)
+    expect(() => resolveProfiles([{ provider: 'not-real' }])).toThrow(/unknown/)
+    expect(() => resolveProfiles([{ provider: 'openai' }, { provider: 'openai' }])).toThrow(/duplicate/)
+    expect(() => resolveProfiles([{ provider: 'openai', apiKey: '' }])).toThrow(/empty apiKey/)
+    expect(() => resolveProfiles([{ provider: 'openai', apiKey: '  ' }])).toThrow(/empty apiKey/)
+    expect(() => resolveProfiles([{ provider: 'openai', baseURL: '' }])).toThrow(/empty baseURL/)
+  })
+
+  it('rejects negative or fractional stream tunables at schema validation', () => {
+    const invalid = [
+      { timeoutMs: -1 },
+      { websocketConnectTimeoutMs: -1 },
+      { maxRetries: -1 },
+      { maxRetries: 0.5 },
+      { maxRetryDelayMs: -1 },
+    ]
+    for (const entry of invalid) {
+      expect(() => new LlmPiAi.Config({ providers: [{ provider: 'openai', ...entry }] })).toThrow()
     }
   })
 
-  it('defaults to the public base URL without config or env', async () => {
-    vi.stubEnv('DEEPSEEK_API_KEY', 'k')
-    vi.stubEnv('DEEPSEEK_BASE_URL', undefined)
-    try {
-      const ctx = new Context()
-      await ctx.plugin(LlmService)
-      await ctx.plugin(LlmPiAi, {})
-      expect(ctx.llm.models().length).toBeGreaterThan(0)
-    } finally {
-      vi.unstubAllEnvs()
-    }
+  it('constructs the adapter directly and rejects routes it does not own', async () => {
+    const adapter = new PiAiAdapter({ profiles: [{ provider: 'openai' }] })
+    await expect(adapter.listModels('anthropic')).rejects.toMatchObject({ code: 'NO_ADAPTER' })
+    await expect((async () => {
+      for await (const _chunk of adapter.stream({ provider: 'anthropic', model: 'claude-sonnet-4', messages: [] })) { /* drain */ }
+    })()).rejects.toMatchObject({ code: 'NO_ADAPTER' })
+    expect(new LlmError('x', 'X')).toBeInstanceOf(Error)
   })
 })
 
-describe('buildModel', () => {
-  it('builds a DeepSeek-compat openai-completions model descriptor', () => {
-    const model = buildModel('deepseek-v4-pro', { apiKey: 'k', baseURL: 'http://x', reasoning: 'high' })
-    expect(model).toMatchObject({
-      id: 'deepseek-v4-pro',
-      api: 'openai-completions',
+describe('abort wiring', () => {
+  it('resolves catalog endpoints without an override before honoring pre-abort', async () => {
+    const adapter = new PiAiAdapter({ profiles: [{ provider: 'deepseek', apiKey: 'test-key', maxRetries: 0 }] })
+    const controller = new AbortController()
+    controller.abort('already stopped')
+    const chunks = []
+    for await (const chunk of adapter.stream({
       provider: 'deepseek',
-      baseUrl: 'http://x',
-      reasoning: true,
-      compat: { thinkingFormat: 'deepseek', requiresReasoningContentOnAssistantMessages: true },
-    })
-  })
-
-  it('keeps reasoning true even for off (pi-ai gates the thinking field on it)', () => {
-    // 'off' yields {thinking: {type: 'disabled'}} on the wire — pi-ai only
-    // emits the field at all when model.reasoning is true.
-    expect(buildModel('m', { apiKey: 'k', baseURL: 'http://x', reasoning: 'off' }).reasoning).toBe(true)
-  })
-
-  it('adapter is constructible directly for embedding', () => {
-    expect(new PiAiAdapter({ apiKey: 'k', baseURL: 'http://x' })).toBeInstanceOf(PiAiAdapter)
-  })
-})
-
-describe('provider reasoning, passback, and early-stream cancellation', () => {
-  it('defaults omitted reasoning config to thinking ENABLED (provider default)', async () => {
-    const server = await mockServer([{ events: textEvents }])
-    const ctx = await harness(server.url) // no reasoning key at all
-    await assemble(ctx,{ model: 'deepseek-v4-flash', messages: [] })
-    const request = server.requests[0] as Record<string, unknown>
-    expect(request.thinking).toEqual({ type: 'enabled' })
-    expect('reasoning_effort' in request).toBe(false)
-  })
-
-  it('replays reasoning_content on assistant tool-call turns (passback rule)', async () => {
-    const server = await mockServer([{ events: textEvents }])
-    const ctx = await harness(server.url)
-    await assemble(ctx,{
-      model: 'deepseek-v4-flash',
-      messages: [
-        { role: 'user', content: [{ type: 'text', text: 'weather?' }] },
-        {
-          role: 'assistant',
-          content: [
-            { type: 'reasoning', text: 'I should check.' },
-            { type: 'tool-call', id: CallId('c1'), name: 'get_weather', arguments: '{"city":"Paris"}' },
-          ],
-        },
-        {
-          role: 'user',
-          content: [{ type: 'tool-result', toolCallId: CallId('c1'), content: [{ type: 'text', text: 'Sunny' }] }],
-        },
-      ],
-    })
-    const request = server.requests[0] as { messages: { role: string; reasoning_content?: string }[] }
-    const assistant = request.messages.find(message => message.role === 'assistant')
-    expect(assistant?.reasoning_content).toBe('I should check.')
-  })
-
-  it('aborts the upstream request when the consumer stops streaming early', async () => {
-    // Slow server: write one chunk, then hold the connection open and record
-    // whether the socket closes (the adapter must cancel on early break).
-    let socketClosed = false
-    const server = createServer((request: IncomingMessage, response: ServerResponse) => {
-      request.on('data', () => undefined)
-      request.on('end', () => {
-        response.writeHead(200, { 'content-type': 'text/event-stream' })
-        response.write(`data: ${textEvents[0]}\n\n`)
-        response.write(`data: ${textEvents[1]}\n\n`)
-        // never finish; rely on client abort
-        request.socket.on('close', () => { socketClosed = true })
-      })
-    })
-    servers.push(server)
-    await new Promise<void>(resolve => server.listen(0, '127.0.0.1', resolve))
-    const address = server.address()
-    if (address === null || typeof address === 'string') throw new Error('no port')
-    const ctx = await harness(`http://127.0.0.1:${address.port}`)
-
-    for await (const chunk of ctx.llm.stream({ model: 'deepseek-v4-flash', messages: [] })) {
-      if (chunk.type === 'text-delta') break // stop early mid-stream
-    }
-    // The finally-abort must reach the server as a closed socket.
-    await vi.waitFor(() => { expect(socketClosed).toBe(true) }, { timeout: 5_000 })
-  })
-})
-
-describe('caller cancellation', () => {
-  it('honors a pre-aborted caller signal', async () => {
-    const ctx = await harness('http://127.0.0.1:1')
-    const controller = new AbortController()
-    controller.abort('already cancelled')
-    // pi-ai surfaces the abort as an in-stream error event → aborted finish.
-    const result = await assemble(ctx,{
       model: 'deepseek-v4-flash',
       messages: [],
       signal: controller.signal,
-    })
+    })) chunks.push(chunk)
+    expect(chunks.at(-1)).toMatchObject({ type: 'finish', reason: { kind: 'aborted' } })
+  })
+
+  it('honors a pre-aborted caller signal', async () => {
+    const server = await mockServer([{ events: textEvents, delayMs: 20 }])
+    const ctx = await harness(server.url)
+    const controller = new AbortController()
+    controller.abort('already stopped')
+    const result = await assemble(ctx, { model: 'deepseek-v4-flash', messages: [], signal: controller.signal })
     expect(result.finish.kind).toBe('aborted')
   })
 
-  it('propagates a mid-stream caller abort to the upstream request', async () => {
-    const server = await mockServer([{ events: textEvents }])
+  it('forwards an abort that arrives while provider streaming is active', async () => {
+    const server = await mockServer([{ events: textEvents, delayMs: 30 }])
     const ctx = await harness(server.url)
     const controller = new AbortController()
-    const pending = assemble(ctx,{
-      model: 'deepseek-v4-flash',
-      messages: [],
-      signal: controller.signal,
+    const resultPromise = assemble(ctx, {
+      model: 'deepseek-v4-flash', messages: [], signal: controller.signal,
     })
-    controller.abort()
-    const result = await pending
-    // Either the abort lands before any chunk (aborted) or after the tiny
-    // mock stream finished (stop) — both are valid races; never a hang.
-    expect(['aborted', 'stop']).toContain(result.finish.kind)
+    setTimeout(() => { controller.abort('stopped during stream') }, 10)
+    const result = await resultPromise
+    expect(result.finish.kind).toBe('aborted')
+  })
+
+  it('aborts upstream when a consumer stops early', async () => {
+    const server = await mockServer([{ events: textEvents, delayMs: 30 }])
+    const ctx = await harness(server.url)
+    for await (const chunk of ctx.llm.stream({ provider: 'deepseek', model: 'deepseek-v4-flash', messages: [] })) {
+      if (chunk.type === 'block-start') break
+    }
+    await new Promise(resolve => setTimeout(resolve, 20))
+    expect(server.requests).toHaveLength(1)
   })
 })
diff --git a/packages/llm/llm-pi-ai/tests/assemble.ts b/packages/llm/llm-pi-ai/tests/assemble.ts
index b0182615e0..494eeac494 100644
--- a/packages/llm/llm-pi-ai/tests/assemble.ts
+++ b/packages/llm/llm-pi-ai/tests/assemble.ts
@@ -15,11 +15,19 @@ export interface AssembledResult {
   finish: FinishReason
 }
 
-export async function assemble(ctx: Context, options: GenerateOptions): Promise<AssembledResult> {
+export async function assemble(ctx: Context, options: Omit<GenerateOptions, 'provider'> & { provider?: string }): Promise<AssembledResult> {
   const assembler = new BlockAssembler()
-  for await (const chunk of ctx.llm.stream(options)) assembler.push(chunk)
+  const request = { provider: 'deepseek', ...options }
+  for await (const chunk of ctx.llm.stream(request)) assembler.push(chunk)
   return {
-    message: assembler.message(),
+    message: {
+      ...assembler.message(),
+      provenance: {
+        provider: request.provider,
+        model: request.model,
+        ...assembler.replayState === undefined ? {} : { replayState: assembler.replayState },
+      },
+    },
     ...assembler.usage !== undefined ? { usage: assembler.usage } : {},
     finish: assembler.finish,
   }
diff --git a/packages/llm/llm-pi-ai/tests/convert.spec.ts b/packages/llm/llm-pi-ai/tests/convert.spec.ts
index 078d2a4d3b..50955f0607 100644
--- a/packages/llm/llm-pi-ai/tests/convert.spec.ts
+++ b/packages/llm/llm-pi-ai/tests/convert.spec.ts
@@ -1,8 +1,8 @@
 import { describe, expect, it } from 'vitest'
-import { CallId } from '@deepseek-ai/dsh-llm'
+import { CallId, LlmError } from '@deepseek-ai/dsh-llm'
 import type { ContentBlock, StreamChunk } from '@deepseek-ai/dsh-llm'
 import type { AssistantMessage, AssistantMessageEvent, Usage } from '@earendil-works/pi-ai'
-import { mapStopReason, mapUsage, toPiContext, toStreamChunks } from '@deepseek-ai/dsh-llm-pi-ai'
+import { mapStopReason, mapUsage, toPiContext, toPiReplayState, toStreamChunks } from '@deepseek-ai/dsh-llm-pi-ai'
 
 function usage(input = 0, output = 0, cacheRead = 0, cacheWrite = 0): Usage {
   return {
@@ -42,6 +42,7 @@ async function collect(stream: AsyncIterable<StreamChunk>): Promise<StreamChunk[
 describe('toPiContext', () => {
   it('maps system prompt, user text, and tools', () => {
     const context = toPiContext({
+      provider: 'deepseek',
       model: 'deepseek-v4-flash',
       system: 'be helpful',
       messages: [{ role: 'user', content: [{ type: 'text', text: 'hi' }] }],
@@ -55,13 +56,14 @@ describe('toPiContext', () => {
   })
 
   it('omits empty tools and absent system prompt', () => {
-    const context = toPiContext({ model: 'm', messages: [], tools: [] })
+    const context = toPiContext({ provider: 'deepseek', model: 'm', messages: [], tools: [] })
     expect(context.systemPrompt).toBeUndefined()
     expect(context.tools).toBeUndefined()
   })
 
   it('maps assistant text/reasoning/tool-call blocks', () => {
     const context = toPiContext({
+      provider: 'deepseek',
       model: 'm',
       messages: [{
         role: 'assistant',
@@ -76,8 +78,7 @@ describe('toPiContext', () => {
     expect(message.role).toBe('assistant')
     expect(message.stopReason).toBe('toolUse')
     expect(message.content).toEqual([
-      // thinkingSignature names the replay field — DeepSeek's passback rule.
-      { type: 'thinking', thinking: 'hmm', thinkingSignature: 'reasoning_content' },
+      { type: 'thinking', thinking: 'hmm' },
       { type: 'text', text: 'calling' },
       { type: 'toolCall', id: 'c1', name: 'f', arguments: { a: 1 } },
     ])
@@ -85,6 +86,7 @@ describe('toPiContext', () => {
 
   it('marks tool-call-free assistant messages with stopReason stop', () => {
     const context = toPiContext({
+      provider: 'deepseek',
       model: 'm',
       messages: [{ role: 'assistant', content: [{ type: 'text', text: 'done' }] }],
     })
@@ -93,6 +95,7 @@ describe('toPiContext', () => {
 
   it('parses malformed tool-call arguments to {}', () => {
     const context = toPiContext({
+      provider: 'deepseek',
       model: 'm',
       messages: [{
         role: 'assistant',
@@ -105,6 +108,7 @@ describe('toPiContext', () => {
 
   it('parses non-object argument JSON (arrays, scalars) to {}', () => {
     const context = toPiContext({
+      provider: 'deepseek',
       model: 'm',
       messages: [{
         role: 'assistant',
@@ -116,6 +120,7 @@ describe('toPiContext', () => {
 
   it('recovers toolName for tool results from the preceding assistant call', () => {
     const context = toPiContext({
+      provider: 'deepseek',
       model: 'm',
       messages: [
         {
@@ -140,6 +145,7 @@ describe('toPiContext', () => {
 
   it('labels unmatched tool results with toolName unknown and keeps isError', () => {
     const context = toPiContext({
+      provider: 'deepseek',
       model: 'm',
       messages: [{
         role: 'user',
@@ -156,6 +162,7 @@ describe('toPiContext', () => {
 
   it('splits mixed user text + tool results and folds history system messages', () => {
     const context = toPiContext({
+      provider: 'deepseek',
       model: 'm',
       messages: [
         { role: 'system', content: [{ type: 'text', text: 'rule' }] },
@@ -173,6 +180,7 @@ describe('toPiContext', () => {
 
   it('skips plugin-added (unknown) blocks in assistant content', () => {
     const context = toPiContext({
+      provider: 'deepseek',
       model: 'm',
       messages: [{
         role: 'assistant',
@@ -184,6 +192,195 @@ describe('toPiContext', () => {
     })
     expect((context.messages[0] as AssistantMessage).content).toEqual([{ type: 'text', text: 'visible' }])
   })
+
+  it('recombines durable content with pi-ai replay metadata across target providers and models', () => {
+    const state = toPiReplayState(assistant({
+      api: 'openai-responses',
+      provider: 'openai',
+      model: 'gpt-5',
+      responseModel: 'gpt-5-2026-01-01',
+      responseId: 'resp_123',
+      stopReason: 'toolUse',
+      content: [
+        { type: 'thinking', thinking: 'private reasoning', thinkingSignature: 'think-sig', redacted: true },
+        { type: 'text', text: 'calling', textSignature: 'text-sig' },
+        { type: 'toolCall', id: 'c1', name: 'f', arguments: { a: 1 }, thoughtSignature: 'tool-sig' },
+      ],
+    }))
+    const context = toPiContext({
+      provider: 'anthropic',
+      model: 'claude-next',
+      messages: [{
+        role: 'assistant',
+        content: [
+          { type: 'reasoning', text: 'private reasoning' },
+          { type: 'text', text: 'calling' },
+          { type: 'tool-call', id: CallId('c1'), name: 'f', arguments: '{"a":1}' },
+        ],
+        provenance: { provider: 'openai', model: 'gpt-5', replayState: state },
+      }],
+    })
+
+    expect(context.messages[0]).toMatchObject({
+      role: 'assistant',
+      api: 'openai-responses',
+      provider: 'openai',
+      model: 'gpt-5',
+      responseModel: 'gpt-5-2026-01-01',
+      responseId: 'resp_123',
+      stopReason: 'toolUse',
+      content: [
+        { type: 'thinking', thinking: 'private reasoning', thinkingSignature: 'think-sig', redacted: true },
+        { type: 'text', text: 'calling', textSignature: 'text-sig' },
+        { type: 'toolCall', id: 'c1', name: 'f', arguments: { a: 1 }, thoughtSignature: 'tool-sig' },
+      ],
+    })
+  })
+
+  it('replays all native block kinds when optional metadata is absent', () => {
+    const state = toPiReplayState(assistant({
+      content: [
+        { type: 'thinking', thinking: 'private reasoning' },
+        { type: 'text', text: 'calling' },
+        { type: 'toolCall', id: 'c1', name: 'f', arguments: { a: 1 } },
+      ],
+    }))
+    const context = toPiContext({
+      provider: 'deepseek',
+      model: 'new-model',
+      messages: [{
+        role: 'assistant',
+        content: [
+          { type: 'reasoning', text: 'private reasoning' },
+          { type: 'text', text: 'calling' },
+          { type: 'tool-call', id: CallId('c1'), name: 'f', arguments: '{"a":1}' },
+        ],
+        provenance: { provider: 'deepseek', model: 'deepseek-v4-flash', replayState: state },
+      }],
+    })
+
+    expect(context.messages[0]).toMatchObject({
+      role: 'assistant',
+      content: [
+        { type: 'thinking', thinking: 'private reasoning' },
+        { type: 'text', text: 'calling' },
+        { type: 'toolCall', id: 'c1', name: 'f', arguments: { a: 1 } },
+      ],
+    })
+    expect(context.messages[0]).not.toHaveProperty('responseModel')
+    expect(context.messages[0]).not.toHaveProperty('responseId')
+  })
+
+  it('rejects unsupported replay-state versions with a stable error code', () => {
+    try {
+      toPiContext({
+        provider: 'deepseek',
+        model: 'm',
+        messages: [{
+          role: 'assistant',
+          content: [{ type: 'text', text: 'done' }],
+          provenance: {
+            provider: 'deepseek',
+            model: 'old',
+            replayState: { kind: 'pi-ai', version: 2 },
+          },
+        }],
+      })
+      expect.fail('expected invalid replay state')
+    } catch (error: unknown) {
+      expect(error).toBeInstanceOf(LlmError)
+      expect((error as LlmError).code).toBe('INVALID_REPLAY_STATE')
+      expect((error as Error).message).toContain('unsupported version 2')
+    }
+  })
+
+  it('rejects replay metadata whose blocks do not match the durable content', () => {
+    const state = toPiReplayState(assistant({ content: [{ type: 'text', text: 'done' }] }))
+    expect(() => toPiContext({
+      provider: 'deepseek',
+      model: 'm',
+      messages: [{
+        role: 'assistant',
+        content: [{ type: 'reasoning', text: 'done' }],
+        provenance: { provider: 'deepseek', model: 'deepseek-v4-flash', replayState: state },
+      }],
+    })).toThrow(/block 0 does not match assistant content/)
+  })
+
+  it('rejects replay metadata whose block count differs from durable content', () => {
+    const state = toPiReplayState(assistant())
+    expect(() => toPiContext({
+      provider: 'deepseek',
+      model: 'm',
+      messages: [{
+        role: 'assistant',
+        content: [{ type: 'text', text: 'done' }],
+        provenance: { provider: 'deepseek', model: 'deepseek-v4-flash', replayState: state },
+      }],
+    })).toThrow(/block count does not match assistant content/)
+  })
+
+  const validReplay = {
+    kind: 'pi-ai',
+    version: 1,
+    api: 'openai-completions',
+    provider: 'deepseek',
+    model: 'deepseek-v4-flash',
+    stopReason: 'stop',
+    blocks: [{ type: 'text' }],
+  }
+
+  it.each([
+    ['provider', { ...validReplay, provider: 'openai' }],
+    ['model', { ...validReplay, model: 'deepseek-v4-pro' }],
+  ])('rejects replay metadata whose %s differs from assistant provenance', (field, replayState) => {
+    try {
+      toPiContext({
+        provider: 'deepseek',
+        model: 'next-model',
+        messages: [{
+          role: 'assistant',
+          content: [{ type: 'text', text: 'done' }],
+          provenance: { provider: 'deepseek', model: 'deepseek-v4-flash', replayState },
+        }],
+      })
+      expect.fail('expected invalid replay state')
+    } catch (error: unknown) {
+      expect(error).toBeInstanceOf(LlmError)
+      expect((error as LlmError).code).toBe('INVALID_REPLAY_STATE')
+      expect((error as Error).message).toContain(`${field} does not match assistant provenance`)
+    }
+  })
+
+  it.each([
+    ['number state', 1, 'expected an object'],
+    ['null state', null, 'expected an object'],
+    ['array state', [], 'expected an object'],
+    ['unknown kind', { ...validReplay, kind: 'other' }, 'unknown state kind'],
+    ['non-string api', { ...validReplay, api: 1 }, 'api must be a non-empty string'],
+    ['empty provider', { ...validReplay, provider: '' }, 'provider must be a non-empty string'],
+    ['missing model', { ...validReplay, model: undefined }, 'model must be a non-empty string'],
+    ['unknown stop reason', { ...validReplay, stopReason: 'pause' }, 'unknown stopReason'],
+    ['non-string response model', { ...validReplay, responseModel: 1 }, 'responseModel must be a string'],
+    ['non-string response id', { ...validReplay, responseId: 1 }, 'responseId must be a string'],
+    ['non-array blocks', { ...validReplay, blocks: 'text' }, 'blocks must be an array'],
+    ['number block', { ...validReplay, blocks: [1] }, 'block 0 must be an object'],
+    ['null block', { ...validReplay, blocks: [null] }, 'block 0 must be an object'],
+    ['array block', { ...validReplay, blocks: [[]] }, 'block 0 must be an object'],
+    ['unknown block type', { ...validReplay, blocks: [{ type: 'audio' }] }, 'block 0 has an unknown type'],
+    ['non-string signature', { ...validReplay, blocks: [{ type: 'text', textSignature: 1 }] }, 'textSignature must be a string'],
+    ['non-boolean redaction', { ...validReplay, blocks: [{ type: 'reasoning', redacted: 'yes' }] }, 'redacted must be boolean'],
+  ])('rejects malformed replay state: %s', (_name, replayState, message) => {
+    expect(() => toPiContext({
+      provider: 'deepseek',
+      model: 'm',
+      messages: [{
+        role: 'assistant',
+        content: [{ type: 'text', text: 'done' }],
+        provenance: { provider: 'deepseek', model: 'deepseek-v4-flash', replayState },
+      }],
+    })).toThrow(message)
+  })
 })
 
 describe('toStreamChunks', () => {
@@ -205,7 +402,19 @@ describe('toStreamChunks', () => {
       { type: 'text-delta', index: 0, text: 'hi' },
       { type: 'block-end', index: 0, block: { type: 'text', text: 'hi' } },
       { type: 'usage', usage: { inputTokens: 3, outputTokens: 2 } },
-      { type: 'finish', reason: { kind: 'stop' } },
+      {
+        type: 'finish',
+        reason: { kind: 'stop' },
+        replayState: {
+          kind: 'pi-ai',
+          version: 1,
+          api: 'openai-completions',
+          provider: 'deepseek',
+          model: 'deepseek-v4-flash',
+          stopReason: 'stop',
+          blocks: [{ type: 'text' }],
+        },
+      },
     ])
   })
 
@@ -234,7 +443,7 @@ describe('toStreamChunks', () => {
         toolCall: { type: 'toolCall', id: 'call-1', name: 'f', arguments: { a: 1 } },
         partial: partialWithToolCall,
       },
-      { type: 'done', reason: 'toolUse', message: assistant({ stopReason: 'toolUse' }) },
+      { type: 'done', reason: 'toolUse', message: assistant({ content: partialWithToolCall.content, stopReason: 'toolUse' }) },
     )))
     expect(chunks).toEqual([
       { type: 'block-start', index: 0, blockType: 'tool-call' },
@@ -242,7 +451,19 @@ describe('toStreamChunks', () => {
       { type: 'tool-call-delta', index: 0, id: 'call-1', name: 'f', argumentsDelta: ':1}' },
       { type: 'block-end', index: 0, block: { type: 'tool-call', id: 'call-1', name: 'f', arguments: '{"a":1}' } },
       { type: 'usage', usage: { inputTokens: 0, outputTokens: 0 } },
-      { type: 'finish', reason: { kind: 'tool-calls' } },
+      {
+        type: 'finish',
+        reason: { kind: 'tool-calls' },
+        replayState: {
+          kind: 'pi-ai',
+          version: 1,
+          api: 'openai-completions',
+          provider: 'deepseek',
+          model: 'deepseek-v4-flash',
+          stopReason: 'toolUse',
+          blocks: [{ type: 'tool-call' }],
+        },
+      },
     ])
   })
 
diff --git a/packages/llm/llm/README.md b/packages/llm/llm/README.md
index 296fd0c3d3..a67859ed19 100644
--- a/packages/llm/llm/README.md
+++ b/packages/llm/llm/README.md
@@ -8,10 +8,13 @@ An adapter registry plus a single streaming call surface, interceptable via a wa
 
 ### Public API
 
-- `ctx.llm.registerAdapter(models: string[], adapter: LlmAdapter): () => void` Register an adapter for the given model names. Disposed with the calling fiber.
-- `ctx.llm.models(): string[]` — model names with a registered adapter.
+- `ctx.llm.registerAdapter(providers: string[], adapter: LlmAdapter): () => void` Register one adapter instance for the given provider routes. Registration is all-or-nothing, and is disposed with the calling fiber.
+- `ctx.llm.listProviders(): LlmProviderInfo[]` Describe registered provider routes in registration order.
+- `ctx.llm.listModels(provider: string): Promise<LlmModelInfo[]>` Discover the models one registered provider currently advertises.
 - `ctx.llm.stream(options: GenerateOptions): AsyncIterable<StreamChunk>` Stream one model call as raw chunks (token-level deltas). Consumers assemble the chunks into blocks/messages with `BlockAssembler`.
 
+Provider and model metadata is a discovery surface, not a routing whitelist. `registerAdapter()` still owns provider exclusivity, while an adapter may accept model ids absent from `listModels()`; consumers must not reject a request because its model is unlisted. Returned metadata is detached and invalid or duplicate adapter entries fail with `INVALID_ADAPTER` or `INVALID_CATALOG`.
+
 ### Events
 
 | Event | Mode | Purpose |
@@ -20,18 +23,18 @@ An adapter registry plus a single streaming call surface, interceptable via a wa
 
 ### Extension points
 
-- Subclass `LlmAdapter` and call `ctx.llm.registerAdapter(models, adapter)` to add a new model provider.
+- Subclass `LlmAdapter` and call `ctx.llm.registerAdapter(providers, adapter)` to add one or more provider routes. `GenerateOptions.provider` selects the adapter; `GenerateOptions.model` is adapter-owned and may be resolved dynamically. Override `providerInfo()` and asynchronous `listModels()` to expose selector metadata; their defaults use the route id as its name and advertise no models.
 - Wrap `llm/stream` via `ctx.on()` waterfall listeners for caching, retry, logging, rate-limiting, etc.
 
 ### Content-block vocabulary (`types.ts`)
 
-Messages are arrays of typed content blocks: `text`, `reasoning`, `tool-call`, `tool-result`. The union is derived from the merge-extensible `ContentBlockMap`, so plugins can add block types via declaration merging. The core set is limited to blocks every shipping path honors — multimodal content (images, audio, …) has no core block type; a feature that needs one adds it via the map together with the adapter/UI/compaction support that honors it.
+Messages are arrays of typed content blocks: `text`, `reasoning`, `tool-call`, `tool-result`. The union is derived from the merge-extensible `ContentBlockMap`, so plugins can add block types via declaration merging. Assistant messages produced by the loop also carry provider/model provenance and optional adapter-private replay state. Before dispatch, `LlmService` retains that state only when the historical provider route and target provider route are currently owned by the exact same adapter instance; the adapter then decides whether it can restore or convert the state across models/providers. The core block set is limited to blocks every shipping path honors — multimodal content (images, audio, …) has no core block type; a feature that needs one adds it via the map together with the adapter/UI/compaction support that honors it.
 
 Streaming is a raw chunk protocol (`block-start`, `text-delta`, `reasoning-delta`, `tool-call-delta`, `block-end`, `usage`, `finish`). `BlockAssembler` is the single shared implementation that assembles chunks into blocks/messages.
 
 ### Call configuration (`call-config.ts`)
 
-`LlmCallConfig` is the model + sampling scalars of one conversation's requests (`model`, `temperature`, `maxTokens`, `stop` — each mapping 1:1 onto the same-named `GenerateOptions` field). It is per-conversation state recorded in the session log as part of the request header (see the dsh-session `request/header` events), never a silently-adjustable per-call knob: the `agent/request` waterfall proposes a replacement and the loop logs a real change. `callConfigEquals(a, b)` is the field-wise real-change detector; `deepFreeze(value)` is the ownership helper the loop applies to every built request before dispatch (`llm/stream` listeners and adapters read, never rewrite).
+`LlmCallConfig` is the provider + model + sampling scalars of one conversation's requests (`provider`, `model`, `temperature`, `maxTokens`, `stop` — each mapping 1:1 onto the same-named `GenerateOptions` field). It is per-conversation state recorded in the session log as part of the request header (see the dsh-session `request/header` events), never a silently-adjustable per-call knob: the `agent/request` waterfall proposes a replacement and the loop logs a real change. `callConfigEquals(a, b)` is the field-wise real-change detector; `deepFreeze(value)` is the ownership helper the loop applies to every built request before dispatch (`llm/stream` listeners and adapters read, never rewrite).
 
 ### App attribution (`attribution.ts`)
 
@@ -46,7 +49,7 @@ Every product adapter sends application identity on provider HTTP requests. `att
 
 ### Real adapters
 
-Two adapters implement `LlmAdapter` on different internals: [`@deepseek-ai/dsh-llm-deepseek`](../llm-deepseek) uses hand-rolled fetch/SSE, while [`@deepseek-ai/dsh-llm-pi-ai`](../llm-pi-ai) uses `@earendil-works/pi-ai`. Both follow the `StreamChunk` conventions in `types.ts`: usage precedes finish, tool arguments remain raw strings, and errors take one of two sanctioned paths. See [the twin LLM adapters](../../../docs/rfc/implemented/architecture/2026-06-13-twin-llm-adapters.md) for the design rationale.
+Two adapters implement `LlmAdapter` on different internals: [`@deepseek-ai/dsh-llm-deepseek`](../llm-deepseek) uses hand-rolled fetch/SSE for the `deepseek` route, while [`@deepseek-ai/dsh-llm-pi-ai`](../llm-pi-ai) dynamically resolves configured provider/model pairs through `@earendil-works/pi-ai`. Both follow the `StreamChunk` conventions in `types.ts`: usage precedes finish, tool arguments remain raw strings, and errors take one of two sanctioned paths. See [the twin LLM adapters](../../../docs/rfc/implemented/architecture/2026-06-13-twin-llm-adapters.md) for the design rationale.
 
 ## Model Experience
 
diff --git a/packages/llm/llm/src/assembler.ts b/packages/llm/llm/src/assembler.ts
index 163bb5679a..8780718c7a 100644
--- a/packages/llm/llm/src/assembler.ts
+++ b/packages/llm/llm/src/assembler.ts
@@ -36,6 +36,7 @@ export class BlockAssembler {
   private order: number[] = []
   private _usage: TokenUsage | undefined
   private _finish: FinishReason | undefined
+  private _replayState: unknown = undefined
 
   /**
    * Feed one chunk. Returns the completed block when the chunk closes one
@@ -85,6 +86,7 @@ export class BlockAssembler {
       }
       case 'finish': {
         this._finish = chunk.reason
+        this._replayState = chunk.replayState
         return
       }
       default: return assertNever(chunk, 'BlockAssembler.push')
@@ -142,6 +144,11 @@ export class BlockAssembler {
     return this._finish ?? { kind: 'stop' }
   }
 
+  /** Adapter-private replay state from the terminal finish chunk, if any. */
+  get replayState(): unknown {
+    return this._replayState
+  }
+
   /**
    * The assembled assistant message.
    * @returns an assistant-role message over `blocks()` (same open-block assembly rules).
diff --git a/packages/llm/llm/src/call-config.ts b/packages/llm/llm/src/call-config.ts
index aa4c871e2f..82eb68615d 100644
--- a/packages/llm/llm/src/call-config.ts
+++ b/packages/llm/llm/src/call-config.ts
@@ -1,17 +1,18 @@
 /**
- * Conversation call configuration and freeze utilities. Model and sampling
- * values are request-header state that can affect cache reuse; request
- * waterfalls replace them and the loop logs changes instead of allowing
- * silent per-call drift.
+ * Conversation call configuration and freeze utilities. Provider routing,
+ * model, and sampling values are request-header state that can affect cache
+ * reuse; request waterfalls replace them and the loop logs changed snapshots
+ * instead of allowing silent per-call drift.
  * @module dsh-llm/call-config
  */
 
 /**
- * Model + sampling scalars of one conversation's requests. Every field maps
+ * Provider + model + sampling scalars of one conversation's requests. Every field maps
  * 1:1 onto the same-named `GenerateOptions` field; the loop builds requests
  * from the logged header rather than accepting these per call.
  */
 export interface LlmCallConfig {
+  provider: string
   model: string
   temperature?: number
   maxTokens?: number
@@ -21,13 +22,13 @@ export interface LlmCallConfig {
 /**
  * Field-wise equality over {@link LlmCallConfig} — the comparison a caller
  * runs to decide whether a proposed configuration is a real change (worth a
- * logged header delta) or the held one restated.
+ * logged header snapshot) or the held one restated.
  * @param a - one configuration.
  * @param b - the other.
  * @returns whether every field (including the `stop` list, element-wise) matches.
  */
 export function callConfigEquals(a: LlmCallConfig, b: LlmCallConfig): boolean {
-  if (a.model !== b.model || a.temperature !== b.temperature || a.maxTokens !== b.maxTokens) return false
+  if (a.provider !== b.provider || a.model !== b.model || a.temperature !== b.temperature || a.maxTokens !== b.maxTokens) return false
   if (a.stop === undefined || b.stop === undefined) return a.stop === b.stop
   return a.stop.length === b.stop.length && a.stop.every((s, i) => s === b.stop?.[i])
 }
diff --git a/packages/llm/llm/src/index.ts b/packages/llm/llm/src/index.ts
index 08f3f54c51..26e9e3b289 100644
--- a/packages/llm/llm/src/index.ts
+++ b/packages/llm/llm/src/index.ts
@@ -7,8 +7,9 @@
  */
 
 import { Context, Service } from 'cordis'
-import type { GenerateOptions, StreamChunk } from './types.ts'
+import type { GenerateOptions, LlmModelInfo, LlmProviderInfo, Message, StreamChunk } from './types.ts'
 import { HarnessError } from './error.ts'
+import { deepFreeze } from './call-config.ts'
 
 export * from './attribution.ts'
 export * from './brand.ts'
@@ -55,11 +56,31 @@ export class LlmError extends HarnessError {
 
 /**
  * Provider-wire adapter for the harness message and stream vocabulary. Register implementations
- * with `ctx.llm.registerAdapter(models, adapter)`. Every provider HTTP request must include
+ * with `ctx.llm.registerAdapter(providers, adapter)`. Every provider HTTP request must include
  * `attributionHeaders()`; prove that at the wire or library header-hook boundary. The hand-rolled
  * DeepSeek and pi-ai adapters intentionally exercise this contract through different internals.
  */
 export abstract class LlmAdapter {
+  /**
+   * Describe one provider route owned by this adapter.
+   * @param provider - a route passed to `registerAdapter()` for this instance.
+   * @returns detached display metadata whose id must equal `provider`.
+   */
+  providerInfo(provider: string): LlmProviderInfo {
+    return { id: provider, name: provider }
+  }
+
+  /**
+   * List models this adapter can currently advertise for one owned provider.
+   * The result is advisory: an adapter may accept unlisted model ids, and
+   * consumers must not turn absence into request rejection.
+   * @param _provider - one provider route owned by this adapter.
+   * @returns discoverable models in adapter-preferred order.
+   */
+  listModels(_provider: string): Promise<readonly LlmModelInfo[]> {
+    return Promise.resolve([])
+  }
+
   /**
    * Stream one model call as raw chunks. The only required method.
    * @param options - the fully-assembled request; implementations must honor `options.signal`.
@@ -73,30 +94,40 @@ export abstract class LlmAdapter {
  * surface, interceptable via the `llm/stream` waterfall.
  */
 export class LlmService extends Service {
-  private adapters = new Map<string, LlmAdapter>()
+  private adapters = new Map<string, { adapter: LlmAdapter; provider: LlmProviderInfo }>()
 
   constructor(ctx: Context) {
     super(ctx, 'llm')
   }
 
   /**
-   * Register an adapter for the given model names. Throws `LlmError` with code
-   * `DUPLICATE_ADAPTER` if any model already has an adapter (all-or-nothing).
+   * Register an adapter for the given provider routes. Throws `LlmError` with code
+   * `DUPLICATE_ADAPTER` if any provider already has an adapter (all-or-nothing).
    * Disposed with the fiber.
-   * @param models - every model name this adapter should serve.
-   * @param adapter - the adapter that streams calls for those models.
+   * @param providers - every provider route this adapter should serve.
+   * @param adapter - the adapter that streams calls for those providers.
    * @returns the disposer that unregisters all of them.
    */
-  registerAdapter(models: string[], adapter: LlmAdapter): () => void {
+  registerAdapter(providers: string[], adapter: LlmAdapter): () => void {
     const dispose = this.ctx.effect(function* (this: LlmService) {
-      for (const model of models) {
-        if (this.adapters.has(model)) {
-          throw new LlmError(`an adapter for model "${model}" is already registered`, 'DUPLICATE_ADAPTER')
+      if (providers.length === 0) throw new LlmError('an adapter must register at least one provider', 'INVALID_ADAPTER')
+      const unique = new Set<string>()
+      const registrations: { adapter: LlmAdapter; provider: LlmProviderInfo }[] = []
+      for (const provider of providers) {
+        if (provider.length === 0) throw new LlmError('adapter provider names must be non-empty', 'INVALID_ADAPTER')
+        if (unique.has(provider) || this.adapters.has(provider)) {
+          throw new LlmError(`an adapter for provider "${provider}" is already registered`, 'DUPLICATE_ADAPTER')
         }
+        const info = adapter.providerInfo(provider)
+        if (typeof info.id !== 'string' || info.id !== provider || typeof info.name !== 'string' || info.name.length === 0) {
+          throw new LlmError(`adapter metadata for provider "${provider}" must preserve its id and have a non-empty name`, 'INVALID_ADAPTER')
+        }
+        unique.add(provider)
+        registrations.push({ adapter, provider: { id: info.id, name: info.name } })
       }
-      for (const model of models) this.adapters.set(model, adapter)
+      for (const registration of registrations) this.adapters.set(registration.provider.id, registration)
       yield () => {
-        for (const model of models) this.adapters.delete(model)
+        for (const provider of providers) this.adapters.delete(provider)
       }
     }.bind(this), 'llm.registerAdapter()')
     // ctx.effect's disposer returns Promise<void>; our disposer API is
@@ -105,29 +136,81 @@ export class LlmService extends Service {
   }
 
   /**
-   * Model names with a registered adapter.
-   * @returns the registered names, in registration order.
+   * Describe provider routes with a registered adapter.
+   * @returns detached provider metadata in registration order.
    */
-  models(): string[] {
-    return [...this.adapters.keys()]
+  listProviders(): LlmProviderInfo[] {
+    return [...this.adapters.values()].map(({ provider }) => ({ ...provider }))
   }
 
-  private adapter(model: string): LlmAdapter {
-    const adapter = this.adapters.get(model)
-    if (!adapter) throw new LlmError(`no adapter registered for model "${model}"`, 'NO_ADAPTER')
-    return adapter
+  /**
+   * Discover models advertised by one registered provider. Catalog membership
+   * is advisory and never changes routing or request validation.
+   * @param provider - registered provider route to inspect.
+   * @returns detached model metadata in adapter-preferred order.
+   */
+  async listModels(provider: string): Promise<LlmModelInfo[]> {
+    const adapter = this.registration(provider).adapter
+    const models = await adapter.listModels(provider)
+    const seen = new Set<string>()
+    return models.map((model) => {
+      if (
+        typeof model.provider !== 'string'
+        || model.provider !== provider
+        || typeof model.id !== 'string'
+        || model.id.length === 0
+        || typeof model.name !== 'string'
+        || model.name.length === 0
+        || (model.description !== undefined && typeof model.description !== 'string')
+        || seen.has(model.id)
+      ) {
+        throw new LlmError(`adapter returned invalid or duplicate model metadata for provider "${provider}"`, 'INVALID_CATALOG')
+      }
+      seen.add(model.id)
+      return {
+        provider: model.provider,
+        id: model.id,
+        name: model.name,
+        ...model.description === undefined ? {} : { description: model.description },
+      }
+    })
+  }
+
+  private registration(provider: string): { adapter: LlmAdapter; provider: LlmProviderInfo } {
+    const registration = this.adapters.get(provider)
+    if (!registration) throw new LlmError(`no adapter registered for provider "${provider}"`, 'NO_ADAPTER')
+    return registration
+  }
+
+  /** Remove replay state whose historical route is owned by another adapter. */
+  private forAdapter(options: GenerateOptions, adapter: LlmAdapter): GenerateOptions {
+    const messages: Message[] = options.messages.map((message) => {
+      const provenance = message.provenance
+      if (message.role !== 'assistant' || provenance?.replayState === undefined) return message
+      if (this.adapters.get(provenance.provider)?.adapter === adapter) return message
+      return {
+        ...message,
+        provenance: { provider: provenance.provider, model: provenance.model },
+      }
+    })
+    if (messages.every((message, index) => message === options.messages[index])) return options
+    const filtered = { ...options, messages }
+    return Object.isFrozen(options) ? deepFreeze(filtered) : filtered
   }
 
   /**
    * Stream one model call as raw chunks (token-level deltas). Throws
    * `LlmError` with code `NO_ADAPTER` if no adapter is registered for
-   * `options.model`. Dispatches through the `llm/stream` waterfall.
-   * @param options - the full request; `options.model` selects the adapter.
+   * `options.provider`. Replay state is retained only when the same adapter
+   * instance owns its historical provider and the target provider. Dispatches
+   * through the `llm/stream` waterfall.
+   * @param options - the full request; `options.provider` selects the adapter.
    * @returns the chunk stream, possibly wrapped by `llm/stream` listeners.
    */
   stream(options: GenerateOptions): AsyncIterable<StreamChunk> {
     return this.ctx.waterfall(this, 'llm/stream', options, () => {
-      return this.adapter(options.model).stream(options)
+      const adapter = this.registration(options.provider).adapter
+      return adapter.stream(this.forAdapter(options, adapter))
     })
   }
 }
diff --git a/packages/llm/llm/src/types.ts b/packages/llm/llm/src/types.ts
index 8f5fcf0d1e..b8054c2046 100644
--- a/packages/llm/llm/src/types.ts
+++ b/packages/llm/llm/src/types.ts
@@ -53,10 +53,29 @@ export type ContentBlockType = keyof ContentBlockMap
 /** Any known content block, derived from {@link ContentBlockMap}; switch on `type` and fall through unknowns (merge-extensible). */
 export type ContentBlock = ContentBlockMap[ContentBlockType]
 
-/** A single message in a conversation history. */
+/** Provider ownership and adapter-private replay data for an assistant message. */
+export interface AssistantProvenance {
+  /** Provider route that produced the message. */
+  provider: string
+  /** Provider model id that produced the message. */
+  model: string
+  /**
+   * Lossless-JSON adapter state needed to replay the provider response.
+   * `LlmService` exposes it to a target adapter only when that adapter instance
+   * currently owns both this historical provider and the target provider.
+   */
+  replayState?: unknown
+}
+
+/**
+ * A single message in a conversation history. Loop-derived assistant messages
+ * always carry provenance; callers may omit it on hand-built foreign history.
+ */
 export interface Message {
   role: 'system' | 'user' | 'assistant'
   content: ContentBlock[]
+  /** Present only on assistant messages produced by a routed adapter. */
+  provenance?: AssistantProvenance
 }
 
 /**
@@ -102,6 +121,26 @@ export interface TokenUsage {
   reasoningTokens?: number
 }
 
+/** Display metadata for one registered provider route. */
+export interface LlmProviderInfo {
+  /** Provider route key used by {@link GenerateOptions.provider}. */
+  id: string
+  /** Human-readable provider name for selectors and diagnostics. */
+  name: string
+}
+
+/** One adapter-discovered model; catalog membership is advisory, not request validation. */
+export interface LlmModelInfo {
+  /** Provider route that owns this model entry. */
+  provider: string
+  /** Model id passed to {@link GenerateOptions.model}. */
+  id: string
+  /** Human-readable model name for selectors. */
+  name: string
+  /** Optional user-facing distinction from otherwise similar models. */
+  description?: string
+}
+
 /**
  * Raw streaming protocol emitted by adapters.
  * Block indexes correlate interleaved deltas, and `block-end` carries the
@@ -116,7 +155,12 @@ export type StreamChunk =
   | { type: 'tool-call-delta'; index: number; id: CallId; name?: string; argumentsDelta: string }
   | { type: 'block-end'; index: number; block: ContentBlock }
   | { type: 'usage'; usage: TokenUsage }
-  | { type: 'finish'; reason: FinishReason }
+  | {
+    type: 'finish'
+    reason: FinishReason
+    /** Adapter-private lossless-JSON state for replaying a successful response. */
+    replayState?: unknown
+  }
 
 /**
  * JSON-schema description of a tool, as sent to the model.
@@ -134,6 +178,8 @@ export interface ToolSchema {
 
 /** A single model request, fully assembled. */
 export interface GenerateOptions {
+  /** Registered provider route selecting the adapter instance. */
+  provider: string
   model: string
   /**
    * Ordered conversation messages, exactly as the provider sees them (after
diff --git a/packages/llm/llm/tests/call-config.spec.ts b/packages/llm/llm/tests/call-config.spec.ts
index 65ff7d7d34..3e656f7563 100644
--- a/packages/llm/llm/tests/call-config.spec.ts
+++ b/packages/llm/llm/tests/call-config.spec.ts
@@ -1,6 +1,6 @@
 /**
  * call-config unit tests: field-wise LlmCallConfig equality (the real-change
- * detector behind logged header deltas) and the deepFreeze ownership helper
+ * detector behind logged changed headers) and the deepFreeze ownership helper
  * the loop applies to every built request.
  */
 
@@ -9,14 +9,16 @@ import { callConfigEquals, deepFreeze } from '../src/call-config.ts'
 
 describe('callConfigEquals', () => {
   it('compares every field, including the stop list element-wise', () => {
-    expect(callConfigEquals({ model: 'm' }, { model: 'm' })).toBe(true)
-    expect(callConfigEquals({ model: 'm' }, { model: 'x' })).toBe(false)
-    expect(callConfigEquals({ model: 'm', temperature: 0.5 }, { model: 'm' })).toBe(false)
-    expect(callConfigEquals({ model: 'm', maxTokens: 1 }, { model: 'm', maxTokens: 2 })).toBe(false)
-    expect(callConfigEquals({ model: 'm', stop: ['a'] }, { model: 'm' })).toBe(false)
-    expect(callConfigEquals({ model: 'm', stop: ['a'] }, { model: 'm', stop: ['a', 'b'] })).toBe(false)
-    expect(callConfigEquals({ model: 'm', stop: ['a'] }, { model: 'm', stop: ['b'] })).toBe(false)
-    expect(callConfigEquals({ model: 'm', stop: ['a', 'b'] }, { model: 'm', stop: ['a', 'b'] })).toBe(true)
+    const base = { provider: 'p', model: 'm' }
+    expect(callConfigEquals(base, base)).toBe(true)
+    expect(callConfigEquals(base, { provider: 'x', model: 'm' })).toBe(false)
+    expect(callConfigEquals(base, { provider: 'p', model: 'x' })).toBe(false)
+    expect(callConfigEquals({ ...base, temperature: 0.5 }, base)).toBe(false)
+    expect(callConfigEquals({ ...base, maxTokens: 1 }, { ...base, maxTokens: 2 })).toBe(false)
+    expect(callConfigEquals({ ...base, stop: ['a'] }, base)).toBe(false)
+    expect(callConfigEquals({ ...base, stop: ['a'] }, { ...base, stop: ['a', 'b'] })).toBe(false)
+    expect(callConfigEquals({ ...base, stop: ['a'] }, { ...base, stop: ['b'] })).toBe(false)
+    expect(callConfigEquals({ ...base, stop: ['a', 'b'] }, { ...base, stop: ['a', 'b'] })).toBe(true)
   })
 })
 
diff --git a/packages/llm/llm/tests/service.spec.ts b/packages/llm/llm/tests/service.spec.ts
index f669069c44..83429ba0cd 100644
--- a/packages/llm/llm/tests/service.spec.ts
+++ b/packages/llm/llm/tests/service.spec.ts
@@ -1,6 +1,7 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
 import LlmService, { GenerateOptions, LlmAdapter, LlmError, StreamChunk } from '@deepseek-ai/dsh-llm'
+import type { LlmModelInfo, LlmProviderInfo } from '@deepseek-ai/dsh-llm'
 
 class ScriptedAdapter extends LlmAdapter {
   constructor(private script: StreamChunk[]) {
@@ -12,6 +13,32 @@ class ScriptedAdapter extends LlmAdapter {
   }
 }
 
+class RecordingAdapter extends ScriptedAdapter {
+  lastOptions: GenerateOptions | undefined
+
+  override async * stream(options: GenerateOptions): AsyncIterable<StreamChunk> {
+    this.lastOptions = options
+    yield * super.stream(options)
+  }
+}
+
+class CatalogAdapter extends ScriptedAdapter {
+  constructor(
+    private readonly provider: LlmProviderInfo,
+    private readonly models: readonly LlmModelInfo[],
+  ) {
+    super(SCRIPT)
+  }
+
+  override providerInfo(_provider: string): LlmProviderInfo {
+    return this.provider
+  }
+
+  override listModels(_provider: string): Promise<readonly LlmModelInfo[]> {
+    return Promise.resolve(this.models)
+  }
+}
+
 const SCRIPT: StreamChunk[] = [
   { type: 'block-start', index: 0, blockType: 'text' },
   { type: 'text-delta', index: 0, text: 'hi' },
@@ -22,18 +49,18 @@ describe('LlmService', () => {
   it('routes stream() to the registered adapter', async () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
-    ctx.llm.registerAdapter(['test-model'], new ScriptedAdapter(SCRIPT))
+    ctx.llm.registerAdapter(['test-provider'], new ScriptedAdapter(SCRIPT))
 
     const chunks: StreamChunk[] = []
-    for await (const chunk of ctx.llm.stream({ model: 'test-model', messages: [] })) chunks.push(chunk)
+    for await (const chunk of ctx.llm.stream({ provider: 'test-provider', model: 'test-model', messages: [] })) chunks.push(chunk)
     expect(chunks).toEqual(SCRIPT)
   })
 
-  it('throws NO_ADAPTER for unregistered models', async () => {
+  it('throws NO_ADAPTER for unregistered providers', async () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
     await expect((async () => {
-      for await (const _ of ctx.llm.stream({ model: 'nope', messages: [] })) { /* drain */ }
+      for await (const _ of ctx.llm.stream({ provider: 'nope', model: 'any-model', messages: [] })) { /* drain */ }
     })()).rejects.toThrow('no adapter registered')
   })
 
@@ -44,10 +71,80 @@ describe('LlmService', () => {
     const fiber = await ctx.plugin(Object.assign((inner: Context) => {
       inner.llm.registerAdapter(['scoped-model'], new ScriptedAdapter(SCRIPT))
     }, { inject: ['llm'] }))
-    expect(ctx.llm.models()).toEqual(['scoped-model'])
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'scoped-model', name: 'scoped-model' }])
 
     await fiber.dispose()
-    expect(ctx.llm.models()).toEqual([])
+    expect(ctx.llm.listProviders()).toEqual([])
+  })
+
+  it('discovers detached provider and advisory model metadata', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    const provider = { id: 'catalog', name: 'Catalog Provider' }
+    const model = { provider: 'catalog', id: 'fast', name: 'Fast', description: 'Low latency' }
+    ctx.llm.registerAdapter(['catalog'], new CatalogAdapter(provider, [model]))
+
+    const providers = ctx.llm.listProviders()
+    const models = await ctx.llm.listModels('catalog')
+    expect(providers).toEqual([provider])
+    expect(models).toEqual([model])
+
+    providers[0]!.name = 'mutated'
+    models[0]!.name = 'mutated'
+    provider.name = 'source mutated'
+    model.name = 'source mutated'
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'catalog', name: 'Catalog Provider' }])
+    await expect(ctx.llm.listModels('catalog')).resolves.toEqual([{
+      provider: 'catalog', id: 'fast', name: 'source mutated', description: 'Low latency',
+    }])
+  })
+
+  it('defaults adapters to their route name and an empty advisory model list', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    ctx.llm.registerAdapter(['plain'], new ScriptedAdapter(SCRIPT))
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'plain', name: 'plain' }])
+    await expect(ctx.llm.listModels('plain')).resolves.toEqual([])
+    await expect(ctx.llm.listModels('missing')).rejects.toMatchObject({ code: 'NO_ADAPTER' })
+  })
+
+  it.each([
+    [{ id: 1, name: 'Name' }, 'non-string id'],
+    [{ id: 'other', name: 'Name' }, 'mismatched id'],
+    [{ id: 'route', name: 1 }, 'non-string name'],
+    [{ id: 'route', name: '' }, 'empty name'],
+  ] as const)('rejects invalid provider metadata atomically (%s: %s)', async (metadata, _label) => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    const adapter = new CatalogAdapter(metadata as unknown as LlmProviderInfo, [])
+    expect(() => ctx.llm.registerAdapter(['route'], adapter)).toThrow(expect.objectContaining({ code: 'INVALID_ADAPTER' }))
+    expect(ctx.llm.listProviders()).toEqual([])
+  })
+
+  it.each([
+    [{ provider: 1, id: 'm', name: 'M' }, 'non-string provider'],
+    [{ provider: 'other', id: 'm', name: 'M' }, 'mismatched provider'],
+    [{ provider: 'route', id: 1, name: 'M' }, 'non-string id'],
+    [{ provider: 'route', id: '', name: 'M' }, 'empty id'],
+    [{ provider: 'route', id: 'm', name: 1 }, 'non-string name'],
+    [{ provider: 'route', id: 'm', name: '' }, 'empty name'],
+    [{ provider: 'route', id: 'm', name: 'M', description: 1 }, 'non-string description'],
+  ] as const)('rejects invalid model metadata (%s: %s)', async (metadata, _label) => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    ctx.llm.registerAdapter(['route'], new CatalogAdapter(
+      { id: 'route', name: 'Route' },
+      [metadata as unknown as LlmModelInfo],
+    ))
+    await expect(ctx.llm.listModels('route')).rejects.toMatchObject({ code: 'INVALID_CATALOG' })
+  })
+
+  it('rejects duplicate model ids in one provider catalog', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    const model = { provider: 'route', id: 'same', name: 'Same' }
+    ctx.llm.registerAdapter(['route'], new CatalogAdapter({ id: 'route', name: 'Route' }, [model, model]))
+    await expect(ctx.llm.listModels('route')).rejects.toMatchObject({ code: 'INVALID_CATALOG' })
   })
 
   it('lets llm/stream waterfall listeners wrap the underlying stream', async () => {
@@ -64,11 +161,90 @@ describe('LlmService', () => {
     })
 
     const chunks: StreamChunk[] = []
-    for await (const chunk of ctx.llm.stream({ model: 'test-model', messages: [] })) chunks.push(chunk)
+    for await (const chunk of ctx.llm.stream({ provider: 'test-model', model: 'dynamic-model', messages: [] })) chunks.push(chunk)
     expect(chunks).toHaveLength(4)
     expect(chunks[0]).toMatchObject({ index: 99 })
   })
 
+  it('resolves the provider after llm/stream listeners have had a chance to route it', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    const adapter = new RecordingAdapter(SCRIPT)
+    ctx.llm.registerAdapter(['routed'], adapter)
+    ctx.on('llm/stream', (options, next) => {
+      options.provider = 'routed'
+      return next()
+    })
+
+    for await (const _chunk of ctx.llm.stream({ provider: 'initial', model: 'm', messages: [] })) { /* drain */ }
+    expect(adapter.lastOptions?.provider).toBe('routed')
+  })
+
+  it('keeps replay state when historical and target providers belong to the same adapter instance', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    const adapter = new RecordingAdapter(SCRIPT)
+    ctx.llm.registerAdapter(['historical', 'target'], adapter)
+    const replayState = { private: 'state' }
+
+    for await (const _chunk of ctx.llm.stream({
+      provider: 'target',
+      model: 'new-model',
+      messages: [{
+        role: 'assistant',
+        content: [{ type: 'text', text: 'old response' }],
+        provenance: { provider: 'historical', model: 'old-model', replayState },
+      }],
+    })) { /* drain */ }
+
+    expect(adapter.lastOptions?.messages[0]?.provenance).toEqual({
+      provider: 'historical', model: 'old-model', replayState,
+    })
+  })
+
+  it('strips replay state but preserves provenance when the target uses a different adapter instance', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    ctx.llm.registerAdapter(['historical'], new RecordingAdapter(SCRIPT))
+    const target = new RecordingAdapter(SCRIPT)
+    ctx.llm.registerAdapter(['target'], target)
+
+    for await (const _chunk of ctx.llm.stream({
+      provider: 'target',
+      model: 'new-model',
+      messages: [{
+        role: 'assistant',
+        content: [{ type: 'text', text: 'old response' }],
+        provenance: { provider: 'historical', model: 'old-model', replayState: { private: 'state' } },
+      }],
+    })) { /* drain */ }
+
+    expect(target.lastOptions?.messages[0]?.provenance).toEqual({ provider: 'historical', model: 'old-model' })
+  })
+
+  it('preserves immutability while stripping replay state from frozen requests', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    ctx.llm.registerAdapter(['historical'], new RecordingAdapter(SCRIPT))
+    const target = new RecordingAdapter(SCRIPT)
+    ctx.llm.registerAdapter(['target'], target)
+    const options = Object.freeze({
+      provider: 'target',
+      model: 'new-model',
+      messages: [{
+        role: 'assistant' as const,
+        content: [{ type: 'text' as const, text: 'old response' }],
+        provenance: { provider: 'historical', model: 'old-model', replayState: { private: 'state' } },
+      }],
+    })
+
+    for await (const _chunk of ctx.llm.stream(options)) { /* drain */ }
+
+    expect(target.lastOptions).not.toBe(options)
+    expect(Object.isFrozen(target.lastOptions)).toBe(true)
+    expect(target.lastOptions?.messages[0]?.provenance).toEqual({ provider: 'historical', model: 'old-model' })
+  })
+
   it('creates LlmError with a code for programmatic handling', () => {
     const err = new LlmError('something went wrong', 'CUSTOM_CODE')
     expect(err).toBeInstanceOf(Error)
@@ -104,9 +280,9 @@ describe('LlmService', () => {
     await ctx.plugin(LlmService)
 
     const dispose = ctx.llm.registerAdapter(['m1'], new ScriptedAdapter(SCRIPT))
-    expect(ctx.llm.models()).toEqual(['m1'])
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'm1', name: 'm1' }])
     dispose()
-    expect(ctx.llm.models()).toEqual([])
+    expect(ctx.llm.listProviders()).toEqual([])
   })
 
   it('rejects duplicate adapter registration with DUPLICATE_ADAPTER code', async () => {
@@ -123,19 +299,30 @@ describe('LlmService', () => {
     }
   })
 
+  it('rejects empty and internally duplicated provider registrations atomically', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    const adapter = new ScriptedAdapter(SCRIPT)
+
+    expect(() => ctx.llm.registerAdapter([], adapter)).toThrow(expect.objectContaining({ code: 'INVALID_ADAPTER' }))
+    expect(() => ctx.llm.registerAdapter([''], adapter)).toThrow(expect.objectContaining({ code: 'INVALID_ADAPTER' }))
+    expect(() => ctx.llm.registerAdapter(['first', 'first'], adapter)).toThrow(expect.objectContaining({ code: 'DUPLICATE_ADAPTER' }))
+    expect(ctx.llm.listProviders()).toEqual([])
+  })
+
   it('re-registers a model after its prior registration is disposed', async () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
 
     const dispose = ctx.llm.registerAdapter(['m1'], new ScriptedAdapter(SCRIPT))
-    expect(ctx.llm.models()).toEqual(['m1'])
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'm1', name: 'm1' }])
     dispose()
-    expect(ctx.llm.models()).toEqual([])
+    expect(ctx.llm.listProviders()).toEqual([])
 
     // The duplicate check is not wedged: the same model registers cleanly again.
     const disposeAgain = ctx.llm.registerAdapter(['m1'], new ScriptedAdapter(SCRIPT))
-    expect(ctx.llm.models()).toEqual(['m1'])
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'm1', name: 'm1' }])
     disposeAgain()
-    expect(ctx.llm.models()).toEqual([])
+    expect(ctx.llm.listProviders()).toEqual([])
   })
 })
diff --git a/packages/llm/token-meter/README.md b/packages/llm/token-meter/README.md
new file mode 100644
index 0000000000..262fc8d152
--- /dev/null
+++ b/packages/llm/token-meter/README.md
@@ -0,0 +1,50 @@
+# @deepseek-ai/dsh-token-meter
+
+Replay-aware token measurement through the singleton `ctx.tokenMeter` service. It advances one isolated fold per session from the durable log, so compaction and other pressure-sensitive plugins can share accounting without depending on `CompactService`.
+
+## Configuration
+
+| Key | Default | Contract |
+|---|---:|---|
+| `contextWindow` | `128000` | Positive integer service-wide context capacity. |
+
+The estimator intentionally uses one fixed heuristic: four characters per token plus structural overhead for roles, blocks, and request-envelope fields. `contextWindow` is the only deployment setting. Direct construction validates it; Loader mounts first apply the package's Schemastery shape validation. Unrecognized top-level keys are rejected.
+
+## Measurement contract
+
+`ctx.tokenMeter` directly exposes two operations:
+
+- `measure(session, requestHeader?)` returns request pressure and the current priced surface at one consumed-log revision.
+- `estimateMessage(message)` prices one message with the fixed heuristic.
+
+`measure()` synchronizes once and returns one detached, deeply immutable snapshot. `totalTokens` is request-and-response pressure, while `surfaceTokens` is the surface-only heuristic total and equals the sum of `nodes[].tokens`. A `requestHeader` override affects pressure fields only; the surface fields still describe the current session. Every call clones the positional nodes, so measurement is O(surface).
+
+The fold tracks full request-header snapshots, step boundaries, surface appends and replacements, successful assistant messages, provider usage, and assistant-chunk provenance. Provider usage is reused only when the latest successful call's canonical request envelope matches the measured envelope and its total is no lower than that call's full heuristic anchor; a later success replaces the earlier anchor. Otherwise the complete current envelope and surface are estimated. Surface changes remain signed relative to a matching anchor, including negative deltas after shrinking replacements.
+
+Usage accounting sums disjoint input, cache-read, cache-write, and output buckets; reasoning is not added again. Every successful call records an assistant anchor, including content-less calls. An explicit empty provenance list means a known empty provider stream, while absent legacy provenance conservatively treats the durable assistant output as provider output.
+
+## Composition
+
+```yaml
+- name: '@deepseek-ai/dsh-token-meter'
+- name: '@deepseek-ai/dsh-compact-basic'
+```
+
+Both plugins have usable defaults. A deployment with a different capacity configures the meter once:
+
+```yaml
+- name: '@deepseek-ai/dsh-token-meter'
+  config:
+    contextWindow: 32768
+```
+
+## Model Experience
+
+Indirectly, through consumers such as `dsh-compact-basic`; the service itself adds no prompt, message, schema, tool, or model call.
+
+## Known Limitations and Deferred Work
+
+- **The fixed heuristic is approximate** — content without reusable provider usage is priced by character count plus structural overhead, not an exact provider tokenizer or request serializer.
+- **Every measurement clones the current surface** — coherent immutable snapshots make reads O(surface), including below-threshold pressure checks.
+- **Provider usage is only reusable for an identical canonical envelope** — prompt, prefix, tools, provider, model, or call-config changes deliberately fall back to full heuristic estimation.
+- **Legacy provenance is conservative** — assistant messages without `sourceEventSeqs` cannot distinguish provider output from listener rewrites, so the fold avoids claiming a known empty or exact chunk stream.
diff --git a/packages/llm/token-meter/package.json b/packages/llm/token-meter/package.json
new file mode 100644
index 0000000000..13fa4e3cdc
--- /dev/null
+++ b/packages/llm/token-meter/package.json
@@ -0,0 +1,37 @@
+{
+  "name": "@deepseek-ai/dsh-token-meter",
+  "description": "Replay-aware token measurement service (ctx.tokenMeter) for the DeepSeek Harness",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "@deepseek-ai/dsh-llm": "^0.0.1",
+    "@deepseek-ai/dsh-session": "^0.0.1",
+    "cordis": "^4.0.0-rc.7"
+  },
+  "dependencies": {
+    "schemastery": "^3.18.0"
+  },
+  "devDependencies": {
+    "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-session": "workspace:^",
+    "cordis": "^4.0.0-rc.7"
+  }
+}
diff --git a/packages/llm/token-meter/src/index.ts b/packages/llm/token-meter/src/index.ts
new file mode 100644
index 0000000000..0c18dee74b
--- /dev/null
+++ b/packages/llm/token-meter/src/index.ts
@@ -0,0 +1,419 @@
+/**
+ * Single replay-aware token-meter service for request and surface pressure.
+ *
+ * @module @deepseek-ai/dsh-token-meter
+ */
+
+import { Context, Service } from 'cordis'
+import z from 'schemastery'
+import { BlockAssembler, deepFreeze } from '@deepseek-ai/dsh-llm'
+import type { ContentBlock, Message, TokenUsage } from '@deepseek-ai/dsh-llm'
+import type { EpochHeader, Session, SessionEvent, SurfaceEvent } from '@deepseek-ai/dsh-session'
+import { canonicalHeader, headerEquals, isSurfaceEvent } from '@deepseek-ai/dsh-session'
+import type {
+  TokenMeasurement,
+  TokenMeasurementBaseline,
+  TokenMeterConfig,
+  TokenSurfaceNode,
+} from './types.ts'
+
+export type * from './types.ts'
+
+/** Default service-wide provider context capacity. */
+const DEFAULT_CONTEXT_WINDOW = 128_000
+
+/** Complete public configuration key set. */
+const TOKEN_METER_CONFIG_KEYS: ReadonlySet<string> = new Set(['contextWindow'])
+
+/** Fixed text-density estimate used until exact tokenization is needed. */
+const CHARS_PER_TOKEN = 4
+
+/** Per-block structural overhead for JSON framing and type tags. */
+const BLOCK_OVERHEAD = 4
+
+/** Role-field framing overhead added to every priced message. */
+const ROLE_OVERHEAD = 4
+
+interface MeasurementAnchor {
+  readonly header: EpochHeader | undefined
+  readonly surfaceTokens: number
+  readonly baseline: Exclude<TokenMeasurementBaseline, { kind: 'none' }>
+}
+
+interface ReplayState {
+  consumedEvents: number
+  header: EpochHeader | undefined
+  surface: TokenSurfaceNode[]
+  surfaceTokens: number
+  stepStart: { turn: number; step: number; surfaceTokens: number } | undefined
+  anchor: MeasurementAnchor | undefined
+}
+
+interface PreparedSurfaceMutation {
+  readonly tokens: number
+  commit(state: ReplayState): void
+}
+
+/** Sum disjoint provider usage buckets without double-counting reasoning output. */
+function usageTokens(usage: TokenUsage): number {
+  return usage.inputTokens
+    + (usage.cacheReadTokens ?? 0)
+    + (usage.cacheWriteTokens ?? 0)
+    + usage.outputTokens
+}
+
+/** Compare optional envelopes so a headerless estimate can track later surface deltas. */
+function optionalHeaderEquals(
+  left: EpochHeader | undefined,
+  right: EpochHeader | undefined,
+): boolean {
+  if (left === undefined || right === undefined) return left === right
+  return headerEquals(left, right)
+}
+
+/** Reject stale or misspelled keys before defaults can hide them. */
+function validateConfigKeys(config: TokenMeterConfig): void {
+  for (const key of Object.keys(config)) {
+    if (!TOKEN_METER_CONFIG_KEYS.has(key)) {
+      throw new Error(
+        `TokenMeterConfig: unknown key "${key}" (allowed: contextWindow)`,
+      )
+    }
+  }
+}
+
+/** Resolve and validate the one service-wide context capacity. */
+function resolveContextWindow(config: TokenMeterConfig): number {
+  validateConfigKeys(config)
+  const contextWindow = config.contextWindow === undefined
+    ? DEFAULT_CONTEXT_WINDOW
+    : config.contextWindow
+  if (!Number.isInteger(contextWindow) || contextWindow <= 0) {
+    throw new Error(
+      `TokenMeterConfig: contextWindow (${contextWindow}) must be a positive integer`,
+    )
+  }
+  return contextWindow
+}
+
+declare module 'cordis' {
+  interface Context {
+    tokenMeter: TokenMeterService
+  }
+}
+
+/** Replay owner for one service-wide estimator and isolated per-session folds. */
+export class TokenMeterService extends Service {
+  static Config: z<TokenMeterConfig> = z.object({
+    contextWindow: z.number().step(1).min(1).default(DEFAULT_CONTEXT_WINDOW),
+  })
+
+  /** Provider context-window capacity used by pressure consumers. */
+  readonly contextWindow: number
+
+  private readonly states = new WeakMap<Session, ReplayState>()
+
+  constructor(ctx: Context, config: TokenMeterConfig = {}) {
+    super(ctx, 'tokenMeter')
+    this.contextWindow = resolveContextWindow(config)
+
+    // Readers catch up independently, while eager observation bounds ordinary
+    // read latency without creating state for sessions no consumer has read.
+    ctx.on('session/event', (session) => {
+      if (this.states.has(session)) this._sync(session)
+    })
+  }
+
+  /**
+   * Measure current request pressure and surface through the durable tail.
+   *
+   * Provider usage is reused only when the latest successful call's canonical
+   * request envelope matches `requestHeader` and its total is no lower than
+   * that call's full heuristic anchor; otherwise the complete envelope and
+   * surface are heuristically repriced.
+   *
+   * `requestHeader` affects request pressure only; surface fields always
+   * describe the current session surface. Every call clones those positional
+   * nodes, so measurement is O(surface).
+   *
+   * @param session - session to replay through its current durable tail.
+   * @param requestHeader - optional effective request envelope replacing the latest logged header.
+   * @returns a detached deeply immutable pressure and surface measurement.
+   */
+  measure(session: Session, requestHeader?: EpochHeader): TokenMeasurement {
+    const state = this._sync(session)
+    const header = requestHeader === undefined
+      ? state.header
+      : canonicalHeader(requestHeader)
+    const anchor = state.anchor
+
+    let baseline: TokenMeasurementBaseline
+    let surfaceDeltaTokens: number
+    if (anchor !== undefined && optionalHeaderEquals(anchor.header, header)) {
+      baseline = anchor.baseline
+      surfaceDeltaTokens = state.surfaceTokens - anchor.surfaceTokens
+    } else if (header === undefined && state.surfaceTokens === 0) {
+      baseline = { kind: 'none', tokens: 0 }
+      surfaceDeltaTokens = 0
+    } else {
+      baseline = {
+        kind: 'estimated',
+        tokens: this._estimateHeader(header) + state.surfaceTokens,
+      }
+      surfaceDeltaTokens = 0
+    }
+
+    return deepFreeze(structuredClone({
+      logRevision: state.consumedEvents,
+      baseline,
+      surfaceDeltaTokens,
+      totalTokens: Math.max(0, baseline.tokens + surfaceDeltaTokens),
+      surfaceTokens: state.surfaceTokens,
+      nodes: state.surface,
+    }))
+  }
+
+  /**
+   * Heuristically price one model-visible message.
+   * @param message - message to price without mutation.
+   * @returns content and role-framing tokens under the fixed service heuristic.
+   */
+  estimateMessage(message: Message): number {
+    return this._estimateContent(message.content) + ROLE_OVERHEAD
+  }
+
+  /** Catch one session's fold up to the current durable tail. */
+  private _sync(session: Session): ReplayState {
+    let state = this.states.get(session)
+    if (state === undefined) {
+      state = {
+        consumedEvents: 0,
+        header: undefined,
+        surface: [],
+        surfaceTokens: 0,
+        stepStart: undefined,
+        anchor: undefined,
+      }
+      this.states.set(session, state)
+    }
+
+    while (state.consumedEvents < session.events.length) {
+      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion -- contiguous session seqs index the durable log
+      const event = session.events[state.consumedEvents]!
+      this._foldEvent(session, state, event)
+      state.consumedEvents += 1
+    }
+    return state
+  }
+
+  /**
+   * Validate and prepare every fallible part before mutating replay state.
+   * A malformed event remains unread on every retry instead of partially
+   * applying the same mutation more than once.
+   */
+  private _foldEvent(session: Session, state: ReplayState, event: SessionEvent): void {
+    let nextHeader = state.header
+    let nextStepStart = state.stepStart
+    let nextAnchor = state.anchor
+
+    switch (event.type) {
+      case 'request/header':
+        nextHeader = canonicalHeader(event.data.header)
+        break
+      case 'step/start':
+        if (state.stepStart !== undefined) {
+          throw new Error(
+            `token meter: step/start at seq ${event.seq} arrived before turn ${state.stepStart.turn}/step ${state.stepStart.step} ended`,
+          )
+        }
+        nextStepStart = { ...event.data, surfaceTokens: state.surfaceTokens }
+        break
+      case 'step/end':
+        if (state.stepStart === undefined
+          || state.stepStart.turn !== event.data.turn
+          || state.stepStart.step !== event.data.step) {
+          throw new Error(`token meter: step/end at seq ${event.seq} has no matching step/start boundary`)
+        }
+        nextStepStart = undefined
+        break
+      default:
+        break
+    }
+
+    const surface = isSurfaceEvent(event)
+      ? this._prepareSurfaceMutation(session, state, event)
+      : undefined
+
+    if (event.type === 'assistant/message') {
+      const stepStart = state.stepStart
+      if (stepStart === undefined
+        || stepStart.turn !== event.data.turn
+        || stepStart.step !== event.data.step) {
+        throw new Error(`token meter: assistant/message at seq ${event.seq} has no matching step/start boundary`)
+      }
+
+      // assistant/message is surface-mandatory at every append/seed boundary.
+      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+      const eventTokens = surface!.tokens
+      if (event.data.usage !== undefined && nextHeader !== undefined) {
+        const providerAssistantTokens = this._estimateProviderAssistant(
+          session,
+          event,
+          eventTokens,
+        )
+        const anchorSurfaceTokens = stepStart.surfaceTokens + providerAssistantTokens
+        const providerTokens = usageTokens(event.data.usage)
+        const estimatedAnchorTokens = this._estimateHeader(nextHeader) + anchorSurfaceTokens
+        nextAnchor = {
+          header: nextHeader,
+          surfaceTokens: anchorSurfaceTokens,
+          // Signed heuristic deltas remain conservative only from an anchor
+          // that is at least as large as the matching full heuristic price.
+          baseline: providerTokens >= estimatedAnchorTokens
+            ? { kind: 'usage', tokens: providerTokens, usage: event.data.usage }
+            : { kind: 'estimated', tokens: estimatedAnchorTokens },
+        }
+      } else {
+        const anchorSurfaceTokens = stepStart.surfaceTokens + eventTokens
+        nextAnchor = {
+          header: nextHeader,
+          surfaceTokens: anchorSurfaceTokens,
+          baseline: {
+            kind: 'estimated',
+            tokens: this._estimateHeader(nextHeader) + anchorSurfaceTokens,
+          },
+        }
+      }
+    }
+
+    state.header = nextHeader
+    state.stepStart = nextStepStart
+    if (surface !== undefined) surface.commit(state)
+    state.anchor = nextAnchor
+  }
+
+  /** Validate one surface operation and return its allocation-light commit. */
+  private _prepareSurfaceMutation(
+    session: Session,
+    state: ReplayState,
+    event: SurfaceEvent,
+  ): PreparedSurfaceMutation {
+    const tokens = this._estimateSurfaceEvent(session, event)
+    const op = event.surfaceOp
+    if (op === 'append') {
+      return {
+        tokens,
+        commit(target) {
+          target.surface.push({ seq: event.seq, tokens })
+          target.surfaceTokens += tokens
+        },
+      }
+    }
+
+    const startIdx = state.surface.findIndex(node => node.seq === op.start)
+    const endIdx = state.surface.findIndex(node => node.seq === op.end)
+    if (startIdx === -1 || endIdx === -1 || startIdx > endIdx) {
+      throw new Error(
+        `token meter: replace at seq ${event.seq} has invalid current range ${op.start}-${op.end}`,
+      )
+    }
+    const removedTokens = state.surface
+      .slice(startIdx, endIdx + 1)
+      .reduce((total, node) => total + node.tokens, 0)
+    return {
+      tokens,
+      commit(target) {
+        target.surface.splice(startIdx, endIdx - startIdx + 1, { seq: event.seq, tokens })
+        target.surfaceTokens += tokens - removedTokens
+      },
+    }
+  }
+
+  /** Price one current surface event exactly as it projects to a request. */
+  private _estimateSurfaceEvent(session: Session, event: SurfaceEvent): number {
+    const message = session.deriveEventMessage(event)
+    return message === null ? 0 : this.estimateMessage(message)
+  }
+
+  /**
+   * Reassemble provider output from exact chunk provenance for a usage anchor.
+   * Missing legacy provenance conservatively treats the durable output as the
+   * provider output; explicit empty provenance prices a known empty stream.
+   */
+  private _estimateProviderAssistant(
+    session: Session,
+    event: SessionEvent<'assistant/message'>,
+    durableEventTokens: number,
+  ): number {
+    const sourceSeqs = event.sourceEventSeqs
+    if (sourceSeqs === undefined) return durableEventTokens
+
+    const assembler = new BlockAssembler()
+    const seen = new Set<number>()
+    for (const seq of sourceSeqs) {
+      if (seq >= event.seq) {
+        throw new Error(`token meter: assistant/message at seq ${event.seq} source seq ${seq} is not earlier`)
+      }
+      if (seen.has(seq)) {
+        throw new Error(`token meter: assistant/message at seq ${event.seq} repeats source seq ${seq}`)
+      }
+      seen.add(seq)
+      // Session construction validates contiguous seqs, and the explicit
+      // earlier-than-assistant check above therefore guarantees existence.
+      const source = session.events[seq]
+      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+      const sourceEvent = source!
+      if (sourceEvent.type !== 'assistant/chunk') {
+        throw new Error(`token meter: assistant/message at seq ${event.seq} source seq ${seq} is not assistant/chunk`)
+      }
+      if (sourceEvent.data.turn !== event.data.turn || sourceEvent.data.step !== event.data.step) {
+        throw new Error(`token meter: assistant/message at seq ${event.seq} source seq ${seq} belongs to another step`)
+      }
+      assembler.push(sourceEvent.data.chunk)
+    }
+    const providerMessage = assembler.message()
+    return providerMessage.content.length === 0 ? 0 : this.estimateMessage(providerMessage)
+  }
+
+  /** Price content blocks recursively under the fixed density heuristic. */
+  private _estimateContent(blocks: readonly ContentBlock[]): number {
+    let tokens = 0
+    for (const block of blocks) {
+      switch (block.type) {
+        case 'text':
+        case 'reasoning':
+          tokens += Math.ceil(block.text.length / CHARS_PER_TOKEN) + BLOCK_OVERHEAD
+          break
+        case 'tool-call':
+          tokens += Math.ceil(block.name.length / CHARS_PER_TOKEN)
+            + Math.ceil(block.arguments.length / CHARS_PER_TOKEN)
+            + BLOCK_OVERHEAD
+          break
+        case 'tool-result':
+          tokens += this._estimateContent(block.content) + BLOCK_OVERHEAD
+          break
+        default:
+          // ContentBlockMap is merge-extensible; unknown blocks retain a
+          // conservative structural JSON price under the fixed heuristic.
+          tokens += BLOCK_OVERHEAD + Math.ceil(JSON.stringify(block).length / CHARS_PER_TOKEN)
+      }
+    }
+    return tokens
+  }
+
+  /** Price the canonical non-surface request envelope. */
+  private _estimateHeader(header: EpochHeader | undefined): number {
+    if (header === undefined) return 0
+    let tokens = 0
+    for (const message of header.messagePrefix ?? []) tokens += this.estimateMessage(message)
+    if (header.system !== undefined) {
+      tokens += Math.ceil(header.system.length / CHARS_PER_TOKEN) + ROLE_OVERHEAD
+    }
+    if (header.tools !== undefined && header.tools.length > 0) {
+      tokens += Math.ceil(JSON.stringify(header.tools).length / CHARS_PER_TOKEN) + BLOCK_OVERHEAD
+    }
+    return tokens
+  }
+}
+
+export default TokenMeterService
diff --git a/packages/llm/token-meter/src/types.ts b/packages/llm/token-meter/src/types.ts
new file mode 100644
index 0000000000..7fb35af997
--- /dev/null
+++ b/packages/llm/token-meter/src/types.ts
@@ -0,0 +1,43 @@
+/**
+ * Public configuration and measurement vocabulary for replay token metering.
+ *
+ * @module @deepseek-ai/dsh-token-meter/types
+ */
+
+import type { TokenUsage } from '@deepseek-ai/dsh-llm'
+
+/** Token-meter plugin configuration. */
+export interface TokenMeterConfig {
+  /** Service-wide context-window capacity in tokens. Defaults to `128000`. */
+  contextWindow?: number
+}
+
+/** The baseline from which a signed surface delta produces current pressure. */
+export type TokenMeasurementBaseline =
+  | { readonly kind: 'none'; readonly tokens: 0 }
+  | { readonly kind: 'estimated'; readonly tokens: number }
+  | { readonly kind: 'usage'; readonly tokens: number; readonly usage: Readonly<TokenUsage> }
+
+/** Detached immutable request-pressure and surface snapshot at one consumed log revision. */
+export interface TokenMeasurement {
+  /** Number of durable events consumed; equal to the next unread event seq. */
+  readonly logRevision: number
+  /** Provider or heuristic anchor used for this measurement. */
+  readonly baseline: TokenMeasurementBaseline
+  /** Signed repricing of current surface content relative to the baseline anchor. */
+  readonly surfaceDeltaTokens: number
+  /** Non-negative current request-and-response pressure. */
+  readonly totalTokens: number
+  /** Total heuristic tokens across the current surface. */
+  readonly surfaceTokens: number
+  /** Current surface nodes in positional head-to-tail order. */
+  readonly nodes: readonly TokenSurfaceNode[]
+}
+
+/** One token-priced node in the current ordered session surface. */
+export interface TokenSurfaceNode {
+  /** Durable sequence number of the surface event. */
+  readonly seq: number
+  /** Heuristic tokens for the exact message projected by this node. */
+  readonly tokens: number
+}
diff --git a/packages/llm/token-meter/tests/token-meter.spec.ts b/packages/llm/token-meter/tests/token-meter.spec.ts
new file mode 100644
index 0000000000..e7ddca6696
--- /dev/null
+++ b/packages/llm/token-meter/tests/token-meter.spec.ts
@@ -0,0 +1,649 @@
+import { describe, expect, it } from 'vitest'
+import { Context } from 'cordis'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import type { ContentBlock, Message, TokenUsage } from '@deepseek-ai/dsh-llm'
+import SessionStore, { Session, SessionId, canonicalHeader } from '@deepseek-ai/dsh-session'
+import type { EpochHeader, SessionEvent } from '@deepseek-ai/dsh-session'
+import TokenMeterService from '@deepseek-ai/dsh-token-meter'
+import type { TokenMeasurement, TokenMeterConfig } from '@deepseek-ai/dsh-token-meter'
+
+function header(model: string, extras: Omit<EpochHeader, 'config'> = {}): EpochHeader {
+  return canonicalHeader({ config: { provider: 'mock', model }, ...extras })
+}
+
+function textMessage(text: string, role: Message['role'] = 'user'): Message {
+  return { role, content: [{ type: 'text', text }] }
+}
+
+function appendHeader(session: Session, value: EpochHeader): void {
+  session.append('request/header', { header: value, reason: 'initial' })
+}
+
+/** Inject malformed persisted history after the live append boundary for defensive replay tests. */
+function appendUnchecked(session: Session, event: SessionEvent): void {
+  const log = (session as unknown as { log: SessionEvent[] }).log
+  log.push(event)
+}
+
+interface SuccessfulCallOptions {
+  turn?: number
+  step?: number
+  providerText?: string
+  durableText?: string
+  usage?: TokenUsage
+  provenance?: 'exact' | 'empty' | 'absent'
+}
+
+function appendSuccessfulCall(
+  session: Session,
+  value: EpochHeader,
+  options: SuccessfulCallOptions = {},
+): void {
+  const turn = options.turn ?? 1
+  const step = options.step ?? 1
+  const providerText = options.providerText ?? 'provider answer'
+  const durableText = options.durableText ?? providerText
+  const provenance = options.provenance ?? 'exact'
+  session.append('step/start', { turn, step })
+  appendHeader(session, value)
+
+  const sources: number[] = []
+  if (provenance === 'exact') {
+    const chunks = [
+      { type: 'block-start' as const, index: 0, blockType: 'text' as const },
+      { type: 'text-delta' as const, index: 0, text: providerText },
+      { type: 'block-end' as const, index: 0, block: { type: 'text' as const, text: providerText } },
+      ...options.usage === undefined ? [] : [{ type: 'usage' as const, usage: options.usage }],
+      { type: 'finish' as const, reason: { kind: 'stop' as const } },
+    ]
+    for (const chunk of chunks) {
+      sources.push(session.append('assistant/chunk', { turn, step, chunk }).seq)
+    }
+  }
+
+  const intent = provenance === 'absent'
+    ? { surfaceOp: 'append' as const }
+    : { surfaceOp: 'append' as const, sourceEventSeqs: provenance === 'empty' ? [] : sources }
+  session.append('assistant/message', {
+    provenance: {
+      provider: value.config.provider,
+      model: value.config.model,
+    },
+    turn,
+    step,
+    content: durableText.length === 0 ? [] : [{ type: 'text', text: durableText }],
+    ...options.usage === undefined ? {} : { usage: options.usage },
+  }, intent)
+  session.append('step/end', { turn, step })
+}
+
+function meter(config: TokenMeterConfig = {}): TokenMeterService {
+  return new TokenMeterService(new Context(), config)
+}
+
+function expectSurfaceTotal(measurement: TokenMeasurement): void {
+  expect(measurement.nodes.reduce((total, node) => total + node.tokens, 0))
+    .toBe(measurement.surfaceTokens)
+}
+
+describe('TokenMeterService configuration and registration', () => {
+  it('provides one zero-config context window', () => {
+    const service = meter()
+    expect(service.contextWindow).toBe(128_000)
+  })
+
+  it('accepts one service-wide context-window override', () => {
+    expect(meter({ contextWindow: 32_000 }).contextWindow).toBe(32_000)
+  })
+
+  it.each(['models', 'contextWidow'])('rejects unknown top-level config key %s', (key) => {
+    expect(() => meter({ [key]: {} }))
+      .toThrow(`TokenMeterConfig: unknown key "${key}"`)
+  })
+
+  it.each([
+    { contextWindow: 0 },
+    { contextWindow: -1 },
+    { contextWindow: 1.5 },
+    { contextWindow: Number.NaN },
+    { contextWindow: null },
+  ] as unknown as TokenMeterConfig[])('rejects invalid context capacity %#', (config) => {
+    expect(() => meter(config)).toThrow(/contextWindow .* positive integer/)
+  })
+
+  it('registers and unregisters ctx.tokenMeter with its plugin fiber', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const fiber = await ctx.plugin(TokenMeterService)
+    expect(ctx.get('tokenMeter')).toBeInstanceOf(TokenMeterService)
+    await fiber.dispose()
+    expect(ctx.get('tokenMeter')).toBeUndefined()
+  })
+})
+
+describe('TokenMeterService pricing', () => {
+  it('prices every built-in content shape and merge-extended blocks with one fixed heuristic', () => {
+    const service = meter({ contextWindow: 100 })
+    const blocks: ContentBlock[] = [
+      { type: 'text', text: 'abcd' },
+      { type: 'reasoning', text: 'ab' },
+      { type: 'tool-call', id: CallId('c'), name: 'read', arguments: '{"x":1}' },
+      {
+        type: 'tool-result',
+        toolCallId: CallId('c'),
+        content: [{ type: 'text', text: 'xy' }],
+        isError: false,
+      },
+      { type: 'future-block', payload: 'abcd' } as unknown as ContentBlock,
+    ]
+    const estimated = service.estimateMessage({ role: 'assistant', content: blocks })
+    expect(estimated).toBeGreaterThan(30)
+    expect(service.estimateMessage(textMessage('abcd'))).toBe(9)
+  })
+
+  it('returns a detached deeply immutable empty measurement', () => {
+    const service = meter()
+    const session = new Session(SessionId('empty'))
+    const result = service.measure(session)
+    expect(result).toEqual({
+      logRevision: 0,
+      baseline: { kind: 'none', tokens: 0 },
+      surfaceDeltaTokens: 0,
+      totalTokens: 0,
+      surfaceTokens: 0,
+      nodes: [],
+    })
+    expect(Object.isFrozen(result)).toBe(true)
+    expect(Object.isFrozen(result.baseline)).toBe(true)
+    expect(Object.isFrozen(result.nodes)).toBe(true)
+    expectSurfaceTotal(result)
+    expect(() => {
+      ;(result as { totalTokens: number }).totalTokens = 1
+    }).toThrow(TypeError)
+  })
+
+  it('keeps an earlier unified snapshot detached from later replay', () => {
+    const service = meter()
+    const session = new Session(SessionId('detached'))
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'first' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+    const snapshot = service.measure(session)
+    const snapshotCopy = structuredClone(snapshot)
+    expect(Object.isFrozen(snapshot.nodes)).toBe(true)
+    expect(Object.isFrozen(snapshot.nodes[0])).toBe(true)
+    expectSurfaceTotal(snapshot)
+    expect(() => {
+      ;(snapshot.nodes as Array<{ seq: number; tokens: number }>).push({ seq: 99, tokens: 1 })
+    }).toThrow(TypeError)
+    expect(() => {
+      ;(snapshot.nodes[0] as { seq: number; tokens: number }).tokens = 1
+    }).toThrow(TypeError)
+
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'second' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+    const advanced = service.measure(session)
+    expect(advanced.logRevision).toBe(2)
+    expect(advanced.nodes).toHaveLength(2)
+    expectSurfaceTotal(advanced)
+    expect(snapshot).toEqual(snapshotCopy)
+    expect(snapshot.logRevision).toBe(1)
+    expect(snapshot.nodes).toHaveLength(1)
+  })
+
+  it('prices header, prefix, tools, and surface when no reusable usage exists', () => {
+    const service = meter()
+    const session = new Session(SessionId('heuristic'))
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'question' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+    appendHeader(session, header('deepseek-v4-flash', {
+      system: 'system',
+      messagePrefix: [textMessage('prefix')],
+      tools: [{ name: 'read', description: 'read', parameters: { type: 'object' } }],
+    }))
+    const result = service.measure(session)
+    expect(result.baseline.kind).toBe('estimated')
+    expect(result.totalTokens).toBeGreaterThan(result.surfaceTokens)
+    expect(result.logRevision).toBe(session.events.length)
+    expectSurfaceTotal(result)
+  })
+
+  it('keeps request-header overrides out of the returned surface', () => {
+    const service = meter()
+    const session = new Session(SessionId('override-surface'))
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'question' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+
+    const logged = service.measure(session)
+    const overridden = service.measure(session, header('another-model', {
+      system: 'large override '.repeat(100),
+    }))
+    expect(overridden.totalTokens).toBeGreaterThan(logged.totalTokens)
+    expect(overridden.surfaceTokens).toBe(logged.surfaceTokens)
+    expect(overridden.nodes).toEqual(logged.nodes)
+    expectSurfaceTotal(overridden)
+  })
+})
+
+describe('replay anchors and surface folds', () => {
+  const USAGE: TokenUsage = {
+    inputTokens: 20,
+    cacheReadTokens: 3,
+    cacheWriteTokens: 4,
+    outputTokens: 7,
+    reasoningTokens: 6,
+  }
+
+  it('uses disjoint provider usage and signed durable-output rewrites', () => {
+    const service = meter()
+    const session = new Session(SessionId('usage'))
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'before' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+    appendSuccessfulCall(session, header('deepseek-v4-flash'), {
+      providerText: 'short',
+      durableText: 'a much longer rewritten durable assistant answer',
+      usage: USAGE,
+    })
+    const result = service.measure(session)
+    expect(result.baseline).toMatchObject({ kind: 'usage', tokens: 34, usage: USAGE })
+    expect(result.surfaceDeltaTokens).toBeGreaterThan(0)
+    expect(result.totalTokens).toBe(34 + result.surfaceDeltaTokens)
+    expect(() => {
+      ;((result.baseline as { usage: { inputTokens: number } }).usage.inputTokens) = 1
+    }).toThrow(TypeError)
+  })
+
+  it('selects a heuristic anchor when provider usage would undercut its scale', () => {
+    const service = meter()
+    const session = new Session(SessionId('low-usage-anchor'))
+    const system = 'system context'
+    const requestHeader = header('deepseek-v4-flash', { system })
+    appendSuccessfulCall(session, requestHeader, {
+      providerText: 'abcd'.repeat(512),
+      usage: { inputTokens: 20, outputTokens: 7 },
+    })
+
+    const anchored = service.measure(session)
+    expect(anchored.baseline.kind).toBe('estimated')
+    const assistant = anchored.nodes[0]!.seq
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'short' }],
+      source: { kind: 'plugin', plugin: 'test' },
+    }, {
+      surfaceOp: { op: 'replace', start: assistant, end: assistant },
+      sourceEventSeqs: [assistant],
+    })
+
+    const shrunken = service.measure(session)
+    expect(27 + shrunken.surfaceDeltaTokens).toBeLessThan(0)
+    expect(shrunken.totalTokens).toBeGreaterThan(0)
+    expect(shrunken.totalTokens).toBe(service.measure(
+      session,
+      header('different-model', { system }),
+    ).totalTokens)
+  })
+
+  it('uses an estimated anchor when provider usage is absent', () => {
+    const service = meter()
+    const session = new Session(SessionId('missing-usage'))
+    appendSuccessfulCall(session, header('deepseek-v4-flash', { system: 's' }), {
+      providerText: 'provider',
+      durableText: 'rewritten',
+    })
+    const anchored = service.measure(session)
+    expect(anchored.baseline.kind).toBe('estimated')
+    expect(anchored.surfaceDeltaTokens).toBe(0)
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'later' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+    const advanced = service.measure(session)
+    expect(advanced.surfaceDeltaTokens).toBeGreaterThan(0)
+  })
+
+  it('distinguishes explicit empty provenance from absent legacy provenance', () => {
+    const explicit = new Session(SessionId('explicit-empty'))
+    const legacy = new Session(SessionId('legacy-absent'))
+    appendSuccessfulCall(explicit, header('deepseek-v4-flash'), {
+      durableText: 'listener injected text',
+      providerText: '',
+      usage: USAGE,
+      provenance: 'empty',
+    })
+    appendSuccessfulCall(legacy, header('deepseek-v4-flash'), {
+      durableText: 'listener injected text',
+      providerText: '',
+      usage: USAGE,
+      provenance: 'absent',
+    })
+    const service = meter()
+    expect(service.measure(explicit).surfaceDeltaTokens).toBeGreaterThan(0)
+    expect(service.measure(legacy).surfaceDeltaTokens).toBe(0)
+  })
+
+  it('keeps only the latest successful request anchor across model switches', () => {
+    const service = meter({ contextWindow: 1_000 })
+    const session = new Session(SessionId('switch'))
+    const alphaHeader = header('alpha', { system: 'same envelope' })
+    appendSuccessfulCall(session, alphaHeader, { usage: USAGE, providerText: 'alpha' })
+    expect(service.measure(session).baseline).toMatchObject({ kind: 'usage', tokens: 34 })
+
+    appendSuccessfulCall(session, header('beta'), {
+      turn: 1,
+      step: 2,
+      usage: { inputTokens: 100, outputTokens: 50 },
+      providerText: 'beta response',
+    })
+    expect(service.measure(session).baseline).toMatchObject({ kind: 'usage', tokens: 150 })
+
+    appendHeader(session, alphaHeader)
+    const switchedBack = service.measure(session)
+    expect(switchedBack.baseline.kind).toBe('estimated')
+    expect(switchedBack.surfaceDeltaTokens).toBe(0)
+  })
+
+  it('invalidates usage for any canonical envelope change or explicit override', () => {
+    const service = meter()
+    const session = new Session(SessionId('envelope'))
+    const anchoredHeader = header('deepseek-v4-flash', { system: 'one' })
+    appendSuccessfulCall(session, anchoredHeader, { usage: USAGE })
+    expect(service.measure(session, { ...anchoredHeader, tools: [] }).baseline.kind).toBe('usage')
+    expect(service.measure(session, header('deepseek-v4-flash', { system: 'two' })).baseline.kind)
+      .toBe('estimated')
+    expect(service.measure(session, header('deepseek-v4-pro', { system: 'one' })).baseline.kind)
+      .toBe('estimated')
+    expect(service.measure(session, {
+      ...anchoredHeader,
+      config: { ...anchoredHeader.config, temperature: 0.2 },
+    }).baseline.kind).toBe('estimated')
+    expect(service.measure(session, {
+      ...anchoredHeader,
+      messagePrefix: [textMessage('prefix')],
+    }).baseline.kind).toBe('estimated')
+    expect(service.measure(session, {
+      ...anchoredHeader,
+      tools: [{ name: 'read', description: 'read', parameters: { type: 'object' } }],
+    }).baseline.kind).toBe('estimated')
+  })
+
+  it('folds the latest full header snapshot into the effective envelope', () => {
+    const session = new Session(SessionId('header-snapshot'))
+    appendHeader(session, header('deepseek-v4-flash'))
+    session.append('request/header', {
+      header: header('deepseek-v4-pro'),
+      reason: 'change',
+    })
+    const result = meter().measure(session)
+    expect(result.baseline.kind).toBe('estimated')
+    expect(result.logRevision).toBe(2)
+  })
+
+  it('replays seeded append and replace operations with signed deltas', () => {
+    const service = meter()
+    const original = new Session(SessionId('surface-original'))
+    appendSuccessfulCall(original, header('deepseek-v4-flash'), {
+      usage: USAGE,
+      providerText: 'long provider answer '.repeat(100),
+    })
+    original.append('user/message', {
+      content: [{ type: 'text', text: 'new tail' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+    const seeded = new Session(SessionId('surface-seeded'), original.events)
+    const before = service.measure(seeded)
+    expect(before.nodes).toHaveLength(2)
+    expect(before.surfaceDeltaTokens).toBeGreaterThan(0)
+    expectSurfaceTotal(before)
+
+    const first = seeded.surface.nodes[0]!
+    seeded.append('user/message', {
+      content: [{ type: 'text', text: 'replacement' }],
+      source: { kind: 'plugin', plugin: 'test' },
+    }, { surfaceOp: { op: 'replace', start: first, end: first }, sourceEventSeqs: [first] })
+    const after = service.measure(seeded)
+    expect(after.nodes).toHaveLength(2)
+    expect(after.nodes[0]!.seq).toBe(seeded.events.length - 1)
+    expect(after.logRevision).toBe(seeded.events.length)
+    expect(Object.isFrozen(after.nodes)).toBe(true)
+    expect(Object.isFrozen(after.nodes[0])).toBe(true)
+    expect(after.surfaceDeltaTokens).toBeLessThan(0)
+    expectSurfaceTotal(after)
+    expect(before.nodes).toHaveLength(2)
+    expect(before.logRevision).toBe(original.events.length)
+    expect(before.surfaceDeltaTokens).toBeGreaterThan(0)
+  })
+
+  it('prices an empty assistant surface anchor as zero', () => {
+    const session = new Session(SessionId('empty-assistant'))
+    appendSuccessfulCall(session, header('deepseek-v4-flash'), {
+      providerText: '',
+      durableText: '',
+      provenance: 'empty',
+    })
+    const measurement = meter().measure(session)
+    const assistant = session.events.find(event => event.type === 'assistant/message')!
+    expect(measurement.nodes).toEqual([{ seq: assistant.seq, tokens: 0 }])
+    expect(measurement.surfaceTokens).toBe(0)
+    expectSurfaceTotal(measurement)
+  })
+})
+
+describe('malformed replay and listener lifecycle', () => {
+  function expectRepeatedFailure(service: TokenMeterService, session: Session, pattern: RegExp): void {
+    expect(() => service.measure(session)).toThrow(pattern)
+    expect(() => service.measure(session)).toThrow(pattern)
+  }
+
+  it('rejects an assistant without its step boundary transactionally', () => {
+    const session = new Session(SessionId('bad-step'))
+    appendHeader(session, header('deepseek-v4-flash'))
+    session.append('assistant/message', {
+      provenance: { provider: 'mock', model: 'deepseek-v4-flash' },
+      turn: 1,
+      step: 1,
+      content: [{ type: 'text', text: 'bad' }],
+    }, { surfaceOp: 'append', sourceEventSeqs: [] })
+    expectRepeatedFailure(meter(), session, /no matching step\/start/)
+  })
+
+  it('clears completed step boundaries and rejects overlapping or late step events', () => {
+    const overlapping = new Session(SessionId('overlapping-step'))
+    overlapping.append('step/start', { turn: 1, step: 1 })
+    overlapping.append('step/start', { turn: 1, step: 2 })
+    expectRepeatedFailure(
+      meter(),
+      overlapping,
+      /arrived before turn 1\/step 1 ended/,
+    )
+
+    const late = new Session(SessionId('late-assistant'))
+    late.append('step/start', { turn: 1, step: 1 })
+    appendHeader(late, header('deepseek-v4-flash'))
+    late.append('step/end', { turn: 1, step: 1 })
+    late.append('assistant/message', {
+      provenance: { provider: 'mock', model: 'deepseek-v4-flash' },
+      turn: 1,
+      step: 1,
+      content: [],
+    }, { surfaceOp: 'append', sourceEventSeqs: [] })
+    expectRepeatedFailure(
+      meter(),
+      late,
+      /no matching step\/start/,
+    )
+
+    const mismatchedEnd = new Session(SessionId('mismatched-end'))
+    mismatchedEnd.append('step/start', { turn: 1, step: 1 })
+    mismatchedEnd.append('step/end', { turn: 1, step: 2 })
+    expectRepeatedFailure(
+      meter(),
+      mismatchedEnd,
+      /step\/end .* no matching step\/start/,
+    )
+  })
+
+  it('rejects invalid assistant provenance', () => {
+    const cases: Array<{
+      name: string
+      appendSource(session: Session): number[]
+      pattern: RegExp
+    }> = [
+      {
+        name: 'non-chunk',
+        appendSource(session) {
+          return [session.append('user/message', {
+            content: [{ type: 'text', text: 'x' }],
+            source: { kind: 'user' },
+          }, { surfaceOp: 'append' }).seq]
+        },
+        pattern: /is not assistant\/chunk/,
+      },
+      {
+        name: 'wrong-step',
+        appendSource(session) {
+          return [session.append('assistant/chunk', {
+            turn: 1,
+            step: 2,
+            chunk: { type: 'finish', reason: { kind: 'stop' } },
+          }).seq]
+        },
+        pattern: /belongs to another step/,
+      },
+    ]
+    for (const testCase of cases) {
+      const session = new Session(SessionId(`bad-source-${testCase.name}`))
+      session.append('step/start', { turn: 1, step: 1 })
+      appendHeader(session, header('deepseek-v4-flash'))
+      const sourceEventSeqs = testCase.appendSource(session)
+      session.append('assistant/message', {
+        provenance: { provider: 'mock', model: 'deepseek-v4-flash' },
+        turn: 1,
+        step: 1,
+        content: [{ type: 'text', text: 'bad' }],
+        usage: { inputTokens: 1, outputTokens: 1 },
+      }, { surfaceOp: 'append', sourceEventSeqs })
+      expect(() => meter().measure(session)).toThrow(testCase.pattern)
+    }
+  })
+
+  it('rejects repeated and non-earlier assistant provenance', () => {
+    const duplicate = new Session(SessionId('duplicate-source'))
+    duplicate.append('step/start', { turn: 1, step: 1 })
+    appendHeader(duplicate, header('deepseek-v4-flash'))
+    const source = duplicate.append('assistant/chunk', {
+      turn: 1,
+      step: 1,
+      chunk: { type: 'finish', reason: { kind: 'stop' } },
+    }).seq
+    appendUnchecked(duplicate, {
+      type: 'assistant/message',
+      seq: duplicate.seq,
+      time: 0,
+      data: {
+        provenance: { provider: 'mock', model: 'deepseek-v4-flash' },
+        turn: 1,
+        step: 1,
+        content: [],
+        usage: { inputTokens: 1, outputTokens: 0 },
+      },
+      surfaceOp: 'append',
+      sourceEventSeqs: [source, source],
+    })
+    expect(() => meter().measure(duplicate)).toThrow(/repeats source seq/)
+
+    const future = new Session(SessionId('future-source'))
+    future.append('step/start', { turn: 1, step: 1 })
+    appendHeader(future, header('deepseek-v4-flash'))
+    appendUnchecked(future, {
+      type: 'assistant/message',
+      seq: future.seq,
+      time: 0,
+      data: {
+        provenance: { provider: 'mock', model: 'deepseek-v4-flash' },
+        turn: 1,
+        step: 1,
+        content: [],
+        usage: { inputTokens: 1, outputTokens: 0 },
+      },
+      surfaceOp: 'append',
+      sourceEventSeqs: [99],
+    })
+    expect(() => meter().measure(future)).toThrow(/is not earlier/)
+  })
+
+  it('does not partially apply a malformed assistant replacement', () => {
+    const session = new Session(SessionId('transactional-replace'))
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'head' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+    appendHeader(session, header('deepseek-v4-flash'))
+    const head = session.events[0]!.seq
+    session.append('assistant/message', {
+      provenance: { provider: 'mock', model: 'deepseek-v4-flash' },
+      turn: 1,
+      step: 1,
+      content: [{ type: 'text', text: 'replacement' }],
+    }, { surfaceOp: { op: 'replace', start: head, end: head }, sourceEventSeqs: [head] })
+    expectRepeatedFailure(
+      meter(),
+      session,
+      /no matching step\/start/,
+    )
+  })
+
+  it('rejects corrupt replacement ranges without advancing the replay cursor', () => {
+    const session = new Session(SessionId('bad-replace'))
+    const head = session.append('user/message', {
+      content: [{ type: 'text', text: 'head' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' }).seq
+    appendUnchecked(session, {
+      type: 'user/message',
+      seq: session.seq,
+      time: 0,
+      data: {
+        content: [{ type: 'text', text: 'bad' }],
+        source: { kind: 'user' },
+      },
+      surfaceOp: { op: 'replace', start: 99, end: 99 },
+      sourceEventSeqs: [head],
+    })
+    expectRepeatedFailure(meter(), session, /invalid current range/)
+  })
+
+  it('handles earlier-reader catch-up, eager observation, and service reload', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    let activeMeter: TokenMeterService | undefined
+    const revisions: number[] = []
+    ctx.on('session/event', (session) => {
+      if (activeMeter !== undefined) revisions.push(activeMeter.measure(session).logRevision)
+    })
+    const firstFiber = await ctx.plugin(TokenMeterService)
+    activeMeter = ctx.tokenMeter
+    const session = ctx.sessions.create(SessionId('listener-order'))
+    activeMeter.measure(session)
+    session.append('user/message', {
+      content: [{ type: 'text', text: 'one' }],
+      source: { kind: 'user' },
+    }, { surfaceOp: 'append' })
+    expect(revisions).toEqual([1])
+    expect(activeMeter.measure(session).logRevision).toBe(1)
+
+    await firstFiber.dispose()
+    const secondFiber = await ctx.plugin(TokenMeterService)
+    activeMeter = ctx.tokenMeter
+    expect(activeMeter.measure(session).logRevision).toBe(1)
+    await secondFiber.dispose()
+  })
+})
diff --git a/packages/llm/token-meter/tsconfig.json b/packages/llm/token-meter/tsconfig.json
new file mode 100644
index 0000000000..5e1604e02f
--- /dev/null
+++ b/packages/llm/token-meter/tsconfig.json
@@ -0,0 +1,27 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": [
+    "src"
+  ],
+  "references": [
+    {
+      "path": "../../../vendor/cosmokit"
+    },
+    {
+      "path": "../../../vendor/cordis"
+    },
+    {
+      "path": "../../../vendor/schemastery"
+    },
+    {
+      "path": "../../llm/llm"
+    },
+    {
+      "path": "../../core/session"
+    }
+  ]
+}
diff --git a/packages/mcp/mcp-client/tests/fixture-server.ts b/packages/mcp/mcp-client/tests/fixture-server.ts
index d127412736..8491b96634 100644
--- a/packages/mcp/mcp-client/tests/fixture-server.ts
+++ b/packages/mcp/mcp-client/tests/fixture-server.ts
@@ -2,7 +2,7 @@
  * Minimal MCP server over stdio for e2e testing of the dsh-mcp-client plugin.
  * Registers controlled tools with predictable behavior for asserting edge cases.
  *
- * Run: node --import tsx fixture-server.ts
+ * Run: node fixture-server.ts
  */
 
 import { McpServer } from '@modelcontextprotocol/sdk/server/mcp.js'
diff --git a/packages/mcp/mcp-client/tests/mcp-client.e2e.ts b/packages/mcp/mcp-client/tests/mcp-client.e2e.ts
index 686d51acea..dc783b3edf 100644
--- a/packages/mcp/mcp-client/tests/mcp-client.e2e.ts
+++ b/packages/mcp/mcp-client/tests/mcp-client.e2e.ts
@@ -26,9 +26,7 @@ import { apply } from '@deepseek-ai/dsh-mcp-client/src/index.ts'
 import { publicToolName } from '@deepseek-ai/dsh-mcp-client/src/tools.ts'
 import type { Config } from '@deepseek-ai/dsh-mcp-client'
 
-const tsxLoader = fileURLToPath(import.meta.resolve('tsx'))
 const fixtureServerPath = fileURLToPath(new URL('./fixture-server.ts', import.meta.url))
-const repoTsconfig = fileURLToPath(new URL('../../../../tsconfig.json', import.meta.url))
 
 // Resolve package-local .bin for pnpm-hoisted MCP server binaries.
 const packageDir = fileURLToPath(new URL('..', import.meta.url))
@@ -86,8 +84,8 @@ describe('fixture server — controlled scenarios', () => {
     transport: 'stdio',
     serverName: 'fixture',
     command: process.execPath,
-    args: ['--import', tsxLoader, fixtureServerPath],
-    env: { TSX_TSCONFIG_PATH: repoTsconfig },
+    args: [fixtureServerPath],
+    env: {},
     cwd: packageDir,
     toolCallTimeoutMs: 15_000,
   }
@@ -170,8 +168,8 @@ describe('fixture server — duplicate serverName', () => {
       transport: 'stdio',
       serverName: 'dup',
       command: process.execPath,
-      args: ['--import', tsxLoader, fixtureServerPath],
-      env: { TSX_TSCONFIG_PATH: repoTsconfig },
+      args: [fixtureServerPath],
+      env: {},
       cwd: packageDir,
       toolCallTimeoutMs: 15_000,
     }
@@ -191,8 +189,8 @@ describe('fixture server — disposal', () => {
       transport: 'stdio',
       serverName: 'fixture',
       command: process.execPath,
-      args: ['--import', tsxLoader, fixtureServerPath],
-      env: { TSX_TSCONFIG_PATH: repoTsconfig },
+      args: [fixtureServerPath],
+      env: {},
       cwd: packageDir,
       toolCallTimeoutMs: 15_000,
     })
diff --git a/packages/sandbox/sandbox-local/README.md b/packages/sandbox/sandbox-local/README.md
index fa0ff7257f..cbd765bbf9 100644
--- a/packages/sandbox/sandbox-local/README.md
+++ b/packages/sandbox/sandbox-local/README.md
@@ -2,6 +2,8 @@
 
 Local implementation of the [`dsh-sandbox`](../sandbox/) seam. It selects and caches one platform runner: Linux prefers a working `bwrap` then Landlock; macOS uses Seatbelt. Multiple candidates are probed in order, while a sole candidate is selected directly.
 
+The package root exports the default and named `LocalSandboxProvider` plugin, `Config`, and its public test-injection seam; platform profile builders stay internal.
+
 Unsupported platforms and unusable runners fail closed with `SANDBOX_UNAVAILABLE`; execution never silently falls through unconfined. Each wrap carries runner-failure signatures so consumers can distinguish a broken sandbox from a command failure. The [sandbox RFC](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md) owns selection rationale and profile differences.
 
 Policy is per call; the provider stores only the mechanism and cached runner verdict. Each wrap reports enforcement completeness plus backend-specific denial and runner-failure signatures. `runnerCommand` is an operator assertion of a bwrap-shaped runner and skips probes, but missing or unexecutable commands still fail closed at execution. Because its mechanism is unknown, it carries both Linux denial dialects. `probeTimeoutMs` bounds functional probes. The [sandbox RFC](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md) owns selection and failure semantics.
diff --git a/packages/sandbox/sandbox-local/src/index.ts b/packages/sandbox/sandbox-local/src/index.ts
index cc464eb5c4..98dc86d23e 100644
--- a/packages/sandbox/sandbox-local/src/index.ts
+++ b/packages/sandbox/sandbox-local/src/index.ts
@@ -7,14 +7,13 @@
  */
 
 import { spawnSync } from 'node:child_process'
-import { realpathSync } from 'node:fs'
-import { tmpdir } from 'node:os'
-import { grantArgs as landlockGrantArgs, LAUNCHER_BIN, launcherPath as landlockLauncherPath, probe as defaultProbeLandlock } from 'node-addon-landlock-run'
+import { LAUNCHER_BIN, launcherPath as landlockLauncherPath, probe as defaultProbeLandlock } from 'node-addon-landlock-run'
 import { Context } from 'cordis'
 import z from 'schemastery'
 import { assertNever } from '@deepseek-ai/dsh-llm'
 import { SandboxProvider, SandboxUnavailableError } from '@deepseek-ai/dsh-sandbox'
 import type { ConfinedArgv, ConfinedSandboxMode, SandboxEnforcement, SandboxPolicy } from '@deepseek-ai/dsh-sandbox'
+import { bwrapProfileArgs, landlockProfileArgs, seatbeltProfileArgs } from './profiles.ts'
 
 /** Plugin config. All optional — `static Config` supplies the defaults. */
 export interface Config {
@@ -38,77 +37,6 @@ export interface Config {
   probeTimeoutMs?: number
 }
 
-/**
- * Build a bwrap profile: the host is read-only with fresh `/dev` and `/proc`;
- * workspace-write overlays writable temp and workspace mounts. PID and network
- * isolation are intentionally outside the file-effect policy.
- *
- * @param policy - the file-effect policy to express as bwrap arguments.
- * @returns the bwrap profile arguments (before the trailing `--` + argv).
- */
-export function bwrapProfileArgs(policy: SandboxPolicy): string[] {
-  const args = ['--ro-bind', '/', '/', '--dev', '/dev', '--proc', '/proc', '--die-with-parent']
-  if (policy.mode === 'workspace-write') {
-    args.push('--tmpfs', '/tmp')
-    args.push('--bind', policy.workspaceRoot, policy.workspaceRoot)
-  }
-  return args
-}
-
-/**
- * Build Landlock grants for the same file policy without synthetic mounts.
- * Read-only grants only `/dev/null` for writes; workspace-write also grants the
- * host temp root and workspace.
- *
- * @param policy - the file-effect policy to express as launcher grants.
- * @returns the launcher grant arguments (before `--` + argv).
- */
-export function landlockProfileArgs(policy: SandboxPolicy): string[] {
-  const readWrite = ['/dev/null']
-  if (policy.mode === 'workspace-write') {
-    readWrite.push('/tmp', policy.workspaceRoot)
-  }
-  return landlockGrantArgs({ readOnly: ['/'], readWrite })
-}
-
-/**
- * Resolve a granted root to the path the kernel actually sees. Seatbelt path
- * filters match the CANONICAL path (symlinks resolved), and the roots this
- * profile grants are symlinked on every macOS: `/tmp` is `/private/tmp` and
- * the user temp dir lives under `/var` → `/private/var` — an as-spelled
- * grant would match nothing.
- */
-function canonicalPath(path: string): string {
-  try {
-    return realpathSync(path)
-  } catch {
-    // An unresolved grant matches nothing until the named path exists; keep its spelling.
-    return path
-  }
-}
-
-/** Quote one path as an SBPL string literal (backslashes and double quotes escaped). */
-function sbplString(path: string): string {
-  return `"${path.replaceAll('\\', String.raw`\\`).replaceAll('"', String.raw`\"`)}"`
-}
-
-/**
- * Build a Seatbelt profile that denies file writes then allows `/dev/null` and,
- * for workspace-write, the canonical workspace, host temp, and per-user macOS
- * temp roots. Network and process visibility remain unrestricted.
- *
- * @param policy - the file-effect policy to express as an SBPL profile.
- * @returns the `sandbox-exec` arguments (`-p` + profile, before `--` + argv).
- */
-export function seatbeltProfileArgs(policy: SandboxPolicy): string[] {
-  const forms = ['(version 1)', '(allow default)', '(deny file-write*)', `(allow file-write* (literal ${sbplString('/dev/null')}))`]
-  if (policy.mode === 'workspace-write') {
-    const roots = [...new Set([policy.workspaceRoot, '/tmp', tmpdir()].map(canonicalPath))]
-    forms.push(`(allow file-write* ${roots.map(root => `(subpath ${sbplString(root)})`).join(' ')})`)
-  }
-  return ['-p', forms.join(' ')]
-}
-
 /** Probe whether `bwrap` can create the profile; the provider caches the bounded result. */
 function defaultProbeBwrap(timeoutMs: number): boolean {
   const probe = spawnSync('bwrap', ['--ro-bind', '/', '/', '--dev', '/dev', '--proc', '/proc', '--die-with-parent', '--', 'true'], {
diff --git a/packages/sandbox/sandbox-local/src/profiles.ts b/packages/sandbox/sandbox-local/src/profiles.ts
new file mode 100644
index 0000000000..9303583cbb
--- /dev/null
+++ b/packages/sandbox/sandbox-local/src/profiles.ts
@@ -0,0 +1,67 @@
+/**
+ * Internal platform-profile builders for the local sandbox provider.
+ *
+ * @module @deepseek-ai/dsh-sandbox-local/profiles
+ */
+
+import { realpathSync } from 'node:fs'
+import { tmpdir } from 'node:os'
+import { grantArgs as landlockGrantArgs } from 'node-addon-landlock-run'
+import type { SandboxPolicy } from '@deepseek-ai/dsh-sandbox'
+
+/**
+ * Build the bwrap profile arguments for one file-effect policy.
+ * @param policy - file-effect policy to express as bwrap mounts.
+ * @returns profile arguments before the trailing separator and command argv.
+ */
+export function bwrapProfileArgs(policy: SandboxPolicy): string[] {
+  const args = ['--ro-bind', '/', '/', '--dev', '/dev', '--proc', '/proc', '--die-with-parent']
+  if (policy.mode === 'workspace-write') {
+    args.push('--tmpfs', '/tmp')
+    args.push('--bind', policy.workspaceRoot, policy.workspaceRoot)
+  }
+  return args
+}
+
+/**
+ * Build the Landlock launcher grants for one file-effect policy.
+ * @param policy - file-effect policy to express as Landlock allow-list grants.
+ * @returns launcher grant arguments before the trailing separator and command argv.
+ */
+export function landlockProfileArgs(policy: SandboxPolicy): string[] {
+  const readWrite = ['/dev/null']
+  if (policy.mode === 'workspace-write') {
+    readWrite.push('/tmp', policy.workspaceRoot)
+  }
+  return landlockGrantArgs({ readOnly: ['/'], readWrite })
+}
+
+/** Resolve a granted root to the canonical path the Seatbelt kernel sees. */
+function canonicalPath(path: string): string {
+  try {
+    return realpathSync(path)
+  } catch {
+    // Missing or unreadable roots stay as spelled; an unresolved root grants
+    // nothing until it exists, which is the conservative outcome.
+    return path
+  }
+}
+
+/** Quote one path as an SBPL string literal. */
+function sbplString(path: string): string {
+  return `"${path.replaceAll('\\', String.raw`\\`).replaceAll('"', String.raw`\"`)}"`
+}
+
+/**
+ * Build the sandbox-exec arguments and SBPL profile for one policy.
+ * @param policy - file-effect policy to express as an SBPL profile.
+ * @returns sandbox-exec arguments before the trailing separator and command argv.
+ */
+export function seatbeltProfileArgs(policy: SandboxPolicy): string[] {
+  const forms = ['(version 1)', '(allow default)', '(deny file-write*)', `(allow file-write* (literal ${sbplString('/dev/null')}))`]
+  if (policy.mode === 'workspace-write') {
+    const roots = [...new Set([policy.workspaceRoot, '/tmp', tmpdir()].map(canonicalPath))]
+    forms.push(`(allow file-write* ${roots.map(root => `(subpath ${sbplString(root)})`).join(' ')})`)
+  }
+  return ['-p', forms.join(' ')]
+}
diff --git a/packages/sandbox/sandbox-local/tests/bwrap.e2e.ts b/packages/sandbox/sandbox-local/tests/bwrap.e2e.ts
index 1e818e1bdf..9a093e4632 100644
--- a/packages/sandbox/sandbox-local/tests/bwrap.e2e.ts
+++ b/packages/sandbox/sandbox-local/tests/bwrap.e2e.ts
@@ -6,7 +6,8 @@ import { join } from 'node:path'
 import { afterEach, describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
 import type { SandboxPolicy } from '@deepseek-ai/dsh-sandbox'
-import { bwrapProfileArgs, LocalSandboxProvider } from '@deepseek-ai/dsh-sandbox-local'
+import { LocalSandboxProvider } from '@deepseek-ai/dsh-sandbox-local'
+import { bwrapProfileArgs } from '../src/profiles.ts'
 
 /**
  * Keyless backend integration through `confine()` and a real bwrap process. With no rung forced,
diff --git a/packages/sandbox/sandbox-local/tests/local.spec.ts b/packages/sandbox/sandbox-local/tests/local.spec.ts
index c221544861..f9efbf992c 100644
--- a/packages/sandbox/sandbox-local/tests/local.spec.ts
+++ b/packages/sandbox/sandbox-local/tests/local.spec.ts
@@ -15,12 +15,10 @@ import { Context } from 'cordis'
 import { SANDBOX_UNAVAILABLE, SandboxUnavailableError } from '@deepseek-ai/dsh-sandbox'
 import type { SandboxPolicy } from '@deepseek-ai/dsh-sandbox'
 import {
-  bwrapProfileArgs,
-  landlockProfileArgs,
   LocalSandboxProvider,
-  seatbeltProfileArgs,
 } from '@deepseek-ai/dsh-sandbox-local'
 import type { Config } from '@deepseek-ai/dsh-sandbox-local'
+import { bwrapProfileArgs, landlockProfileArgs, seatbeltProfileArgs } from '../src/profiles.ts'
 
 const RO: SandboxPolicy = { mode: 'read-only', workspaceRoot: '/ws' }
 const WW: SandboxPolicy = { mode: 'workspace-write', workspaceRoot: '/ws' }
diff --git a/packages/sandbox/sandbox-local/tests/seatbelt.e2e.ts b/packages/sandbox/sandbox-local/tests/seatbelt.e2e.ts
index 2c5a026ddb..6d645b1a3b 100644
--- a/packages/sandbox/sandbox-local/tests/seatbelt.e2e.ts
+++ b/packages/sandbox/sandbox-local/tests/seatbelt.e2e.ts
@@ -6,7 +6,8 @@ import { join } from 'node:path'
 import { afterEach, describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
 import type { SandboxPolicy } from '@deepseek-ai/dsh-sandbox'
-import { LocalSandboxProvider, seatbeltProfileArgs } from '@deepseek-ai/dsh-sandbox-local'
+import { LocalSandboxProvider } from '@deepseek-ai/dsh-sandbox-local'
+import { seatbeltProfileArgs } from '../src/profiles.ts'
 
 /**
  * Keyless backend integration through `confine()` and a real macOS Seatbelt process, with Linux
diff --git a/packages/sdk/README.md b/packages/sdk/README.md
new file mode 100644
index 0000000000..953d52ea76
--- /dev/null
+++ b/packages/sdk/README.md
@@ -0,0 +1,15 @@
+# SDK packages
+
+Developer tooling for creating, editing, building, and running DeepSeek Harness projects.
+
+The [feature RFC](../../docs/rfc/proposed/feature/2026-07-14-sdk-developer-projects.md) owns the developer workflow; the [architecture RFC](../../docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.md) owns the package and project-editing boundaries.
+
+| Package | Role |
+|---|---|
+| [`helper`](helper/README.md) | Project aggregate, edit session, builtin features, project documents, templates, package managers, and prompt abstraction |
+| [`scripts`](scripts/README.md) | The `dsh-sdk` launcher: `start`, `dev`, `build`, and interactive `config` |
+| [`create-sdk`](create-sdk/README.md) | The `npm create @deepseek-ai/sdk` initializer |
+
+`@deepseek-ai/create-sdk` is the one package-name exception to the repository's `@deepseek-ai/dsh-*` rule: npm's scoped initializer convention requires that name for `npm create @deepseek-ai/sdk`.
+
+Generated projects keep `cordis.yml` as the only runtime plugin tree. `dsh-sdk dev` adds TypeScript and local-workspace resolution around that same file; it does not create a development-only config.
diff --git a/packages/sdk/create-sdk/README.md b/packages/sdk/create-sdk/README.md
new file mode 100644
index 0000000000..66cddc030a
--- /dev/null
+++ b/packages/sdk/create-sdk/README.md
@@ -0,0 +1,19 @@
+# `@deepseek-ai/create-sdk`
+
+Interactive initializer for `npm create @deepseek-ai/sdk [directory]`. Directory/name/description have visible editable defaults. A tree picker selects features and configures finite options with Right/Left navigation; secret text follows only for selected options. Local plugin creation is one none/plugin/tool choice.
+
+The supported package surface is the `create-sdk` bin. The package root exports no symbols, and workflow, bin, source, and package-manifest subpaths are not exported.
+
+The initializer rejects every existing target path, creates one `SdkProject` edit session, validates and commits it, then asks whether to install NPM dependencies and build. Install or build failures keep the generated project and print a retry command.
+
+Public flags are `[directory]`, `--description`, `--provider`, `--base-url`, `--api-key`, `--model`, `--interface`, `--pm`, and `--install`/`--no-install`. Flags prefill matching questions, but creation always requires a TTY.
+
+The provider choice is DeepSeek or a custom endpoint backed by `llm-pi-ai`. DeepSeek asks only for an API key and uses the public endpoint plus `deepseek-v4-flash`; custom also asks for a base URL. An empty key requires confirmation and creates a commented empty `.env` variable so provider startup fails clearly until it is filled. Existing plugin defaults are omitted; required SDK presets remain typed against the owning package's Config.
+
+## Model Experience
+
+Indirectly, through the generated project composition and its selected runtime plugins.
+
+## Known Limitations and Deferred Work
+
+- **TTY-only creation** — flags prefill questions, but the wizard still requires an interactive terminal before it writes a project.
diff --git a/packages/sdk/create-sdk/package.json b/packages/sdk/create-sdk/package.json
new file mode 100644
index 0000000000..e23415f8dc
--- /dev/null
+++ b/packages/sdk/create-sdk/package.json
@@ -0,0 +1,37 @@
+{
+  "name": "@deepseek-ai/create-sdk",
+  "description": "Create a DeepSeek Harness SDK project with npm create @deepseek-ai/sdk",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "bin": {
+    "create-sdk": "lib/bin.js"
+  },
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    }
+  },
+  "files": [
+    "lib/index.js",
+    "lib/bin.js",
+    "lib/assets",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "dependencies": {
+    "@deepseek-ai/dsh-helper": "workspace:^",
+    "commander": "^15.0.0"
+  },
+  "peerDependencies": {
+    "cordis": "^4.0.0-rc.7"
+  },
+  "devDependencies": {
+    "cordis": "^4.0.0-rc.7"
+  }
+}
diff --git a/packages/sdk/create-sdk/src/args.ts b/packages/sdk/create-sdk/src/args.ts
new file mode 100644
index 0000000000..521132caa9
--- /dev/null
+++ b/packages/sdk/create-sdk/src/args.ts
@@ -0,0 +1,84 @@
+/**
+ * Commander adapter for the create-sdk command surface.
+ *
+ * @module @deepseek-ai/create-sdk/args
+ */
+
+import { Command, Option } from 'commander'
+import type { PackageManagerName, RunInterface } from '@deepseek-ai/dsh-helper'
+
+/** Parsed create command flags before interactive resolution. */
+export interface CreateArgs {
+  directory?: string
+  description?: string
+  provider?: 'deepseek' | 'custom'
+  baseURL?: string
+  apiKey?: string
+  model?: string
+  runInterface?: RunInterface
+  packageManager?: PackageManagerName
+  install?: boolean
+  linkWorkspace?: boolean
+  help: boolean
+}
+
+interface CommanderCreateOptions {
+  description?: string
+  provider?: 'deepseek' | 'custom'
+  baseUrl?: string
+  apiKey?: string
+  model?: string
+  interface?: RunInterface
+  pm?: PackageManagerName
+  install?: boolean
+  linkWorkspace?: boolean
+  help?: boolean
+}
+
+function createProgram(): Command {
+  return new Command()
+    .name('create-sdk')
+    .description('Create a DeepSeek Harness SDK project')
+    .helpOption(false)
+    .showHelpAfterError(false)
+    .exitOverride()
+    .configureOutput({
+      /* v8 ignore next -- the command wrapper renders the package-owned usage template */
+      writeOut: () => {},
+      /* v8 ignore next -- Commander output is deliberately suppressed; errors are returned to the bin wrapper */
+      writeErr: () => {},
+    })
+    .argument('[directory]')
+    .option('-h, --help')
+    .option('--description <text>')
+    .addOption(new Option('--provider <name>').choices(['deepseek', 'custom']))
+    .option('--base-url <url>')
+    .option('--api-key <key>')
+    .option('--model <name>')
+    .addOption(new Option('--interface <name>').choices(['acp', 'stdio', 'embed']))
+    .addOption(new Option('--pm <name>').choices(['npm', 'pnpm', 'yarn']))
+    .addOption(new Option('--install').default(undefined))
+    .addOption(new Option('--no-install').default(undefined))
+    .option('--link-workspace')
+}
+
+/** Parse create-sdk positionals/options through Commander into a domain-neutral value. */
+export function parseCreateArgs(argv: readonly string[]): CreateArgs {
+  const program = createProgram()
+  program.parse([...argv], { from: 'user' })
+  const options = program.opts<CommanderCreateOptions>()
+  const directory = program.processedArgs[0] as string | undefined
+  return {
+    ...directory === undefined ? {} : { directory },
+    ...options.description === undefined ? {} : { description: options.description },
+    ...options.provider === undefined ? {} : { provider: options.provider },
+    ...options.baseUrl === undefined ? {} : { baseURL: options.baseUrl },
+    ...options.apiKey === undefined ? {} : { apiKey: options.apiKey },
+    ...options.model === undefined ? {} : { model: options.model },
+    ...options.interface === undefined ? {} : { runInterface: options.interface },
+    ...options.pm === undefined ? {} : { packageManager: options.pm },
+    ...options.install === undefined ? {} : { install: options.install },
+    ...options.linkWorkspace ? { linkWorkspace: true } : {},
+    help: options.help ?? false,
+  }
+}
diff --git a/packages/sdk/create-sdk/src/bin.ts b/packages/sdk/create-sdk/src/bin.ts
new file mode 100644
index 0000000000..7a04af8b84
--- /dev/null
+++ b/packages/sdk/create-sdk/src/bin.ts
@@ -0,0 +1,10 @@
+#!/usr/bin/env node
+/**
+ * Self-executing create-sdk command.
+ *
+ * @module @deepseek-ai/create-sdk/bin
+ */
+
+import { runCreateCommand } from './command.ts'
+
+process.exitCode = await runCreateCommand()
diff --git a/packages/sdk/create-sdk/src/command.ts b/packages/sdk/create-sdk/src/command.ts
new file mode 100644
index 0000000000..0c7076c90b
--- /dev/null
+++ b/packages/sdk/create-sdk/src/command.ts
@@ -0,0 +1,111 @@
+/**
+ * Internal create-sdk command composition used by the package bin.
+ *
+ * @module @deepseek-ai/create-sdk/command
+ */
+
+import { readFile } from 'node:fs/promises'
+import {
+  ClackPromptPort,
+  PromptCancelledError,
+  type PackageManagerVersionProbe,
+  type PromptPort,
+} from '@deepseek-ai/dsh-helper'
+import { parseCreateArgs } from './args.ts'
+import { CreateWizard, type ResolvedCreateRequest } from './create-wizard.ts'
+import { scaffoldProject, type ScaffoldResult } from './project-scaffolder.ts'
+import { CREATE_TEMPLATES, packageManagerTemplateModel } from './templates/create-templates.ts'
+
+/** Process and terminal slice used by the initializer. */
+export interface CreateCommandContext {
+  cwd: string
+  stdin: NodeJS.ReadStream
+  stdout: NodeJS.WriteStream
+  stderr: NodeJS.WriteStream
+  releaseVersion?: string
+  versionProbe?: PackageManagerVersionProbe
+  port?: PromptPort
+  setup?: (request: ResolvedCreateRequest) => Promise<void>
+}
+
+/** Read this initializer package's release version in source and built layouts. */
+export async function readCreateSdkVersion(): Promise<string> {
+  const manifest = JSON.parse(await readFile(new URL('../package.json', import.meta.url), 'utf8')) as { version?: unknown }
+  /* v8 ignore next -- this package's checked-in manifest always carries its version */
+  if (typeof manifest.version !== 'string') throw new Error('create-sdk package version is missing')
+  return manifest.version
+}
+
+/** Resolve, write, optionally install, and build one new project. */
+export async function createProject(
+  argv: readonly string[],
+  context: CreateCommandContext,
+): Promise<ScaffoldResult | undefined> {
+  const args = parseCreateArgs(argv)
+  if (args.help) {
+    context.stdout.write(CREATE_TEMPLATES.usage.render({}))
+    return undefined
+  }
+  if (!context.port && (!context.stdin.isTTY || !context.stdout.isTTY)) {
+    throw new Error('create-sdk requires an interactive TTY')
+  }
+  const wizard = new CreateWizard({
+    args,
+    /* v8 ignore next -- production TTY wiring is exercised by the built-bin smoke */
+    port: context.port ?? new ClackPromptPort(context.stdin, context.stdout),
+    cwd: context.cwd,
+    releaseVersion: context.releaseVersion ?? await readCreateSdkVersion(),
+    ...context.versionProbe ? { versionProbe: context.versionProbe } : {},
+  })
+  const resolved = await wizard.run()
+  const result = await scaffoldProject(resolved.directory, resolved.request)
+  context.stdout.write(CREATE_TEMPLATES.created.render({
+    name: resolved.request.name,
+    directory: resolved.directory,
+  }))
+  if (resolved.install) {
+    try {
+      if (context.setup) await context.setup(resolved)
+      else {
+        await resolved.request.packageManager.install(resolved.directory)
+        await resolved.request.packageManager.build(resolved.directory)
+      }
+    } catch (error) {
+      context.stderr.write(CREATE_TEMPLATES.setupFailure.render({
+        directory: resolved.directory,
+        error: String(error),
+        ...packageManagerTemplateModel(resolved.request.packageManager),
+      }))
+      throw error
+    }
+  }
+  context.stdout.write(CREATE_TEMPLATES.nextSteps.render({
+    directory: resolved.directory,
+    setupRequired: !resolved.install,
+    ...packageManagerTemplateModel(resolved.request.packageManager),
+  }))
+  return result
+}
+
+/** Run the create command with process defaults and convert cancellation to a clean exit. */
+export async function runCreateCommand(
+  argv: readonly string[] = process.argv.slice(2),
+  context: CreateCommandContext = {
+    cwd: process.cwd(),
+    stdin: process.stdin,
+    stdout: process.stdout,
+    stderr: process.stderr,
+  },
+): Promise<number> {
+  try {
+    await createProject(argv, context)
+    return 0
+  } catch (error) {
+    if (error instanceof PromptCancelledError) {
+      context.stderr.write('create-sdk: cancelled\n')
+      return 1
+    }
+    context.stderr.write(`create-sdk: ${error instanceof Error ? error.message : String(error)}\n`)
+    return 1
+  }
+}
diff --git a/packages/sdk/create-sdk/src/create-questions.ts b/packages/sdk/create-sdk/src/create-questions.ts
new file mode 100644
index 0000000000..c53e6e227e
--- /dev/null
+++ b/packages/sdk/create-sdk/src/create-questions.ts
@@ -0,0 +1,205 @@
+/**
+ * Static create-sdk question sequence; dynamic feature/plugin loops remain
+ * in the wizard orchestrator.
+ *
+ * @module @deepseek-ai/create-sdk/create-questions
+ */
+
+import { existsSync } from 'node:fs'
+import { basename, resolve } from 'node:path'
+import {
+  ConfirmQuestion,
+  SecretQuestion,
+  SelectQuestion,
+  TextQuestion,
+  requireAnswer,
+  type PromptPort,
+  type Question,
+  type RunInterface,
+} from '@deepseek-ai/dsh-helper'
+import type { CreateArgs } from './args.ts'
+
+/** Answers that establish project identity and feature applicability. */
+export interface ProjectAnswers {
+  directory: string
+  name: string
+  description: string
+  provider: 'deepseek' | 'custom'
+  baseURL: string
+  apiKey: string
+  model: string
+  runInterface: RunInterface
+}
+
+interface ProjectAnswerState extends Partial<ProjectAnswers> {
+  readonly args: CreateArgs
+  readonly cwd: string
+}
+
+interface WizardStep<TState> {
+  run(port: PromptPort, state: TState): Promise<void>
+}
+
+function questionStep<TState, TValue>(options: {
+  question: (state: TState) => Question<TValue>
+  when?: (state: TState) => boolean
+  prefilled?: (state: TState) => TValue | undefined
+  apply: (state: TState, value: TValue) => void
+}): WizardStep<TState> {
+  return {
+    async run(port, state) {
+      if (options.when && !options.when(state)) return
+      const value = requireAnswer(await options.question(state).resolve(port, options.prefilled?.(state)))
+      options.apply(state, value)
+    },
+  }
+}
+
+/** Validate one required text answer. */
+function nonEmpty(value: string): string | undefined {
+  return value.trim().length === 0 ? 'A value is required' : undefined
+}
+
+function packageName(value: string): string | undefined {
+  if (!/^(?:@[a-z0-9][a-z0-9._-]*\/)?[a-z0-9][a-z0-9._-]*$/.test(value)) {
+    return 'Use a lowercase npm package name'
+  }
+  return undefined
+}
+
+function projectDirectory(value: string, cwd: string): string | undefined {
+  const empty = nonEmpty(value)
+  if (empty) return empty
+  return existsSync(resolve(cwd, value)) ? 'Target already exists' : undefined
+}
+
+const API_KEY_STEP: WizardStep<ProjectAnswerState> = {
+  async run(port, state) {
+    let prefilled = state.args.apiKey
+    while (true) {
+      const apiKey = requireAnswer(await new SecretQuestion({
+        id: 'apiKey',
+        message: state.provider === 'custom' ? 'Custom provider API key' : 'DeepSeek API key',
+      }).resolve(port, prefilled))
+      if (apiKey.length > 0) {
+        state.apiKey = apiKey
+        return
+      }
+      const keepEmpty = requireAnswer(await new ConfirmQuestion({
+        id: 'apiKey.empty',
+        message: 'Keep the API key empty and fill .env later?',
+        initialValue: false,
+        tone: 'warning',
+      }).resolve(port))
+      if (keepEmpty) {
+        state.apiKey = ''
+        return
+      }
+      prefilled = undefined
+    }
+  },
+}
+
+const PROJECT_QUESTION_STEPS: readonly WizardStep<ProjectAnswerState>[] = [
+  questionStep({
+    question: state => new TextQuestion({
+      id: 'directory',
+      message: 'Where should the project be created?',
+      placeholder: 'my-agent',
+      defaultValue: 'my-agent',
+      validate: value => projectDirectory(value, state.cwd),
+    }),
+    prefilled: state => state.args.directory,
+    apply: (state, value) => { state.directory = resolve(state.cwd, value) },
+  }),
+  questionStep({
+    question: (state) => {
+      /* v8 ignore next -- the preceding directory step always populates this state */
+      if (!state.directory) throw new Error('directory must resolve before package name')
+      return new TextQuestion({
+        id: 'name',
+        message: 'Package name',
+        placeholder: basename(state.directory),
+        defaultValue: basename(state.directory),
+        validate: packageName,
+      })
+    },
+    apply: (state, value) => { state.name = value },
+  }),
+  questionStep({
+    question: (state) => {
+      /* v8 ignore next -- the preceding package-name step always populates this state */
+      if (!state.name) throw new Error('package name must resolve before description')
+      return new TextQuestion({
+        id: 'description',
+        message: 'Project description',
+        placeholder: `A DeepSeek Harness agent named ${state.name}`,
+        defaultValue: `A DeepSeek Harness agent named ${state.name}`,
+        validate: nonEmpty,
+      })
+    },
+    prefilled: state => state.args.description,
+    apply: (state, value) => { state.description = value },
+  }),
+  questionStep({
+    question: () => new SelectQuestion<'deepseek' | 'custom'>({
+      id: 'provider',
+      message: 'Model provider',
+      options: [
+        { value: 'deepseek', label: 'DeepSeek' },
+        { value: 'custom', label: 'Custom endpoint (pi-ai)' },
+      ],
+      initialValue: 'deepseek',
+    }),
+    prefilled: state => state.args.provider,
+    apply: (state, value) => { state.provider = value },
+  }),
+  questionStep({
+    question: () => new TextQuestion({
+      id: 'baseURL', message: 'Custom provider base URL', validate: nonEmpty,
+    }),
+    when: state => state.provider === 'custom' || state.args.baseURL !== undefined,
+    prefilled: state => state.args.baseURL,
+    apply: (state, value) => { state.baseURL = value },
+  }),
+  API_KEY_STEP,
+  questionStep({
+    question: () => new SelectQuestion<RunInterface>({
+      id: 'interface',
+      message: 'Run interface',
+      options: [
+        { value: 'acp', label: 'ACP server' },
+        { value: 'stdio', label: 'Terminal REPL' },
+        { value: 'embed', label: 'Embedded context' },
+      ],
+      initialValue: 'stdio',
+    }),
+    prefilled: state => state.args.runInterface,
+    apply: (state, value) => { state.runInterface = value },
+  }),
+]
+
+function completeAnswers(state: ProjectAnswerState): ProjectAnswers {
+  const keys = ['directory', 'name', 'description', 'provider', 'baseURL', 'apiKey', 'model', 'runInterface'] as const
+  for (const key of keys) {
+    /* v8 ignore next -- the fixed step list above populates every key or throws/cancels first */
+    if (state[key] === undefined) throw new Error(`create question did not resolve ${key}`)
+  }
+  return state as ProjectAnswerState & ProjectAnswers
+}
+
+/** Run the fixed project-context sequence in declaration order. */
+export async function collectProjectAnswers(
+  port: PromptPort,
+  args: CreateArgs,
+  cwd: string,
+): Promise<ProjectAnswers> {
+  const state: ProjectAnswerState = {
+    args,
+    cwd,
+    baseURL: args.baseURL ?? '',
+    model: args.model ?? 'deepseek-v4-flash',
+  }
+  for (const step of PROJECT_QUESTION_STEPS) await step.run(port, state)
+  return completeAnswers(state)
+}
diff --git a/packages/sdk/create-sdk/src/create-wizard.ts b/packages/sdk/create-sdk/src/create-wizard.ts
new file mode 100644
index 0000000000..8391fae1e4
--- /dev/null
+++ b/packages/sdk/create-sdk/src/create-wizard.ts
@@ -0,0 +1,222 @@
+/**
+ * Declarative create questions with dynamic feature and plugin orchestration.
+ *
+ * @module @deepseek-ai/create-sdk/create-wizard
+ */
+
+import { resolve } from 'node:path'
+import { fileURLToPath } from 'node:url'
+import {
+  FeatureConfigurator,
+  ConfirmQuestion,
+  LocalPluginBlueprint,
+  NpmPackageManager,
+  SelectQuestion,
+  featureId,
+  createBuiltinRegistry,
+  createPackageManager,
+  inferPackageManagerName,
+  probePackageManagerVersion,
+  requireAnswer,
+  type FeatureRegistry,
+  type FeatureSelection,
+  type LocalPluginKind,
+  type PackageManager,
+  type PackageManagerName,
+  type PackageManagerVersionProbe,
+  type ProjectCreationRequest,
+  type ProjectProfile,
+  type PromptPort,
+} from '@deepseek-ai/dsh-helper'
+import type { CreateArgs } from './args.ts'
+import { collectProjectAnswers, type ProjectAnswers } from './create-questions.ts'
+import { CREATE_TEMPLATES, packageManagerTemplateModel } from './templates/create-templates.ts'
+
+/** Fully resolved initializer request and post-create choice. */
+export interface ResolvedCreateRequest {
+  directory: string
+  request: ProjectCreationRequest
+  install: boolean
+}
+
+/** Create-specific orchestration around declarative questions and dynamic selections. */
+export class CreateWizard {
+  private readonly args: CreateArgs
+  private readonly port: PromptPort
+  private readonly cwd: string
+  private readonly releaseVersion: string
+  private readonly versionProbe: PackageManagerVersionProbe
+  private readonly userAgent: string
+  private readonly linkWorkspaceRoot: string | undefined
+
+  /** Bind parsed args and infrastructure to one wizard run. */
+  constructor(options: {
+    args: CreateArgs
+    port: PromptPort
+    cwd?: string
+    releaseVersion: string
+    versionProbe?: PackageManagerVersionProbe
+    userAgent?: string
+  }) {
+    this.args = options.args
+    this.port = options.port
+    this.cwd = resolve(options.cwd ?? process.cwd())
+    this.releaseVersion = options.releaseVersion
+    this.versionProbe = options.versionProbe ?? probePackageManagerVersion
+    /* v8 ignore next -- pnpm supplies npm_config_user_agent while direct invocations may omit it */
+    this.userAgent = options.userAgent ?? process.env.npm_config_user_agent ?? ''
+    this.linkWorkspaceRoot = options.args.linkWorkspace
+      ? fileURLToPath(new URL('../../../../', import.meta.url))
+      : undefined
+  }
+
+  /** Collect all answers before constructing any project files. */
+  async run(): Promise<ResolvedCreateRequest> {
+    const answers = await this.collectProjectAnswers()
+    const profile = this.provisionalProfile(answers)
+    const registry = createBuiltinRegistry(profile)
+    const features = await this.collectFeatures(profile, registry, answers)
+    const localPlugins = await this.collectPlugins()
+    const { manager, install } = await this.collectPackageManager()
+    return {
+      directory: answers.directory,
+      install,
+      request: {
+        name: answers.name,
+        description: answers.description,
+        runtime: { model: answers.model },
+        packageManager: manager,
+        releaseVersion: this.releaseVersion,
+        ...this.linkWorkspaceRoot ? { linkWorkspaceRoot: this.linkWorkspaceRoot } : {},
+        features,
+        localPlugins,
+      },
+    }
+  }
+
+  private async collectProjectAnswers(): Promise<ProjectAnswers> {
+    return collectProjectAnswers(this.port, this.args, this.cwd)
+  }
+
+  private provisionalProfile(answers: ProjectAnswers): ProjectProfile {
+    return {
+      name: answers.name,
+      description: answers.description,
+      runtime: { model: answers.model },
+      runInterface: answers.runInterface,
+      packageManager: new NpmPackageManager('10.0.0'),
+      releaseVersion: this.releaseVersion,
+      ...this.linkWorkspaceRoot ? { linkWorkspaceRoot: this.linkWorkspaceRoot } : {},
+    }
+  }
+
+  private async collectFeatures(
+    profile: ProjectProfile,
+    registry: FeatureRegistry,
+    answers: ProjectAnswers,
+  ): Promise<FeatureSelection[]> {
+    const configurator = new FeatureConfigurator(this.port)
+    const selections: FeatureSelection[] = [
+      {
+        id: featureId('provider'),
+        options: [answers.provider],
+        ...answers.baseURL ? { values: { baseURL: answers.baseURL } } : {},
+        secrets: { apiKey: answers.apiKey },
+      },
+      { id: featureId('spine'), options: ['default'] },
+      { id: featureId('app'), options: [answers.runInterface] },
+    ]
+    const configurable = registry.all().filter(feature => feature.id === 'bash'
+      || feature.id === 'persistence'
+      || (!feature.required && feature.isApplicable(profile)))
+    const selected = [...requireAnswer(await this.port.nestedMultiselect({
+      message: 'Select features',
+      options: configurable.map((feature) => {
+        const nested = feature.mode !== 'single'
+        const defaults = new Set(feature.defaultOptions(profile))
+        return {
+          value: feature.id,
+          label: feature.summary,
+          required: feature.required,
+          default: feature.required || feature.id === 'hmr' || feature.id === 'fs' || feature.id === 'todo'
+            || feature.id === 'skill',
+          ...nested ? {
+            choiceMode: feature.mode === 'multiple' ? 'multiple' as const : 'exclusive' as const,
+            choices: feature.options.map(option => ({
+              value: option.id,
+              label: option.label,
+              default: defaults.has(option.id),
+            })),
+          } : {},
+        }
+      }),
+    }))]
+    for (const { value: id } of [...selected]) {
+      const feature = registry.get(id)
+      for (const suggestedId of feature.suggests) {
+        if (selected.some(item => item.value === suggestedId)) continue
+        const suggested = registry.get(suggestedId)
+        const add = requireAnswer(await new ConfirmQuestion({
+          id: `${feature.id}.${suggested.id}`,
+          message: `Add the recommended ${suggested.summary.toLowerCase()} for ${feature.summary.toLowerCase()}?`,
+          initialValue: true,
+        }).resolve(this.port))
+        if (add) selected.push({ value: suggested.id, choices: suggested.defaultOptions(profile) })
+      }
+    }
+    const fixed = new Set(selections.map(selection => selection.id))
+    const choices = new Map<FeatureSelection['id'], readonly string[] | undefined>()
+    for (const feature of registry.all()) {
+      if (feature.required && feature.isApplicable(profile) && !fixed.has(feature.id)) {
+        choices.set(feature.id, feature.defaultOptions(profile))
+      }
+    }
+    for (const choice of selected) {
+      choices.set(choice.value, choice.choices.length > 0 ? choice.choices : undefined)
+    }
+    for (const [id, options] of choices) {
+      selections.push(await configurator.configure(
+        registry.get(id),
+        profile,
+        undefined,
+        options,
+      ))
+    }
+    return selections
+  }
+
+  private async collectPlugins(): Promise<LocalPluginBlueprint[]> {
+    const kind = requireAnswer(await new SelectQuestion<LocalPluginKind | 'none'>({
+      id: 'plugins.kind',
+      message: 'Local plugin',
+      options: [
+        { value: 'none', label: 'No local plugin' },
+        { value: 'plugin', label: 'Cordis plugin' },
+        { value: 'tool', label: 'Model-facing tool' },
+      ],
+      initialValue: 'none',
+    }).resolve(this.port))
+    return kind === 'none' ? [] : [new LocalPluginBlueprint(kind, kind)]
+  }
+
+  private async collectPackageManager(): Promise<{ manager: PackageManager; install: boolean }> {
+    const inferred = inferPackageManagerName(this.args.packageManager, this.userAgent)
+    const name = requireAnswer(await new SelectQuestion<PackageManagerName>({
+      id: 'packageManager',
+      message: 'Package manager',
+      options: [
+        { value: 'npm', label: 'npm' },
+        { value: 'pnpm', label: 'pnpm' },
+        { value: 'yarn', label: 'Yarn' },
+      ],
+      initialValue: inferred ?? 'npm',
+    }).resolve(this.port, inferred))
+    const manager = createPackageManager(name, await this.versionProbe(name, this.cwd))
+    const install = requireAnswer(await new ConfirmQuestion({
+      id: 'install',
+      message: CREATE_TEMPLATES.installQuestion.render(packageManagerTemplateModel(manager)).trimEnd(),
+      initialValue: true,
+    }).resolve(this.port, this.args.install))
+    return { manager, install }
+  }
+}
diff --git a/packages/sdk/create-sdk/src/index.ts b/packages/sdk/create-sdk/src/index.ts
new file mode 100644
index 0000000000..f931d956f6
--- /dev/null
+++ b/packages/sdk/create-sdk/src/index.ts
@@ -0,0 +1,7 @@
+/**
+ * The create-sdk package is a CLI initializer; its library entry exports no symbols.
+ *
+ * @module @deepseek-ai/create-sdk
+ */
+
+export {}
diff --git a/packages/sdk/create-sdk/src/project-scaffolder.ts b/packages/sdk/create-sdk/src/project-scaffolder.ts
new file mode 100644
index 0000000000..f17f04317d
--- /dev/null
+++ b/packages/sdk/create-sdk/src/project-scaffolder.ts
@@ -0,0 +1,41 @@
+/**
+ * Project creation use case over the shared SDK aggregate and edit session.
+ *
+ * @module @deepseek-ai/create-sdk/project-scaffolder
+ */
+
+import { stat } from 'node:fs/promises'
+import {
+  SdkProject,
+  createBuiltinRegistry,
+  type ChangeSet,
+  type ProjectCreationRequest,
+} from '@deepseek-ai/dsh-helper'
+
+/** Result of writing one new SDK project. */
+export interface ScaffoldResult {
+  project: SdkProject
+  changes: ChangeSet
+}
+
+/** Create a project entirely in memory, then validate and commit it once. */
+export async function scaffoldProject(root: string, request: ProjectCreationRequest): Promise<ScaffoldResult> {
+  let targetExists = true
+  try {
+    await stat(root)
+  } catch (error) {
+    /* v8 ignore else -- the other arm requires a filesystem permission/IO fault from stat */
+    if ((error as NodeJS.ErrnoException).code === 'ENOENT') targetExists = false
+    /* v8 ignore next -- paired with the ignored defensive stat-error arm above */
+    else throw error
+  }
+  if (targetExists) throw new Error(`target already exists: ${root}`)
+  const project = SdkProject.create(root, request)
+  const registry = createBuiltinRegistry(project.profile)
+  const edit = project.edit(registry)
+  for (const selection of request.features) {
+    edit.installFeature(registry.get(selection.id), selection)
+  }
+  for (const plugin of request.localPlugins) edit.addPlugin(plugin)
+  return edit.commit()
+}
diff --git a/packages/sdk/create-sdk/src/templates/assets/created.txt.tpl b/packages/sdk/create-sdk/src/templates/assets/created.txt.tpl
new file mode 100644
index 0000000000..9d4eff4e92
--- /dev/null
+++ b/packages/sdk/create-sdk/src/templates/assets/created.txt.tpl
@@ -0,0 +1 @@
+Created {{name}} in {{directory}}
diff --git a/packages/sdk/create-sdk/src/templates/assets/install-question.txt.tpl b/packages/sdk/create-sdk/src/templates/assets/install-question.txt.tpl
new file mode 100644
index 0000000000..4872f63df0
--- /dev/null
+++ b/packages/sdk/create-sdk/src/templates/assets/install-question.txt.tpl
@@ -0,0 +1 @@
+Run {{packageManager}} {{installArgs}} and then build the project?
diff --git a/packages/sdk/create-sdk/src/templates/assets/next-steps.txt.tpl b/packages/sdk/create-sdk/src/templates/assets/next-steps.txt.tpl
new file mode 100644
index 0000000000..c3f237deb6
--- /dev/null
+++ b/packages/sdk/create-sdk/src/templates/assets/next-steps.txt.tpl
@@ -0,0 +1,5 @@
+{{#if setupRequired}}
+Next: cd {{directory}} && {{packageManager}} {{installArgs}} && {{packageManager}} {{buildArgs}} && {{packageManager}} start
+{{else}}
+Next: cd {{directory}} && {{packageManager}} start
+{{/if}}
diff --git a/packages/sdk/create-sdk/src/templates/assets/setup-failure.txt.tpl b/packages/sdk/create-sdk/src/templates/assets/setup-failure.txt.tpl
new file mode 100644
index 0000000000..7d1e06c820
--- /dev/null
+++ b/packages/sdk/create-sdk/src/templates/assets/setup-failure.txt.tpl
@@ -0,0 +1,2 @@
+Project files are ready, but setup failed: {{error}}
+Retry: cd {{directory}} && {{packageManager}} {{installArgs}} && {{packageManager}} {{buildArgs}}
diff --git a/packages/sdk/create-sdk/src/templates/assets/usage.txt.tpl b/packages/sdk/create-sdk/src/templates/assets/usage.txt.tpl
new file mode 100644
index 0000000000..2b734eb753
--- /dev/null
+++ b/packages/sdk/create-sdk/src/templates/assets/usage.txt.tpl
@@ -0,0 +1,11 @@
+Usage: create-sdk [directory] [options]
+
+Options:
+  --description <text>
+  --provider <deepseek|custom>
+  --base-url <url>
+  --api-key <key>
+  --model <name>
+  --interface <acp|stdio|embed>
+  --pm <npm|pnpm|yarn>
+  --install / --no-install
diff --git a/packages/sdk/create-sdk/src/templates/create-templates.ts b/packages/sdk/create-sdk/src/templates/create-templates.ts
new file mode 100644
index 0000000000..efeaf759c6
--- /dev/null
+++ b/packages/sdk/create-sdk/src/templates/create-templates.ts
@@ -0,0 +1,59 @@
+/**
+ * Package-owned terminal templates for create-sdk.
+ *
+ * @module @deepseek-ai/create-sdk/templates/create-templates
+ */
+
+import {
+  TextTemplate,
+  type PackageManager,
+  type PackageManagerName,
+} from '@deepseek-ai/dsh-helper'
+
+interface CreatedTemplateModel {
+  name: string
+  directory: string
+}
+
+interface NextStepsTemplateModel extends PackageManagerTemplateModel {
+  directory: string
+  setupRequired: boolean
+}
+
+interface SetupFailureTemplateModel extends PackageManagerTemplateModel {
+  directory: string
+  error: string
+}
+
+/** Package-manager execution data consumed by create-sdk templates. */
+export interface PackageManagerTemplateModel {
+  packageManager: PackageManagerName
+  installArgs: string
+  buildArgs: string
+}
+
+/**
+ * Map package-manager execution data into terminal-template fields.
+ * @param manager - selected package-manager strategy.
+ * @returns executable name and operation arguments.
+ */
+export function packageManagerTemplateModel(manager: PackageManager): PackageManagerTemplateModel {
+  return {
+    packageManager: manager.name,
+    installArgs: manager.installCommand().join(' '),
+    buildArgs: manager.buildCommand().join(' '),
+  }
+}
+
+/** Compiled create-sdk terminal templates. */
+export const CREATE_TEMPLATES = {
+  usage: TextTemplate.fromFile<Record<string, never>>(new URL('./assets/usage.txt.tpl', import.meta.url)),
+  created: TextTemplate.fromFile<CreatedTemplateModel>(new URL('./assets/created.txt.tpl', import.meta.url)),
+  nextSteps: TextTemplate.fromFile<NextStepsTemplateModel>(new URL('./assets/next-steps.txt.tpl', import.meta.url)),
+  setupFailure: TextTemplate.fromFile<SetupFailureTemplateModel>(
+    new URL('./assets/setup-failure.txt.tpl', import.meta.url),
+  ),
+  installQuestion: TextTemplate.fromFile<PackageManagerTemplateModel>(
+    new URL('./assets/install-question.txt.tpl', import.meta.url),
+  ),
+} as const
diff --git a/packages/sdk/create-sdk/tests/built-artifacts.e2e.ts b/packages/sdk/create-sdk/tests/built-artifacts.e2e.ts
new file mode 100644
index 0000000000..4b2de078c7
--- /dev/null
+++ b/packages/sdk/create-sdk/tests/built-artifacts.e2e.ts
@@ -0,0 +1,28 @@
+import { execFile } from 'node:child_process'
+import { existsSync } from 'node:fs'
+import { join } from 'node:path'
+import { fileURLToPath } from 'node:url'
+import { promisify } from 'node:util'
+import { describe, expect, it } from 'vitest'
+
+const execFileAsync = promisify(execFile)
+const repoRoot = fileURLToPath(new URL('../../../../', import.meta.url))
+const createBin = join(repoRoot, 'packages/sdk/create-sdk/lib/bin.js')
+const scriptsBin = join(repoRoot, 'packages/sdk/scripts/lib/bin.js')
+
+describe.skipIf(!existsSync(createBin) || !existsSync(scriptsBin))(
+  'SDK built artifacts',
+  () => {
+    it('runs the published dsh-sdk bin help path under plain Node', async () => {
+      const result = await execFileAsync(process.execPath, [scriptsBin, '--help'], { encoding: 'utf8' })
+      expect(result.stdout).toContain('Usage: dsh-sdk <command>')
+      expect(result.stderr).toBe('')
+    })
+
+    it('runs the published create-sdk bin help path under plain Node', async () => {
+      const result = await execFileAsync(process.execPath, [createBin, '--help'], { encoding: 'utf8' })
+      expect(result.stdout).toContain('Usage: create-sdk [directory]')
+      expect(result.stderr).toBe('')
+    })
+  },
+)
diff --git a/packages/sdk/create-sdk/tests/create.snapshot.ts b/packages/sdk/create-sdk/tests/create.snapshot.ts
new file mode 100644
index 0000000000..a5ea46db53
--- /dev/null
+++ b/packages/sdk/create-sdk/tests/create.snapshot.ts
@@ -0,0 +1,398 @@
+import { describe, expect, it } from 'vitest'
+import {
+  featureId,
+  createPackageManager,
+  type NestedMultiSelectValue,
+  type PromptPort,
+} from '@deepseek-ai/dsh-helper'
+import type {
+  ConfirmPromptRequest,
+  MultiSelectPromptRequest,
+  NestedMultiSelectRequest,
+  PromptOutcome,
+  SecretPromptRequest,
+  SelectPromptRequest,
+  TextPromptRequest,
+} from '../../helper/src/questions/prompt-port.ts'
+import { parseCreateArgs } from '../src/args.ts'
+import { CreateWizard } from '../src/create-wizard.ts'
+import { CREATE_TEMPLATES, packageManagerTemplateModel } from '../src/templates/create-templates.ts'
+
+class RecordingPort implements PromptPort {
+  readonly transcript: unknown[] = []
+  readonly #answers: unknown[]
+
+  constructor(answers: unknown[]) { this.#answers = [...answers] }
+
+  answer<T>(record: unknown): Promise<PromptOutcome<T>> {
+    this.transcript.push(record)
+    return Promise.resolve({ status: 'answered', value: this.#answers.shift() as T })
+  }
+
+  text(request: TextPromptRequest): Promise<PromptOutcome<string>> {
+    return this.answer({
+      kind: 'text',
+      message: request.message,
+      defaultValue: request.defaultValue,
+      initialValue: request.initialValue,
+    })
+  }
+  secret(request: SecretPromptRequest): Promise<PromptOutcome<string>> {
+    return this.answer({ kind: 'secret', message: request.message })
+  }
+  select<T>(request: SelectPromptRequest<T>): Promise<PromptOutcome<T>> {
+    return this.answer({
+      kind: 'select', message: request.message, options: request.options.map(option => option.label),
+      initialValue: request.initialValue,
+    })
+  }
+  multiselect<T>(request: MultiSelectPromptRequest<T>): Promise<PromptOutcome<readonly T[]>> {
+    return this.answer({
+      kind: 'multiselect', message: request.message, options: request.options.map(option => option.label),
+      initialValues: request.initialValues,
+    })
+  }
+  confirm(request: ConfirmPromptRequest): Promise<PromptOutcome<boolean>> {
+    return this.answer({ kind: 'confirm', message: request.message, initialValue: request.initialValue })
+  }
+  nestedMultiselect<TValue, TChoice>(
+    request: NestedMultiSelectRequest<TValue, TChoice>,
+  ): Promise<PromptOutcome<readonly NestedMultiSelectValue<TValue, TChoice>[]>> {
+    return this.answer({
+      kind: 'nested-multiselect',
+      message: request.message,
+      options: request.options.map(option => ({
+        label: option.label,
+        required: option.required,
+        default: option.default,
+        choices: option.choices?.map(choice => choice.label),
+      })),
+    })
+  }
+}
+
+describe('create-sdk terminal contract', () => {
+  it('renders package-manager-specific setup commands', () => {
+    const model = packageManagerTemplateModel(createPackageManager('yarn', '4.0.0'))
+    expect(CREATE_TEMPLATES.installQuestion.render(model)).toBe('Run yarn install and then build the project?\n')
+    expect(CREATE_TEMPLATES.setupFailure.render({
+      directory: '/workspace/agent',
+      error: 'offline',
+      ...model,
+    })).toContain('yarn install && yarn build')
+  })
+
+  it('pins the full unresolved question order and completion messages', async () => {
+    const port = new RecordingPort([
+      'my-agent',
+      'my-agent',
+      'Snapshot agent',
+      'deepseek',
+      'secret-key',
+      'acp',
+      [
+        { value: featureId('persistence'), choices: ['jsonl'] },
+        { value: featureId('hmr'), choices: [] },
+        { value: featureId('web'), choices: ['exa'] },
+        { value: featureId('workflow'), choices: [] },
+      ],
+      true,
+      'exa-key',
+      'none',
+      'npm',
+      false,
+    ])
+    const resolved = await new CreateWizard({
+      args: parseCreateArgs([]),
+      port,
+      cwd: '/workspace',
+      releaseVersion: '0.0.1',
+      userAgent: '',
+      versionProbe: async () => '10.0.0',
+    }).run()
+    expect({
+      prompts: port.transcript,
+      result: {
+        directory: resolved.directory,
+        name: resolved.request.name,
+        manager: resolved.request.packageManager.name,
+        install: resolved.install,
+        features: resolved.request.features.map(item => ({ id: item.id, options: item.options })),
+      },
+      messages: {
+        created: CREATE_TEMPLATES.created.render({
+          name: resolved.request.name,
+          directory: resolved.directory,
+        }),
+        next: CREATE_TEMPLATES.nextSteps.render({
+          directory: resolved.directory,
+          setupRequired: false,
+          ...packageManagerTemplateModel(resolved.request.packageManager),
+        }),
+        failure: CREATE_TEMPLATES.setupFailure.render({
+          directory: resolved.directory,
+          error: String(new Error('offline')),
+          ...packageManagerTemplateModel(resolved.request.packageManager),
+        }),
+      },
+    }).toMatchInlineSnapshot(`
+      {
+        "messages": {
+          "created": "Created my-agent in /workspace/my-agent
+      ",
+          "failure": "Project files are ready, but setup failed: Error: offline
+      Retry: cd /workspace/my-agent && npm install && npm run build
+      ",
+          "next": "Next: cd /workspace/my-agent && npm start
+      ",
+        },
+        "prompts": [
+          {
+            "defaultValue": "my-agent",
+            "initialValue": undefined,
+            "kind": "text",
+            "message": "Where should the project be created?",
+          },
+          {
+            "defaultValue": "my-agent",
+            "initialValue": undefined,
+            "kind": "text",
+            "message": "Package name",
+          },
+          {
+            "defaultValue": "A DeepSeek Harness agent named my-agent",
+            "initialValue": undefined,
+            "kind": "text",
+            "message": "Project description",
+          },
+          {
+            "initialValue": "deepseek",
+            "kind": "select",
+            "message": "Model provider",
+            "options": [
+              "DeepSeek",
+              "Custom endpoint (pi-ai)",
+            ],
+          },
+          {
+            "kind": "secret",
+            "message": "DeepSeek API key",
+          },
+          {
+            "initialValue": "stdio",
+            "kind": "select",
+            "message": "Run interface",
+            "options": [
+              "ACP server",
+              "Terminal REPL",
+              "Embedded context",
+            ],
+          },
+          {
+            "kind": "nested-multiselect",
+            "message": "Select features",
+            "options": [
+              {
+                "choices": [
+                  "Local executor",
+                  "Sandboxed executor",
+                ],
+                "default": true,
+                "label": "Command execution",
+                "required": true,
+              },
+              {
+                "choices": [
+                  "JSONL files",
+                  "SQLite database",
+                ],
+                "default": true,
+                "label": "Durable session storage",
+                "required": true,
+              },
+              {
+                "choices": undefined,
+                "default": true,
+                "label": "Hot-module reload",
+                "required": false,
+              },
+              {
+                "choices": undefined,
+                "default": true,
+                "label": "Read, write, and edit local files",
+                "required": false,
+              },
+              {
+                "choices": undefined,
+                "default": true,
+                "label": "Model-facing task tracking",
+                "required": false,
+              },
+              {
+                "choices": undefined,
+                "default": true,
+                "label": "Local skill discovery",
+                "required": false,
+              },
+              {
+                "choices": [
+                  "DeepSeek search",
+                  "Exa search",
+                  "Perplexity search",
+                  "Fetch only",
+                ],
+                "default": false,
+                "label": "Web search and fetch tools",
+                "required": false,
+              },
+              {
+                "choices": [
+                  "Fresh child agent",
+                  "Fork parent history",
+                ],
+                "default": false,
+                "label": "Delegate work to child agents",
+                "required": false,
+              },
+              {
+                "choices": undefined,
+                "default": false,
+                "label": "Scripted multi-agent workflows",
+                "required": false,
+              },
+              {
+                "choices": undefined,
+                "default": false,
+                "label": "Automatic context compaction",
+                "required": false,
+              },
+              {
+                "choices": [
+                  "Claude Code hooks",
+                  "Codex hooks",
+                ],
+                "default": false,
+                "label": "Run Claude Code or Codex hooks",
+                "required": false,
+              },
+              {
+                "choices": undefined,
+                "default": false,
+                "label": "Loop-hygiene reminders",
+                "required": false,
+              },
+              {
+                "choices": undefined,
+                "default": false,
+                "label": "Tool timeout policy",
+                "required": false,
+              },
+              {
+                "choices": undefined,
+                "default": false,
+                "label": "Ask the user from the model loop",
+                "required": false,
+              },
+            ],
+          },
+          {
+            "initialValue": true,
+            "kind": "confirm",
+            "message": "Add the recommended tool timeout policy for web search and fetch tools?",
+          },
+          {
+            "kind": "secret",
+            "message": "Exa API key",
+          },
+          {
+            "initialValue": "none",
+            "kind": "select",
+            "message": "Local plugin",
+            "options": [
+              "No local plugin",
+              "Cordis plugin",
+              "Model-facing tool",
+            ],
+          },
+          {
+            "initialValue": "npm",
+            "kind": "select",
+            "message": "Package manager",
+            "options": [
+              "npm",
+              "pnpm",
+              "Yarn",
+            ],
+          },
+          {
+            "initialValue": true,
+            "kind": "confirm",
+            "message": "Run npm install and then build the project?",
+          },
+        ],
+        "result": {
+          "directory": "/workspace/my-agent",
+          "features": [
+            {
+              "id": "provider",
+              "options": [
+                "deepseek",
+              ],
+            },
+            {
+              "id": "spine",
+              "options": [
+                "default",
+              ],
+            },
+            {
+              "id": "app",
+              "options": [
+                "acp",
+              ],
+            },
+            {
+              "id": "bash",
+              "options": [
+                "local",
+              ],
+            },
+            {
+              "id": "persistence",
+              "options": [
+                "jsonl",
+              ],
+            },
+            {
+              "id": "hmr",
+              "options": [
+                "default",
+              ],
+            },
+            {
+              "id": "web",
+              "options": [
+                "exa",
+              ],
+            },
+            {
+              "id": "workflow",
+              "options": [
+                "workerthread",
+              ],
+            },
+            {
+              "id": "timeout-policy",
+              "options": [
+                "default",
+              ],
+            },
+          ],
+          "install": false,
+          "manager": "npm",
+          "name": "my-agent",
+        },
+      }
+    `)
+  })
+})
diff --git a/packages/sdk/create-sdk/tests/create.spec.ts b/packages/sdk/create-sdk/tests/create.spec.ts
new file mode 100644
index 0000000000..9d75700bb9
--- /dev/null
+++ b/packages/sdk/create-sdk/tests/create.spec.ts
@@ -0,0 +1,502 @@
+import { mkdir, mkdtemp, readFile, rm, writeFile } from 'node:fs/promises'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { PassThrough, Writable } from 'node:stream'
+import { fileURLToPath } from 'node:url'
+import { afterEach, describe, expect, it, vi } from 'vitest'
+import {
+  LocalPluginBlueprint,
+  featureId,
+  NpmPackageManager,
+  type NestedMultiSelectValue,
+  type PromptPort,
+} from '@deepseek-ai/dsh-helper'
+import type {
+  ConfirmPromptRequest,
+  MultiSelectPromptRequest,
+  NestedMultiSelectRequest,
+  PromptOutcome,
+  SecretPromptRequest,
+  SelectPromptRequest,
+  TextPromptRequest,
+} from '../../helper/src/questions/prompt-port.ts'
+import { parseCreateArgs } from '../src/args.ts'
+import {
+  createProject,
+  readCreateSdkVersion,
+  runCreateCommand,
+  type CreateCommandContext,
+} from '../src/command.ts'
+import { CreateWizard } from '../src/create-wizard.ts'
+import { scaffoldProject } from '../src/project-scaffolder.ts'
+
+class ScriptedPort implements PromptPort {
+  readonly requests: string[] = []
+  readonly #answers: unknown[]
+
+  constructor(answers: unknown[]) {
+    this.#answers = [...answers]
+  }
+
+  answer<T>(message: string): Promise<PromptOutcome<T>> {
+    this.requests.push(message)
+    const value = this.#answers.shift()
+    return Promise.resolve(value === ScriptedPort.cancel
+      ? { status: 'cancelled' }
+      : { status: 'answered', value: value as T })
+  }
+
+  async text(request: TextPromptRequest): Promise<PromptOutcome<string>> {
+    const outcome = await this.answer<string>(request.message)
+    if (outcome.status === 'cancelled') return outcome
+    const value = outcome.value || request.defaultValue || ''
+    const diagnostic = request.validate?.(value)
+    if (diagnostic) throw new Error(diagnostic)
+    return { status: 'answered', value }
+  }
+  secret(request: SecretPromptRequest): Promise<PromptOutcome<string>> { return this.answer(request.message) }
+  select<T>(request: SelectPromptRequest<T>): Promise<PromptOutcome<T>> { return this.answer(request.message) }
+  multiselect<T>(request: MultiSelectPromptRequest<T>): Promise<PromptOutcome<readonly T[]>> {
+    return this.answer(request.message)
+  }
+  confirm(request: ConfirmPromptRequest): Promise<PromptOutcome<boolean>> { return this.answer(request.message) }
+  nestedMultiselect<TValue, TChoice>(
+    request: NestedMultiSelectRequest<TValue, TChoice>,
+  ): Promise<PromptOutcome<readonly NestedMultiSelectValue<TValue, TChoice>[]>> {
+    return this.answer(request.message)
+  }
+
+  static readonly cancel = Symbol('cancel')
+}
+
+const temporary: string[] = []
+const repoRoot = fileURLToPath(new URL('../../../../', import.meta.url))
+
+interface GeneratedPackageManifest {
+  scripts?: Record<string, string>
+  dependencies?: Record<string, string>
+  devDependencies?: Record<string, string>
+}
+
+interface GeneratedTsConfig {
+  compilerOptions: {
+    types?: readonly string[]
+  }
+}
+
+function parseGeneratedPackageManifest(text: string): GeneratedPackageManifest {
+  return JSON.parse(text) as GeneratedPackageManifest
+}
+
+function parseGeneratedTsConfig(text: string): GeneratedTsConfig {
+  return JSON.parse(text) as GeneratedTsConfig
+}
+
+function commandContext(
+  cwd: string,
+  port?: PromptPort,
+  setup?: CreateCommandContext['setup'],
+): CreateCommandContext & { readStdout: () => string; readStderr: () => string } {
+  let stdout = ''
+  let stderr = ''
+  const input = Object.assign(new PassThrough(), { isTTY: true }) as unknown as NodeJS.ReadStream
+  const output = Object.assign(new Writable({
+    write(chunk, _encoding, callback) { stdout += String(chunk); callback() },
+  }), { isTTY: true }) as unknown as NodeJS.WriteStream
+  const error = new Writable({
+    write(chunk, _encoding, callback) { stderr += String(chunk); callback() },
+  }) as unknown as NodeJS.WriteStream
+  return {
+    cwd,
+    stdin: input,
+    stdout: output,
+    stderr: error,
+    releaseVersion: '0.0.1',
+    versionProbe: async () => '10.0.0',
+    ...port ? { port } : {},
+    ...setup ? { setup } : {},
+    readStdout: () => stdout,
+    readStderr: () => stderr,
+  }
+}
+
+afterEach(async () => {
+  await Promise.all(temporary.splice(0).map(path => rm(path, { recursive: true, force: true })))
+})
+
+describe('create arguments', () => {
+  it('parses public options and the private repository link mode', () => {
+    expect(parseCreateArgs([
+      'agent', '--description=demo', '--provider', 'deepseek', '--base-url=https://api.example',
+      '--api-key', 'key', '--model=m', '--interface', 'acp', '--pm=pnpm', '--no-install',
+      '--link-workspace',
+    ])).toEqual({
+      directory: 'agent',
+      description: 'demo',
+      provider: 'deepseek',
+      baseURL: 'https://api.example',
+      apiKey: 'key',
+      model: 'm',
+      runInterface: 'acp',
+      packageManager: 'pnpm',
+      install: false,
+      linkWorkspace: true,
+      help: false,
+    })
+    expect(parseCreateArgs(['--link-workspace']).linkWorkspace).toBe(true)
+    expect(() => parseCreateArgs(['--link-packages-workspace'])).toThrow("unknown option '--link-packages-workspace'")
+    expect(parseCreateArgs(['--provider=custom']).provider).toBe('custom')
+    expect(parseCreateArgs(['--help']).help).toBe(true)
+    expect(() => parseCreateArgs(['--interface=bad'])).toThrow('Allowed choices are acp, stdio, embed')
+    expect(() => parseCreateArgs(['--unknown'])).toThrow("unknown option '--unknown'")
+    expect(() => parseCreateArgs(['one', 'two'])).toThrow('too many arguments')
+  })
+
+  it('validates empty directories and package names', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'create-validation-'))
+    temporary.push(root)
+    await expect(new CreateWizard({
+      args: parseCreateArgs(['']), port: new ScriptedPort([]), cwd: root,
+      releaseVersion: '0.0.1', versionProbe: async () => '10.0.0',
+    }).run()).rejects.toThrow('A value is required')
+    await expect(new CreateWizard({
+      args: parseCreateArgs(['agent']), port: new ScriptedPort(['Invalid Name']), cwd: root,
+      releaseVersion: '0.0.1', versionProbe: async () => '10.0.0',
+    }).run()).rejects.toThrow('lowercase npm package name')
+  })
+
+  it('rejects an existing target before asking project questions', async () => {
+    const cwd = await mkdtemp(join(tmpdir(), 'create-existing-target-'))
+    temporary.push(cwd)
+    await mkdir(join(cwd, 'taken'))
+    const port = new ScriptedPort([])
+    const wizard = new CreateWizard({
+      args: parseCreateArgs(['taken']),
+      port,
+      cwd,
+      releaseVersion: '0.0.1',
+      versionProbe: async () => '10.0.0',
+    })
+    await expect(wizard.run()).rejects.toThrow('directory: Target already exists')
+    expect(port.requests).toEqual([])
+  })
+})
+
+describe('CreateWizard and scaffolder', () => {
+  it('asks only unresolved questions in requirement-safe order', async () => {
+    const cwd = await mkdtemp(join(tmpdir(), 'create-wizard-'))
+    temporary.push(cwd)
+    const port = new ScriptedPort([
+      'my-agent',
+      [
+        { value: featureId('persistence'), choices: ['sqlite'] },
+        { value: featureId('hmr'), choices: [] },
+        { value: featureId('fs'), choices: [] },
+        { value: featureId('web'), choices: ['exa'] },
+      ],
+      false,
+      'exa-key',
+      'tool',
+    ])
+    const args = parseCreateArgs([
+      'my-agent',
+      '--description=demo',
+      '--provider=deepseek',
+      '--api-key=deepseek-key',
+      '--model=deepseek-v4-flash',
+      '--interface=stdio',
+      '--pm=npm',
+      '--no-install',
+      '--link-workspace',
+    ])
+    const resolved = await new CreateWizard({
+      args,
+      port,
+      cwd,
+      releaseVersion: '0.0.1',
+      versionProbe: async () => '10.0.0',
+    }).run()
+    expect(port.requests).toEqual([
+      'Package name',
+      'Select features',
+      'Add the recommended tool timeout policy for web search and fetch tools?',
+      'Exa API key',
+      'Local plugin',
+    ])
+    expect(resolved.install).toBe(false)
+    expect(resolved.request.packageManager.name).toBe('npm')
+    expect(resolved.request.linkWorkspaceRoot).toBe(repoRoot)
+    expect(resolved.request.localPlugins[0]).toMatchObject({ name: 'tool', kind: 'tool' })
+    expect(resolved.request.features.find(item => item.id === 'web')).toMatchObject({
+      options: ['exa'], secrets: { apiKey: 'exa-key' },
+    })
+    expect(resolved.request.features.find(item => item.id === 'hmr')).toMatchObject({ options: ['default'] })
+  })
+
+  it('writes the project once and refuses every existing target', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'create-scaffold-'))
+    temporary.push(root)
+    const request = {
+      name: 'agent',
+      description: 'demo',
+      runtime: { model: 'deepseek-v4-flash' },
+      packageManager: new NpmPackageManager('10.0.0'),
+      releaseVersion: '0.0.1',
+      features: [
+        { id: featureId('provider'), options: ['deepseek'], secrets: { apiKey: 'key' } },
+        { id: featureId('bash'), options: ['local'] },
+        { id: featureId('app'), options: ['embed'] },
+        { id: featureId('persistence'), options: ['jsonl'] },
+      ],
+      localPlugins: [new LocalPluginBlueprint('plugin', 'plugin')],
+    }
+    const target = join(root, 'project')
+    const result = await scaffoldProject(target, request)
+    expect(result.changes.changedFiles).toContain('README.md')
+    const index = await readFile(join(target, 'index.ts'), 'utf8')
+    expect(index).toContain('SdkBootContext')
+    expect(index).toContain('ctx.agents.create')
+    expect(index).toContain('agentOptions: { model: "deepseek-v4-flash" }')
+    const tsconfig = parseGeneratedTsConfig(await readFile(join(target, 'tsconfig.base.json'), 'utf8'))
+    const manifest = parseGeneratedPackageManifest(await readFile(join(target, 'package.json'), 'utf8'))
+    expect(tsconfig.compilerOptions.types).toEqual(['node'])
+    expect(manifest.scripts).toEqual({
+      dev: 'dsh-sdk dev index.ts',
+      build: 'dsh-sdk build',
+      typecheck: 'tsc -b',
+      start: 'dsh-sdk start index.js',
+      config: 'dsh-sdk config',
+    })
+    expect(manifest.dependencies).not.toHaveProperty('node-addon-require-builtin')
+    expect(manifest.devDependencies?.['@types/node']).toBe('^22.20.0')
+    expect(await readFile(join(target, 'plugins/plugin/src/index.ts'), 'utf8')).toContain('export function apply')
+    const cordis = await readFile(join(target, 'cordis.yml'), 'utf8')
+    expect(cordis).toMatch(/^- id:/)
+    expect(cordis).not.toMatch(/^\[/)
+    const occupied = join(root, 'occupied')
+    await mkdir(occupied)
+    await expect(scaffoldProject(occupied, request)).rejects.toThrow('already exists')
+    await writeFile(join(occupied, 'keep'), 'x')
+    await expect(scaffoldProject(occupied, request)).rejects.toThrow('already exists')
+  })
+
+  it('installs workflow requirements before validating the next feature', async () => {
+    const cwd = await mkdtemp(join(tmpdir(), 'create-workflow-requires-'))
+    temporary.push(cwd)
+    const port = new ScriptedPort([
+      'workflow-agent',
+      [
+        { value: featureId('persistence'), choices: ['jsonl'] },
+        { value: featureId('workflow'), choices: [] },
+      ],
+      'none',
+    ])
+    const resolved = await new CreateWizard({
+      args: parseCreateArgs([
+        'workflow-agent', '--description=test', '--provider=deepseek', '--api-key=key',
+        '--interface=embed', '--pm=npm', '--no-install',
+      ]),
+      port,
+      cwd,
+      releaseVersion: '0.0.1',
+      versionProbe: async () => '10.0.0',
+    }).run()
+    const result = await scaffoldProject(resolved.directory, resolved.request)
+    expect(result.project.cordis.entry('subagent-spawn')).toBeDefined()
+    expect(result.project.cordis.entry('tool-subagent')).toBeDefined()
+  })
+
+  it('confirms an empty provider key and leaves a documented .env placeholder', async () => {
+    const cwd = await mkdtemp(join(tmpdir(), 'create-empty-key-'))
+    temporary.push(cwd)
+    const port = new ScriptedPort([
+      'empty-key-agent',
+      '',
+      true,
+      [{ value: featureId('persistence'), choices: ['jsonl'] }],
+      'none',
+    ])
+    const resolved = await new CreateWizard({
+      args: parseCreateArgs([
+        'empty-key-agent', '--description=test', '--provider=deepseek',
+        '--interface=embed', '--pm=npm', '--no-install',
+      ]),
+      port,
+      cwd,
+      releaseVersion: '0.0.1',
+      versionProbe: async () => '10.0.0',
+    }).run()
+    await scaffoldProject(resolved.directory, resolved.request)
+    expect(await readFile(join(resolved.directory, '.env'), 'utf8')).toBe(
+      '# Required before start; an empty value makes provider startup fail.\nDEEPSEEK_API_KEY=\n',
+    )
+    expect(port.requests).toContain('Keep the API key empty and fill .env later?')
+  })
+
+  it('collects custom provider inputs, retries an empty key, and accepts a recommendation', async () => {
+    const cwd = await mkdtemp(join(tmpdir(), 'create-custom-inputs-'))
+    temporary.push(cwd)
+    const port = new ScriptedPort([
+      'custom-agent',
+      'test custom provider',
+      'custom',
+      'https://provider.example/v1',
+      '', false, 'custom-key',
+      'embed',
+      [
+        { value: featureId('persistence'), choices: ['jsonl'] },
+        { value: featureId('web'), choices: ['deepseek'] },
+      ],
+      true,
+      'none',
+      'npm',
+      false,
+    ])
+    const resolved = await new CreateWizard({
+      args: parseCreateArgs(['custom-agent']),
+      port,
+      cwd,
+      releaseVersion: '0.0.1',
+      versionProbe: async () => '10.0.0',
+      userAgent: '',
+    }).run()
+    expect(resolved.request.features.find(item => item.id === 'provider')).toMatchObject({
+      options: ['custom'], values: { baseURL: 'https://provider.example/v1' }, secrets: { apiKey: 'custom-key' },
+    })
+    expect(resolved.request.features.some(item => item.id === 'timeout-policy')).toBe(true)
+  })
+
+  it('does not re-suggest an already selected feature', async () => {
+    const cwd = await mkdtemp(join(tmpdir(), 'create-selected-suggestion-'))
+    temporary.push(cwd)
+    const port = new ScriptedPort([
+      'agent',
+      [
+        { value: featureId('persistence'), choices: ['jsonl'] },
+        { value: featureId('web'), choices: ['deepseek'] },
+        { value: featureId('timeout-policy'), choices: ['default'] },
+      ],
+      'none',
+    ])
+    const resolved = await new CreateWizard({
+      args: parseCreateArgs([
+        'agent', '--description=test', '--provider=deepseek', '--api-key=key',
+        '--interface=embed', '--pm=npm', '--no-install',
+      ]),
+      port,
+      cwd,
+      releaseVersion: '0.0.1',
+      versionProbe: async () => '10.0.0',
+    }).run()
+    expect(resolved.request.features.filter(item => item.id === 'timeout-policy')).toHaveLength(1)
+  })
+
+  it('uses process defaults when constructor infrastructure is omitted', async () => {
+    const name = `default-infra-${String(process.pid)}`
+    const port = new ScriptedPort([
+      name, [{ value: featureId('persistence'), choices: ['jsonl'] }], 'none',
+    ])
+    const resolved = await new CreateWizard({
+      args: parseCreateArgs([
+        name, '--description=test', '--provider=deepseek', '--api-key=key',
+        '--interface=embed', '--pm=npm', '--no-install',
+      ]),
+      port,
+      releaseVersion: '0.0.1',
+    }).run()
+    expect(resolved.request.packageManager.name).toBe('npm')
+  })
+
+  it('reads the release batch from the initializer package', async () => {
+    await expect(readCreateSdkVersion()).resolves.toBe('0.0.1')
+  })
+})
+
+describe('create command composition', () => {
+  const argv = (directory: string, install: boolean): string[] => [
+    directory, '--description=test', '--provider=deepseek', '--api-key=key',
+    '--interface=embed', '--pm=npm', install ? '--install' : '--no-install',
+  ]
+
+  it('prints help before requiring a TTY and rejects non-interactive creation', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'create-command-help-'))
+    temporary.push(root)
+    const context = commandContext(root)
+    context.stdin.isTTY = false
+    context.stdout.isTTY = false
+    await expect(createProject(['--help'], context)).resolves.toBeUndefined()
+    expect(context.readStdout()).toContain('Usage: create-sdk')
+    expect(context.readStdout()).not.toContain('--link-workspace')
+    await expect(createProject(argv('agent', false), context)).rejects.toThrow('interactive TTY')
+    context.stdin.isTTY = true
+    await expect(createProject(argv('agent', false), context)).rejects.toThrow('interactive TTY')
+  })
+
+  it('creates through an injected prompt port and delegates optional setup', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'create-command-success-'))
+    temporary.push(root)
+    const port = new ScriptedPort([
+      'agent', [{ value: featureId('persistence'), choices: ['jsonl'] }], 'none',
+    ])
+    let setupDirectory = ''
+    const context = commandContext(root, port, async (request) => { setupDirectory = request.directory })
+    const result = await createProject(argv('agent', true), context)
+    expect(result?.project.root).toBe(join(root, 'agent'))
+    expect(setupDirectory).toBe(join(root, 'agent'))
+    expect(context.readStdout()).toContain('Created agent')
+    expect(context.readStdout()).toContain('Next: cd')
+    const noInstall = commandContext(root, new ScriptedPort([
+      'next', [{ value: featureId('persistence'), choices: ['jsonl'] }], 'none',
+    ]))
+    await expect(createProject(argv('next', false), noInstall)).resolves.toBeDefined()
+    expect(noInstall.readStdout()).toContain('npm install && npm run build && npm start')
+  })
+
+  it('uses the package manager setup path when no setup override is supplied', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'create-command-default-setup-'))
+    temporary.push(root)
+    const port = new ScriptedPort([
+      'agent', [{ value: featureId('persistence'), choices: ['jsonl'] }], 'none',
+    ])
+    const install = vi.spyOn(NpmPackageManager.prototype, 'install').mockResolvedValue()
+    const build = vi.spyOn(NpmPackageManager.prototype, 'build').mockResolvedValue()
+    const context = commandContext(root, port)
+    delete context.releaseVersion
+    delete context.versionProbe
+    await createProject(argv('agent', true), context)
+    expect(install).toHaveBeenCalledOnce()
+    expect(build).toHaveBeenCalledOnce()
+    install.mockRestore()
+    build.mockRestore()
+  })
+
+  it('reports setup failures after preserving generated files', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'create-command-failure-'))
+    temporary.push(root)
+    const port = new ScriptedPort([
+      'agent', [{ value: featureId('persistence'), choices: ['jsonl'] }], 'none',
+    ])
+    const context = commandContext(root, port, async () => { throw new Error('offline') })
+    await expect(createProject(argv('agent', true), context)).rejects.toThrow('offline')
+    expect(context.readStderr()).toContain('Project files are ready, but setup failed')
+    expect(context.readStderr()).toContain('npm install && npm run build')
+    const stringFailure = commandContext(root, new ScriptedPort([
+      'next', [{ value: featureId('persistence'), choices: ['jsonl'] }], 'none',
+    ]), async () => { throw 'offline-string' })
+    await expect(runCreateCommand(argv('next', true), stringFailure)).resolves.toBe(1)
+    expect(stringFailure.readStderr()).toContain('offline-string')
+  })
+
+  it('maps cancellation and ordinary errors to command exit codes', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'create-command-exit-'))
+    temporary.push(root)
+    const cancelled = commandContext(root, new ScriptedPort([ScriptedPort.cancel]))
+    await expect(runCreateCommand([], cancelled)).resolves.toBe(1)
+    expect(cancelled.readStderr()).toContain('cancelled')
+    const invalid = commandContext(root)
+    await expect(runCreateCommand(['--unknown'], invalid)).resolves.toBe(1)
+    expect(invalid.readStderr()).toContain('unknown option')
+    const help = commandContext(root)
+    await expect(runCreateCommand(['--help'], help)).resolves.toBe(0)
+  })
+})
diff --git a/packages/sdk/create-sdk/tests/link-workspace.e2e.ts b/packages/sdk/create-sdk/tests/link-workspace.e2e.ts
new file mode 100644
index 0000000000..6e0b404dc6
--- /dev/null
+++ b/packages/sdk/create-sdk/tests/link-workspace.e2e.ts
@@ -0,0 +1,112 @@
+import { execFile } from 'node:child_process'
+import { existsSync } from 'node:fs'
+import { mkdtemp, readFile, rm, writeFile } from 'node:fs/promises'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { fileURLToPath } from 'node:url'
+import { promisify } from 'node:util'
+import { afterEach, describe, expect, it } from 'vitest'
+import {
+  LocalPluginBlueprint,
+  featureId,
+  createPackageManager,
+  type PackageManagerName,
+} from '@deepseek-ai/dsh-helper'
+import { scrubEnvironment } from '../../helper/src/package-managers/package-manager.ts'
+import { scaffoldProject } from '../src/project-scaffolder.ts'
+
+const execFileAsync = promisify(execFile)
+const repoRoot = fileURLToPath(new URL('../../../../', import.meta.url))
+const builtScripts = join(repoRoot, 'packages/sdk/scripts/lib/bin.js')
+const temporary: string[] = []
+
+afterEach(async () => {
+  await Promise.all(temporary.splice(0).map(path => rm(path, { recursive: true, force: true })))
+})
+
+async function managerVersion(name: PackageManagerName): Promise<string | undefined> {
+  try {
+    return (await execFileAsync(name, ['--version'], { encoding: 'utf8' })).stdout.trim()
+  } catch {
+    // An unavailable optional manager skips only its own live-link case.
+    return undefined
+  }
+}
+
+const managers: PackageManagerName[] = ['npm', 'pnpm', 'yarn']
+
+describe.skipIf(!existsSync(builtScripts))('live-linked generated projects', () => {
+  for (const name of managers) {
+    it(`${name}: installs the local closure and resolves plugin TypeScript in dev`, async (context) => {
+      const version = await managerVersion(name)
+      if (!version) {
+        context.skip()
+        return
+      }
+      const parent = await mkdtemp(join(tmpdir(), `dsh-link-${name}-`))
+      const root = join(parent, 'project')
+      temporary.push(parent)
+      const manager = createPackageManager(name, version)
+      await scaffoldProject(root, {
+        name: `linked-${name}`,
+        description: 'link e2e',
+        runtime: { model: 'deepseek-v4-flash' },
+        packageManager: manager,
+        releaseVersion: '0.0.1',
+        linkWorkspaceRoot: repoRoot,
+        features: [
+          { id: featureId('provider'), options: ['deepseek'], secrets: { apiKey: 'test-key' } },
+          { id: featureId('bash'), options: ['local'] },
+          { id: featureId('app'), options: ['embed'] },
+          { id: featureId('persistence'), options: ['jsonl'] },
+        ],
+        localPlugins: [new LocalPluginBlueprint('probe', 'plugin')],
+      })
+      await writeFile(join(root, 'plugins/probe/src/index.ts'), `
+        import { writeFileSync } from 'node:fs'
+        import type { Context } from 'cordis'
+        export const name = 'probe'
+        export function apply(_ctx: Context): void {
+          writeFileSync(new URL('../../../plugin-loaded', import.meta.url), 'loaded\\n')
+        }
+      `)
+      const cacheRoot = join(tmpdir(), 'dsh-sdk-link-cache', name)
+      const commandEnvironment = {
+        ...scrubEnvironment(),
+        COREPACK_HOME: join(cacheRoot, 'corepack'),
+        XDG_CACHE_HOME: join(cacheRoot, 'cache'),
+        XDG_DATA_HOME: join(cacheRoot, 'data'),
+        npm_config_cache: join(cacheRoot, 'npm'),
+        pnpm_config_store_dir: join(cacheRoot, 'pnpm-store'),
+      }
+      await execFileAsync(name, manager.installCommand(), {
+        cwd: root,
+        env: commandEnvironment,
+        encoding: 'utf8',
+        timeout: 120_000,
+      })
+      await execFileAsync(name, manager.buildCommand(), {
+        cwd: root,
+        env: commandEnvironment,
+        encoding: 'utf8',
+        timeout: 120_000,
+      })
+      expect(existsSync(join(root, 'index.js'))).toBe(true)
+      expect(existsSync(join(root, 'plugins/probe/lib/index.js'))).toBe(true)
+      const dshSdk = join(root, 'node_modules/@deepseek-ai/dsh-scripts/lib/bin.js')
+      const run = await execFileAsync(process.execPath, [dshSdk, 'dev', 'index.ts'], {
+        cwd: root,
+        env: { ...commandEnvironment, DEEPSEEK_API_KEY: 'test-key' },
+        encoding: 'utf8',
+        timeout: 30_000,
+      })
+      expect(run.stderr).not.toContain('without inject')
+      expect(await readFile(join(root, 'plugin-loaded'), 'utf8')).toBe('loaded\n')
+      const manifest = JSON.parse(await readFile(join(root, 'package.json'), 'utf8')) as {
+        dependencies: Record<string, string>
+      }
+      expect(manifest.dependencies.cordis).toMatch(name === 'npm' ? /^file:/ : name === 'pnpm' ? /^link:/ : /^portal:/)
+      expect(manifest.dependencies).not.toHaveProperty('node-addon-require-builtin')
+    }, 180_000)
+  }
+})
diff --git a/packages/sdk/create-sdk/tsconfig.json b/packages/sdk/create-sdk/tsconfig.json
new file mode 100644
index 0000000000..f77f711880
--- /dev/null
+++ b/packages/sdk/create-sdk/tsconfig.json
@@ -0,0 +1,12 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": ["src"],
+  "references": [
+    { "path": "../helper" },
+    { "path": "../../../vendor/cordis" }
+  ]
+}
diff --git a/packages/sdk/create-sdk/tsdown.config.ts b/packages/sdk/create-sdk/tsdown.config.ts
new file mode 100644
index 0000000000..cdf9ef6d4f
--- /dev/null
+++ b/packages/sdk/create-sdk/tsdown.config.ts
@@ -0,0 +1,14 @@
+import { defineConfig } from 'tsdown'
+
+/** Bundle the library and create bin, then mirror package-owned terminal templates. */
+export default defineConfig({
+  entry: ['lib/types/index.js', 'lib/types/bin.js'],
+  outDir: 'lib',
+  format: ['esm'],
+  platform: 'node',
+  target: 'es2024',
+  fixedExtension: false,
+  dts: false,
+  clean: false,
+  copy: [{ from: 'src/templates/assets/*', to: 'lib/assets' }],
+})
diff --git a/packages/sdk/helper/README.md b/packages/sdk/helper/README.md
new file mode 100644
index 0000000000..5b5608cdf0
--- /dev/null
+++ b/packages/sdk/helper/README.md
@@ -0,0 +1,23 @@
+# `@deepseek-ai/dsh-helper`
+
+Shared project domain and infrastructure for `create-sdk` and `dsh-sdk config`. `SdkProject` is a read-only snapshot; `ProjectEditSession` is the only mutation and commit boundary. The [SDK architecture RFC](../../../docs/rfc/proposed/architecture/2026-07-15-sdk-project-editing-architecture.md) owns the rationale.
+
+The package owns the builtin typed-spec catalog, provider/app behavior entities, structured project file objects, helper-owned project templates, the shared typed `TextTemplate` renderer, package-manager strategies, local-plugin blueprints, typed questions, and the clack prompt adapter. It never boots a Cordis application.
+
+All business and document validation completes before commit writes any affected file. Commit detects external edits made after the session opened, but deliberately provides no cross-file rollback after writing starts.
+
+Builtin features are provider, bash, app, persistence, HMR, filesystem, todo, skill, web, subagent, workflow, compaction, hooks, repeat-tool guard, timeout policy, and ask-user. The catalog owns feature options, required and non-default Cordis plugin config, feature requirements, resource contribution, and round-trip markers; create and config use the same registry and configurator.
+
+`SdkProject.open()` requires only readable root `package.json` and `cordis.yml`. A Cordis config entry anchors feature installation; a package present only through a linked NPM dependency closure leaves the feature absent. Once an owned Cordis config entry exists, an incomplete resource shape is `inconsistent` and cannot be modified automatically.
+
+`.env.example` follows the currently selected features. `.env` is append-only: helper may add a missing differently named variable, but never updates or removes existing content.
+
+The package root explicitly exports only the objects consumed by `create-sdk` and `dsh-scripts`; internal modules have no `src/*` or package-manifest subpath export.
+
+## Model Experience
+
+None, as the project domain edits files and never mounts a live agent or model request.
+
+## Known Limitations and Deferred Work
+
+- **Commit is not transactional across files** — external edits are detected before each write, but a later failure does not roll back files already written.
diff --git a/packages/sdk/helper/package.json b/packages/sdk/helper/package.json
new file mode 100644
index 0000000000..8d95dafe66
--- /dev/null
+++ b/packages/sdk/helper/package.json
@@ -0,0 +1,44 @@
+{
+  "name": "@deepseek-ai/dsh-helper",
+  "description": "Domain model and infrastructure for creating and editing DeepSeek Harness SDK projects",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    }
+  },
+  "files": [
+    "lib/index.js",
+    "lib/assets",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "dependencies": {
+    "@clack/core": "^1.4.3",
+    "@clack/prompts": "^1.7.0",
+    "handlebars": "^4.7.9",
+    "jsonc-parser": "^3.3.1",
+    "yaml": "^2.9.0"
+  },
+  "peerDependencies": {
+    "@deepseek-ai/dsh-brand": "^0.0.1",
+    "cordis": "^4.0.0-rc.7"
+  },
+  "devDependencies": {
+    "@deepseek-ai/dsh-brand": "workspace:^",
+    "@deepseek-ai/dsh-hooks-claude": "workspace:^",
+    "@deepseek-ai/dsh-hooks-codex": "workspace:^",
+    "@deepseek-ai/dsh-session-persistence-jsonl": "workspace:^",
+    "@deepseek-ai/dsh-session-persistence-sqlite": "workspace:^",
+    "@deepseek-ai/dsh-tool-subagent": "workspace:^",
+    "@deepseek-ai/dsh-tool-web": "workspace:^",
+    "cordis": "^4.0.0-rc.7"
+  }
+}
diff --git a/packages/sdk/helper/src/documents/cordis-yaml-file.ts b/packages/sdk/helper/src/documents/cordis-yaml-file.ts
new file mode 100644
index 0000000000..39fd8fd596
--- /dev/null
+++ b/packages/sdk/helper/src/documents/cordis-yaml-file.ts
@@ -0,0 +1,190 @@
+/**
+ * Comment-preserving Cordis YAML document and `!!js` expression value.
+ *
+ * @module @deepseek-ai/dsh-helper/documents/cordis-yaml-file
+ */
+
+import {
+  Document, isMap, isSeq, parseDocument, visit, YAMLMap, YAMLSeq,
+  type ScalarTag,
+} from 'yaml'
+import { ProjectFile, withTrailingNewline } from './project-file.ts'
+
+/** Explicit JavaScript expression serialized with Cordis' `!!js` YAML tag. */
+export class JsExpression {
+  /** Expression source evaluated by the Cordis include loader. */
+  readonly source: string
+
+  /** Create an expression value. */
+  constructor(source: string) {
+    if (source.trim().length === 0) throw new Error('JavaScript expression must not be empty')
+    this.source = source
+  }
+
+  /** Return expression source for YAML scalar stringification. */
+  toString(): string {
+    return this.source
+  }
+}
+
+const JS_EXPRESSION_TAG: ScalarTag = {
+  tag: 'tag:yaml.org,2002:js',
+  identify: value => value instanceof JsExpression,
+  resolve: value => new JsExpression(value),
+  stringify: item => String(item.value),
+}
+
+/** Plain domain representation of one top-level Cordis config entry. */
+export interface CordisConfigEntry {
+  id: string
+  name: string
+  config?: Record<string, unknown>
+  disabled?: boolean
+}
+
+function parseYaml(text: string): Document.Parsed {
+  const document = parseDocument(text, {
+    customTags: [JS_EXPRESSION_TAG],
+    keepSourceTokens: true,
+    prettyErrors: true,
+  })
+  if (document.errors.length > 0) {
+    throw new Error(`invalid cordis.yml: ${document.errors.map(error => error.message).join('; ')}`)
+  }
+  if (!isSeq(document.contents)) throw new Error('invalid cordis.yml: root must be a sequence')
+  visit(document, { Collection: (_key, collection) => { collection.flow = false } })
+  return document
+}
+
+function entryFromValue(value: unknown): CordisConfigEntry {
+  /* v8 ignore next -- entries() calls this only after requiring a YAMLMap, whose JSON value is an object */
+  if (value === null || Array.isArray(value) || typeof value !== 'object') {
+    throw new Error('invalid cordis.yml entry: expected an object')
+  }
+  const entry = value as Record<string, unknown>
+  if (typeof entry.id !== 'string' || entry.id.length === 0) {
+    throw new Error('invalid cordis.yml entry: id must be a non-empty string')
+  }
+  if (typeof entry.name !== 'string' || entry.name.length === 0) {
+    throw new Error(`invalid cordis.yml entry ${entry.id}: name must be a non-empty string`)
+  }
+  if (entry.config !== undefined
+    && (entry.config === null || Array.isArray(entry.config) || typeof entry.config !== 'object')) {
+    throw new Error(`invalid cordis.yml entry ${entry.id}: plugin config must be an object`)
+  }
+  if (entry.disabled !== undefined && typeof entry.disabled !== 'boolean') {
+    throw new Error(`invalid cordis.yml entry ${entry.id}: disabled must be boolean`)
+  }
+  return {
+    id: entry.id,
+    name: entry.name,
+    ...entry.config !== undefined ? { config: entry.config as Record<string, unknown> } : {},
+    ...entry.disabled !== undefined ? { disabled: entry.disabled } : {},
+  }
+}
+
+/** Editable top-level cordis.yml using YAML's document API. */
+export class CordisYamlFile extends ProjectFile {
+  private readonly document: Document.Parsed
+
+  private constructor(document: Document.Parsed, originalText?: string) {
+    super('cordis.yml', originalText)
+    this.document = document
+  }
+
+  /** Create an empty Cordis config entry list. */
+  static create(): CordisYamlFile {
+    return new CordisYamlFile(parseYaml('[]\n'))
+  }
+
+  /** Parse an existing cordis.yml while retaining comments and scalar styles. */
+  static parse(text: string): CordisYamlFile {
+    return new CordisYamlFile(parseYaml(text), text)
+  }
+
+  /** Clone through YAML text so the edit session owns an independent AST. */
+  override clone(): CordisYamlFile {
+    return new CordisYamlFile(parseYaml(this.serialize()), this.originalText)
+  }
+
+  private sequence(): YAMLSeq {
+    /* v8 ignore next -- parseYaml and create both establish a sequence root */
+    if (!isSeq(this.document.contents)) throw new Error('cordis.yml root is not a sequence')
+    return this.document.contents
+  }
+
+  private entryNode(id: string): YAMLMap | undefined {
+    for (const item of this.sequence().items) {
+      if (!isMap(item)) continue
+      if (item.get('id') === id) return item
+    }
+    return undefined
+  }
+
+  /** Return defensive plain entry values in file order. */
+  entries(): CordisConfigEntry[] {
+    return this.sequence().items.map((item) => {
+      if (!isMap(item)) throw new Error('invalid cordis.yml: every entry must be a mapping')
+      return entryFromValue(item.toJSON())
+    })
+  }
+
+  /** Find one entry by stable id. */
+  entry(id: string): CordisConfigEntry | undefined {
+    return this.entries().find(entry => entry.id === id)
+  }
+
+  /** Add one new top-level entry, rejecting duplicate ids. */
+  addEntry(entry: CordisConfigEntry, commentedExample?: string): void {
+    if (this.entryNode(entry.id)) throw new Error(`Cordis config entry already exists: ${entry.id}`)
+    const node = this.document.createNode(entry)
+    if (commentedExample) node.comment = commentedExample.split('\n').map(line => ` ${line}`).join('\n')
+    this.sequence().items.push(node)
+  }
+
+  /** Remove an entry by id and report whether it existed. */
+  removeEntry(id: string): boolean {
+    const sequence = this.sequence()
+    const index = sequence.items.findIndex(item => isMap(item) && item.get('id') === id)
+    if (index < 0) return false
+    sequence.items.splice(index, 1)
+    return true
+  }
+
+  /** Enable or disable an entry through the Loader-native field. */
+  setDisabled(id: string, disabled: boolean): void {
+    const node = this.entryNode(id)
+    if (!node) throw new Error(`Cordis config entry does not exist: ${id}`)
+    if (disabled) node.set('disabled', true)
+    else node.delete('disabled')
+  }
+
+  /** Replace only owned plugin config keys while retaining unknown user keys. */
+  updateOwnedConfig(id: string, ownedKeys: readonly string[], next: Record<string, unknown>): void {
+    const entry = this.entryNode(id)
+    if (!entry) throw new Error(`Cordis config entry does not exist: ${id}`)
+    let config: unknown = entry.get('config', true)
+    if (config === undefined || config === null) {
+      config = new YAMLMap()
+      entry.set('config', config)
+    }
+    if (!isMap(config)) throw new Error(`Cordis config entry ${id} plugin config is not a mapping`)
+    for (const key of ownedKeys) config.delete(key)
+    for (const [key, value] of Object.entries(next)) config.set(key, this.document.createNode(value))
+    if (config.items.length === 0) entry.delete('config')
+  }
+
+  /** Validate ids, names, plugin config maps, and id uniqueness. */
+  override validate(): void {
+    const seen = new Set<string>()
+    for (const entry of this.entries()) {
+      if (seen.has(entry.id)) throw new Error(`duplicate Cordis config entry id: ${entry.id}`)
+      seen.add(entry.id)
+    }
+  }
+
+  /** Serialize through the YAML document while retaining untouched trivia. */
+  override serialize(): string {
+    return withTrailingNewline(this.document.toString({ lineWidth: 0 }))
+  }
+}
diff --git a/packages/sdk/helper/src/documents/env-file.ts b/packages/sdk/helper/src/documents/env-file.ts
new file mode 100644
index 0000000000..d0e6ca8473
--- /dev/null
+++ b/packages/sdk/helper/src/documents/env-file.ts
@@ -0,0 +1,111 @@
+/**
+ * Ownership-aware, line-preserving dotenv document.
+ *
+ * @module @deepseek-ai/dsh-helper/documents/env-file
+ */
+
+import { ProjectFile, withTrailingNewline } from './project-file.ts'
+
+interface ParsedVariable {
+  index: number
+  value: string
+}
+
+const VARIABLE = /^(?:export\s+)?([A-Za-z_][A-Za-z0-9_]*)=(.*)$/
+const VARIABLE_NAME = /^[A-Za-z_][A-Za-z0-9_]*$/
+
+/** `.env` appends missing variables; `.env.example` supports managed replacement and removal. */
+export class EnvFile extends ProjectFile {
+  private readonly lines: string[]
+
+  private constructor(relativePath: '.env' | '.env.example', lines: string[], originalText?: string) {
+    super(relativePath, originalText, relativePath === '.env' ? 0o600 : undefined)
+    this.lines = [...lines]
+  }
+
+  /** Create an empty environment file. */
+  static create(relativePath: '.env' | '.env.example'): EnvFile {
+    return new EnvFile(relativePath, [])
+  }
+
+  /** Parse an existing environment file without rewriting unknown lines. */
+  static parse(relativePath: '.env' | '.env.example', text: string): EnvFile {
+    const normalized = text.replace(/\n$/, '')
+    return new EnvFile(relativePath, normalized.length === 0 ? [] : normalized.split('\n'), text)
+  }
+
+  /** Clone the current line model. */
+  override clone(): EnvFile {
+    return new EnvFile(this.relativePath as '.env' | '.env.example', this.lines, this.originalText)
+  }
+
+  private variables(): Map<string, ParsedVariable[]> {
+    const values = new Map<string, ParsedVariable[]>()
+    this.lines.forEach((line, index) => {
+      const match = VARIABLE.exec(line)
+      if (!match) return
+      const name = match[1]
+      const value = match[2]
+      /* v8 ignore next -- both captures are mandatory in VARIABLE */
+      if (name === undefined || value === undefined) return
+      const occurrences = values.get(name) ?? []
+      occurrences.push({ index, value })
+      values.set(name, occurrences)
+    })
+    return values
+  }
+
+  /** Read the effective value; append-only `.env` accepts duplicates and uses the last declaration. */
+  get(name: string): string | undefined {
+    const occurrences = this.variables().get(name) ?? []
+    if (this.relativePath === '.env.example' && occurrences.length > 1) {
+      throw new Error(`${this.relativePath} contains duplicate variable ${name}`)
+    }
+    return occurrences.at(-1)?.value
+  }
+
+  /** Add or replace one SDK-managed `.env.example` variable while preserving unrelated lines. */
+  set(name: string, value: string): void {
+    if (this.relativePath !== '.env.example') throw new Error('.env is append-only')
+    if (!VARIABLE_NAME.test(name)) throw new Error(`invalid environment variable name: ${name}`)
+    const occurrences = this.variables().get(name) ?? []
+    if (occurrences.length > 1) throw new Error(`${this.relativePath} contains duplicate variable ${name}`)
+    const line = `${name}=${value}`
+    if (occurrences[0]) this.lines[occurrences[0].index] = line
+    else this.lines.push(line)
+  }
+
+  /** Append a missing `.env` variable and optional comment without changing any existing declaration. */
+  append(name: string, value: string, comment?: string): boolean {
+    if (this.relativePath !== '.env') throw new Error('.env.example is SDK-managed')
+    if (!VARIABLE_NAME.test(name)) throw new Error(`invalid environment variable name: ${name}`)
+    if (comment !== undefined && (!comment || comment.includes('\n'))) {
+      throw new Error('environment comment must be one non-empty line')
+    }
+    if (this.variables().has(name)) return false
+    if (comment) this.lines.push(`# ${comment}`)
+    this.lines.push(`${name}=${value}`)
+    return true
+  }
+
+  /** Remove one SDK-managed `.env.example` variable while retaining every other line. */
+  remove(name: string): void {
+    if (this.relativePath !== '.env.example') throw new Error('.env is append-only')
+    const occurrences = this.variables().get(name) ?? []
+    if (occurrences.length > 1) throw new Error(`${this.relativePath} contains duplicate variable ${name}`)
+    if (occurrences[0]) this.lines.splice(occurrences[0].index, 1)
+  }
+
+  /** Validate the managed placeholder file; append-only `.env` accepts duplicate declarations. */
+  override validate(): void {
+    if (this.relativePath === '.env') return
+    for (const [name, occurrences] of this.variables()) {
+      if (occurrences.length > 1) throw new Error(`${this.relativePath} contains duplicate variable ${name}`)
+    }
+  }
+
+  /** Serialize all retained lines with one trailing newline. */
+  override serialize(): string {
+    return withTrailingNewline(this.lines.join('\n'))
+  }
+}
diff --git a/packages/sdk/helper/src/documents/package-json-file.ts b/packages/sdk/helper/src/documents/package-json-file.ts
new file mode 100644
index 0000000000..f565f383d2
--- /dev/null
+++ b/packages/sdk/helper/src/documents/package-json-file.ts
@@ -0,0 +1,169 @@
+/**
+ * Structured package.json document owned by an SDK project.
+ *
+ * @module @deepseek-ai/dsh-helper/documents/package-json-file
+ */
+
+import { ProjectFile, withTrailingNewline } from './project-file.ts'
+
+/** NPM dependency sections managed by the SDK. */
+export type NpmDependencySection = 'dependencies' | 'devDependencies'
+
+/** JSON shape retained by {@link PackageJsonFile}. */
+export interface PackageManifest {
+  name?: string
+  version?: string
+  private?: boolean
+  description?: string
+  type?: string
+  packageManager?: string
+  scripts?: Record<string, string>
+  dependencies?: Record<string, string>
+  devDependencies?: Record<string, string>
+  workspaces?: string[]
+  resolutions?: Record<string, string>
+  [key: string]: unknown
+}
+
+function parseManifest(text: string): PackageManifest {
+  let value: unknown
+  try {
+    value = JSON.parse(text)
+  } catch (error) {
+    throw new Error(`invalid package.json: ${String(error)}`)
+  }
+  if (value === null || Array.isArray(value) || typeof value !== 'object') {
+    throw new Error('invalid package.json: root must be an object')
+  }
+  return value as PackageManifest
+}
+
+function sortedRecord(value: Record<string, string>): Record<string, string> {
+  return Object.fromEntries(Object.entries(value).sort(([left], [right]) => left.localeCompare(right)))
+}
+
+/** Editable, deterministic package.json representation. */
+export class PackageJsonFile extends ProjectFile {
+  private readonly manifest: PackageManifest
+
+  private constructor(manifest: PackageManifest, originalText?: string) {
+    super('package.json', originalText)
+    this.manifest = structuredClone(manifest)
+  }
+
+  /** Create a new package manifest from a complete rendered template. */
+  static create(text: string): PackageJsonFile {
+    return new PackageJsonFile(parseManifest(text))
+  }
+
+  /** Parse an existing package.json document. */
+  static parse(text: string): PackageJsonFile {
+    return new PackageJsonFile(parseManifest(text), text)
+  }
+
+  /** Clone this document and its nested manifest data. */
+  override clone(): PackageJsonFile {
+    return new PackageJsonFile(this.manifest, this.originalText)
+  }
+
+  /** Return a defensive copy of the manifest. */
+  value(): Readonly<PackageManifest> {
+    return structuredClone(this.manifest)
+  }
+
+  /** Set one package script. */
+  setScript(name: string, command: string): void {
+    this.manifest.scripts ??= {}
+    this.manifest.scripts[name] = command
+  }
+
+  /** Read one package script. */
+  script(name: string): string | undefined {
+    return this.manifest.scripts?.[name]
+  }
+
+  /** Remove one package script. */
+  removeScript(name: string): void {
+    delete this.manifest.scripts?.[name]
+  }
+
+  /** Set one NPM dependency in its runtime or development section. */
+  setNpmDependency(section: NpmDependencySection, name: string, spec: string): void {
+    this.manifest[section] ??= {}
+    this.manifest[section][name] = spec
+  }
+
+  /** Remove one NPM dependency from a section. */
+  removeNpmDependency(section: NpmDependencySection, name: string): void {
+    delete this.manifest[section]?.[name]
+  }
+
+  /** Read an NPM dependency spec from either managed section. */
+  npmDependency(name: string): { section: NpmDependencySection; spec: string } | undefined {
+    for (const section of ['dependencies', 'devDependencies'] as const) {
+      const spec = this.manifest[section]?.[name]
+      if (spec !== undefined) return { section, spec }
+    }
+    return undefined
+  }
+
+  /** Return all managed NPM dependency names. */
+  npmDependencyNames(): string[] {
+    return [...new Set([
+      ...Object.keys(this.manifest.dependencies ?? {}),
+      ...Object.keys(this.manifest.devDependencies ?? {}),
+    ])].sort()
+  }
+
+  /** Add a package-manager workspace glob. */
+  addWorkspace(pattern: string): void {
+    const workspaces = this.manifest.workspaces ??= []
+    if (!workspaces.includes(pattern)) workspaces.push(pattern)
+  }
+
+  /** Set or remove the packageManager field. */
+  setPackageManager(value: string | undefined): void {
+    if (value === undefined) delete this.manifest.packageManager
+    else this.manifest.packageManager = value
+  }
+
+  /** Pin a Yarn resolution used by live-link projects. */
+  setResolution(name: string, spec: string): void {
+    this.manifest.resolutions ??= {}
+    this.manifest.resolutions[name] = spec
+  }
+
+  /** Validate the fields the SDK relies on. */
+  override validate(): void {
+    if (!this.manifest.name || typeof this.manifest.name !== 'string') {
+      throw new Error('package.json name must be a non-empty string')
+    }
+    for (const section of ['scripts', 'dependencies', 'devDependencies'] as const) {
+      const value: unknown = this.manifest[section]
+      if (value === undefined) continue
+      if (value === null || Array.isArray(value) || typeof value !== 'object') {
+        throw new Error(`package.json ${section} must be an object`)
+      }
+      for (const [key, item] of Object.entries(value)) {
+        if (typeof item !== 'string' || item.length === 0) {
+          throw new Error(`package.json ${section}.${key} must be a non-empty string`)
+        }
+      }
+    }
+    if (this.manifest.workspaces !== undefined
+      && (!Array.isArray(this.manifest.workspaces) || this.manifest.workspaces.some(item => typeof item !== 'string'))) {
+      throw new Error('package.json workspaces must be an array of strings')
+    }
+  }
+
+  /** Serialize with deterministic managed maps and two-space JSON formatting. */
+  override serialize(): string {
+    const value: PackageManifest = structuredClone(this.manifest)
+    if (this.manifest.scripts) value.scripts = sortedRecord(this.manifest.scripts)
+    if (this.manifest.dependencies) value.dependencies = sortedRecord(this.manifest.dependencies)
+    if (this.manifest.devDependencies) value.devDependencies = sortedRecord(this.manifest.devDependencies)
+    if (this.manifest.workspaces) value.workspaces = [...this.manifest.workspaces].sort()
+    if (this.manifest.resolutions) value.resolutions = sortedRecord(this.manifest.resolutions)
+    return withTrailingNewline(JSON.stringify(value, null, 2))
+  }
+}
diff --git a/packages/sdk/helper/src/documents/pnpm-workspace-file.ts b/packages/sdk/helper/src/documents/pnpm-workspace-file.ts
new file mode 100644
index 0000000000..b980434bad
--- /dev/null
+++ b/packages/sdk/helper/src/documents/pnpm-workspace-file.ts
@@ -0,0 +1,96 @@
+/**
+ * Structured pnpm workspace configuration for generated SDK projects.
+ *
+ * @module @deepseek-ai/dsh-helper/documents/pnpm-workspace-file
+ */
+
+import {
+  isMap, isScalar, isSeq, parseDocument,
+  type Document, type Scalar, type YAMLMap, type YAMLSeq,
+} from 'yaml'
+import { ProjectFile, withTrailingNewline } from './project-file.ts'
+
+function parseYaml(text: string): Document.Parsed {
+  const document = parseDocument(text, { keepSourceTokens: true, prettyErrors: true })
+  if (document.errors.length > 0) {
+    throw new Error(`invalid pnpm-workspace.yaml: ${document.errors.map(error => error.message).join('; ')}`)
+  }
+  if (!isMap(document.contents)) throw new Error('pnpm-workspace.yaml root must be an object')
+  return document
+}
+
+/** Generated pnpm-workspace.yaml model. */
+export class PnpmWorkspaceFile extends ProjectFile {
+  private readonly document: Document.Parsed
+
+  private constructor(document: Document.Parsed, originalText?: string) {
+    super('pnpm-workspace.yaml', originalText)
+    this.document = document
+  }
+
+  /** Create a pnpm workspace document. */
+  static create(): PnpmWorkspaceFile {
+    const document = new PnpmWorkspaceFile(parseYaml('{}\n'))
+    document.mapping().set('packages', document.document.createNode([]))
+    document.mapping().set('allowBuilds', document.document.createNode({ esbuild: true }))
+    return document
+  }
+
+  /** Parse the workspace fields the SDK owns while retaining all other YAML. */
+  static parse(text: string): PnpmWorkspaceFile {
+    const document = new PnpmWorkspaceFile(parseYaml(text), text)
+    document.packageSequence()
+    const autoInstallPeers = document.mapping().get('autoInstallPeers')
+    if (autoInstallPeers !== undefined && typeof autoInstallPeers !== 'boolean') {
+      throw new Error('pnpm-workspace.yaml autoInstallPeers must be boolean')
+    }
+    return document
+  }
+
+  /** Clone the complete comment-preserving workspace document. */
+  override clone(): PnpmWorkspaceFile {
+    return new PnpmWorkspaceFile(parseYaml(this.serialize()), this.originalText)
+  }
+
+  /** Add one package workspace glob. */
+  addPackage(pattern: string): void {
+    const packages = this.packageSequence()
+    if (packages.items.some(item => item.value === pattern)) return
+    packages.add(this.document.createNode(pattern))
+  }
+
+  /** Disable registry peer auto-installation for live-link projects. */
+  disableAutoInstallPeers(): void {
+    this.mapping().set('autoInstallPeers', false)
+  }
+
+  /** Validate workspace globs. */
+  override validate(): void {
+    for (const pattern of this.packageValues()) {
+      if (pattern.trim().length === 0) throw new Error('pnpm workspace pattern must not be empty')
+    }
+  }
+
+  /** Serialize the workspace while retaining unknown settings and comments. */
+  override serialize(): string {
+    return withTrailingNewline(this.document.toString({ lineWidth: 0 }))
+  }
+
+  private mapping(): YAMLMap {
+    /* v8 ignore next -- parseYaml and create both establish a mapping root */
+    if (!isMap(this.document.contents)) throw new Error('pnpm-workspace.yaml root must be an object')
+    return this.document.contents
+  }
+
+  private packageSequence(): YAMLSeq<Scalar<string>> {
+    const packages = this.mapping().get('packages', true)
+    if (!isSeq(packages) || packages.items.some(item => !isScalar(item) || typeof item.value !== 'string')) {
+      throw new Error('pnpm-workspace.yaml packages must be an array of strings')
+    }
+    return packages as YAMLSeq<Scalar<string>>
+  }
+
+  private packageValues(): string[] {
+    return this.packageSequence().items.map(item => item.value)
+  }
+}
diff --git a/packages/sdk/helper/src/documents/project-file.ts b/packages/sdk/helper/src/documents/project-file.ts
new file mode 100644
index 0000000000..9be20c3395
--- /dev/null
+++ b/packages/sdk/helper/src/documents/project-file.ts
@@ -0,0 +1,64 @@
+/**
+ * Base abstraction for one file in an SDK project snapshot.
+ *
+ * @module @deepseek-ai/dsh-helper/documents/project-file
+ */
+
+/** Return text with exactly one trailing newline. */
+export function withTrailingNewline(text: string): string {
+  return text.replace(/\n*$/, '') + '\n'
+}
+
+/** One cloneable, validatable project file. */
+export abstract class ProjectFile {
+  /** Project-relative POSIX path. */
+  readonly relativePath: string
+
+  /** Text observed when the document entered the snapshot; absent for a new file. */
+  readonly originalText: string | undefined
+
+  /** Permission bits used only when the file is first created. */
+  readonly createMode: number | undefined
+
+  protected constructor(relativePath: string, originalText?: string, createMode?: number) {
+    if (relativePath.startsWith('/') || relativePath.split('/').includes('..')) {
+      throw new Error(`project document path must stay inside the project: ${relativePath}`)
+    }
+    this.relativePath = relativePath
+    this.originalText = originalText
+    this.createMode = createMode
+  }
+
+  /** Clone the document for an isolated edit session. */
+  abstract clone(): ProjectFile
+
+  /** Validate the document's complete current state. */
+  abstract validate(): void
+
+  /** Serialize the complete current file. */
+  abstract serialize(): string
+}
+
+/** Immutable complete-text file used by one-shot artifacts. */
+export class TextProjectFile extends ProjectFile {
+  private readonly text: string
+
+  /** Create a complete-text project document. */
+  constructor(relativePath: string, text: string, originalText?: string) {
+    super(relativePath, originalText)
+    this.text = withTrailingNewline(text)
+  }
+
+  /** Clone this immutable document. */
+  override clone(): TextProjectFile {
+    return new TextProjectFile(this.relativePath, this.text, this.originalText)
+  }
+
+  /** Complete text artifacts have no extra structural validation. */
+  override validate(): void {}
+
+  /** Return the complete artifact text. */
+  override serialize(): string {
+    return this.text
+  }
+}
diff --git a/packages/sdk/helper/src/documents/tsconfig-file.ts b/packages/sdk/helper/src/documents/tsconfig-file.ts
new file mode 100644
index 0000000000..368b93eb43
--- /dev/null
+++ b/packages/sdk/helper/src/documents/tsconfig-file.ts
@@ -0,0 +1,89 @@
+/**
+ * Comment-preserving root tsconfig editor for local plugin references.
+ *
+ * @module @deepseek-ai/dsh-helper/documents/tsconfig-file
+ */
+
+import { applyEdits, modify, parse, type ParseError } from 'jsonc-parser'
+import { ProjectFile, withTrailingNewline } from './project-file.ts'
+
+const FORMAT = { insertSpaces: true, tabSize: 2, eol: '\n' }
+
+function parseConfig(text: string): Record<string, unknown> {
+  const errors: ParseError[] = []
+  const value: unknown = parse(text, errors, { allowTrailingComma: true, disallowComments: false })
+  if (errors.length > 0 || value === null || Array.isArray(value) || typeof value !== 'object') {
+    throw new Error('tsconfig.json is not a valid JSONC object')
+  }
+  return value as Record<string, unknown>
+}
+
+/** Root tsconfig document edited with jsonc-parser patches. */
+export class TsConfigFile extends ProjectFile {
+  private text: string
+
+  private constructor(text: string, originalText?: string) {
+    super('tsconfig.json', originalText)
+    this.text = withTrailingNewline(text)
+  }
+
+  /** Create the root project-reference config. */
+  static create(): TsConfigFile {
+    return new TsConfigFile(JSON.stringify({
+      extends: './tsconfig.base.json',
+      compilerOptions: { noEmit: true },
+      include: ['index.ts'],
+      references: [],
+    }, null, 2))
+  }
+
+  /** Parse an existing root tsconfig. */
+  static parse(text: string): TsConfigFile {
+    parseConfig(text)
+    return new TsConfigFile(text, text)
+  }
+
+  /** Clone the current JSONC text. */
+  override clone(): TsConfigFile {
+    return new TsConfigFile(this.text, this.originalText)
+  }
+
+  /** Add one project reference while retaining comments and formatting. */
+  addReference(path: string): void {
+    const value = parseConfig(this.text)
+    const references = value.references
+    if (references !== undefined && !Array.isArray(references)) {
+      throw new Error('tsconfig.json references must be an array')
+    }
+    const typed = (references ?? []) as unknown[]
+    for (const item of typed) {
+      if (item === null || Array.isArray(item) || typeof item !== 'object' || typeof (item as { path?: unknown }).path !== 'string') {
+        throw new Error('tsconfig.json references must contain { path: string } objects')
+      }
+    }
+    if (typed.some(item => (item as { path: string }).path === path)) return
+    this.text = applyEdits(this.text, modify(
+      this.text,
+      ['references', typed.length],
+      { path },
+      { formattingOptions: FORMAT, isArrayInsertion: true },
+    ))
+  }
+
+  /** Validate JSONC and the project-reference shape. */
+  override validate(): void {
+    const value = parseConfig(this.text)
+    if (value.references === undefined) return
+    if (!Array.isArray(value.references)) throw new Error('tsconfig.json references must be an array')
+    for (const item of value.references) {
+      if (item === null || Array.isArray(item) || typeof item !== 'object' || typeof (item as { path?: unknown }).path !== 'string') {
+        throw new Error('tsconfig.json references must contain { path: string } objects')
+      }
+    }
+  }
+
+  /** Return patched JSONC text. */
+  override serialize(): string {
+    return withTrailingNewline(this.text)
+  }
+}
diff --git a/packages/sdk/helper/src/features/builtin/app.ts b/packages/sdk/helper/src/features/builtin/app.ts
new file mode 100644
index 0000000000..a89f2c84ae
--- /dev/null
+++ b/packages/sdk/helper/src/features/builtin/app.ts
@@ -0,0 +1,126 @@
+/**
+ * Required run-interface app feature.
+ *
+ * @module @deepseek-ai/dsh-helper/features/builtin/app
+ */
+
+import { featureId } from '../../ids.ts'
+import type { ProjectProfile } from '../../project/types.ts'
+import {
+  createAppPackageScripts,
+  createAppProjectArtifacts,
+  createProjectTemplateContext,
+} from '../../templates/project-template.ts'
+import {
+  FeatureOption,
+  ExclusiveOptionFeature,
+} from '../feature.ts'
+import { ProjectContribution, type ProjectResource } from '../resources.ts'
+import {
+  npmCordisConfigEntry,
+  optionalString,
+  ownedTextFile,
+  packageScript,
+  requiredString,
+} from './helpers.ts'
+
+const ID = featureId('app')
+
+function appProjectResources(
+  profile: ProjectProfile,
+  runInterface: 'acp' | 'stdio' | 'embed',
+): readonly ProjectResource[] {
+  const context = createProjectTemplateContext(profile, runInterface)
+  const scripts = createAppPackageScripts(context)
+  return [
+    ...createAppProjectArtifacts(context).map(document => (
+      ownedTextFile(ID, document.relativePath, document.serialize())
+    )),
+    packageScript(ID, 'dev', scripts.dev),
+    packageScript(ID, 'start', scripts.start),
+  ]
+}
+
+class AppOption extends FeatureOption {
+  override readonly id: 'acp' | 'stdio' | 'embed'
+  override readonly label: string
+
+  constructor(id: 'acp' | 'stdio' | 'embed', label: string) {
+    super()
+    this.id = id
+    this.label = label
+  }
+
+  /** Identify options by their unique front door, not the shared interaction service. */
+  override markerConfigEntries(): readonly { id: string; name: string }[] {
+    switch (this.id) {
+      case 'acp': return [{ id: 'acp', name: '@deepseek-ai/dsh-acp' }]
+      case 'stdio': return [{ id: 'stdio', name: '@deepseek-ai/dsh-stdio' }]
+      case 'embed': return []
+    }
+  }
+
+  /** Embed is identified by the configured loop with no external front door. */
+  override matchesConfigEntries(entries: readonly { id: string; name: string }[], profile: ProjectProfile): boolean {
+    if (this.id !== 'embed') return super.matchesConfigEntries(entries, profile)
+    return entries.some(entry => entry.id === 'agent-loop' && entry.name === '@deepseek-ai/dsh-agent-loop')
+      && !entries.some(entry => entry.name === '@deepseek-ai/dsh-acp' || entry.name === '@deepseek-ai/dsh-stdio')
+  }
+
+  override contribution(profile: ProjectProfile): ProjectContribution {
+    switch (this.id) {
+      case 'acp':
+        return new ProjectContribution([
+          ...appProjectResources(profile, this.id),
+          ...npmCordisConfigEntry(ID, {
+            id: 'user-interaction',
+            name: '@deepseek-ai/dsh-user-interaction',
+          }),
+          ...npmCordisConfigEntry(ID, {
+            id: 'acp',
+            name: '@deepseek-ai/dsh-acp',
+            config: { model: profile.runtime.model },
+          }, ['model'], config => requiredString(config, 'model')),
+        ])
+      case 'stdio':
+        return new ProjectContribution([
+          ...appProjectResources(profile, this.id),
+          ...npmCordisConfigEntry(ID, {
+            id: 'user-interaction',
+            name: '@deepseek-ai/dsh-user-interaction',
+          }),
+          ...npmCordisConfigEntry(ID, {
+            id: 'stdio',
+            name: '@deepseek-ai/dsh-stdio',
+            config: {
+              welcome: 'agent REPL ready. Give it a coding task.',
+              agent: 'main',
+            },
+          }, ['welcome', 'agent'], config => [
+            ...optionalString(config, 'welcome'),
+            ...requiredString(config, 'agent'),
+          ]),
+        ])
+      case 'embed':
+        return new ProjectContribution(appProjectResources(profile, this.id))
+    }
+  }
+}
+
+/** Required app selection represented by acp, stdio, or embed options. */
+export class AppFeature extends ExclusiveOptionFeature {
+  override readonly id = ID
+  override readonly summary = 'Run interface'
+  override readonly required = true
+  override readonly requires = [featureId('spine')]
+  override readonly options = [
+    new AppOption('acp', 'ACP server'),
+    new AppOption('stdio', 'Terminal REPL'),
+    new AppOption('embed', 'Embedded context'),
+  ]
+
+  /** Default to the profile's already selected front door. */
+  override defaultOptions(profile: ProjectProfile): readonly string[] {
+    return [profile.runInterface]
+  }
+}
diff --git a/packages/sdk/helper/src/features/builtin/helpers.ts b/packages/sdk/helper/src/features/builtin/helpers.ts
new file mode 100644
index 0000000000..60f7530c61
--- /dev/null
+++ b/packages/sdk/helper/src/features/builtin/helpers.ts
@@ -0,0 +1,111 @@
+/**
+ * Small resource constructors shared by builtin feature modules.
+ *
+ * @module @deepseek-ai/dsh-helper/features/builtin/helpers
+ */
+
+import type { CordisConfigEntry } from '../../documents/cordis-yaml-file.ts'
+import { TextProjectFile } from '../../documents/project-file.ts'
+import { resourceKey } from '../../ids.ts'
+import type {
+  CordisConfigEntryResource,
+  EnvironmentResource,
+  OwnedFileResource,
+  NpmDependencyResource,
+  PackageScriptResource,
+} from '../resources.ts'
+
+/** Create a runtime NPM dependency resource. */
+function npmDependency(_owner: string, name: string): NpmDependencyResource {
+  return {
+    kind: 'npm-dependency',
+    key: resourceKey(`npm-dependency:${name}`),
+    name,
+    section: 'dependencies',
+  }
+}
+
+/** Create a feature-owned package script that is replaceable only while unchanged. */
+export function packageScript(_owner: string, name: string, command: string): PackageScriptResource {
+  return {
+    kind: 'package-script',
+    key: resourceKey(`package-script:${name}`),
+    name,
+    command,
+    removeOnlyWhenUnchanged: true,
+  }
+}
+
+/** Create a Cordis config entry resource with explicitly owned config keys. */
+export function cordisConfigEntry(
+  _owner: string,
+  value: CordisConfigEntry,
+  ownedConfigKeys: readonly string[] = Object.keys(value.config ?? {}),
+  validateConfig?: CordisConfigEntryResource['validateConfig'],
+): CordisConfigEntryResource {
+  return {
+    kind: 'cordis-config-entry',
+    key: resourceKey(`cordis-config-entry:${value.id}`),
+    entry: value,
+    ownedConfigKeys,
+    ...validateConfig ? { validateConfig } : {},
+  }
+}
+
+/** Couple one bare-package Cordis config entry to its mandatory runtime NPM dependency. */
+export function npmCordisConfigEntry(
+  owner: string,
+  value: CordisConfigEntry,
+  ownedConfigKeys: readonly string[] = Object.keys(value.config ?? {}),
+  validateConfig?: CordisConfigEntryResource['validateConfig'],
+): readonly [NpmDependencyResource, CordisConfigEntryResource] {
+  return [
+    npmDependency(owner, value.name),
+    cordisConfigEntry(owner, value, ownedConfigKeys, validateConfig),
+  ]
+}
+
+/** Create a secret/environment binding resource. */
+export function environment(
+  _owner: string,
+  name: string,
+  value: string | undefined,
+  comment?: string,
+): EnvironmentResource {
+  return {
+    kind: 'environment',
+    key: resourceKey(`environment:${name}`),
+    name,
+    ...value === undefined ? {} : { value },
+    exampleValue: '',
+    ...comment === undefined ? {} : { comment },
+  }
+}
+
+/** Create an owned complete-text file that is removable only while unchanged. */
+export function ownedTextFile(_owner: string, path: string, text: string): OwnedFileResource {
+  return {
+    kind: 'owned-file',
+    key: resourceKey(`file:${path}`),
+    document: new TextProjectFile(path, text),
+    removeOnlyWhenUnchanged: true,
+  }
+}
+
+/** Validate a config key as a string when present. */
+export function optionalString(config: Readonly<Record<string, unknown>>, key: string): string[] {
+  return config[key] === undefined || typeof config[key] === 'string' ? [] : [`${key} must be a string`]
+}
+
+/** Validate a config key as a non-empty string when required. */
+export function requiredString(config: Readonly<Record<string, unknown>>, key: string): string[] {
+  return typeof config[key] === 'string' && config[key].length > 0 ? [] : [`${key} must be a non-empty string`]
+}
+
+/** Validate a config key as an array of strings. */
+export function stringArray(config: Readonly<Record<string, unknown>>, key: string): string[] {
+  const value = config[key]
+  return Array.isArray(value) && value.every(item => typeof item === 'string')
+    ? []
+    : [`${key} must be an array of strings`]
+}
diff --git a/packages/sdk/helper/src/features/builtin/index.ts b/packages/sdk/helper/src/features/builtin/index.ts
new file mode 100644
index 0000000000..c4043e7d59
--- /dev/null
+++ b/packages/sdk/helper/src/features/builtin/index.ts
@@ -0,0 +1,363 @@
+/**
+ * Ordered builtin feature catalog: behavior entities only where project
+ * context changes the contribution, typed specs everywhere else.
+ *
+ * @module @deepseek-ai/dsh-helper/features/builtin
+ */
+
+import type { Config as ClaudeHooksConfig } from '@deepseek-ai/dsh-hooks-claude'
+import type { Config as CodexHooksConfig } from '@deepseek-ai/dsh-hooks-codex'
+import type { Config as JsonlConfig } from '@deepseek-ai/dsh-session-persistence-jsonl'
+import type { Config as SqliteConfig } from '@deepseek-ai/dsh-session-persistence-sqlite'
+import type { Config as ToolSubagentConfig } from '@deepseek-ai/dsh-tool-subagent'
+import type { Config as ToolWebConfig } from '@deepseek-ai/dsh-tool-web'
+import type { ProjectProfile } from '../../project/types.ts'
+import { defineFeatures } from '../define-feature.ts'
+import { FeatureRegistry } from '../registry.ts'
+import { AppFeature } from './app.ts'
+import { ProviderFeature } from './provider.ts'
+import { SpineFeature } from './spine.ts'
+
+/**
+ * Build and definition-check the complete builtin set for one project profile.
+ * @param profile - project context used to validate conditional contributions.
+ * @returns ordered builtin feature registry.
+ */
+export function createBuiltinRegistry(profile: ProjectProfile): FeatureRegistry {
+  return new FeatureRegistry(defineFeatures([
+    new ProviderFeature(),
+    new SpineFeature(),
+    {
+      id: 'bash',
+      summary: 'Command execution',
+      mode: 'exclusive',
+      required: true,
+      baseResources: [{ kind: 'npm-cordis-config-entry', id: 'tool-bash', package: '@deepseek-ai/dsh-tool-bash' }],
+      options: [
+        {
+          id: 'local',
+          label: 'Local executor',
+          default: true,
+          resources: [{ kind: 'npm-cordis-config-entry', id: 'bash', package: '@deepseek-ai/dsh-bash-local' }],
+        },
+        {
+          id: 'sandbox',
+          label: 'Sandboxed executor',
+          resources: [
+            { kind: 'npm-cordis-config-entry', id: 'sandbox', package: '@deepseek-ai/dsh-sandbox-local' },
+            {
+              kind: 'npm-cordis-config-entry',
+              id: 'bash',
+              package: '@deepseek-ai/dsh-bash-sandbox',
+              commentedExample: `Uncomment to allow writes under the project workspace.
+config:
+  mode: workspace-write
+  workspaceRoot: !!js process.cwd()`,
+            },
+          ],
+        },
+      ],
+    },
+    new AppFeature(),
+    {
+      id: 'persistence',
+      summary: 'Durable session storage',
+      mode: 'exclusive',
+      required: true,
+      options: [
+        {
+          id: 'jsonl',
+          label: 'JSONL files',
+          default: true,
+          resources: [{
+            kind: 'npm-cordis-config-entry',
+            id: 'session-persistence',
+            package: '@deepseek-ai/dsh-session-persistence-jsonl',
+            config: { root: './.sessions' } satisfies JsonlConfig,
+          }],
+        },
+        {
+          id: 'sqlite',
+          label: 'SQLite database',
+          resources: [{
+            kind: 'npm-cordis-config-entry',
+            id: 'session-persistence',
+            package: '@deepseek-ai/dsh-session-persistence-sqlite',
+            config: { path: './.sessions/sessions.sqlite' } satisfies SqliteConfig,
+          }],
+        },
+      ],
+    },
+    {
+      id: 'hmr',
+      summary: 'Hot-module reload',
+      mode: 'single',
+      options: [{
+        id: 'default',
+        label: 'Cordis HMR',
+        default: true,
+        resources: [{ kind: 'npm-cordis-config-entry', id: 'hmr', package: '@cordisjs/plugin-hmr' }],
+      }],
+    },
+    {
+      id: 'fs',
+      summary: 'Read, write, and edit local files',
+      mode: 'single',
+      options: [{
+        id: 'local',
+        label: 'Local filesystem',
+        default: true,
+        resources: [
+          { kind: 'npm-cordis-config-entry', id: 'fs-local', package: '@deepseek-ai/dsh-fs-local' },
+          { kind: 'npm-cordis-config-entry', id: 'fs-policy', package: '@deepseek-ai/dsh-fs-policy' },
+          { kind: 'npm-cordis-config-entry', id: 'tool-fs', package: '@deepseek-ai/dsh-tool-fs' },
+        ],
+      }],
+    },
+    {
+      id: 'todo',
+      summary: 'Model-facing task tracking',
+      mode: 'single',
+      options: [{
+        id: 'default',
+        label: 'todo_write tool',
+        default: true,
+        resources: [{ kind: 'npm-cordis-config-entry', id: 'tool-todo', package: '@deepseek-ai/dsh-tool-todo' }],
+      }],
+    },
+    {
+      id: 'skill',
+      summary: 'Local skill discovery',
+      mode: 'single',
+      options: [{
+        id: 'default',
+        label: 'Local skills and skill tool',
+        default: true,
+        resources: [
+          { kind: 'npm-cordis-config-entry', id: 'skill', package: '@deepseek-ai/dsh-skill' },
+          { kind: 'npm-cordis-config-entry', id: 'skill-local', package: '@deepseek-ai/dsh-skill-local' },
+          { kind: 'npm-cordis-config-entry', id: 'tool-skill', package: '@deepseek-ai/dsh-tool-skill' },
+        ],
+      }],
+    },
+    {
+      id: 'web',
+      summary: 'Web search and fetch tools',
+      mode: 'exclusive',
+      suggests: ['timeout-policy'],
+      baseResources: [
+        { kind: 'npm-cordis-config-entry', id: 'web', package: '@deepseek-ai/dsh-web' },
+        { kind: 'npm-cordis-config-entry', id: 'web-fetch-local', package: '@deepseek-ai/dsh-web-fetch-local' },
+      ],
+      options: [
+        {
+          id: 'deepseek',
+          label: 'DeepSeek search',
+          default: true,
+          markers: [{ id: 'web-search-deepseek', name: '@deepseek-ai/dsh-web-search-deepseek' }],
+          resources: [
+            { kind: 'npm-cordis-config-entry', id: 'web-search-deepseek', package: '@deepseek-ai/dsh-web-search-deepseek' },
+            { kind: 'npm-cordis-config-entry', id: 'tool-web', package: '@deepseek-ai/dsh-tool-web' },
+          ],
+        },
+        {
+          id: 'exa',
+          label: 'Exa search',
+          secrets: [{ id: 'apiKey', environment: 'EXA_API_KEY', message: 'Exa API key', required: true }],
+          markers: [{ id: 'web-search-exa', name: '@deepseek-ai/dsh-web-search-exa' }],
+          resources: [
+            { kind: 'npm-cordis-config-entry', id: 'web-search-exa', package: '@deepseek-ai/dsh-web-search-exa' },
+            { kind: 'npm-cordis-config-entry', id: 'tool-web', package: '@deepseek-ai/dsh-tool-web' },
+          ],
+        },
+        {
+          id: 'perplexity',
+          label: 'Perplexity search',
+          secrets: [{
+            id: 'apiKey',
+            environment: 'PERPLEXITY_API_KEY',
+            message: 'Perplexity API key',
+            required: true,
+          }],
+          markers: [{ id: 'web-search-perplexity', name: '@deepseek-ai/dsh-web-search-perplexity' }],
+          resources: [
+            {
+              kind: 'npm-cordis-config-entry',
+              id: 'web-search-perplexity',
+              package: '@deepseek-ai/dsh-web-search-perplexity',
+            },
+            { kind: 'npm-cordis-config-entry', id: 'tool-web', package: '@deepseek-ai/dsh-tool-web' },
+          ],
+        },
+        {
+          id: 'fetch-only',
+          label: 'Fetch only',
+          markers: [{ id: 'tool-web', name: '@deepseek-ai/dsh-tool-web', config: { search: false } }],
+          resources: [{
+            kind: 'npm-cordis-config-entry',
+            id: 'tool-web',
+            package: '@deepseek-ai/dsh-tool-web',
+            config: { search: false } satisfies ToolWebConfig,
+          }],
+        },
+      ],
+    },
+    {
+      id: 'subagent',
+      summary: 'Delegate work to child agents',
+      mode: 'multiple',
+      baseResources: [{ kind: 'npm-cordis-config-entry', id: 'subagent', package: '@deepseek-ai/dsh-subagent' }],
+      options: [
+        {
+          id: 'spawn',
+          label: 'Fresh child agent',
+          default: true,
+          resources: [
+            { kind: 'npm-cordis-config-entry', id: 'subagent-spawn', package: '@deepseek-ai/dsh-subagent-spawn' },
+            {
+              kind: 'npm-cordis-config-entry',
+              id: 'tool-subagent',
+              package: '@deepseek-ai/dsh-tool-subagent',
+              config: { provider: 'spawn' } satisfies ToolSubagentConfig,
+            },
+          ],
+        },
+        {
+          id: 'fork',
+          label: 'Fork parent history',
+          resources: [
+            { kind: 'npm-cordis-config-entry', id: 'subagent-fork', package: '@deepseek-ai/dsh-subagent-fork' },
+            {
+              kind: 'npm-cordis-config-entry',
+              id: 'tool-subagent-fork',
+              package: '@deepseek-ai/dsh-tool-subagent',
+              config: { provider: 'fork', toolName: 'subagent_fork' } satisfies ToolSubagentConfig,
+            },
+          ],
+        },
+      ],
+    },
+    {
+      id: 'workflow',
+      summary: 'Scripted multi-agent workflows',
+      mode: 'single',
+      options: [{
+        id: 'workerthread',
+        label: 'Worker thread engine',
+        default: true,
+        requires: [{ id: 'subagent', options: ['spawn'] }],
+        resources: [
+          {
+            kind: 'npm-cordis-config-entry',
+            id: 'workflow-workerthread',
+            package: '@deepseek-ai/dsh-workflow-workerthread',
+          },
+          { kind: 'npm-cordis-config-entry', id: 'tool-workflow', package: '@deepseek-ai/dsh-tool-workflow' },
+        ],
+      }],
+    },
+    {
+      id: 'compact',
+      summary: 'Automatic context compaction',
+      mode: 'single',
+      options: [{
+        id: 'basic',
+        label: 'Basic compaction',
+        default: true,
+        resources: [
+          {
+            kind: 'npm-cordis-config-entry',
+            id: 'token-meter',
+            package: '@deepseek-ai/dsh-token-meter',
+          },
+          {
+            kind: 'npm-cordis-config-entry',
+            id: 'compact-basic',
+            package: '@deepseek-ai/dsh-compact-basic',
+          },
+        ],
+      }],
+    },
+    {
+      id: 'hooks',
+      summary: 'Run Claude Code or Codex hooks',
+      mode: 'multiple',
+      requires: [{ id: 'bash' }],
+      options: [
+        {
+          id: 'claude',
+          label: 'Claude Code hooks',
+          default: true,
+          resources: [
+            {
+              kind: 'npm-cordis-config-entry',
+              id: 'hooks-claude',
+              package: '@deepseek-ai/dsh-hooks-claude',
+              config: { configPath: './hooks.json' } satisfies ClaudeHooksConfig,
+            },
+            { kind: 'owned-file', path: 'hooks.json', text: '{}' },
+          ],
+        },
+        {
+          id: 'codex',
+          label: 'Codex hooks',
+          resources: [
+            {
+              kind: 'npm-cordis-config-entry',
+              id: 'hooks-codex',
+              package: '@deepseek-ai/dsh-hooks-codex',
+              config: { configPath: './codex-hooks.json' } satisfies CodexHooksConfig,
+            },
+            { kind: 'owned-file', path: 'codex-hooks.json', text: '{}' },
+          ],
+        },
+      ],
+    },
+    {
+      id: 'guard',
+      summary: 'Loop-hygiene reminders',
+      mode: 'single',
+      options: [{
+        id: 'repeat-tool',
+        label: 'Repeat-tool reminders',
+        default: true,
+        resources: [{
+          kind: 'npm-cordis-config-entry',
+          id: 'repeat-tool-guard',
+          package: '@deepseek-ai/dsh-repeat-tool-guard',
+        }],
+      }],
+    },
+    {
+      id: 'timeout-policy',
+      summary: 'Tool timeout policy',
+      mode: 'single',
+      options: [{
+        id: 'default',
+        label: 'Timeout policy',
+        default: true,
+        resources: [{
+          kind: 'npm-cordis-config-entry',
+          id: 'timeout-policy',
+          package: '@deepseek-ai/dsh-timeout-policy',
+        }],
+      }],
+    },
+    {
+      id: 'ask-user',
+      summary: 'Ask the user from the model loop',
+      mode: 'single',
+      supportedInterfaces: ['acp', 'stdio'],
+      options: [{
+        id: 'default',
+        label: 'ask_user_question tool',
+        default: true,
+        resources: [{
+          kind: 'npm-cordis-config-entry',
+          id: 'tool-ask-user',
+          package: '@deepseek-ai/dsh-tool-ask-user',
+        }],
+      }],
+    },
+  ]), profile)
+}
diff --git a/packages/sdk/helper/src/features/builtin/provider.ts b/packages/sdk/helper/src/features/builtin/provider.ts
new file mode 100644
index 0000000000..227d80088c
--- /dev/null
+++ b/packages/sdk/helper/src/features/builtin/provider.ts
@@ -0,0 +1,111 @@
+/**
+ * Required hand-rolled DeepSeek and custom pi-ai provider behavior.
+ *
+ * @module @deepseek-ai/dsh-helper/features/builtin/provider
+ */
+
+import { JsExpression } from '../../documents/cordis-yaml-file.ts'
+import { featureId } from '../../ids.ts'
+import type { FeatureSelection, ProjectProfile } from '../../project/types.ts'
+import {
+  FeatureOption,
+  ExclusiveOptionFeature,
+  type FeatureProjectView,
+} from '../feature.ts'
+import { ProjectContribution } from '../resources.ts'
+import { npmCordisConfigEntry, environment } from './helpers.ts'
+
+const ID = featureId('provider')
+const DEFAULT_MODEL = 'deepseek-v4-flash'
+const API_KEY_COMMENT = 'Required before start; an empty value makes provider startup fail.'
+
+class DeepSeekOption extends FeatureOption {
+  override readonly id = 'deepseek'
+  override readonly label = 'DeepSeek'
+  override readonly secrets = [{
+    id: 'apiKey',
+    environment: 'DEEPSEEK_API_KEY',
+    message: 'DeepSeek API key',
+    required: true,
+  }]
+
+  override contribution(_profile: ProjectProfile, secrets: Readonly<Record<string, string>>): ProjectContribution {
+    return new ProjectContribution([
+      ...npmCordisConfigEntry(ID, {
+        id: 'llm-deepseek',
+        name: '@deepseek-ai/dsh-llm-deepseek',
+        config: { apiKey: new JsExpression('process.env.DEEPSEEK_API_KEY') },
+      }, ['apiKey', 'baseURL', 'models']),
+      environment(ID, 'DEEPSEEK_API_KEY', secrets.apiKey, API_KEY_COMMENT),
+    ])
+  }
+}
+
+class CustomOption extends FeatureOption {
+  override readonly id = 'custom'
+  override readonly label = 'Custom endpoint (pi-ai)'
+  override readonly secrets = [{
+    id: 'apiKey',
+    environment: 'DEEPSEEK_API_KEY',
+    message: 'Custom provider API key',
+    required: true,
+  }]
+  override readonly inputs = [{
+    id: 'baseURL',
+    message: 'Custom provider base URL',
+  }]
+
+  override contribution(_profile: ProjectProfile, secrets: Readonly<Record<string, string>>): ProjectContribution {
+    return new ProjectContribution([
+      ...npmCordisConfigEntry(ID, {
+        id: 'llm-pi-ai',
+        name: '@deepseek-ai/dsh-llm-pi-ai',
+        config: { apiKey: new JsExpression('process.env.DEEPSEEK_API_KEY') },
+      }, ['apiKey', 'baseURL', 'models']),
+      environment(ID, 'DEEPSEEK_API_KEY', secrets.apiKey, API_KEY_COMMENT),
+    ])
+  }
+}
+
+/** Required provider feature with DeepSeek and custom pi-ai options. */
+export class ProviderFeature extends ExclusiveOptionFeature {
+  override readonly id = ID
+  override readonly summary = 'Model provider'
+  override readonly required = true
+  override readonly options = [new DeepSeekOption(), new CustomOption()]
+
+  /** Prefer the hand-rolled adapter and its public endpoint defaults. */
+  override defaultOptions(): readonly string[] {
+    return ['deepseek']
+  }
+
+  /** Recover literal endpoint overrides from either provider entry. */
+  override readSelection(project: FeatureProjectView, selection: FeatureSelection): FeatureSelection {
+    const base = super.readSelection(project, selection)
+    const entry = project.cordisConfigEntries().find(item => item.id === 'llm-deepseek' || item.id === 'llm-pi-ai')
+    const baseURL = entry?.config?.baseURL
+    return typeof baseURL === 'string' ? { ...base, values: { baseURL } } : base
+  }
+
+  /** Apply explicit endpoint/model overrides while omitting provider defaults. */
+  override contribution(selection: FeatureSelection, profile: ProjectProfile): ProjectContribution {
+    const contribution = super.contribution(selection, profile)
+    const baseURL = selection.values?.baseURL
+    if (baseURL !== undefined && typeof baseURL !== 'string') throw new Error('provider baseURL must be a string')
+    return new ProjectContribution(contribution.resources.map((resource) => {
+      if (resource.kind !== 'cordis-config-entry' || (resource.entry.id !== 'llm-deepseek'
+        && resource.entry.id !== 'llm-pi-ai')) return resource
+      return {
+        ...resource,
+        entry: {
+          ...resource.entry,
+          config: {
+            ...resource.entry.config,
+            ...baseURL ? { baseURL } : {},
+            ...profile.runtime.model === DEFAULT_MODEL ? {} : { models: [profile.runtime.model] },
+          },
+        },
+      }
+    }))
+  }
+}
diff --git a/packages/sdk/helper/src/features/builtin/spine.ts b/packages/sdk/helper/src/features/builtin/spine.ts
new file mode 100644
index 0000000000..caf066865e
--- /dev/null
+++ b/packages/sdk/helper/src/features/builtin/spine.ts
@@ -0,0 +1,55 @@
+/**
+ * Required agent-spine feature expressed as top-level Cordis config entries.
+ *
+ * @module @deepseek-ai/dsh-helper/features/builtin/spine
+ */
+
+import { featureId } from '../../ids.ts'
+import type { ProjectProfile } from '../../project/types.ts'
+import { loadHelperTemplate } from '../../templates/template-assets.ts'
+import { FeatureOption, FixedFeature } from '../feature.ts'
+import { ProjectContribution } from '../resources.ts'
+import { npmCordisConfigEntry, requiredString } from './helpers.ts'
+
+const ID = featureId('spine')
+const PERSONA = loadHelperTemplate<Record<string, never>>('persona.txt.tpl').render({}).trimEnd()
+
+function emptyAgentsDiagnostics(config: Readonly<Record<string, unknown>>): string[] {
+  const agents = config.agents
+  if (!Array.isArray(agents)) return ['agents must be an array']
+  return agents.length === 0 ? [] : ['agents must be empty']
+}
+
+class SpineOption extends FeatureOption {
+  override readonly id = 'default'
+  override readonly label = 'Default agent spine'
+
+  override contribution(_profile: ProjectProfile): ProjectContribution {
+    return new ProjectContribution([
+      ...npmCordisConfigEntry(ID, { id: 'timer', name: '@cordisjs/plugin-timer' }),
+      ...npmCordisConfigEntry(ID, { id: 'llm', name: '@deepseek-ai/dsh-llm' }),
+      ...npmCordisConfigEntry(ID, { id: 'session', name: '@deepseek-ai/dsh-session' }),
+      ...npmCordisConfigEntry(ID, {
+        id: 'system-prompt',
+        name: '@deepseek-ai/dsh-system-prompt',
+        config: { persona: PERSONA },
+      }, ['persona'], config => requiredString(config, 'persona')),
+      ...npmCordisConfigEntry(ID, { id: 'tools', name: '@deepseek-ai/dsh-tools' }, []),
+      ...npmCordisConfigEntry(ID, { id: 'agent', name: '@deepseek-ai/dsh-agent' }),
+      ...npmCordisConfigEntry(ID, { id: 'invariants', name: '@deepseek-ai/dsh-invariants' }),
+      ...npmCordisConfigEntry(ID, {
+        id: 'agent-loop',
+        name: '@deepseek-ai/dsh-agent-loop',
+        config: { agents: [] },
+      }, ['agents'], emptyAgentsDiagnostics),
+    ])
+  }
+}
+
+/** Required providerless agent spine without a composition bundle entry. */
+export class SpineFeature extends FixedFeature {
+  override readonly id = ID
+  override readonly summary = 'Agent runtime spine'
+  override readonly required = true
+  override readonly options = [new SpineOption()]
+}
diff --git a/packages/sdk/helper/src/features/define-feature.ts b/packages/sdk/helper/src/features/define-feature.ts
new file mode 100644
index 0000000000..7c90a2853a
--- /dev/null
+++ b/packages/sdk/helper/src/features/define-feature.ts
@@ -0,0 +1,286 @@
+/**
+ * Typed declarative definitions for features whose behavior is entirely
+ * the shared resource lifecycle.
+ *
+ * @module @deepseek-ai/dsh-helper/features/define-feature
+ */
+
+import type { CordisConfigEntry } from '../documents/cordis-yaml-file.ts'
+import { TextProjectFile } from '../documents/project-file.ts'
+import { featureId, resourceKey, type FeatureId } from '../ids.ts'
+import type { FeatureSelection, ProjectProfile, RunInterface } from '../project/types.ts'
+import {
+  Feature,
+  FeatureOption,
+  type FeatureRequirement,
+  type FeatureSecret,
+} from './feature.ts'
+import { ProjectContribution, type ProjectResource } from './resources.ts'
+
+/** Static NPM dependency in a declarative feature. */
+interface NpmDependencySpec {
+  kind: 'npm-dependency'
+  name: string
+  section?: 'dependencies' | 'devDependencies'
+}
+
+/** Bare-package Cordis config entry that also contributes its NPM dependency. */
+interface NpmCordisConfigEntrySpec {
+  kind: 'npm-cordis-config-entry'
+  id: string
+  package: string
+  config?: Readonly<Record<string, unknown>>
+  ownedConfigKeys?: readonly string[]
+  commentedExample?: string
+}
+
+/** Relative or absolute file Cordis config entry with no NPM dependency. */
+interface FileCordisConfigEntrySpec {
+  kind: 'file-cordis-config-entry'
+  id: string
+  path: string
+  config?: Readonly<Record<string, unknown>>
+  ownedConfigKeys?: readonly string[]
+  commentedExample?: string
+}
+
+/** Static complete file owned by one feature option. */
+interface OwnedFileSpec {
+  kind: 'owned-file'
+  path: string
+  text: string
+  removeOnlyWhenUnchanged?: boolean
+}
+
+/** Resource forms that require no feature-specific imperative code. */
+type FeatureResourceSpec =
+  | NpmDependencySpec
+  | NpmCordisConfigEntrySpec
+  | FileCordisConfigEntrySpec
+  | OwnedFileSpec
+
+/** Cordis config entry identity and optional plugin-config subset that identifies an option. */
+interface FeatureOptionMarkerSpec {
+  id: string
+  name: string
+  config?: Readonly<Record<string, unknown>>
+}
+
+/** Declarative requirement converted to branded domain identity at the boundary. */
+interface FeatureRequirementSpec {
+  id: string
+  options?: readonly string[]
+}
+
+/** One static option inside a typed feature definition. */
+interface FeatureOptionSpec {
+  id: string
+  label: string
+  default?: boolean
+  resources: readonly FeatureResourceSpec[]
+  secrets?: readonly FeatureSecret[]
+  markers?: readonly FeatureOptionMarkerSpec[]
+  requires?: readonly FeatureRequirementSpec[]
+}
+
+/** Complete declarative feature definition. */
+export interface FeatureSpec {
+  id: string
+  summary: string
+  mode: 'single' | 'exclusive' | 'multiple'
+  options: readonly FeatureOptionSpec[]
+  baseResources?: readonly FeatureResourceSpec[]
+  required?: boolean
+  requires?: readonly FeatureRequirementSpec[]
+  suggests?: readonly string[]
+  supportedInterfaces?: readonly RunInterface[]
+}
+
+function sameShape(expected: unknown, actual: unknown): boolean {
+  if (expected === null || actual === null) return expected === actual
+  if (Array.isArray(expected)) {
+    return Array.isArray(actual) && (expected.length === 0 || actual.every(item => sameShape(expected[0], item)))
+  }
+  if (typeof expected !== 'object') return typeof expected === typeof actual
+  if (typeof actual !== 'object' || Array.isArray(actual)) return false
+  return Object.entries(expected as Record<string, unknown>).every(
+    ([key, value]) => sameShape(value, (actual as Record<string, unknown>)[key]),
+  )
+}
+
+function configDiagnostics(
+  expected: Readonly<Record<string, unknown>> | undefined,
+): ((config: Readonly<Record<string, unknown>>) => readonly string[]) | undefined {
+  if (!expected || Object.keys(expected).length === 0) return undefined
+  return config => Object.entries(expected).flatMap(([key, value]) => sameShape(value, config[key])
+    ? []
+    : [`${key} has an incompatible value shape`])
+}
+
+function resourcesFromSpec(spec: FeatureResourceSpec): ProjectResource[] {
+  switch (spec.kind) {
+    case 'npm-dependency':
+      return [{
+        kind: 'npm-dependency',
+        key: resourceKey(`npm-dependency:${spec.name}`),
+        name: spec.name,
+        section: spec.section ?? 'dependencies',
+      }]
+    case 'npm-cordis-config-entry':
+    case 'file-cordis-config-entry': {
+      const config = spec.config ? { ...spec.config } : undefined
+      const validateConfig = configDiagnostics(config)
+      const name = spec.kind === 'npm-cordis-config-entry' ? spec.package : spec.path
+      return [
+        ...spec.kind === 'npm-cordis-config-entry'
+          ? [{
+            kind: 'npm-dependency' as const,
+            key: resourceKey(`npm-dependency:${spec.package}`),
+            name: spec.package,
+            section: 'dependencies' as const,
+          }]
+          : [],
+        {
+          kind: 'cordis-config-entry',
+          key: resourceKey(`cordis-config-entry:${spec.id}`),
+          entry: {
+            id: spec.id,
+            name,
+            ...config ? { config } : {},
+          },
+          ownedConfigKeys: spec.ownedConfigKeys ?? Object.keys(config ?? {}),
+          ...spec.commentedExample ? { commentedExample: spec.commentedExample } : {},
+          ...validateConfig ? { validateConfig } : {},
+        },
+      ]
+    }
+    case 'owned-file':
+      return [{
+        kind: 'owned-file',
+        key: resourceKey(`file:${spec.path}`),
+        document: new TextProjectFile(spec.path, spec.text),
+        removeOnlyWhenUnchanged: spec.removeOnlyWhenUnchanged ?? true,
+      }]
+  }
+}
+
+function isSubset(expected: Readonly<Record<string, unknown>>, actual: Readonly<Record<string, unknown>>): boolean {
+  return Object.entries(expected).every(([key, value]) => Object.is(actual[key], value))
+}
+
+class DefinedFeatureOption extends FeatureOption {
+  override readonly id: string
+  override readonly label: string
+  override readonly secrets: readonly FeatureSecret[]
+  private readonly spec: FeatureOptionSpec
+
+  constructor(spec: FeatureOptionSpec) {
+    super()
+    this.spec = spec
+    this.id = spec.id
+    this.label = spec.label
+    this.secrets = spec.secrets ?? []
+  }
+
+  override contribution(_profile: ProjectProfile, secrets: Readonly<Record<string, string>>): ProjectContribution {
+    return new ProjectContribution([
+      ...this.spec.resources.flatMap(resourcesFromSpec),
+      ...this.secrets.map(secret => ({
+        kind: 'environment' as const,
+        key: resourceKey(`environment:${secret.environment}`),
+        name: secret.environment,
+        ...secrets[secret.id] === undefined ? {} : { value: secrets[secret.id] },
+        exampleValue: '',
+      })),
+    ])
+  }
+
+  override markerConfigEntries(): readonly Pick<CordisConfigEntry, 'id' | 'name'>[] {
+    const markers = this.spec.markers ?? this.spec.resources.flatMap((resource) => {
+      switch (resource.kind) {
+        case 'npm-cordis-config-entry': return [{ id: resource.id, name: resource.package }]
+        case 'file-cordis-config-entry': return [{ id: resource.id, name: resource.path }]
+        default: return []
+      }
+    })
+    return markers.map(marker => ({ id: marker.id, name: marker.name }))
+  }
+
+  override matchesConfigEntries(entries: readonly CordisConfigEntry[]): boolean {
+    const markers = this.spec.markers
+    if (!markers) return this.markerConfigEntries().some(marker => entries.some(
+      entry => entry.id === marker.id && entry.name === marker.name,
+    ))
+    return markers.some(marker => entries.some(entry => entry.id === marker.id
+      && entry.name === marker.name
+      && (!marker.config || isSubset(marker.config, entry.config ?? {}))))
+  }
+}
+
+/** Feature entity backed by a typed static definition. */
+class DefinedFeature extends Feature {
+  override readonly id: FeatureId
+  override readonly summary: string
+  override readonly mode: FeatureSpec['mode']
+  override readonly options: readonly FeatureOption[]
+  override readonly required: boolean
+  override readonly requires: readonly FeatureId[]
+  override readonly suggests: readonly FeatureId[]
+  override readonly supportedInterfaces: readonly RunInterface[]
+  private readonly spec: FeatureSpec
+
+  /** Validate and materialize one declarative definition. */
+  constructor(spec: FeatureSpec) {
+    super()
+    this.spec = spec
+    this.id = featureId(spec.id)
+    this.summary = spec.summary
+    this.mode = spec.mode
+    this.options = spec.options.map(option => new DefinedFeatureOption(option))
+    const defaultCount = spec.options.filter(option => option.default).length
+    if (spec.mode === 'single' && (spec.options.length !== 1 || defaultCount !== 1)) {
+      throw new Error(`single feature ${spec.id} requires one default option`)
+    }
+    if (spec.mode === 'exclusive' && defaultCount !== 1) {
+      throw new Error(`exclusive feature ${spec.id} requires exactly one default option`)
+    }
+    if (spec.mode === 'multiple' && defaultCount === 0) {
+      throw new Error(`multiple feature ${spec.id} requires at least one default option`)
+    }
+    this.required = spec.required ?? false
+    this.requires = (spec.requires ?? []).map(requirement => featureId(requirement.id))
+    this.suggests = (spec.suggests ?? []).map(featureId)
+    this.supportedInterfaces = spec.supportedInterfaces ?? ['acp', 'stdio', 'embed']
+  }
+
+  override defaultOptions(): readonly string[] {
+    return this.spec.options.filter(option => option.default).map(option => option.id)
+  }
+
+  override baseContribution(): ProjectContribution {
+    return new ProjectContribution((this.spec.baseResources ?? []).flatMap(resourcesFromSpec))
+  }
+
+  override requirements(selection: FeatureSelection): readonly FeatureRequirement[] {
+    const selected = new Set(selection.options)
+    return [
+      ...(this.spec.requires ?? []),
+      ...this.spec.options.filter(option => selected.has(option.id)).flatMap(option => option.requires ?? []),
+    ].map(requirement => ({
+      id: featureId(requirement.id),
+      ...requirement.options ? { options: requirement.options } : {},
+    }))
+  }
+}
+
+/** Construct the shared lifecycle entity from a typed declarative definition. */
+export function defineFeature(spec: FeatureSpec): Feature {
+  return new DefinedFeature(spec)
+}
+
+/** Materialize one ordered catalog containing static specs and behavior entities. */
+export function defineFeatures(definitions: readonly (Feature | FeatureSpec)[]): Feature[] {
+  return definitions.map(definition => definition instanceof Feature
+    ? definition
+    : defineFeature(definition))
+}
diff --git a/packages/sdk/helper/src/features/feature-configurator.ts b/packages/sdk/helper/src/features/feature-configurator.ts
new file mode 100644
index 0000000000..47cc011d16
--- /dev/null
+++ b/packages/sdk/helper/src/features/feature-configurator.ts
@@ -0,0 +1,104 @@
+/**
+ * Shared option and secret question flow for create and config.
+ *
+ * @module @deepseek-ai/dsh-helper/features/feature-configurator
+ */
+
+import type { Feature } from './feature.ts'
+import type { FeatureSelection, ProjectProfile } from '../project/types.ts'
+import type { PromptPort } from '../questions/prompt-port.ts'
+import { requireAnswer } from '../questions/prompt-port.ts'
+import { MultiSelectQuestion, SecretQuestion, SelectQuestion, TextQuestion } from '../questions/question.ts'
+
+/** Resolve one feature selection without knowing which workflow requested it. */
+export class FeatureConfigurator {
+  private readonly port: PromptPort
+
+  /** Bind the configurator to the shared prompt boundary. */
+  constructor(port: PromptPort) {
+    this.port = port
+  }
+
+  /**
+   * Ask option and input questions, preserving current secrets on empty input.
+   * @param feature - feature whose options and inputs are collected.
+   * @param profile - target project context.
+   * @param current - currently installed selection, when configuring.
+   * @param prefilledOptions - options already chosen by a tree picker.
+   * @param prefilledSecrets - non-interactive secret values supplied by creation.
+   * @returns normalized selection with captured values and secrets.
+   */
+  async configure(
+    feature: Feature,
+    profile: ProjectProfile,
+    current?: FeatureSelection,
+    prefilledOptions?: readonly string[],
+    prefilledSecrets: Readonly<Record<string, string>> = {},
+  ): Promise<FeatureSelection> {
+    let options: readonly string[]
+    switch (feature.mode) {
+      case 'single':
+        options = feature.defaultOptions(profile)
+        break
+      case 'exclusive': {
+        const initialValue = current?.options[0] ?? feature.defaultOptions(profile)[0]
+        if (initialValue === undefined) throw new Error(`feature ${feature.id} has no default option`)
+        const question = new SelectQuestion({
+          id: `${feature.id}.option`,
+          message: `Choose ${feature.summary.toLowerCase()}`,
+          options: feature.options.map(option => ({ value: option.id, label: option.label })),
+          initialValue,
+        })
+        const prefilled = prefilledOptions?.[0]
+        options = [requireAnswer(await question.resolve(this.port, prefilled))]
+        break
+      }
+      case 'multiple': {
+        const question = new MultiSelectQuestion({
+          id: `${feature.id}.options`,
+          message: `Choose ${feature.summary.toLowerCase()}`,
+          options: feature.options.map(option => ({ value: option.id, label: option.label })),
+          initialValues: current?.options ?? feature.defaultOptions(profile),
+          required: true,
+        })
+        options = requireAnswer(await question.resolve(this.port, prefilledOptions))
+        break
+      }
+    }
+    const selected: FeatureSelection = {
+      id: feature.id,
+      options,
+    }
+    const values: Record<string, string> = {}
+    for (const input of feature.valueInputs(selected, profile)) {
+      const existing = current?.values?.[input.id]
+      if (existing !== undefined && typeof existing !== 'string') {
+        throw new Error(`${feature.id}.${input.id} current value must be a string`)
+      }
+      const question = new TextQuestion({
+        id: `${feature.id}.${input.id}`,
+        message: input.message,
+        ...existing === undefined ? {} : { initialValue: existing },
+        validate: value => value.trim().length === 0 ? 'A value is required' : undefined,
+      })
+      values[input.id] = requireAnswer(await question.resolve(this.port))
+    }
+    const base: FeatureSelection = Object.keys(values).length === 0
+      ? selected
+      : { ...selected, values }
+    const secrets = { ...current?.secrets }
+    for (const secret of feature.secrets(base, profile)) {
+      const existing = secrets[secret.id]
+      const question = new SecretQuestion({
+        id: `${feature.id}.${secret.id}`,
+        message: existing === undefined ? secret.message : `${secret.message} (leave empty to keep current)`,
+        validate: value => secret.required && existing === undefined && value.length === 0
+          ? 'A value is required'
+          : undefined,
+      })
+      const answer = requireAnswer(await question.resolve(this.port, prefilledSecrets[secret.id]))
+      if (answer.length > 0) secrets[secret.id] = answer
+    }
+    return Object.keys(secrets).length === 0 ? base : { ...base, secrets }
+  }
+}
diff --git a/packages/sdk/helper/src/features/feature.ts b/packages/sdk/helper/src/features/feature.ts
new file mode 100644
index 0000000000..1335d8e29b
--- /dev/null
+++ b/packages/sdk/helper/src/features/feature.ts
@@ -0,0 +1,345 @@
+/**
+ * Stateful builtin feature and option domain objects.
+ *
+ * @module @deepseek-ai/dsh-helper/features/feature
+ */
+
+import type { CordisConfigEntry } from '../documents/cordis-yaml-file.ts'
+import type { PackageManifest } from '../documents/package-json-file.ts'
+import type { FeatureId } from '../ids.ts'
+import type { FeatureSelection, ProjectProfile, RunInterface } from '../project/types.ts'
+import { ProjectContribution, type CordisConfigEntryResource, type ProjectResource } from './resources.ts'
+
+/** Read-only project surface used by feature inspection. */
+export interface FeatureProjectView {
+  readonly profile: ProjectProfile
+  cordisConfigEntries(): readonly CordisConfigEntry[]
+  packageManifest(): Readonly<PackageManifest>
+  hasDocument(path: string): boolean
+  readEnvironment(path: '.env' | '.env.example', name: string): string | undefined
+}
+
+/** Installation state visible to create/config workflows. */
+type FeatureInstallationState = 'absent' | 'enabled' | 'disabled' | 'inconsistent'
+
+/** Result of round-tripping one feature from a project snapshot. */
+export interface FeatureInstallation {
+  id: FeatureId
+  state: FeatureInstallationState
+  options: readonly string[]
+  selection?: FeatureSelection
+  diagnostics: readonly string[]
+}
+
+/** One final-state requirement on another builtin feature. */
+export interface FeatureRequirement {
+  id: FeatureId
+  options?: readonly string[]
+}
+
+/** One secret captured into an environment binding rather than Cordis plugin config. */
+export interface FeatureSecret {
+  id: string
+  environment: string
+  message: string
+  required: boolean
+}
+
+/** One visible string value requested only by options that own it. */
+export interface FeatureValueInput {
+  id: string
+  message: string
+}
+
+/** One selectable behavior option owned by a feature. */
+export abstract class FeatureOption {
+  abstract readonly id: string
+  abstract readonly label: string
+  readonly secrets: readonly FeatureSecret[] = []
+  readonly inputs: readonly FeatureValueInput[] = []
+
+  /** Contribute this option's project resources. */
+  abstract contribution(profile: ProjectProfile, secrets: Readonly<Record<string, string>>): ProjectContribution
+
+  /** Every Cordis config entry package owned by this option during inspection. */
+  ownedConfigEntries(profile: ProjectProfile): readonly Pick<CordisConfigEntry, 'id' | 'name'>[] {
+    return this.contribution(profile, {}).resources
+      .filter((resource): resource is CordisConfigEntryResource => resource.kind === 'cordis-config-entry')
+      .map(resource => ({ id: resource.entry.id, name: resource.entry.name }))
+  }
+
+  /** Cordis config entry identities that distinguish this option during inspection. */
+  markerConfigEntries(profile: ProjectProfile): readonly Pick<CordisConfigEntry, 'id' | 'name'>[] {
+    return this.ownedConfigEntries(profile)
+  }
+
+  /** Whether current owned Cordis config entries identify this option. */
+  matchesConfigEntries(entries: readonly CordisConfigEntry[], profile: ProjectProfile): boolean {
+    return this.markerConfigEntries(profile).some(marker => entries.some(
+      entry => entry.id === marker.id && entry.name === marker.name,
+    ))
+  }
+}
+
+/** How a feature's options compose. */
+export type FeatureOptionMode = 'single' | 'exclusive' | 'multiple'
+
+function packageNames(resources: readonly ProjectResource[]): Set<string> {
+  return new Set(resources
+    .filter((resource): resource is CordisConfigEntryResource => resource.kind === 'cordis-config-entry')
+    .map(resource => resource.entry.name))
+}
+
+function configDiagnostics(resource: CordisConfigEntryResource, entry: CordisConfigEntry): string[] {
+  /* v8 ignore next -- entries without validators have no diagnostics to compute */
+  if (!resource.validateConfig) return []
+  return [...resource.validateConfig(entry.config ?? {})].map(message => `${entry.id}: ${message}`)
+}
+
+/** A behavior-owning builtin feature with shallow option composition. */
+export abstract class Feature {
+  /** Stable registry identity. */
+  abstract readonly id: FeatureId
+  /** User-facing feature summary. */
+  abstract readonly summary: string
+  /** Option-selection rule. */
+  abstract readonly mode: FeatureOptionMode
+  /** Available behavior options. */
+  abstract readonly options: readonly FeatureOption[]
+  /** Whether every valid project must enable this feature. */
+  readonly required: boolean = false
+  /** Unconditional feature requirements. */
+  readonly requires: readonly FeatureId[] = []
+  /** Features recommended during creation. */
+  readonly suggests: readonly FeatureId[] = []
+  /** Front doors under which this feature is meaningful. */
+  readonly supportedInterfaces: readonly RunInterface[] = ['acp', 'stdio', 'embed']
+
+  /**
+   * Options selected when installation has no override.
+   * @param profile - project context controlling applicable defaults.
+   * @returns selected option ids.
+   */
+  abstract defaultOptions(profile: ProjectProfile): readonly string[]
+
+  /**
+   * Shared resources present for every installed option set.
+   * @param _profile - project context available to behavior features.
+   * @returns shared project contribution.
+   */
+  baseContribution(_profile: ProjectProfile): ProjectContribution {
+    return new ProjectContribution([])
+  }
+
+  /**
+   * Additional final-state requirements depending on selected options.
+   * @param _selection - normalized feature selection.
+   * @returns required features and option constraints.
+   */
+  requirements(_selection: FeatureSelection): readonly FeatureRequirement[] {
+    return this.requires.map(id => ({ id }))
+  }
+
+  /**
+   * Whether the feature may be selected for this project front door.
+   * @param profile - project context to check.
+   * @returns whether the feature applies.
+   */
+  isApplicable(profile: ProjectProfile): boolean {
+    return this.supportedInterfaces.includes(profile.runInterface)
+  }
+
+  /**
+   * Validate and normalize one requested option set.
+   * @param selection - requested feature and options.
+   * @param profile - project context for applicability and defaults.
+   * @returns deduplicated, sorted selection.
+   */
+  normalizeSelection(selection: FeatureSelection, profile: ProjectProfile): FeatureSelection {
+    if (selection.id !== this.id) throw new Error(`selection ${selection.id} does not belong to feature ${this.id}`)
+    if (!this.isApplicable(profile)) {
+      throw new Error(`feature ${this.id} is not available for ${profile.runInterface}`)
+    }
+    const available = new Set(this.options.map(option => option.id))
+    const options = [...new Set(selection.options.length > 0 ? selection.options : this.defaultOptions(profile))]
+    for (const option of options) {
+      if (!available.has(option)) throw new Error(`unknown ${this.id} option: ${option}`)
+    }
+    if (this.mode === 'single' && (options.length !== 1 || this.options.length !== 1)) {
+      throw new Error(`feature ${this.id} has one fixed option`)
+    }
+    if (this.mode === 'exclusive' && options.length !== 1) {
+      throw new Error(`feature ${this.id} requires exactly one option`)
+    }
+    if (this.mode === 'multiple' && options.length === 0) {
+      throw new Error(`feature ${this.id} requires at least one option`)
+    }
+    return { ...selection, options: options.sort() }
+  }
+
+  /**
+   * Build the complete selected resource contribution.
+   * @param selection - selected options and captured inputs.
+   * @param profile - target project context.
+   * @returns merged base and option resources.
+   */
+  contribution(selection: FeatureSelection, profile: ProjectProfile): ProjectContribution {
+    const normalized = this.normalizeSelection(selection, profile)
+    const selected = this.selectedOptions(normalized)
+      .map(option => option.contribution(profile, normalized.secrets ?? {}))
+    return ProjectContribution.merge(this.baseContribution(profile), ...selected)
+  }
+
+  /**
+   * All secret definitions required by one selected option set.
+   * @param selection - selected options.
+   * @param profile - target project context.
+   * @returns selected secret definitions.
+   */
+  secrets(selection: FeatureSelection, profile: ProjectProfile): readonly FeatureSecret[] {
+    const normalized = this.normalizeSelection(selection, profile)
+    return this.selectedOptions(normalized).flatMap(option => option.secrets)
+  }
+
+  /**
+   * All visible value definitions required by one selected option set.
+   * @param selection - selected options.
+   * @param profile - target project context.
+   * @returns selected visible-input definitions.
+   */
+  valueInputs(selection: FeatureSelection, profile: ProjectProfile): readonly FeatureValueInput[] {
+    const normalized = this.normalizeSelection(selection, profile)
+    return this.selectedOptions(normalized).flatMap(option => option.inputs)
+  }
+
+  private selectedOptions(selection: FeatureSelection): readonly FeatureOption[] {
+    return selection.options.map((id) => {
+      const option = this.options.find(candidate => candidate.id === id)
+      /* v8 ignore next -- normalizeSelection already membership-checks every selected id */
+      if (!option) throw new Error(`unknown ${this.id} option: ${id}`)
+      return option
+    })
+  }
+
+  /**
+   * Recover input and secret values after structural inspection.
+   * @param project - project snapshot being inspected.
+   * @param selection - structurally detected selection.
+   * @returns selection enriched with readable values.
+   */
+  readSelection(project: FeatureProjectView, selection: FeatureSelection): FeatureSelection {
+    const secrets = Object.fromEntries(this.secrets(selection, project.profile).flatMap((secret) => {
+      const value = project.readEnvironment('.env', secret.environment)
+      return value === undefined ? [] : [[secret.id, value]]
+    }))
+    return Object.keys(secrets).length === 0 ? selection : { ...selection, secrets }
+  }
+
+  /**
+   * Inspect current files and reject any partial or ambiguous owned shape.
+   * @param project - project snapshot to inspect.
+   * @returns installation state, selection, and diagnostics.
+   */
+  inspect(project: FeatureProjectView): FeatureInstallation {
+    const profile = project.profile
+    const allPackages = new Set<string>()
+    for (const option of this.options) {
+      for (const entry of option.ownedConfigEntries(profile)) allPackages.add(entry.name)
+    }
+    for (const name of packageNames(this.baseContribution(profile).resources)) allPackages.add(name)
+    const configEntries = project.cordisConfigEntries()
+    const ownedConfigEntries = configEntries.filter(entry => allPackages.has(entry.name))
+
+    const options = this.options
+      .filter(option => option.matchesConfigEntries(configEntries, profile))
+      .map(option => option.id)
+    if (ownedConfigEntries.length === 0 && options.length === 0) {
+      return { id: this.id, state: 'absent', options: [], diagnostics: [] }
+    }
+    let selection: FeatureSelection
+    try {
+      selection = this.normalizeSelection({ id: this.id, options }, profile)
+    } catch (error) {
+      return { id: this.id, state: 'inconsistent', options, diagnostics: [String(error)] }
+    }
+    selection = this.readSelection(project, selection)
+    const expected = this.contribution(selection, profile)
+    const expectedEntries = expected.resources
+      .filter((resource): resource is CordisConfigEntryResource => resource.kind === 'cordis-config-entry')
+    const diagnostics: string[] = []
+    for (const resource of expectedEntries) {
+      const actual = ownedConfigEntries.find(entry => entry.id === resource.entry.id && entry.name === resource.entry.name)
+      if (!actual) diagnostics.push(`missing Cordis config entry ${resource.entry.id} (${resource.entry.name})`)
+      else diagnostics.push(...configDiagnostics(resource, actual))
+    }
+    for (const actual of ownedConfigEntries) {
+      if (!expectedEntries.some(resource => resource.entry.id === actual.id && resource.entry.name === actual.name)) {
+        diagnostics.push(`unexpected owned Cordis config entry ${actual.id} (${actual.name})`)
+      }
+    }
+    const manifest = project.packageManifest()
+    for (const resource of expected.resources) {
+      switch (resource.kind) {
+        case 'npm-dependency':
+          if (!manifest[resource.section]?.[resource.name]) {
+            diagnostics.push(`missing package.json ${resource.section} entry ${resource.name}`)
+          }
+          break
+        case 'package-script':
+          if (!manifest.scripts?.[resource.name]) {
+            diagnostics.push(`missing package.json script ${resource.name}`)
+          }
+          break
+        case 'owned-file':
+          if (!project.hasDocument(resource.document.relativePath)) diagnostics.push(`missing owned file ${resource.document.relativePath}`)
+          break
+        case 'environment':
+          try {
+            if (project.readEnvironment('.env.example', resource.name) === undefined) {
+              diagnostics.push(`missing .env.example variable ${resource.name}`)
+            }
+          } catch (error) {
+            diagnostics.push(String(error))
+          }
+          break
+        case 'cordis-config-entry': break
+      }
+    }
+    const disabled = ownedConfigEntries.map(entry => entry.disabled === true)
+    if (disabled.some(Boolean) && disabled.some(value => !value)) {
+      diagnostics.push('owned Cordis config entries have mixed enabled states')
+    }
+    if (diagnostics.length > 0) {
+      return { id: this.id, state: 'inconsistent', options, diagnostics }
+    }
+    return {
+      id: this.id,
+      state: ownedConfigEntries.length > 0 && disabled.every(Boolean) ? 'disabled' : 'enabled',
+      options,
+      selection,
+      diagnostics: [],
+    }
+  }
+}
+
+/** Fixed one-option feature base. */
+export abstract class FixedFeature extends Feature {
+  override readonly mode = 'single'
+
+  /** Select the sole option. */
+  override defaultOptions(): readonly string[] {
+    const option = this.options[0]
+    if (!option) throw new Error(`simple feature ${this.id} has no option`)
+    return [option.id]
+  }
+}
+
+/** Mutually exclusive option feature base. */
+export abstract class ExclusiveOptionFeature extends Feature {
+  override readonly mode = 'exclusive'
+}
+
+/** Additive multi-option feature base. */
+export abstract class MultiOptionFeature extends Feature {
+  override readonly mode = 'multiple'
+}
diff --git a/packages/sdk/helper/src/features/registry.ts b/packages/sdk/helper/src/features/registry.ts
new file mode 100644
index 0000000000..5135f2e40c
--- /dev/null
+++ b/packages/sdk/helper/src/features/registry.ts
@@ -0,0 +1,87 @@
+/**
+ * Builtin feature registry and definition-time conflict checks.
+ *
+ * @module @deepseek-ai/dsh-helper/features/registry
+ */
+
+import type { FeatureId, ResourceKey } from '../ids.ts'
+import type { ProjectProfile } from '../project/types.ts'
+import type { Feature, FeatureProjectView } from './feature.ts'
+import type { CordisConfigEntryResource } from './resources.ts'
+
+/** Compile-time builtin feature collection. */
+export class FeatureRegistry {
+  private readonly features = new Map<FeatureId, Feature>()
+
+  /** Register and validate a complete builtin set. */
+  constructor(features: readonly Feature[], validationProfile: ProjectProfile) {
+    const owners = new Map<ResourceKey, FeatureId>()
+    for (const feature of features) {
+      if (this.features.has(feature.id)) throw new Error(`duplicate feature id: ${feature.id}`)
+      this.features.set(feature.id, feature)
+      const validationInterface = feature.supportedInterfaces[0]
+      if (!validationInterface) throw new Error(`feature ${feature.id} supports no run interface`)
+      const selections = feature.options.map(option => ({ id: feature.id, options: [option.id] }))
+      for (const selection of selections) {
+        const contribution = feature.contribution(selection, {
+          ...validationProfile,
+          runInterface: validationInterface,
+        })
+        for (const resource of contribution.resources) {
+          const owner = owners.get(resource.key)
+          if (owner && owner !== feature.id) {
+            throw new Error(`resource ${resource.key} is declared by both ${owner} and ${feature.id}`)
+          }
+          owners.set(resource.key, feature.id)
+        }
+      }
+    }
+  }
+
+  /**
+   * Return all builtins in display order.
+   * @returns all registered features.
+   */
+  all(): readonly Feature[] {
+    return [...this.features.values()]
+  }
+
+  /**
+   * Resolve one builtin or fail loud.
+   * @param id - stable feature identity.
+   * @returns registered feature.
+   */
+  get(id: FeatureId): Feature {
+    const feature = this.features.get(id)
+    if (!feature) throw new Error(`unknown feature: ${id}`)
+    return feature
+  }
+
+  /**
+   * Inspect every applicable builtin in display order.
+   * @param project - project view to inspect.
+   * @returns installation snapshots for applicable features.
+   */
+  inspect(project: FeatureProjectView): ReturnType<Feature['inspect']>[] {
+    return this.all()
+      .filter(feature => feature.isApplicable(project.profile))
+      .map(feature => feature.inspect(project))
+  }
+
+  /**
+   * Resolve the builtin that owns a Cordis package name for this profile.
+   * @param name - Loader package name.
+   * @param profile - project context controlling applicability.
+   * @returns owning feature, if the package is builtin-owned.
+   */
+  ownerOfPackage(name: string, profile: ProjectProfile): Feature | undefined {
+    return this.all().find((feature) => {
+      if (!feature.isApplicable(profile)) return false
+      const selections = feature.options.map(option => ({ id: feature.id, options: [option.id] }))
+      return selections.some(selection => feature.contribution(selection, profile).resources.some(
+        (resource): resource is CordisConfigEntryResource => resource.kind === 'cordis-config-entry'
+          && resource.entry.name === name,
+      ))
+    })
+  }
+}
diff --git a/packages/sdk/helper/src/features/resources.ts b/packages/sdk/helper/src/features/resources.ts
new file mode 100644
index 0000000000..14b76db57a
--- /dev/null
+++ b/packages/sdk/helper/src/features/resources.ts
@@ -0,0 +1,97 @@
+/**
+ * Resource vocabulary contributed by builtin SDK features.
+ *
+ * @module @deepseek-ai/dsh-helper/features/resources
+ */
+
+import type { CordisConfigEntry } from '../documents/cordis-yaml-file.ts'
+import type { ProjectFile } from '../documents/project-file.ts'
+import type { ResourceKey } from '../ids.ts'
+
+/** Runtime or development NPM dependency contribution. */
+export interface NpmDependencyResource {
+  kind: 'npm-dependency'
+  key: ResourceKey
+  name: string
+  section: 'dependencies' | 'devDependencies'
+}
+
+/** Feature-owned package script. */
+export interface PackageScriptResource {
+  kind: 'package-script'
+  key: ResourceKey
+  name: string
+  command: string
+  removeOnlyWhenUnchanged: boolean
+}
+
+/** Owned Cordis config entry plus the config keys safe to update in place. */
+export interface CordisConfigEntryResource {
+  kind: 'cordis-config-entry'
+  key: ResourceKey
+  entry: CordisConfigEntry
+  ownedConfigKeys: readonly string[]
+  commentedExample?: string
+  validateConfig?: (config: Readonly<Record<string, unknown>>) => readonly string[]
+}
+
+/** Environment variable reference and dotenv material. */
+export interface EnvironmentResource {
+  kind: 'environment'
+  key: ResourceKey
+  name: string
+  value?: string
+  exampleValue: string
+  comment?: string
+}
+
+/** Feature-exclusive complete file. */
+export interface OwnedFileResource {
+  kind: 'owned-file'
+  key: ResourceKey
+  document: ProjectFile
+  removeOnlyWhenUnchanged: boolean
+}
+
+/** Any resource a feature can add to a project. */
+export type ProjectResource =
+  | NpmDependencyResource
+  | PackageScriptResource
+  | CordisConfigEntryResource
+  | EnvironmentResource
+  | OwnedFileResource
+
+/** Complete resource contribution for one selected feature state. */
+export class ProjectContribution {
+  readonly resources: readonly ProjectResource[]
+
+  /** Validate and retain one feature-owned resource set. */
+  constructor(resources: readonly ProjectResource[]) {
+    const seen = new Set<ResourceKey>()
+    for (const resource of resources) {
+      if (seen.has(resource.key)) throw new Error(`duplicate contribution resource key: ${resource.key}`)
+      seen.add(resource.key)
+    }
+    this.resources = resources
+  }
+
+  /** Merge base and option contributions by stable key. */
+  static merge(...contributions: readonly ProjectContribution[]): ProjectContribution {
+    const resources = new Map<ResourceKey, ProjectResource>()
+    for (const contribution of contributions) {
+      for (const resource of contribution.resources) {
+        const previous = resources.get(resource.key)
+        if (previous && JSON.stringify(previous) !== JSON.stringify(resource)) {
+          throw new Error(`resource ${resource.key} has conflicting definitions inside one feature`)
+        }
+        resources.set(resource.key, resource)
+      }
+    }
+    return new ProjectContribution([...resources.values()])
+  }
+
+  /** Index resources by stable key. */
+  byKey(): ReadonlyMap<ResourceKey, ProjectResource> {
+    return new Map(this.resources.map(resource => [resource.key, resource]))
+  }
+}
diff --git a/packages/sdk/helper/src/ids.ts b/packages/sdk/helper/src/ids.ts
new file mode 100644
index 0000000000..4d671753ca
--- /dev/null
+++ b/packages/sdk/helper/src/ids.ts
@@ -0,0 +1,31 @@
+/**
+ * Branded identities owned by the SDK project domain.
+ *
+ * @module @deepseek-ai/dsh-helper/ids
+ */
+
+import type { Branded } from '@deepseek-ai/dsh-brand'
+
+/** Stable identity of a builtin SDK feature. */
+export type FeatureId = Branded<'FeatureId'>
+
+/**
+ * Construct a feature identity from its registry key.
+ * @param value - lowercase kebab-case registry key.
+ * @returns branded feature identity.
+ */
+export function featureId(value: string): FeatureId {
+  if (!/^[a-z][a-z0-9-]*$/.test(value)) {
+    throw new Error(`invalid feature id: ${JSON.stringify(value)}`)
+  }
+  return value as FeatureId
+}
+
+/** Stable identity of a resource contributed to an SDK project. */
+export type ResourceKey = Branded<'ResourceKey'>
+
+/** Construct a resource key from its owner-qualified value. */
+export function resourceKey(value: string): ResourceKey {
+  if (value.length === 0) throw new Error('resource key must not be empty')
+  return value as ResourceKey
+}
diff --git a/packages/sdk/helper/src/index.ts b/packages/sdk/helper/src/index.ts
new file mode 100644
index 0000000000..98c55c3f8b
--- /dev/null
+++ b/packages/sdk/helper/src/index.ts
@@ -0,0 +1,45 @@
+/**
+ * Shared domain and infrastructure for DeepSeek Harness SDK project tooling.
+ *
+ * @module @deepseek-ai/dsh-helper
+ */
+
+export { featureId } from './ids.ts'
+export { TextTemplate } from './templates/text-template.ts'
+export type {
+  FeatureSelection,
+  ProjectCreationRequest,
+  ProjectProfile,
+  RunInterface,
+} from './project/types.ts'
+export type { ChangeSet, ProjectCommitResult } from './project/change-set.ts'
+export { SdkProject } from './project/sdk-project.ts'
+export {
+  NodeCommandRunner,
+  NpmPackageManager,
+  createPackageManager,
+  inferPackageManagerName,
+  probePackageManagerVersion,
+} from './package-managers/package-manager.ts'
+export type {
+  CommandRunner,
+  PackageManager,
+  PackageManagerName,
+  PackageManagerVersionProbe,
+} from './package-managers/package-manager.ts'
+export { LocalPluginBlueprint } from './plugins/local-plugin-blueprint.ts'
+export type { LocalPluginKind } from './plugins/local-plugin-blueprint.ts'
+export type { Feature, FeatureInstallation } from './features/feature.ts'
+export type { FeatureRegistry } from './features/registry.ts'
+export { FeatureConfigurator } from './features/feature-configurator.ts'
+export { createBuiltinRegistry } from './features/builtin/index.ts'
+export { PromptCancelledError, requireAnswer } from './questions/prompt-port.ts'
+export type { NestedMultiSelectValue, PromptPort } from './questions/prompt-port.ts'
+export {
+  ConfirmQuestion,
+  SecretQuestion,
+  SelectQuestion,
+  TextQuestion,
+} from './questions/question.ts'
+export type { Question } from './questions/question.ts'
+export { ClackPromptPort } from './questions/clack-prompt-port.ts'
diff --git a/packages/sdk/helper/src/package-managers/link-workspace.ts b/packages/sdk/helper/src/package-managers/link-workspace.ts
new file mode 100644
index 0000000000..1a6b518fa9
--- /dev/null
+++ b/packages/sdk/helper/src/package-managers/link-workspace.ts
@@ -0,0 +1,137 @@
+/**
+ * Repository package discovery and NPM dependency-closure rewriting for live links.
+ *
+ * @module @deepseek-ai/dsh-helper/package-managers/link-workspace
+ */
+
+import { readFile, readdir } from 'node:fs/promises'
+import { existsSync, realpathSync } from 'node:fs'
+import { basename, dirname, join, relative, resolve, sep } from 'node:path'
+import type { PackageJsonFile, PackageManifest } from '../documents/package-json-file.ts'
+import { PnpmWorkspaceFile } from '../documents/pnpm-workspace-file.ts'
+import type { ProjectFile } from '../documents/project-file.ts'
+import type { PackageManager } from './package-manager.ts'
+
+interface WorkspacePackage {
+  directory: string
+  manifest: PackageManifest
+}
+
+function posixPath(path: string): string {
+  return path.split(sep).join('/')
+}
+
+function canonicalPath(path: string): string {
+  let existing = resolve(path)
+  const suffix: string[] = []
+  while (!existsSync(existing)) {
+    const parent = dirname(existing)
+    /* v8 ignore next -- every absolute path reaches the existing filesystem root */
+    if (parent === existing) throw new Error(`cannot resolve an existing ancestor for ${path}`)
+    suffix.unshift(basename(existing))
+    existing = parent
+  }
+  return resolve(realpathSync(existing), ...suffix)
+}
+
+async function packageDirectories(root: string): Promise<string[]> {
+  const result: string[] = []
+  for (const vendor of await readdir(join(root, 'vendor'), { withFileTypes: true })) {
+    if (vendor.isDirectory()) result.push(join(root, 'vendor', vendor.name))
+  }
+  for (const group of await readdir(join(root, 'packages'), { withFileTypes: true })) {
+    if (!group.isDirectory()) continue
+    for (const pkg of await readdir(join(root, 'packages', group.name), { withFileTypes: true })) {
+      if (pkg.isDirectory()) result.push(join(root, 'packages', group.name, pkg.name))
+    }
+  }
+  return result
+}
+
+/** Index of repository packages used by `--link-workspace`. */
+export class LinkWorkspace {
+  readonly root: string
+  private readonly packages: Map<string, WorkspacePackage>
+
+  private constructor(root: string, packages: Map<string, WorkspacePackage>) {
+    this.root = root
+    this.packages = packages
+  }
+
+  /** Scan vendor and package workspaces from a repository root. */
+  static async open(root: string): Promise<LinkWorkspace> {
+    const absolute = resolve(root)
+    const packages = new Map<string, WorkspacePackage>()
+    for (const directory of await packageDirectories(absolute)) {
+      let manifest: PackageManifest
+      try {
+        manifest = JSON.parse(await readFile(join(directory, 'package.json'), 'utf8')) as PackageManifest
+      } catch (error) {
+        throw new Error(`cannot read linked package at ${directory}: ${String(error)}`)
+      }
+      if (!manifest.name || typeof manifest.name !== 'string') continue
+      if (packages.has(manifest.name)) throw new Error(`duplicate linked package name: ${manifest.name}`)
+      packages.set(manifest.name, { directory, manifest })
+    }
+    if (!packages.has('cordis') || !packages.has('@deepseek-ai/dsh-scripts')) {
+      throw new Error(`not a DeepSeek Harness repository root: ${absolute}`)
+    }
+    return new LinkWorkspace(absolute, packages)
+  }
+
+  /** Expand direct NPM dependencies through all repository-local NPM dependency edges. */
+  closure(names: Iterable<string>): string[] {
+    const pending = [...names]
+    const result = new Set<string>()
+    while (pending.length > 0) {
+      const name = pending.pop()
+      if (!name || result.has(name)) continue
+      const pkg = this.packages.get(name)
+      /* v8 ignore next -- closure() only returns names present in this package map */
+      if (!pkg) continue
+      result.add(name)
+      const edges = {
+        ...pkg.manifest.dependencies,
+        ...pkg.manifest.peerDependencies as Record<string, string> | undefined,
+      }
+      for (const dependencyName of Object.keys(edges)) {
+        if (this.packages.has(dependencyName) && !result.has(dependencyName)) pending.push(dependencyName)
+      }
+    }
+    return [...result].sort()
+  }
+
+  /** Rewrite the full local closure to manager-specific live-link specs. */
+  apply(
+    projectRoot: string,
+    manifest: PackageJsonFile,
+    manager: PackageManager,
+    documents: readonly ProjectFile[],
+  ): void {
+    const canonicalProjectRoot = canonicalPath(projectRoot)
+    const names = this.closure(manifest.npmDependencyNames())
+    for (const name of names) {
+      const pkg = this.packages.get(name)
+      /* v8 ignore next -- closure() only returns names present in this package map */
+      if (!pkg) continue
+      const relativePath = posixPath(relative(canonicalProjectRoot, realpathSync(pkg.directory)))
+      const spec = manager.linkSpec(relativePath)
+      const current = manifest.npmDependency(name)
+      manifest.setNpmDependency(current?.section ?? 'dependencies', name, spec)
+      if (manager.name === 'yarn') manifest.setResolution(name, spec)
+    }
+    if (manager.name === 'pnpm') {
+      const workspace = documents.find((item): item is PnpmWorkspaceFile => item instanceof PnpmWorkspaceFile)
+      if (!workspace) throw new Error('pnpm link mode requires pnpm-workspace.yaml')
+      workspace.disableAutoInstallPeers()
+    }
+  }
+
+  /** Resolve a package directory for diagnostics and tests. */
+  packageDirectory(name: string): string | undefined {
+    const directory = this.packages.get(name)?.directory
+    return directory
+      ? resolve(dirname(directory), directory.split(sep).at(-1) as string)
+      : undefined
+  }
+}
diff --git a/packages/sdk/helper/src/package-managers/package-manager.ts b/packages/sdk/helper/src/package-managers/package-manager.ts
new file mode 100644
index 0000000000..a6f2fdd103
--- /dev/null
+++ b/packages/sdk/helper/src/package-managers/package-manager.ts
@@ -0,0 +1,280 @@
+/**
+ * Package-manager strategies for SDK project workspaces and child commands.
+ *
+ * @module @deepseek-ai/dsh-helper/package-managers/package-manager
+ */
+
+import { execFile, spawn } from 'node:child_process'
+import { promisify } from 'node:util'
+import type { PackageJsonFile } from '../documents/package-json-file.ts'
+import { PnpmWorkspaceFile } from '../documents/pnpm-workspace-file.ts'
+import type { ProjectFile } from '../documents/project-file.ts'
+
+/** Supported generated-project package managers. */
+export type PackageManagerName = 'npm' | 'pnpm' | 'yarn'
+
+/** Result from one child package-manager process. */
+export interface CommandResult {
+  exitCode: number | null
+  signal: NodeJS.Signals | null
+}
+
+/** Injectable subprocess boundary used by package-manager strategies. */
+export interface CommandRunner {
+  /** Run one executable without a shell and await process exit. */
+  run(command: string, args: readonly string[], cwd: string): Promise<CommandResult>
+}
+
+/** Injectable package-manager version probe used by project creation. */
+export type PackageManagerVersionProbe = (name: PackageManagerName, cwd: string) => Promise<string>
+
+const execFileAsync = promisify(execFile)
+
+/**
+ * Read a manager version without forwarding ambient credentials.
+ * @param name - package-manager executable.
+ * @param cwd - working directory used for resolution.
+ * @returns trimmed version output.
+ */
+export async function probePackageManagerVersion(name: PackageManagerName, cwd: string): Promise<string> {
+  try {
+    const { stdout } = await execFileAsync(name, ['--version'], {
+      cwd,
+      env: scrubEnvironment(),
+      encoding: 'utf8',
+    })
+    const version = stdout.trim()
+    if (!version) throw new Error('empty version output')
+    return version
+  } catch (error) {
+    throw new Error(`cannot run ${name} --version: ${String(error)}`)
+  }
+}
+
+/** Remove credential-shaped environment variables from spawned commands. */
+export function scrubEnvironment(environment: NodeJS.ProcessEnv = process.env): NodeJS.ProcessEnv {
+  return Object.fromEntries(Object.entries(environment).filter(([name]) => !/(?:KEY|SECRET|TOKEN)/i.test(name)))
+}
+
+/** Node child-process command runner with inherited stdio and quiescent completion. */
+export class NodeCommandRunner implements CommandRunner {
+  /** Spawn one child and settle only after its exit. */
+  run(command: string, args: readonly string[], cwd: string): Promise<CommandResult> {
+    return new Promise((resolve, reject) => {
+      const child = spawn(command, [...args], {
+        cwd,
+        env: scrubEnvironment(),
+        stdio: 'inherit',
+        shell: false,
+      })
+      child.once('error', reject)
+      child.once('exit', (exitCode, signal) => { resolve({ exitCode, signal }) })
+    })
+  }
+}
+
+function major(version: string): number {
+  const match = /^(\d+)/.exec(version)
+  if (!match?.[1]) throw new Error(`invalid package manager version: ${JSON.stringify(version)}`)
+  return Number(match[1])
+}
+
+/** Behavior owned by one generated-project package manager. */
+export abstract class PackageManager {
+  /** Manager executable and project identity. */
+  abstract readonly name: PackageManagerName
+
+  /** Detected concrete manager version. */
+  readonly version: string
+
+  constructor(version: string) {
+    this.version = version
+  }
+
+  /** Validate the detected version against this SDK's supported floor. */
+  abstract validateVersion(): void
+
+  /**
+   * Configure root manifest fields and return manager-specific files.
+   * @param manifest - generated root manifest to update.
+   * @returns manager-specific companion documents.
+   */
+  abstract configureWorkspace(manifest: PackageJsonFile): ProjectFile[]
+
+  /**
+   * Build the NPM dependency spec for a local workspace plugin.
+   * @returns manager-specific local NPM dependency spec.
+   */
+  abstract localPluginSpec(): string
+
+  /**
+   * Resolve a repository live-link NPM dependency.
+   * @param relativePath - relative path from generated project to package.
+   * @returns manager-specific NPM dependency spec.
+   */
+  abstract linkSpec(relativePath: string): string
+
+  /**
+   * Build install command arguments.
+   * @returns arguments following the manager executable.
+   */
+  installCommand(): readonly string[] {
+    return ['install']
+  }
+
+  /**
+   * Build project-build command arguments.
+   * @returns arguments following the manager executable.
+   */
+  buildCommand(): readonly string[] {
+    return ['run', 'build']
+  }
+
+  /**
+   * Run NPM dependency installation and fail on non-zero or signalled exit.
+   * @param cwd - generated project directory.
+   * @param runner - optional subprocess boundary.
+   */
+  async install(cwd: string, runner: CommandRunner = new NodeCommandRunner()): Promise<void> {
+    await this.runChecked(runner, this.installCommand(), cwd, 'install')
+  }
+
+  /**
+   * Run the project build and fail on non-zero or signalled exit.
+   * @param cwd - generated project directory.
+   * @param runner - optional subprocess boundary.
+   */
+  async build(cwd: string, runner: CommandRunner = new NodeCommandRunner()): Promise<void> {
+    await this.runChecked(runner, this.buildCommand(), cwd, 'build')
+  }
+
+  private async runChecked(runner: CommandRunner, args: readonly string[], cwd: string, operation: string): Promise<void> {
+    const result = await runner.run(this.name, args, cwd)
+    if (result.signal !== null) {
+      throw new Error(`${this.name} ${operation} was killed by ${result.signal}`)
+    }
+    if (result.exitCode !== 0) {
+      throw new Error(`${this.name} ${operation} exited with code ${String(result.exitCode)}`)
+    }
+  }
+}
+
+/** npm workspace behavior. */
+export class NpmPackageManager extends PackageManager {
+  override readonly name = 'npm'
+
+  /** npm 10 is the supported floor at the repository's Node floor. */
+  override validateVersion(): void {
+    if (major(this.version) < 10) throw new Error(`npm >=10 is required, got ${this.version}`)
+  }
+
+  /** Configure package.json workspaces; npm needs no companion file. */
+  override configureWorkspace(manifest: PackageJsonFile): ProjectFile[] {
+    manifest.addWorkspace('plugins/*')
+    manifest.setPackageManager(undefined)
+    return []
+  }
+
+  /** npm resolves workspace packages through its ordinary wildcard. */
+  override localPluginSpec(): string {
+    return '*'
+  }
+
+  /** npm live links use file NPM dependencies. */
+  override linkSpec(relativePath: string): string {
+    return `file:${relativePath}`
+  }
+}
+
+/** pnpm workspace behavior. */
+export class PnpmPackageManager extends PackageManager {
+  override readonly name = 'pnpm'
+
+  /** pnpm 10 is the supported floor for strict NPM dependency-build policy. */
+  override validateVersion(): void {
+    if (major(this.version) < 10) throw new Error(`pnpm >=10 is required, got ${this.version}`)
+  }
+
+  /** Configure packageManager and a structured pnpm workspace file. */
+  override configureWorkspace(manifest: PackageJsonFile): ProjectFile[] {
+    manifest.setPackageManager(`pnpm@${this.version}`)
+    const workspace = PnpmWorkspaceFile.create()
+    workspace.addPackage('plugins/*')
+    return [workspace]
+  }
+
+  /** pnpm uses its explicit workspace protocol. */
+  override localPluginSpec(): string {
+    return 'workspace:*'
+  }
+
+  /** pnpm live links use link NPM dependencies. */
+  override linkSpec(relativePath: string): string {
+    return `link:${relativePath}`
+  }
+}
+
+/** Yarn Berry-compatible workspace behavior. */
+export class YarnPackageManager extends PackageManager {
+  override readonly name = 'yarn'
+
+  /** Yarn classic is excluded because the generated project relies on modern workspaces. */
+  override validateVersion(): void {
+    if (major(this.version) < 2) throw new Error(`Yarn >=2 is required, got ${this.version}`)
+  }
+
+  /** Configure packageManager and package.json workspaces. */
+  override configureWorkspace(manifest: PackageJsonFile): ProjectFile[] {
+    manifest.addWorkspace('plugins/*')
+    manifest.setPackageManager(`yarn@${this.version}`)
+    return []
+  }
+
+  /** Modern Yarn uses the workspace protocol. */
+  override localPluginSpec(): string {
+    return 'workspace:*'
+  }
+
+  /** Yarn live links use portal NPM dependencies to preserve package identity. */
+  override linkSpec(relativePath: string): string {
+    return `portal:${relativePath}`
+  }
+
+  /** Yarn runs scripts without the `run` token. */
+  override buildCommand(): readonly string[] {
+    return ['build']
+  }
+}
+
+/**
+ * Construct and validate one package-manager strategy.
+ * @param name - selected manager.
+ * @param version - detected concrete version.
+ * @returns validated strategy.
+ */
+export function createPackageManager(name: PackageManagerName, version: string): PackageManager {
+  let manager: PackageManager
+  switch (name) {
+    case 'npm': manager = new NpmPackageManager(version); break
+    case 'pnpm': manager = new PnpmPackageManager(version); break
+    case 'yarn': manager = new YarnPackageManager(version); break
+  }
+  manager.validateVersion()
+  return manager
+}
+
+/**
+ * Infer a package manager from an explicit choice or npm user-agent value.
+ * @param explicit - explicit CLI selection.
+ * @param userAgent - npm-compatible user-agent string.
+ * @returns selected or inferred manager name.
+ */
+export function inferPackageManagerName(
+  explicit: PackageManagerName | undefined,
+  userAgent: string | undefined = process.env.npm_config_user_agent,
+): PackageManagerName | undefined {
+  if (explicit) return explicit
+  const token = userAgent?.split(' ')[0]?.split('/')[0]
+  if (token === 'npm' || token === 'pnpm' || token === 'yarn') return token
+  return undefined
+}
diff --git a/packages/sdk/helper/src/plugins/local-plugin-blueprint.ts b/packages/sdk/helper/src/plugins/local-plugin-blueprint.ts
new file mode 100644
index 0000000000..616d0cedd8
--- /dev/null
+++ b/packages/sdk/helper/src/plugins/local-plugin-blueprint.ts
@@ -0,0 +1,123 @@
+/**
+ * Source blueprints for local Cordis plugins generated under `plugins/*`.
+ *
+ * @module @deepseek-ai/dsh-helper/plugins/local-plugin-blueprint
+ */
+
+import { TextProjectFile } from '../documents/project-file.ts'
+import type { CordisConfigEntry } from '../documents/cordis-yaml-file.ts'
+import { resolveNpmDependency } from '../project/npm-dependency-policy.ts'
+import { loadHelperTemplate } from '../templates/template-assets.ts'
+
+/** Supported generated local-plugin shapes. */
+export type LocalPluginKind = 'plugin' | 'tool'
+
+function kebab(value: string): string {
+  const result = value.trim().toLowerCase().replace(/[^a-z0-9]+/g, '-').replace(/^-|-$/g, '')
+  if (!result || !/^[a-z]/.test(result)) throw new Error(`invalid local plugin name: ${JSON.stringify(value)}`)
+  return result
+}
+
+function packageName(projectName: string, pluginName: string): string {
+  if (projectName.startsWith('@')) {
+    const separator = projectName.indexOf('/')
+    if (separator > 1 && separator < projectName.length - 1) {
+      return `${projectName.slice(0, separator)}/${projectName.slice(separator + 1)}-${pluginName}`
+    }
+  }
+  return `${projectName}-${pluginName}`
+}
+
+interface LocalPluginTemplateContext {
+  pluginName: string
+  toolName: string
+  toolTitle: string
+}
+
+const PLUGIN_SOURCE = loadHelperTemplate<LocalPluginTemplateContext>('local-plugin.ts.tpl')
+const TOOL_SOURCE = loadHelperTemplate<LocalPluginTemplateContext>('local-tool.ts.tpl')
+const PLUGIN_TSDOWN = loadHelperTemplate<LocalPluginTemplateContext>('local-plugin-tsdown.config.ts.tpl')
+
+/** One local plugin's derived package, source, build, and runtime entry. */
+export class LocalPluginBlueprint {
+  /** Normalized local package and Cordis config entry name. */
+  readonly name: string
+  /** Generated plugin source shape. */
+  readonly kind: LocalPluginKind
+
+  /** Normalize and validate one local plugin request. */
+  constructor(name: string, kind: LocalPluginKind) {
+    this.name = kebab(name)
+    this.kind = kind
+  }
+
+  /** Root-relative plugin directory. */
+  get directory(): string {
+    return `plugins/${this.name}`
+  }
+
+  /**
+   * Derive an npm package name from the root project identity.
+   * @param projectName - generated root package name.
+   * @returns local plugin package name.
+   */
+  packageName(projectName: string): string {
+    return packageName(projectName, this.name)
+  }
+
+  /**
+   * Build the runtime Cordis config entry for this local package.
+   * @param projectName - generated root package name.
+   * @returns Loader entry referencing the local package.
+   */
+  cordisConfigEntry(projectName: string): CordisConfigEntry {
+    return { id: this.name, name: this.packageName(projectName) }
+  }
+
+  /**
+   * Render the complete local package files.
+   * @param projectName - generated root package name.
+   * @param releaseVersion - SDK dependency version.
+   * @returns local manifest, configs, and source documents.
+   */
+  documents(projectName: string, releaseVersion: string): TextProjectFile[] {
+    const name = this.packageName(projectName)
+    const toolName = this.name.replaceAll('-', '_')
+    const cordisSpec = resolveNpmDependency('cordis', 'devDependencies', releaseVersion).spec
+    const manifest = {
+      name,
+      version: '0.0.0',
+      private: true,
+      type: 'module',
+      main: 'lib/index.js',
+      types: 'lib/index.d.ts',
+      exports: { '.': { types: './lib/index.d.ts', default: './lib/index.js' } },
+      peerDependencies: {
+        ...this.kind === 'tool' ? { '@deepseek-ai/dsh-tools': `^${releaseVersion}` } : {},
+        cordis: cordisSpec,
+      },
+      devDependencies: {
+        cordis: cordisSpec,
+      },
+    }
+    const tsconfig = {
+      extends: '../../tsconfig.base.json',
+      compilerOptions: { rootDir: 'src', outDir: 'lib/types' },
+      include: ['src'],
+    }
+    const context: LocalPluginTemplateContext = {
+      pluginName: this.name,
+      toolName,
+      toolTitle: toolName.replaceAll('_', ' '),
+    }
+    return [
+      new TextProjectFile(`${this.directory}/package.json`, JSON.stringify(manifest, null, 2)),
+      new TextProjectFile(`${this.directory}/tsconfig.json`, JSON.stringify(tsconfig, null, 2)),
+      new TextProjectFile(`${this.directory}/tsdown.config.ts`, PLUGIN_TSDOWN.render(context)),
+      new TextProjectFile(
+        `${this.directory}/src/index.ts`,
+        (this.kind === 'tool' ? TOOL_SOURCE : PLUGIN_SOURCE).render(context),
+      ),
+    ]
+  }
+}
diff --git a/packages/sdk/helper/src/project/change-set.ts b/packages/sdk/helper/src/project/change-set.ts
new file mode 100644
index 0000000000..d6207a8df1
--- /dev/null
+++ b/packages/sdk/helper/src/project/change-set.ts
@@ -0,0 +1,26 @@
+/**
+ * Result summary for one SDK project edit session.
+ *
+ * @module @deepseek-ai/dsh-helper/project/change-set
+ */
+
+import type { FeatureId } from '../ids.ts'
+
+/** Immutable description of committed or pending project changes. */
+export interface ChangeSet {
+  addedFeatures: readonly FeatureId[]
+  enabledFeatures: readonly FeatureId[]
+  disabledFeatures: readonly FeatureId[]
+  configuredFeatures: readonly FeatureId[]
+  addedPlugins: readonly string[]
+  enabledPlugins: readonly string[]
+  disabledPlugins: readonly string[]
+  changedFiles: readonly string[]
+  npmDependenciesChanged: boolean
+}
+
+/** Result of committing one project edit session. */
+export interface ProjectCommitResult<TProject> {
+  project: TProject
+  changes: ChangeSet
+}
diff --git a/packages/sdk/helper/src/project/npm-dependency-policy.ts b/packages/sdk/helper/src/project/npm-dependency-policy.ts
new file mode 100644
index 0000000000..a8877826e2
--- /dev/null
+++ b/packages/sdk/helper/src/project/npm-dependency-policy.ts
@@ -0,0 +1,62 @@
+/**
+ * NPM dependency baseline and version policy for generated SDK projects.
+ *
+ * @module @deepseek-ai/dsh-helper/project/npm-dependency-policy
+ */
+
+import type { NpmDependencySection } from '../documents/package-json-file.ts'
+
+/** One NPM dependency spec selected by the SDK release policy. */
+export interface ResolvedNpmDependency {
+  section: NpmDependencySection
+  spec: string
+}
+
+/** NPM dependency maps rendered into a newly created root package.json. */
+export interface BaselineNpmDependencies {
+  dependencies: Readonly<Record<string, string>>
+  devDependencies: Readonly<Record<string, string>>
+}
+
+const EXTERNAL_NPM_DEPENDENCY_SPECS: Readonly<Record<string, string>> = {
+  '@cordisjs/plugin-hmr': '^1.0.15',
+  '@cordisjs/plugin-timer': '^1.1.2',
+  '@types/node': '^22.20.0',
+  cordis: '^4.0.0-rc.7',
+  tsdown: '^0.22.2',
+  tsx: '^4.22.4',
+  typescript: '^6.0.3',
+}
+
+const BASELINE_NPM_DEPENDENCY_NAMES: Readonly<Record<NpmDependencySection, readonly string[]>> = {
+  dependencies: ['@deepseek-ai/dsh-scripts', 'cordis'],
+  devDependencies: ['@types/node', 'tsdown', 'tsx', 'typescript'],
+}
+
+/** Resolve one package to its generated-project section and version spec. */
+export function resolveNpmDependency(
+  name: string,
+  requestedSection: NpmDependencySection,
+  releaseVersion: string,
+): ResolvedNpmDependency {
+  if (name.startsWith('@deepseek-ai/dsh-')) {
+    return { section: requestedSection, spec: `^${releaseVersion}` }
+  }
+  const spec = EXTERNAL_NPM_DEPENDENCY_SPECS[name]
+  if (spec) return { section: requestedSection, spec }
+  throw new Error(`no generated-project NPM dependency policy for ${name}`)
+}
+
+/** Build the root package.json NPM dependency maps from the shared version policy. */
+export function baselineNpmDependencies(releaseVersion: string): BaselineNpmDependencies {
+  return {
+    dependencies: Object.fromEntries(BASELINE_NPM_DEPENDENCY_NAMES.dependencies.map((name) => {
+      const dependency = resolveNpmDependency(name, 'dependencies', releaseVersion)
+      return [name, dependency.spec]
+    })),
+    devDependencies: Object.fromEntries(BASELINE_NPM_DEPENDENCY_NAMES.devDependencies.map((name) => {
+      const dependency = resolveNpmDependency(name, 'devDependencies', releaseVersion)
+      return [name, dependency.spec]
+    })),
+  }
+}
diff --git a/packages/sdk/helper/src/project/project-edit-session.ts b/packages/sdk/helper/src/project/project-edit-session.ts
new file mode 100644
index 0000000000..9b2c885ac5
--- /dev/null
+++ b/packages/sdk/helper/src/project/project-edit-session.ts
@@ -0,0 +1,597 @@
+/**
+ * Isolated domain-command and commit boundary for SDK project changes.
+ *
+ * @module @deepseek-ai/dsh-helper/project/project-edit-session
+ */
+
+import { mkdir, readFile, unlink, writeFile } from 'node:fs/promises'
+import { dirname, resolve } from 'node:path'
+import type {
+  Feature,
+  FeatureInstallation,
+  FeatureProjectView,
+  FeatureRequirement,
+} from '../features/feature.ts'
+import type { FeatureRegistry } from '../features/registry.ts'
+import type { ProjectResource } from '../features/resources.ts'
+import { CordisYamlFile, type CordisConfigEntry } from '../documents/cordis-yaml-file.ts'
+import { EnvFile } from '../documents/env-file.ts'
+import { PackageJsonFile, type PackageManifest } from '../documents/package-json-file.ts'
+import { ProjectFile } from '../documents/project-file.ts'
+import { TsConfigFile } from '../documents/tsconfig-file.ts'
+import { featureId, type FeatureId, type ResourceKey } from '../ids.ts'
+import { LinkWorkspace } from '../package-managers/link-workspace.ts'
+import type { LocalPluginBlueprint } from '../plugins/local-plugin-blueprint.ts'
+import type { FeatureSelection, ProjectProfile } from './types.ts'
+import { resolveNpmDependency } from './npm-dependency-policy.ts'
+import type { ChangeSet, ProjectCommitResult } from './change-set.ts'
+import type { SdkProject } from './sdk-project.ts'
+
+interface MutableFeatureState {
+  selection?: FeatureSelection
+  state: FeatureInstallation['state']
+}
+
+function sameText(left: ProjectFile, right: ProjectFile | undefined): boolean {
+  return right !== undefined && left.serialize() === right.serialize()
+}
+
+function npmDependencyShape(manifest: Readonly<PackageManifest>): string {
+  return JSON.stringify({
+    /* v8 ignore next -- generated manifests always carry the managed dependency maps */
+    dependencies: manifest.dependencies ?? {},
+    /* v8 ignore next -- generated manifests always carry the managed dependency maps */
+    devDependencies: manifest.devDependencies ?? {},
+  })
+}
+
+function asError(error: unknown): Error {
+  /* v8 ignore else -- node:fs promise APIs reject Error objects */
+  if (error instanceof Error) return error
+  /* v8 ignore next -- node:fs promise APIs reject Error objects */
+  return new Error(String(error))
+}
+
+function canUpdateResource(previous: ProjectResource, next: ProjectResource): boolean {
+  if (previous.kind !== next.kind) return false
+  switch (previous.kind) {
+    case 'npm-dependency': return previous.name === (next as typeof previous).name
+    case 'package-script': return previous.name === (next as typeof previous).name
+    case 'cordis-config-entry': {
+      const candidate = next as typeof previous
+      return previous.entry.name === candidate.entry.name
+    }
+    case 'environment': return previous.name === (next as typeof previous).name
+    case 'owned-file': return previous.document.relativePath === (next as typeof previous).document.relativePath
+  }
+}
+
+/** Mutable working copy that applies feature and local-plugin domain commands. */
+export class ProjectEditSession implements FeatureProjectView {
+  readonly profile: ProjectProfile
+  private readonly source: SdkProject
+  private readonly registry: FeatureRegistry
+  private readonly documents: Map<string, ProjectFile>
+  private readonly removed = new Map<string, ProjectFile>()
+  private readonly states = new Map<FeatureId, MutableFeatureState>()
+  private readonly added = new Set<FeatureId>()
+  private readonly enabled = new Set<FeatureId>()
+  private readonly disabled = new Set<FeatureId>()
+  private readonly configured = new Set<FeatureId>()
+  private readonly addedPlugins = new Set<string>()
+  private readonly enabledPlugins = new Set<string>()
+  private readonly disabledPlugins = new Set<string>()
+  private committed = false
+
+  /** Clone one project snapshot into an isolated working copy. */
+  constructor(source: SdkProject, registry: FeatureRegistry) {
+    this.source = source
+    this.registry = registry
+    this.profile = source.profile
+    this.documents = source.cloneDocuments()
+    for (const feature of registry.all()) {
+      if (!feature.isApplicable(this.profile)) continue
+      const installation = feature.inspect(this)
+      this.states.set(feature.id, {
+        state: installation.state,
+        ...installation.selection ? { selection: installation.selection } : {},
+      })
+    }
+  }
+
+  /** Root manifest value for feature inspection. */
+  packageManifest(): Readonly<PackageManifest> {
+    return this.manifest().value()
+  }
+
+  /** Cordis config entries for feature and custom-plugin inspection. */
+  cordisConfigEntries(): readonly CordisConfigEntry[] {
+    return this.cordis().entries()
+  }
+
+  /** Whether one managed document exists in the working copy. */
+  /* jscpd:ignore-start -- FeatureProjectView deliberately has symmetric snapshot/edit implementations. */
+  hasDocument(path: string): boolean {
+    return this.documents.has(path)
+  }
+
+  /** Read one unique working-copy environment variable. */
+  readEnvironment(path: '.env' | '.env.example', name: string): string | undefined {
+    const document = this.documents.get(path)
+    if (!document) return undefined
+    if (!(document instanceof EnvFile)) throw new Error(`${path} is not an environment document`)
+    return document.get(name)
+  }
+  /* jscpd:ignore-end */
+
+  /** Inspect every applicable builtin against the current working copy. */
+  inspections(): readonly FeatureInstallation[] {
+    return this.registry.inspect(this)
+  }
+
+  /** Install a builtin and recursively satisfy its declared requirements. */
+  installFeature(feature: Feature, selection: FeatureSelection): void {
+    this.assertOpen()
+    this.installFeatureRecursive(feature, selection, new Set())
+  }
+
+  /** Replace one installed builtin's feature-option and captured-input selection. */
+  configureFeature(feature: Feature, selection: FeatureSelection): void {
+    this.assertOpen()
+    const current = this.state(feature)
+    if (current.state === 'inconsistent') throw new Error(`feature ${feature.id} is inconsistent`)
+    if (current.state === 'absent' || !current.selection) {
+      this.installFeature(feature, selection)
+      return
+    }
+    const normalized = feature.normalizeSelection(selection, this.profile)
+    this.ensureRequirements(feature, normalized, new Set([feature.id]))
+    this.replaceContribution(
+      feature.contribution(current.selection, this.profile),
+      feature.contribution(normalized, this.profile),
+    )
+    current.selection = normalized
+    current.state = current.state === 'disabled' ? 'disabled' : 'enabled'
+    if (current.state === 'disabled') this.setFeatureDisabled(feature, normalized, true)
+    this.assertFeatureConsistent(feature)
+    this.configured.add(feature.id)
+  }
+
+  /** Enable all entries owned by one installed feature. */
+  enableFeature(feature: Feature): void {
+    this.assertOpen()
+    const current = this.state(feature)
+    if (current.state === 'inconsistent') throw new Error(`feature ${feature.id} is inconsistent`)
+    if (current.state === 'absent' || !current.selection) {
+      throw new Error(`feature ${feature.id} is not installed`)
+    }
+    this.setFeatureDisabled(feature, current.selection, false)
+    current.state = 'enabled'
+    this.assertFeatureConsistent(feature)
+    this.disabled.delete(feature.id)
+    this.enabled.add(feature.id)
+  }
+
+  /** Disable an optional feature without removing its configuration. */
+  disableFeature(feature: Feature): void {
+    this.assertOpen()
+    if (feature.required) throw new Error(`required feature ${feature.id} cannot be disabled`)
+    const current = this.state(feature)
+    if (current.state === 'inconsistent') throw new Error(`feature ${feature.id} is inconsistent`)
+    if (current.state === 'absent' || !current.selection) {
+      throw new Error(`feature ${feature.id} is not installed`)
+    }
+    const dependent = this.registry.all().find((candidate) => {
+      const state = this.states.get(candidate.id)
+      return state?.state === 'enabled' && state.selection
+        && candidate.requirements(state.selection).some(requirement => requirement.id === feature.id)
+    })
+    if (dependent) throw new Error(`feature ${feature.id} is required by ${dependent.id}`)
+    this.setFeatureDisabled(feature, current.selection, true)
+    current.state = 'disabled'
+    this.assertFeatureConsistent(feature)
+    this.enabled.delete(feature.id)
+    this.disabled.add(feature.id)
+  }
+
+  /** Add a generated local plugin and all four of its project registrations. */
+  addPlugin(blueprint: LocalPluginBlueprint): void {
+    this.assertOpen()
+    const manifest = this.manifest()
+    const cordis = this.cordis()
+    const tsconfig = this.documents.get('tsconfig.json')
+    if (!(tsconfig instanceof TsConfigFile)) {
+      throw new Error('adding a local plugin requires a valid tsconfig.json')
+    }
+    const packageName = blueprint.packageName(this.profile.name)
+    if (manifest.npmDependency(packageName)) throw new Error(`root NPM dependency already exists: ${packageName}`)
+    const entry = blueprint.cordisConfigEntry(this.profile.name)
+    if (cordis.entry(entry.id)) throw new Error(`Cordis config entry already exists: ${entry.id}`)
+    const documents = blueprint.documents(this.profile.name, this.profile.releaseVersion)
+    for (const document of documents) {
+      if (this.documents.has(document.relativePath)) {
+        throw new Error(`local plugin file already exists: ${document.relativePath}`)
+      }
+    }
+    for (const document of documents) this.documents.set(document.relativePath, document)
+    manifest.setNpmDependency('dependencies', packageName, this.profile.packageManager.localPluginSpec())
+    tsconfig.addReference(`./${blueprint.directory}`)
+    cordis.addEntry(entry)
+    this.addedPlugins.add(entry.id)
+  }
+
+  /** Enable or disable one custom/manual Cordis config entry by stable id. */
+  setCustomPluginDisabled(id: string, disabled: boolean): void {
+    this.assertOpen()
+    const entry = this.cordis().entry(id)
+    if (!entry) throw new Error(`Cordis config entry does not exist: ${id}`)
+    if (this.registry.ownerOfPackage(entry.name, this.profile)) {
+      throw new Error(`Cordis config entry ${id} belongs to a builtin feature`)
+    }
+    this.cordis().setDisabled(id, disabled)
+    if (disabled) {
+      this.enabledPlugins.delete(id)
+      this.disabledPlugins.add(id)
+    } else {
+      this.disabledPlugins.delete(id)
+      this.enabledPlugins.add(id)
+    }
+  }
+
+  /** Summarize all pending domain and file changes. */
+  changes(): ChangeSet {
+    const changedFiles = new Set<string>()
+    for (const [path, document] of this.documents) {
+      if (this.source.origin === 'create' || !sameText(document, this.source.document(path))) changedFiles.add(path)
+    }
+    for (const path of this.removed.keys()) changedFiles.add(path)
+    return {
+      addedFeatures: [...this.added].sort(),
+      enabledFeatures: [...this.enabled].sort(),
+      disabledFeatures: [...this.disabled].sort(),
+      configuredFeatures: [...this.configured].sort(),
+      addedPlugins: [...this.addedPlugins].sort(),
+      enabledPlugins: [...this.enabledPlugins].sort(),
+      disabledPlugins: [...this.disabledPlugins].sort(),
+      changedFiles: [...changedFiles].sort(),
+      npmDependenciesChanged: npmDependencyShape(this.manifest().value())
+        !== npmDependencyShape(this.source.packageManifest()),
+    }
+  }
+
+  /** Validate, detect external edits, write affected files, and return a fresh snapshot. */
+  async commit(): Promise<ProjectCommitResult<SdkProject>> {
+    this.assertOpen()
+    if (this.profile.linkWorkspaceRoot) {
+      const workspace = await LinkWorkspace.open(this.profile.linkWorkspaceRoot)
+      workspace.apply(
+        this.source.root,
+        this.manifest(),
+        this.profile.packageManager,
+        [...this.documents.values()],
+      )
+    }
+    this.validateFinalState()
+    const changes = this.changes()
+    await this.assertUnchanged(changes.changedFiles)
+    await mkdir(this.source.root, { recursive: true })
+    for (const path of changes.changedFiles) {
+      const document = this.documents.get(path)
+      const absolute = resolve(this.source.root, path)
+      if (!document) {
+        await unlink(absolute)
+        continue
+      }
+      await mkdir(dirname(absolute), { recursive: true })
+      await writeFile(absolute, document.serialize(), {
+        encoding: 'utf8',
+        ...document.createMode === undefined ? {} : { mode: document.createMode },
+      })
+    }
+    this.committed = true
+    return { project: await this.source.reopen(), changes }
+  }
+
+  private installFeatureRecursive(
+    feature: Feature,
+    selection: FeatureSelection,
+    stack: Set<FeatureId>,
+  ): void {
+    if (stack.has(feature.id)) throw new Error(`cyclic feature requirement involving ${feature.id}`)
+    const current = this.state(feature)
+    if (current.state === 'inconsistent') throw new Error(`feature ${feature.id} is inconsistent`)
+    if (current.state !== 'absent' && current.selection) {
+      this.configureFeature(feature, selection)
+      if (current.state === 'disabled') this.enableFeature(feature)
+      return
+    }
+    const normalized = feature.normalizeSelection(selection, this.profile)
+    const nextStack = new Set(stack).add(feature.id)
+    this.ensureRequirements(feature, normalized, nextStack)
+    this.replaceContribution(undefined, feature.contribution(normalized, this.profile))
+    current.selection = normalized
+    current.state = 'enabled'
+    this.assertFeatureConsistent(feature)
+    this.added.add(feature.id)
+  }
+
+  private ensureRequirements(feature: Feature, selection: FeatureSelection, stack: Set<FeatureId>): void {
+    for (const requirement of feature.requirements(selection)) {
+      const required = this.registry.get(requirement.id)
+      const state = this.state(required)
+      if (state.state === 'inconsistent') throw new Error(`required feature ${required.id} is inconsistent`)
+      if (state.state === 'absent' || !state.selection) {
+        this.installFeatureRecursive(required, {
+          id: required.id,
+          options: requirement.options ?? required.defaultOptions(this.profile),
+        }, stack)
+      } else {
+        const next = this.selectionWithRequiredOptions(required, state.selection, requirement)
+        if (next !== state.selection) this.configureFeature(required, next)
+        if (state.state === 'disabled') this.enableFeature(required)
+      }
+    }
+  }
+
+  private selectionWithRequiredOptions(
+    feature: Feature,
+    selection: FeatureSelection,
+    requirement: FeatureRequirement,
+  ): FeatureSelection {
+    if (!requirement.options || requirement.options.every(option => selection.options.includes(option))) {
+      return selection
+    }
+    if (feature.mode !== 'multiple') {
+      throw new Error(`${feature.id} does not satisfy the option requirement from another feature`)
+    }
+    return { ...selection, options: [...new Set([...selection.options, ...requirement.options])] }
+  }
+
+  private replaceContribution(
+    previous: ReturnType<Feature['contribution']> | undefined,
+    next: ReturnType<Feature['contribution']>,
+  ): void {
+    const previousByKey = previous?.byKey() ?? new Map<ResourceKey, ProjectResource>()
+    const nextByKey = next.byKey()
+    for (const [key, resource] of previousByKey) {
+      const replacement = nextByKey.get(key)
+      if (!replacement || !canUpdateResource(resource, replacement)) this.removeResource(resource)
+    }
+    for (const [key, resource] of nextByKey) {
+      const previousResource = previousByKey.get(key)
+      this.applyResource(
+        resource,
+        previousResource && canUpdateResource(previousResource, resource) ? previousResource : undefined,
+      )
+    }
+  }
+
+  private applyResource(resource: ProjectResource, previous: ProjectResource | undefined): void {
+    switch (resource.kind) {
+      case 'npm-dependency': {
+        const dependency = resolveNpmDependency(resource.name, resource.section, this.profile.releaseVersion)
+        this.manifest().setNpmDependency(dependency.section, resource.name, dependency.spec)
+        return
+      }
+      case 'package-script': {
+        const manifest = this.manifest()
+        const current = manifest.script(resource.name)
+        if (!previous || previous.kind !== 'package-script') {
+          if (current !== undefined) throw new Error(`feature-owned package script already exists: ${resource.name}`)
+          manifest.setScript(resource.name, resource.command)
+          return
+        }
+        if (current === resource.command) return
+        if (current !== previous.command) {
+          throw new Error(`feature-owned package script was modified: ${resource.name}`)
+        }
+        manifest.setScript(resource.name, resource.command)
+        return
+      }
+      case 'cordis-config-entry': {
+        const current = this.cordis().entry(resource.entry.id)
+        if (!current) this.cordis().addEntry(resource.entry, resource.commentedExample)
+        else {
+          if (current.name !== resource.entry.name) {
+            throw new Error(`Cordis config entry ${resource.entry.id} is owned by ${current.name}, not ${resource.entry.name}`)
+          }
+          this.cordis().updateOwnedConfig(
+            resource.entry.id,
+            resource.ownedConfigKeys,
+            resource.entry.config ?? {},
+          )
+          this.cordis().setDisabled(resource.entry.id, false)
+        }
+        return
+      }
+      case 'environment': {
+        this.environment('.env.example').set(resource.name, resource.exampleValue)
+        /* v8 ignore else -- an omitted secret intentionally materializes only its example placeholder */
+        if (resource.value !== undefined) {
+          const environment = this.environment('.env')
+          environment.append(
+            resource.name,
+            resource.value,
+            resource.value === '' ? resource.comment : undefined,
+          )
+        }
+        return
+      }
+      case 'owned-file': {
+        const existing = this.documents.get(resource.document.relativePath)
+        if (!existing) {
+          this.documents.set(resource.document.relativePath, resource.document.clone())
+          this.removed.delete(resource.document.relativePath)
+          return
+        }
+        if (!previous || previous.kind !== 'owned-file') {
+          throw new Error(`feature-owned file already exists: ${resource.document.relativePath}`)
+        }
+        if (previous.document.serialize() === resource.document.serialize()) return
+        if (existing.serialize() !== previous.document.serialize()) {
+          throw new Error(`feature-owned file was modified: ${resource.document.relativePath}`)
+        }
+        this.documents.set(resource.document.relativePath, resource.document.clone())
+        this.removed.delete(resource.document.relativePath)
+        return
+      }
+    }
+  }
+
+  private removeResource(resource: ProjectResource): void {
+    switch (resource.kind) {
+      case 'npm-dependency':
+        this.manifest().removeNpmDependency(resource.section, resource.name)
+        return
+      case 'package-script': {
+        const manifest = this.manifest()
+        const current = manifest.script(resource.name)
+        if (current === undefined) throw new Error(`owned package script is missing: ${resource.name}`)
+        if (resource.removeOnlyWhenUnchanged && current !== resource.command) {
+          throw new Error(`feature-owned package script was modified: ${resource.name}`)
+        }
+        manifest.removeScript(resource.name)
+        return
+      }
+      case 'cordis-config-entry': {
+        const entry = this.cordis().entry(resource.entry.id)
+        if (!entry || entry.name !== resource.entry.name) {
+          throw new Error(`cannot confirm old Cordis resource ${resource.entry.id}`)
+        }
+        this.cordis().removeEntry(resource.entry.id)
+        return
+      }
+      case 'environment':
+        this.environment('.env.example').remove(resource.name)
+        return
+      case 'owned-file': {
+        const document = this.documents.get(resource.document.relativePath)
+        if (!document) throw new Error(`owned file is missing: ${resource.document.relativePath}`)
+        if (resource.removeOnlyWhenUnchanged && document.serialize() !== resource.document.serialize()) {
+          throw new Error(`owned file was modified: ${resource.document.relativePath}`)
+        }
+        this.documents.delete(resource.document.relativePath)
+        if (this.source.document(resource.document.relativePath)) {
+          this.removed.set(resource.document.relativePath, document)
+        }
+      }
+    }
+  }
+
+  private setFeatureDisabled(feature: Feature, selection: FeatureSelection, disabled: boolean): void {
+    for (const resource of feature.contribution(selection, this.profile).resources) {
+      if (resource.kind === 'cordis-config-entry') this.cordis().setDisabled(resource.entry.id, disabled)
+    }
+  }
+
+  private validateFinalState(): void {
+    for (const document of this.documents.values()) document.validate()
+    const profile = this.finalProfile()
+    const view = this.projectView(profile)
+    for (const feature of this.registry.all()) {
+      const state = this.states.get(feature.id)
+      if (!feature.isApplicable(profile)) {
+        if (state?.state === 'enabled') {
+          throw new Error(`feature ${feature.id} is not available for ${profile.runInterface}`)
+        }
+        continue
+      }
+      const installation = feature.inspect(view)
+      /* v8 ignore next 3 -- public domain commands assert feature consistency before final validation */
+      if (installation.state === 'inconsistent') {
+        throw new Error(`feature ${feature.id} is inconsistent: ${installation.diagnostics.join('; ')}`)
+      }
+      /* v8 ignore next 3 -- required features are installed by creation and cannot be disabled by public commands */
+      if (feature.required && installation.state !== 'enabled') {
+        throw new Error(`required feature ${feature.id} must be installed and enabled`)
+      }
+      if (installation.state !== 'enabled' || !installation.selection) continue
+      for (const requirement of feature.requirements(installation.selection)) {
+        const required = this.registry.get(requirement.id).inspect(view)
+        /* v8 ignore next 3 -- ensureRequirements establishes enabled requirements before contributions change */
+        if (required.state !== 'enabled') {
+          throw new Error(`feature ${feature.id} requires enabled ${requirement.id}`)
+        }
+        for (const option of requirement.options ?? []) {
+          /* v8 ignore next 3 -- selectionWithRequiredOptions establishes required options before commit */
+          if (!required.options.includes(option)) {
+            throw new Error(`feature ${feature.id} requires ${requirement.id} option ${option}`)
+          }
+        }
+      }
+    }
+  }
+
+  private assertFeatureConsistent(feature: Feature): void {
+    const installation = feature.inspect(this)
+    /* v8 ignore next 3 -- resource application either succeeds completely or throws at the owning operation */
+    if (installation.state === 'inconsistent') {
+      throw new Error(`feature ${feature.id} is inconsistent: ${installation.diagnostics.join('; ')}`)
+    }
+  }
+
+  private finalProfile(): ProjectProfile {
+    const runInterface = this.states.get(featureId('app'))?.selection?.options[0]
+    if (runInterface !== 'acp' && runInterface !== 'stdio' && runInterface !== 'embed') return this.profile
+    return { ...this.profile, runInterface }
+  }
+
+  private projectView(profile: ProjectProfile): FeatureProjectView {
+    return {
+      profile,
+      cordisConfigEntries: () => this.cordisConfigEntries(),
+      packageManifest: () => this.packageManifest(),
+      hasDocument: path => this.hasDocument(path),
+      readEnvironment: (path, name) => this.readEnvironment(path, name),
+    }
+  }
+
+  private async assertUnchanged(paths: readonly string[]): Promise<void> {
+    for (const path of paths) {
+      const source = this.source.document(path)
+      const absolute = resolve(this.source.root, path)
+      try {
+        const current = await readFile(absolute, 'utf8')
+        if (source?.originalText === undefined || current !== source.originalText) {
+          throw new Error(`project file changed outside this edit session: ${path}`)
+        }
+      } catch (error) {
+        const code = (error as NodeJS.ErrnoException).code
+        if (code === 'ENOENT' && source?.originalText === undefined) continue
+        if (error instanceof Error && error.message.startsWith('project file changed outside')) throw error
+        throw new Error(`cannot verify project file ${path}: ${asError(error).message}`)
+      }
+    }
+  }
+
+  private state(feature: Feature): MutableFeatureState {
+    const state = this.states.get(feature.id)
+    if (!state) throw new Error(`feature ${feature.id} is not applicable to this project`)
+    return state
+  }
+
+  private manifest(): PackageJsonFile {
+    const document = this.documents.get('package.json')
+    if (!(document instanceof PackageJsonFile)) throw new Error('project package.json is missing')
+    return document
+  }
+
+  private cordis(): CordisYamlFile {
+    const document = this.documents.get('cordis.yml')
+    if (!(document instanceof CordisYamlFile)) throw new Error('project cordis.yml is missing')
+    return document
+  }
+
+  private environment(path: '.env' | '.env.example'): EnvFile {
+    const existing = this.documents.get(path)
+    if (existing instanceof EnvFile) return existing
+    if (existing) throw new Error(`${path} is not an environment document`)
+    const document = EnvFile.create(path)
+    this.documents.set(path, document)
+    return document
+  }
+
+  private assertOpen(): void {
+    if (this.committed) throw new Error('project edit session has already committed')
+  }
+}
diff --git a/packages/sdk/helper/src/project/sdk-project.ts b/packages/sdk/helper/src/project/sdk-project.ts
new file mode 100644
index 0000000000..cd55ffe2e5
--- /dev/null
+++ b/packages/sdk/helper/src/project/sdk-project.ts
@@ -0,0 +1,305 @@
+/**
+ * Read-only aggregate for one generated or existing SDK project.
+ *
+ * @module @deepseek-ai/dsh-helper/project/sdk-project
+ */
+
+import { access, readFile } from 'node:fs/promises'
+import { basename, resolve } from 'node:path'
+import { CordisYamlFile, type CordisConfigEntry } from '../documents/cordis-yaml-file.ts'
+import { EnvFile } from '../documents/env-file.ts'
+import { PackageJsonFile, type PackageManifest } from '../documents/package-json-file.ts'
+import { PnpmWorkspaceFile } from '../documents/pnpm-workspace-file.ts'
+import { ProjectFile, TextProjectFile } from '../documents/project-file.ts'
+import { TsConfigFile } from '../documents/tsconfig-file.ts'
+import {
+  createPackageManager,
+  type PackageManager,
+  type PackageManagerName,
+} from '../package-managers/package-manager.ts'
+import {
+  createBaselineProjectArtifacts,
+  createPackageJsonDoc,
+  createProjectTemplateContext,
+} from '../templates/project-template.ts'
+import type { ProjectCreationRequest, ProjectProfile, RunInterface } from './types.ts'
+import type { FeatureRegistry } from '../features/registry.ts'
+import { ProjectEditSession } from './project-edit-session.ts'
+
+/** Whether a project snapshot describes uncommitted creation or files on disk. */
+export type ProjectOrigin = 'create' | 'disk'
+
+const OPTIONAL_DOCUMENTS = [
+  '.env',
+  '.env.example',
+  'tsconfig.json',
+  'pnpm-workspace.yaml',
+  'hooks.json',
+  'codex-hooks.json',
+  'README.md',
+  'index.ts',
+] as const
+
+function runInterface(entries: readonly CordisConfigEntry[]): RunInterface {
+  if (entries.some(entry => entry.name === '@deepseek-ai/dsh-acp')) return 'acp'
+  if (entries.some(entry => entry.name === '@deepseek-ai/dsh-stdio')) return 'stdio'
+  return 'embed'
+}
+
+function runtimeModel(entries: readonly CordisConfigEntry[]): string {
+  const acp = entries.find(entry => entry.name === '@deepseek-ai/dsh-acp')
+  if (typeof acp?.config?.model === 'string' && acp.config.model.length > 0) return acp.config.model
+  const provider = entries.find(entry => entry.name === '@deepseek-ai/dsh-llm-deepseek'
+    || entry.name === '@deepseek-ai/dsh-llm-pi-ai')
+  const models = provider?.config?.models
+  if (Array.isArray(models) && typeof models[0] === 'string') return models[0]
+  return 'deepseek-v4-flash'
+}
+
+function releaseVersion(manifest: Readonly<PackageManifest>): string {
+  const spec = manifest.dependencies?.['@deepseek-ai/dsh-scripts']
+  const match = spec && /(?:^|[^0-9])(\d+\.\d+\.\d+(?:-[0-9A-Za-z.-]+)?)/.exec(spec)
+  return match?.[1] ?? '0.0.1'
+}
+
+async function pathExists(path: string): Promise<boolean> {
+  try {
+    await access(path)
+    return true
+  } catch (error) {
+    /* v8 ignore else -- the other arm requires a filesystem permission/IO fault from access */
+    if ((error as NodeJS.ErrnoException).code === 'ENOENT') return false
+    /* v8 ignore next -- paired with the ignored defensive access-error arm above */
+    throw error
+  }
+}
+
+async function detectPackageManager(root: string, manifest: Readonly<PackageManifest>): Promise<PackageManager> {
+  let name: PackageManagerName = 'npm'
+  let version = '10.0.0'
+  const field = manifest.packageManager
+  if (field) {
+    const match = /^(npm|pnpm|yarn)@(.+)$/.exec(field)
+    if (!match?.[1] || !match[2]) throw new Error(`invalid packageManager field: ${field}`)
+    name = match[1] as PackageManagerName
+    version = match[2]
+  } else if (await pathExists(resolve(root, 'pnpm-lock.yaml'))) {
+    name = 'pnpm'
+    version = '10.0.0'
+  } else if (await pathExists(resolve(root, 'yarn.lock'))) {
+    name = 'yarn'
+    version = '2.0.0'
+  }
+  return createPackageManager(name, version)
+}
+
+function linkedRepositoryRoot(root: string, manifest: Readonly<PackageManifest>): string | undefined {
+  const spec = manifest.dependencies?.['@deepseek-ai/dsh-scripts']
+  const match = /^(?:file|link|portal):(.+)\/packages\/sdk\/scripts\/?$/.exec(spec ?? '')
+  return match?.[1] ? resolve(root, match[1]) : undefined
+}
+
+function parseOptionalDocument(path: string, text: string): ProjectFile {
+  try {
+    switch (path) {
+      case '.env': return EnvFile.parse('.env', text)
+      case '.env.example': return EnvFile.parse('.env.example', text)
+      case 'tsconfig.json': return TsConfigFile.parse(text)
+      case 'pnpm-workspace.yaml': return PnpmWorkspaceFile.parse(text)
+      default: return new TextProjectFile(path, text, text)
+    }
+  } catch {
+    // Optional malformed resources do not invalidate the project aggregate;
+    // an operation that needs their structure checks the concrete document type.
+    return new TextProjectFile(path, text, text)
+  }
+}
+
+/** A project snapshot whose documents can only be changed through {@link ProjectEditSession}. */
+export class SdkProject {
+  /** Absolute project directory. */
+  readonly root: string
+  /** Whether this snapshot is an uncommitted blueprint or disk state. */
+  readonly origin: ProjectOrigin
+  /** Project identity, runtime, interface, and package-manager context. */
+  readonly profile: ProjectProfile
+  private readonly documents: ReadonlyMap<string, ProjectFile>
+
+  private constructor(
+    root: string,
+    origin: ProjectOrigin,
+    profile: ProjectProfile,
+    documents: ReadonlyMap<string, ProjectFile>,
+  ) {
+    this.root = resolve(root)
+    this.origin = origin
+    this.profile = profile
+    this.documents = documents
+  }
+
+  /**
+   * Build an in-memory project blueprint without touching the target directory.
+   * @param root - target project directory.
+   * @param request - complete creation request.
+   * @returns uncommitted project snapshot.
+   */
+  static create(root: string, request: ProjectCreationRequest): SdkProject {
+    const app = request.features.find(selection => selection.id === 'app')
+    const selectedInterface = app?.options[0]
+    if (selectedInterface !== 'acp' && selectedInterface !== 'stdio' && selectedInterface !== 'embed') {
+      throw new Error('project creation requires one app feature option')
+    }
+    const profile: ProjectProfile = {
+      name: request.name,
+      description: request.description,
+      runtime: request.runtime,
+      runInterface: selectedInterface,
+      packageManager: request.packageManager,
+      releaseVersion: request.releaseVersion,
+      ...request.linkWorkspaceRoot ? { linkWorkspaceRoot: resolve(request.linkWorkspaceRoot) } : {},
+    }
+    const templates = createProjectTemplateContext(profile)
+    const manifest = createPackageJsonDoc(templates)
+    const documents = new Map<string, ProjectFile>()
+    documents.set(manifest.relativePath, manifest)
+    documents.set('cordis.yml', CordisYamlFile.create())
+    documents.set('.env.example', EnvFile.create('.env.example'))
+    documents.set('tsconfig.json', TsConfigFile.create())
+    for (const document of request.packageManager.configureWorkspace(manifest)) {
+      documents.set(document.relativePath, document)
+    }
+    for (const document of createBaselineProjectArtifacts(templates)) {
+      documents.set(document.relativePath, document)
+    }
+    return new SdkProject(root, 'create', profile, documents)
+  }
+
+  /**
+   * Load an existing project from required and SDK-managed optional files.
+   * @param root - existing project directory.
+   * @returns disk-backed project snapshot.
+   */
+  static async open(root: string): Promise<SdkProject> {
+    const absolute = resolve(root)
+    const [manifestText, cordisText] = await Promise.all([
+      readFile(resolve(absolute, 'package.json'), 'utf8'),
+      readFile(resolve(absolute, 'cordis.yml'), 'utf8'),
+    ])
+    const manifest = PackageJsonFile.parse(manifestText)
+    const cordis = CordisYamlFile.parse(cordisText)
+    const value = manifest.value()
+    const manager = await detectPackageManager(absolute, value)
+    const entries = cordis.entries()
+    const linkWorkspaceRoot = linkedRepositoryRoot(absolute, value)
+    const profile: ProjectProfile = {
+      name: value.name ?? basename(absolute),
+      description: typeof value.description === 'string' ? value.description : '',
+      runtime: { model: runtimeModel(entries) },
+      runInterface: runInterface(entries),
+      packageManager: manager,
+      releaseVersion: releaseVersion(value),
+      ...linkWorkspaceRoot ? { linkWorkspaceRoot } : {},
+    }
+    const documents = new Map<string, ProjectFile>([
+      ['package.json', manifest],
+      ['cordis.yml', cordis],
+    ])
+    await Promise.all(OPTIONAL_DOCUMENTS.map(async (path) => {
+      try {
+        const text = await readFile(resolve(absolute, path), 'utf8')
+        documents.set(path, parseOptionalDocument(path, text))
+      } catch (error) {
+        /* v8 ignore next -- optional-file reads fail normally only with ENOENT; other IO faults surface */
+        if ((error as NodeJS.ErrnoException).code !== 'ENOENT') throw error
+      }
+    }))
+    return new SdkProject(absolute, 'disk', profile, documents)
+  }
+
+  /**
+   * Read the root package manifest defensively.
+   * @returns cloned manifest value.
+   */
+  packageManifest(): Readonly<PackageManifest> {
+    return this.packageJson.value()
+  }
+
+  /**
+   * Read Cordis config entries defensively in file order.
+   * @returns cloned Cordis config entries.
+   */
+  cordisConfigEntries(): readonly CordisConfigEntry[] {
+    return this.cordis.entries()
+  }
+
+  /**
+   * Check whether this snapshot contains one managed document.
+   * @param path - project-relative document path.
+   * @returns whether the document is loaded.
+   */
+  hasDocument(path: string): boolean {
+    return this.documents.has(path)
+  }
+
+  /**
+   * Read one environment variable from a loaded dotenv document.
+   * @param path - environment file to read.
+   * @param name - variable name.
+   * @returns variable value when present.
+   */
+  readEnvironment(path: '.env' | '.env.example', name: string): string | undefined {
+    const document = this.documents.get(path)
+    if (!document) return undefined
+    if (!(document instanceof EnvFile)) throw new Error(`${path} is not an environment document`)
+    return document.get(name)
+  }
+
+  /** Read the root package document. */
+  get packageJson(): PackageJsonFile {
+    const document = this.documents.get('package.json')
+    if (!(document instanceof PackageJsonFile)) throw new Error('project package.json is missing or invalid')
+    return document
+  }
+
+  /** Read the root Cordis document. */
+  get cordis(): CordisYamlFile {
+    const document = this.documents.get('cordis.yml')
+    if (!(document instanceof CordisYamlFile)) throw new Error('project cordis.yml is missing or invalid')
+    return document
+  }
+
+  /**
+   * Return one managed document without exposing the aggregate map.
+   * @param path - project-relative document path.
+   * @returns loaded document when present.
+   */
+  document(path: string): ProjectFile | undefined {
+    return this.documents.get(path)
+  }
+
+  /**
+   * Create the only mutable boundary for this snapshot.
+   * @param registry - feature catalog governing edits.
+   * @returns isolated edit session.
+   */
+  edit(registry: FeatureRegistry): ProjectEditSession {
+    return new ProjectEditSession(this, registry)
+  }
+
+  /**
+   * Clone every managed document for an isolated edit session.
+   * @returns project-relative document map.
+   */
+  cloneDocuments(): Map<string, ProjectFile> {
+    return new Map([...this.documents].map(([path, document]) => [path, document.clone()]))
+  }
+
+  /**
+   * Reload this aggregate from committed disk state.
+   * @returns fresh disk-backed snapshot.
+   */
+  reopen(): Promise<SdkProject> {
+    return SdkProject.open(this.root)
+  }
+
+}
diff --git a/packages/sdk/helper/src/project/types.ts b/packages/sdk/helper/src/project/types.ts
new file mode 100644
index 0000000000..44d06d508c
--- /dev/null
+++ b/packages/sdk/helper/src/project/types.ts
@@ -0,0 +1,48 @@
+/**
+ * Shared creation and project-profile values for SDK project editing.
+ *
+ * @module @deepseek-ai/dsh-helper/project/types
+ */
+
+import type { PackageManager } from '../package-managers/package-manager.ts'
+import type { LocalPluginBlueprint } from '../plugins/local-plugin-blueprint.ts'
+import type { FeatureId } from '../ids.ts'
+
+/** Runtime front door selected for a generated project. */
+export type RunInterface = 'acp' | 'stdio' | 'embed'
+
+/** Values shared by the required provider and app features. */
+interface ProjectRuntimeOptions {
+  model: string
+}
+
+/** Selected options and captured secrets for one feature. */
+export interface FeatureSelection {
+  id: FeatureId
+  options: readonly string[]
+  values?: Readonly<Record<string, unknown>>
+  secrets?: Readonly<Record<string, string>>
+}
+
+/** Stable context available to project and feature objects. */
+export interface ProjectProfile {
+  name: string
+  description: string
+  runtime: ProjectRuntimeOptions
+  runInterface: RunInterface
+  packageManager: PackageManager
+  releaseVersion: string
+  linkWorkspaceRoot?: string
+}
+
+/** Fully collected create request; it contains intent, never rendered file text. */
+export interface ProjectCreationRequest {
+  name: string
+  description: string
+  runtime: ProjectRuntimeOptions
+  packageManager: PackageManager
+  releaseVersion: string
+  linkWorkspaceRoot?: string
+  features: readonly FeatureSelection[]
+  localPlugins: readonly LocalPluginBlueprint[]
+}
diff --git a/packages/sdk/helper/src/questions/clack-nested-multiselect.ts b/packages/sdk/helper/src/questions/clack-nested-multiselect.ts
new file mode 100644
index 0000000000..1633bdedf7
--- /dev/null
+++ b/packages/sdk/helper/src/questions/clack-nested-multiselect.ts
@@ -0,0 +1,304 @@
+/**
+ * Tree-shaped Clack picker for root checkboxes with finite child options.
+ *
+ * @module @deepseek-ai/dsh-helper/questions/clack-nested-multiselect
+ */
+
+import { styleText } from 'node:util'
+import type { Readable, Writable } from 'node:stream'
+import { Prompt, isCancel } from '@clack/core'
+import {
+  S_BAR,
+  S_BAR_END,
+  S_CHECKBOX_ACTIVE,
+  S_CHECKBOX_INACTIVE,
+  S_CHECKBOX_SELECTED,
+  S_RADIO_ACTIVE,
+  S_RADIO_INACTIVE,
+  symbol,
+  symbolBar,
+} from '@clack/prompts'
+import type {
+  NestedMultiSelectOption,
+  NestedMultiSelectRequest,
+  NestedMultiSelectValue,
+  PromptOutcome,
+} from './prompt-port.ts'
+
+interface NestedPromptOptions<TValue, TChoice> extends NestedMultiSelectRequest<TValue, TChoice> {
+  input: Readable
+  output: Writable
+}
+
+class NestedPrompt<TValue, TChoice> extends Prompt<readonly NestedMultiSelectValue<TValue, TChoice>[]> {
+  readonly options: readonly NestedMultiSelectOption<TValue, TChoice>[]
+  private readonly selected = new Set<TValue>()
+  private readonly selectedChoices = new Map<TValue, Set<TChoice>>()
+  private readonly initialSelected: Set<TValue>
+  private readonly initialChoices: Map<TValue, Set<TChoice>>
+  private readonly showChanges: boolean
+  private layer: 'root' | 'choices' = 'root'
+  private rootCursor = 0
+  private choiceCursor = 0
+
+  constructor(options: NestedPromptOptions<TValue, TChoice>) {
+    super({
+      input: options.input,
+      output: options.output,
+      validate: value => NestedPrompt.validate(options.options, value),
+      render(this: Prompt<readonly NestedMultiSelectValue<TValue, TChoice>[]>) {
+        return (this as NestedPrompt<TValue, TChoice>).renderFrame(options.message)
+      },
+    }, false)
+    this.options = options.options
+    this.showChanges = options.showChanges ?? false
+    for (const option of options.options) {
+      if (option.required || option.default) this.selected.add(option.value)
+      this.selectedChoices.set(option.value, new Set(
+        option.choices?.filter(choice => choice.default).map(choice => choice.value) ?? [],
+      ))
+    }
+    this.initialSelected = new Set(this.selected)
+    this.initialChoices = new Map([...this.selectedChoices].map(([value, choices]) => [
+      value, new Set(choices),
+    ]))
+    this.updateValue()
+    this.on('cursor', (action) => { this.handleAction(action) })
+  }
+
+  private static validate<TValue, TChoice>(
+    options: readonly NestedMultiSelectOption<TValue, TChoice>[],
+    value: readonly NestedMultiSelectValue<TValue, TChoice>[] | undefined,
+  ): string | undefined {
+    /* v8 ignore next -- NestedPrompt initializes its value before submission validation */
+    const selected = new Map(value?.map(item => [item.value, item.choices]) ?? [])
+    for (const option of options) {
+      if (option.disabled) continue
+      /* v8 ignore next -- required options initialize selected and cannot be toggled off */
+      if (option.required && !selected.has(option.value)) return `${option.label} is required`
+      if (!selected.has(option.value) || !option.choiceMode) continue
+      const choices = selected.get(option.value)
+      /* v8 ignore next -- selected.has above guarantees the map value exists */
+      if (!choices) continue
+      const count = choices.length
+      if (option.choiceMode === 'exclusive' && count !== 1) return `Choose one ${option.label} option`
+      if (option.choiceMode === 'multiple' && count === 0) return `Choose at least one ${option.label} option`
+    }
+    return undefined
+  }
+
+  protected override _shouldSubmit(): boolean {
+    if (this.layer === 'choices') {
+      this.leaveChoices()
+      return false
+    }
+    return true
+  }
+
+  private handleAction(action: string | undefined): void {
+    if (this.layer === 'root') this.handleRootAction(action)
+    else this.handleChoiceAction(action)
+    this.updateValue()
+  }
+
+  private handleRootAction(action: string | undefined): void {
+    if (action === 'up') this.rootCursor = this.move(this.rootCursor, -1, this.options.length)
+    if (action === 'down') this.rootCursor = this.move(this.rootCursor, 1, this.options.length)
+    const option = this.options[this.rootCursor]
+    /* v8 ignore next -- Clack cannot emit a cursor action when the option list is empty */
+    if (!option) return
+    if (action === 'space' && !option.required && !option.disabled) {
+      if (this.selected.has(option.value)) this.selected.delete(option.value)
+      else this.selected.add(option.value)
+    }
+    if (action === 'right' && !option.disabled && option.choices && option.choices.length > 0) {
+      this.selected.add(option.value)
+      this.layer = 'choices'
+      const selected = this.selectedChoices.get(option.value)
+      const selectedIndex = option.choices.findIndex(choice => selected?.has(choice.value))
+      this.choiceCursor = Math.max(selectedIndex, 0)
+    }
+  }
+
+  private handleChoiceAction(action: string | undefined): void {
+    const rootOption = this.options[this.rootCursor]
+    /* v8 ignore next -- the choices layer is entered only from a concrete root option */
+    if (!rootOption) return
+    /* v8 ignore next -- the choices layer is entered only for a non-empty choices array */
+    const choices = rootOption.choices ?? []
+    if (action === 'left') {
+      this.leaveChoices()
+      return
+    }
+    if (action === 'up') this.choiceCursor = this.move(this.choiceCursor, -1, choices.length)
+    if (action === 'down') this.choiceCursor = this.move(this.choiceCursor, 1, choices.length)
+    if ((action === 'up' || action === 'down') && rootOption.choiceMode === 'exclusive') {
+      const choice = choices[this.choiceCursor]
+      /* v8 ignore else -- a cursor in the non-empty choices layer always addresses a choice */
+      if (choice) this.selectedChoices.set(rootOption.value, new Set([choice.value]))
+    }
+    if (action !== 'space' && action !== 'right') return
+    const choice = choices[this.choiceCursor]
+    /* v8 ignore next -- the choices layer requires a non-empty choice list */
+    if (!choice) return
+    /* v8 ignore next -- every root option initializes its choice set in the constructor */
+    const selected = this.selectedChoices.get(rootOption.value) ?? new Set<TChoice>()
+    if (rootOption.choiceMode === 'exclusive') {
+      selected.clear()
+      selected.add(choice.value)
+    } else if (selected.has(choice.value)) selected.delete(choice.value)
+    else selected.add(choice.value)
+    this.selectedChoices.set(rootOption.value, selected)
+  }
+
+  private move(cursor: number, offset: number, length: number): number {
+    /* v8 ignore next -- cursor movement is emitted only for a non-empty displayed list */
+    if (length === 0) return 0
+    return (cursor + offset + length) % length
+  }
+
+  private updateValue(): void {
+    this._setValue(this.options.filter(option => this.selected.has(option.value)).map(option => ({
+      value: option.value,
+      /* v8 ignore next -- every root option initializes its choice set in the constructor */
+      choices: [...this.selectedChoices.get(option.value) ?? []],
+    })))
+  }
+
+  private renderFrame(message: string): string {
+    const header = `${symbolBar(this.state)}  ${message}`
+    if (this.state === 'submit') {
+      /* v8 ignore next -- NestedPrompt initializes its value before it can submit */
+      const summary = (this.value ?? []).map(item => this.options.find(option => option.value === item.value)?.label)
+        .filter(Boolean).join(', ') || 'none'
+      return `${symbol(this.state)}  ${message}\n${styleText('gray', S_BAR)}  ${styleText('dim', summary)}`
+    }
+    if (this.state === 'cancel') return `${symbol(this.state)}  ${message}`
+    const body = this.layer === 'root' ? this.renderRoot() : this.renderChoices()
+    const instructions = this.layer === 'root'
+      ? `${styleText('dim', '↑/↓')} navigate  ${styleText('dim', 'Space')} select  ${styleText('dim', '→')} configure  ${styleText('dim', 'Enter')} confirm`
+      : `${styleText('dim', '↑/↓')} navigate  ${styleText('dim', 'Space/→')} select  ${styleText('dim', '←/Enter')} back`
+    const error = this.state === 'error' ? `\n${styleText('yellow', `${S_BAR_END}  ${this.error}`)}` : ''
+    return `${header}\n${styleText('cyan', S_BAR)}  ${body.join(`\n${styleText('cyan', S_BAR)}  `)}\n${styleText('cyan', S_BAR_END)}  ${instructions}${error}`
+  }
+
+  private renderRoot(): string[] {
+    return this.options.map((option, index) => {
+      const active = index === this.rootCursor
+      const selected = this.selected.has(option.value)
+      const focus = active ? styleText('cyan', '›') : ' '
+      const checkbox = selected
+        ? styleText('green', S_CHECKBOX_SELECTED)
+        : styleText('dim', active ? S_CHECKBOX_ACTIVE : S_CHECKBOX_INACTIVE)
+      const choices = option.choices?.filter(choice => this.selectedChoices.get(option.value)?.has(choice.value))
+        .map(choice => choice.label).join(', ')
+      const suffix = option.choices?.length
+        ? ` ${styleText('dim', `* →${choices ? ` ${choices}` : ''}`)}`
+        : ''
+      const required = option.required ? ` ${styleText('yellow', '(required)')}` : ''
+      const issue = this.choiceIssue(option)
+      const warningText = option.warning ?? issue
+      const warning = warningText ? ` ${styleText('yellow', `▲ ${warningText}`)}` : ''
+      const changed = this.optionChanged(option)
+      const change = changed ? ` ${styleText('yellow', '● changed')}` : ''
+      const label = active
+        ? styleText('cyan', option.label)
+        : changed
+          ? styleText('yellow', option.label)
+          : selected ? styleText('green', option.label) : styleText('dim', option.label)
+      const line = `${focus} ${checkbox} ${label}${required}${suffix}${warning}${change}`
+      return option.disabled ? styleText('gray', line) : line
+    })
+  }
+
+  private renderChoices(): string[] {
+    const rootOption = this.options[this.rootCursor]
+    /* v8 ignore next -- renderChoices runs only after entering from a concrete root option */
+    if (!rootOption) return []
+    /* v8 ignore next -- every root option initializes its choice set in the constructor */
+    const selected = this.selectedChoices.get(rootOption.value) ?? new Set<TChoice>()
+    const issue = this.choiceIssue(rootOption)
+    const changed = this.optionChanged(rootOption)
+    const header = styleText('dim', `${rootOption.label} options`)
+      + (issue ? ` ${styleText('yellow', `▲ ${issue}`)}` : '')
+      + (changed ? ` ${styleText('yellow', '● changed')}` : '')
+    const choices = rootOption.choices
+    /* v8 ignore next -- the choices layer is entered only for a non-empty choices array */
+    if (!choices) return [header]
+    return [
+      header,
+      ...choices.map((choice, index) => {
+        const active = index === this.choiceCursor
+        const checked = selected.has(choice.value)
+        const choiceChanged = this.choiceChanged(rootOption.value, choice.value)
+        const focus = active ? styleText('cyan', '›') : ' '
+        const marker = rootOption.choiceMode === 'exclusive'
+          ? checked ? styleText('green', S_RADIO_ACTIVE) : styleText('dim', S_RADIO_INACTIVE)
+          : checked ? styleText('green', S_CHECKBOX_SELECTED) : styleText('dim', S_CHECKBOX_INACTIVE)
+        const label = active
+          ? styleText('cyan', choice.label)
+          : choiceChanged
+            ? styleText('yellow', choice.label)
+            : checked ? styleText('green', choice.label) : styleText('dim', choice.label)
+        const change = choiceChanged ? ` ${styleText('yellow', '●')}` : ''
+        return `${focus} ${marker} ${label}${change}`
+      }),
+    ]
+  }
+
+  private optionChanged(option: NestedMultiSelectOption<TValue, TChoice>): boolean {
+    if (!this.showChanges) return false
+    const selected = this.selected.has(option.value)
+    const initiallySelected = this.initialSelected.has(option.value)
+    if (selected !== initiallySelected) return true
+    if (!selected) return false
+    /* v8 ignore next -- every root option initializes both current and baseline option sets */
+    const current = this.selectedChoices.get(option.value) ?? new Set<TChoice>()
+    /* v8 ignore next -- every root option initializes both current and baseline option sets */
+    const initial = this.initialChoices.get(option.value) ?? new Set<TChoice>()
+    return current.size !== initial.size || [...current].some(value => !initial.has(value))
+  }
+
+  private choiceChanged(value: TValue, choice: TChoice): boolean {
+    if (!this.showChanges) return false
+    return this.selectedChoices.get(value)?.has(choice) !== this.initialChoices.get(value)?.has(choice)
+  }
+
+  private choiceIssue(option: NestedMultiSelectOption<TValue, TChoice>): string | undefined {
+    if (option.disabled || !this.selected.has(option.value) || !option.choiceMode) return undefined
+    /* v8 ignore next -- every root option initializes its choice set in the constructor */
+    const count = this.selectedChoices.get(option.value)?.size ?? 0
+    if (option.choiceMode === 'exclusive' && count !== 1) return 'choose one'
+    if (option.choiceMode === 'multiple' && count === 0) return 'choose at least one'
+    return undefined
+  }
+
+  private leaveChoices(): boolean {
+    const option = this.options[this.rootCursor]
+    /* v8 ignore next -- leaveChoices runs only after entering from a concrete root option */
+    if (!option) return false
+    const issue = this.choiceIssue(option)
+    if (issue) {
+      this.error = `${option.label}: ${issue}`
+      this.state = 'error'
+      return false
+    }
+    this.error = ''
+    this.layer = 'root'
+    return true
+  }
+}
+
+/** Run the nested picker with Clack's standard cancellation symbol. */
+export async function clackNestedMultiselect<TValue, TChoice>(
+  request: NestedPromptOptions<TValue, TChoice>,
+): Promise<PromptOutcome<readonly NestedMultiSelectValue<TValue, TChoice>[]>> {
+  const value = await new NestedPrompt(request).prompt()
+  return isCancel(value)
+    ? { status: 'cancelled' }
+    : {
+      status: 'answered',
+      /* v8 ignore next -- NestedPrompt initializes its value before it can submit */
+      value: value ?? [],
+    }
+}
diff --git a/packages/sdk/helper/src/questions/clack-prompt-port.ts b/packages/sdk/helper/src/questions/clack-prompt-port.ts
new file mode 100644
index 0000000000..c5ebff1427
--- /dev/null
+++ b/packages/sdk/helper/src/questions/clack-prompt-port.ts
@@ -0,0 +1,127 @@
+/**
+ * Thin @clack/prompts adapter for the shared prompt port.
+ *
+ * @module @deepseek-ai/dsh-helper/questions/clack-prompt-port
+ */
+
+import type { Readable, Writable } from 'node:stream'
+import { styleText } from 'node:util'
+import {
+  confirm,
+  isCancel,
+  multiselect,
+  password,
+  select,
+  text,
+  S_WARN,
+} from '@clack/prompts'
+import type { Option } from '@clack/prompts'
+import type {
+  ConfirmPromptRequest,
+  MultiSelectPromptRequest,
+  NestedMultiSelectRequest,
+  NestedMultiSelectValue,
+  PromptOutcome,
+  PromptPort,
+  SecretPromptRequest,
+  SelectPromptRequest,
+  TextPromptRequest,
+} from './prompt-port.ts'
+import { clackNestedMultiselect } from './clack-nested-multiselect.ts'
+
+function outcome<T>(value: T | symbol): PromptOutcome<T> {
+  return isCancel(value) ? { status: 'cancelled' } : { status: 'answered', value }
+}
+
+function clackOptions<T>(values: readonly import('./prompt-port.ts').PromptOption<T>[]): Option<T>[] {
+  return values.map(value => ({
+    value: value.value,
+    label: value.label,
+    ...value.hint === undefined ? {} : { hint: value.hint },
+    ...value.disabled === undefined ? {} : { disabled: value.disabled },
+  })) as Option<T>[]
+}
+
+/** Clack-backed prompt adapter with injectable streams for snapshots and tests. */
+export class ClackPromptPort implements PromptPort {
+  private readonly input: Readable
+  private readonly output: Writable
+
+  /** Bind all prompts to one input/output pair. */
+  constructor(input: Readable = process.stdin, output: Writable = process.stdout) {
+    this.input = input
+    this.output = output
+  }
+
+  /** Ask for visible text through clack. */
+  async text(request: TextPromptRequest): Promise<PromptOutcome<string>> {
+    return outcome(await text({
+      message: request.message,
+      ...request.placeholder === undefined ? {} : { placeholder: request.placeholder },
+      ...request.initialValue === undefined ? {} : { initialValue: request.initialValue },
+      ...request.defaultValue === undefined ? {} : { defaultValue: request.defaultValue },
+      ...request.validate === undefined
+        ? {}
+        : {
+          /* v8 ignore next -- value/default precedence is exercised through the adapter contract tests */
+          validate: value => request.validate?.(value || request.defaultValue || ''),
+        },
+      input: this.input,
+      output: this.output,
+    }))
+  }
+
+  /** Ask for a masked secret through clack. */
+  async secret(request: SecretPromptRequest): Promise<PromptOutcome<string>> {
+    return outcome(await password({
+      message: request.message,
+      ...request.validate === undefined ? {} : {
+        /* v8 ignore next -- @clack/password always calls validation with a string; fallback is defensive */
+        validate: value => request.validate?.(value ?? ''),
+      },
+      input: this.input,
+      output: this.output,
+    }))
+  }
+
+  /** Ask for one option through clack. */
+  async select<T>(request: SelectPromptRequest<T>): Promise<PromptOutcome<T>> {
+    return outcome(await select({
+      ...request,
+      options: clackOptions(request.options),
+      input: this.input,
+      output: this.output,
+    }))
+  }
+
+  /** Ask for multiple options through clack. */
+  async multiselect<T>(request: MultiSelectPromptRequest<T>): Promise<PromptOutcome<readonly T[]>> {
+    return outcome(await multiselect({
+      message: request.message,
+      options: clackOptions(request.options),
+      ...request.initialValues === undefined ? {} : { initialValues: [...request.initialValues] },
+      ...request.required === undefined ? {} : { required: request.required },
+      input: this.input,
+      output: this.output,
+    }))
+  }
+
+  /** Ask for confirmation through clack. */
+  async confirm(request: ConfirmPromptRequest): Promise<PromptOutcome<boolean>> {
+    return outcome(await confirm({
+      message: request.tone === 'warning'
+        ? styleText('yellow', `${S_WARN} ${request.message}`)
+        : request.message,
+      ...request.initialValue === undefined ? {} : { initialValue: request.initialValue },
+      input: this.input,
+      output: this.output,
+    }))
+  }
+
+  /** Select root values and finite child options in one tree prompt. */
+  nestedMultiselect<TValue, TChoice>(
+    request: NestedMultiSelectRequest<TValue, TChoice>,
+  ): Promise<PromptOutcome<readonly NestedMultiSelectValue<TValue, TChoice>[]>> {
+    return clackNestedMultiselect({ ...request, input: this.input, output: this.output })
+  }
+}
diff --git a/packages/sdk/helper/src/questions/prompt-port.ts b/packages/sdk/helper/src/questions/prompt-port.ts
new file mode 100644
index 0000000000..f37fdff025
--- /dev/null
+++ b/packages/sdk/helper/src/questions/prompt-port.ts
@@ -0,0 +1,124 @@
+/**
+ * Terminal-prompt port shared by create and config workflows.
+ *
+ * @module @deepseek-ai/dsh-helper/questions/prompt-port
+ */
+
+/** One selectable prompt option. */
+export interface PromptOption<T> {
+  value: T
+  label: string
+  hint?: string
+  disabled?: boolean
+}
+
+/** Answer or explicit cancellation returned by every prompt. */
+export type PromptOutcome<T> =
+  | { status: 'answered'; value: T }
+  | { status: 'cancelled' }
+
+/** Input for one text prompt. */
+export interface TextPromptRequest {
+  message: string
+  placeholder?: string
+  initialValue?: string
+  defaultValue?: string
+  validate?: (value: string) => string | undefined
+}
+
+/** Input for one masked secret prompt. */
+export interface SecretPromptRequest {
+  message: string
+  validate?: (value: string) => string | undefined
+}
+
+/** Input for one single-choice prompt. */
+export interface SelectPromptRequest<T> {
+  message: string
+  options: readonly PromptOption<T>[]
+  initialValue?: T
+}
+
+/** Input for one additive multi-choice prompt. */
+export interface MultiSelectPromptRequest<T> {
+  message: string
+  options: readonly PromptOption<T>[]
+  initialValues?: readonly T[]
+  required?: boolean
+}
+
+/** Input for one yes/no prompt. */
+export interface ConfirmPromptRequest {
+  message: string
+  initialValue?: boolean
+  tone?: 'default' | 'warning'
+}
+
+/** One nested choice under a multi-select option. */
+interface NestedSelectChoice<T> {
+  value: T
+  label: string
+  default?: boolean
+}
+
+/** One root option with optional child option configuration. */
+export interface NestedMultiSelectOption<TValue, TChoice> {
+  value: TValue
+  label: string
+  required?: boolean
+  default?: boolean
+  disabled?: boolean
+  warning?: string
+  choiceMode?: 'exclusive' | 'multiple'
+  choices?: readonly NestedSelectChoice<TChoice>[]
+}
+
+/** Input for a tree-shaped feature-style picker. */
+export interface NestedMultiSelectRequest<TValue, TChoice> {
+  message: string
+  options: readonly NestedMultiSelectOption<TValue, TChoice>[]
+  showChanges?: boolean
+}
+
+/** One selected root option and its child options. */
+export interface NestedMultiSelectValue<TValue, TChoice> {
+  value: TValue
+  choices: readonly TChoice[]
+}
+
+/** Interaction boundary consumed by typed question objects. */
+export interface PromptPort {
+  /** Ask for one line of visible text. */
+  text(request: TextPromptRequest): Promise<PromptOutcome<string>>
+  /** Ask for one masked value. */
+  secret(request: SecretPromptRequest): Promise<PromptOutcome<string>>
+  /** Ask for exactly one option. */
+  select<T>(request: SelectPromptRequest<T>): Promise<PromptOutcome<T>>
+  /** Ask for zero or more options. */
+  multiselect<T>(request: MultiSelectPromptRequest<T>): Promise<PromptOutcome<readonly T[]>>
+  /** Ask for a boolean confirmation. */
+  confirm(request: ConfirmPromptRequest): Promise<PromptOutcome<boolean>>
+  /** Select root options and configure finite child options in one tree prompt. */
+  nestedMultiselect<TValue, TChoice>(
+    request: NestedMultiSelectRequest<TValue, TChoice>,
+  ): Promise<PromptOutcome<readonly NestedMultiSelectValue<TValue, TChoice>[]>>
+}
+
+/** Error used when a workflow chooses to turn prompt cancellation into command cancellation. */
+export class PromptCancelledError extends Error {
+  /** Create a stable cancellation error. */
+  constructor(message = 'operation cancelled') {
+    super(message)
+    this.name = 'PromptCancelledError'
+  }
+}
+
+/**
+ * Return an answered value or throw the shared cancellation error.
+ * @param outcome - prompt result to unwrap.
+ * @returns answered value.
+ */
+export function requireAnswer<T>(outcome: PromptOutcome<T>): T {
+  if (outcome.status === 'cancelled') throw new PromptCancelledError()
+  return outcome.value
+}
diff --git a/packages/sdk/helper/src/questions/question.ts b/packages/sdk/helper/src/questions/question.ts
new file mode 100644
index 0000000000..45697d7eb5
--- /dev/null
+++ b/packages/sdk/helper/src/questions/question.ts
@@ -0,0 +1,206 @@
+/**
+ * Typed question objects with prefill, validation, and prompt behavior together.
+ *
+ * @module @deepseek-ai/dsh-helper/questions/question
+ */
+
+import type { PromptOption, PromptOutcome, PromptPort } from './prompt-port.ts'
+
+function resolvePrefilled(
+  id: string,
+  value: string | undefined,
+  validate: ((value: string) => string | undefined) | undefined,
+): PromptOutcome<string> | undefined {
+  if (value === undefined) return undefined
+  const diagnostic = validate?.(value)
+  if (diagnostic) throw new Error(`${id}: ${diagnostic}`)
+  return { status: 'answered', value }
+}
+
+/** A typed business question resolved from prefilled input or one prompt call. */
+export abstract class Question<T> {
+  /** Stable question identity used in diagnostics. */
+  readonly id: string
+  /** User-facing prompt text. */
+  readonly message: string
+
+  protected constructor(id: string, message: string) {
+    this.id = id
+    this.message = message
+  }
+
+  /**
+   * Resolve a prefilled answer without prompting, or ask through the port.
+   * @param port - prompt interaction boundary.
+   * @param prefilled - optional value supplied by CLI or current project state.
+   * @returns answered or cancelled prompt outcome.
+   */
+  abstract resolve(port: PromptPort, prefilled?: T): Promise<PromptOutcome<T>>
+}
+
+/** Visible single-line text question. */
+export class TextQuestion extends Question<string> {
+  /** Light hint displayed when no text has been entered. */
+  readonly placeholder: string | undefined
+  /** Editable value displayed in the input. */
+  readonly initialValue: string | undefined
+  /** Value accepted when the user submits an empty input. */
+  readonly defaultValue: string | undefined
+  private readonly validate: ((value: string) => string | undefined) | undefined
+
+  /** Configure one text question. */
+  constructor(options: {
+    id: string
+    message: string
+    placeholder?: string
+    initialValue?: string
+    defaultValue?: string
+    validate?: (value: string) => string | undefined
+  }) {
+    super(options.id, options.message)
+    this.placeholder = options.placeholder
+    this.initialValue = options.initialValue
+    this.defaultValue = options.defaultValue
+    this.validate = options.validate
+  }
+
+  /** Validate prefilled text or ask for it. */
+  override async resolve(port: PromptPort, prefilled?: string): Promise<PromptOutcome<string>> {
+    const resolved = resolvePrefilled(this.id, prefilled, this.validate)
+    if (resolved) return resolved
+    return port.text({
+      message: this.message,
+      ...this.placeholder === undefined ? {} : { placeholder: this.placeholder },
+      ...this.initialValue === undefined ? {} : { initialValue: this.initialValue },
+      ...this.defaultValue === undefined ? {} : { defaultValue: this.defaultValue },
+      ...this.validate === undefined ? {} : { validate: this.validate },
+    })
+  }
+}
+
+/** Masked secret question whose empty-input semantics are set by its caller. */
+export class SecretQuestion extends Question<string> {
+  private readonly validate: ((value: string) => string | undefined) | undefined
+
+  /** Configure one secret question. */
+  constructor(options: {
+    id: string
+    message: string
+    validate?: (value: string) => string | undefined
+  }) {
+    super(options.id, options.message)
+    this.validate = options.validate
+  }
+
+  /** Validate a prefilled secret or ask for a masked value. */
+  override async resolve(port: PromptPort, prefilled?: string): Promise<PromptOutcome<string>> {
+    const resolved = resolvePrefilled(this.id, prefilled, this.validate)
+    if (resolved) return resolved
+    return port.secret({
+      message: this.message,
+      ...this.validate === undefined ? {} : { validate: this.validate },
+    })
+  }
+}
+
+/** Single-choice question. */
+export class SelectQuestion<T> extends Question<T> {
+  /** Available choices in display order. */
+  readonly options: readonly PromptOption<T>[]
+  /** Initially focused choice. */
+  readonly initialValue: T | undefined
+
+  /** Configure one single-choice question. */
+  constructor(options: {
+    id: string
+    message: string
+    options: readonly PromptOption<T>[]
+    initialValue?: T
+  }) {
+    super(options.id, options.message)
+    this.options = options.options
+    this.initialValue = options.initialValue
+  }
+
+  /** Validate a prefilled option or ask for one choice. */
+  override async resolve(port: PromptPort, prefilled?: T): Promise<PromptOutcome<T>> {
+    if (prefilled !== undefined) {
+      if (!this.options.some(option => Object.is(option.value, prefilled) && !option.disabled)) {
+        throw new Error(`${this.id}: unknown or disabled option ${String(prefilled)}`)
+      }
+      return { status: 'answered', value: prefilled }
+    }
+    return port.select({
+      message: this.message,
+      options: this.options,
+      ...this.initialValue === undefined ? {} : { initialValue: this.initialValue },
+    })
+  }
+}
+
+/** Additive multi-choice question. */
+export class MultiSelectQuestion<T> extends Question<readonly T[]> {
+  readonly options: readonly PromptOption<T>[]
+  readonly initialValues: readonly T[]
+  readonly required: boolean
+
+  /** Configure one multi-choice question. */
+  constructor(options: {
+    id: string
+    message: string
+    options: readonly PromptOption<T>[]
+    initialValues?: readonly T[]
+    required?: boolean
+  }) {
+    super(options.id, options.message)
+    this.options = options.options
+    this.initialValues = options.initialValues ?? []
+    this.required = options.required ?? false
+  }
+
+  /** Validate prefilled values or ask for an additive selection. */
+  override async resolve(port: PromptPort, prefilled?: readonly T[]): Promise<PromptOutcome<readonly T[]>> {
+    if (prefilled !== undefined) {
+      for (const value of prefilled) {
+        /* v8 ignore next -- unknown, disabled, and accepted values are each pinned by the question tests */
+        if (!this.options.some(option => Object.is(option.value, value) && option.disabled !== true)) {
+          throw new Error(`${this.id}: unknown or disabled option ${String(value)}`)
+        }
+      }
+      if (this.required && prefilled.length === 0) throw new Error(`${this.id}: choose at least one option`)
+      return { status: 'answered', value: prefilled }
+    }
+    return port.multiselect({
+      message: this.message,
+      options: this.options,
+      initialValues: this.initialValues,
+      required: this.required,
+    })
+  }
+}
+
+/** Boolean confirmation question. */
+export class ConfirmQuestion extends Question<boolean> {
+  /** Answer selected by pressing Enter. */
+  readonly initialValue: boolean
+  /** Visual severity used by the prompt adapter. */
+  readonly tone: 'default' | 'warning'
+
+  /** Configure one confirmation question. */
+  constructor(options: {
+    id: string
+    message: string
+    initialValue?: boolean
+    tone?: 'default' | 'warning'
+  }) {
+    super(options.id, options.message)
+    this.initialValue = options.initialValue ?? true
+    this.tone = options.tone ?? 'default'
+  }
+
+  /** Return a prefilled boolean or ask for confirmation. */
+  override async resolve(port: PromptPort, prefilled?: boolean): Promise<PromptOutcome<boolean>> {
+    if (prefilled !== undefined) return { status: 'answered', value: prefilled }
+    return port.confirm({ message: this.message, initialValue: this.initialValue, tone: this.tone })
+  }
+}
diff --git a/packages/sdk/helper/src/templates/assets/README.md.tpl b/packages/sdk/helper/src/templates/assets/README.md.tpl
new file mode 100644
index 0000000000..0033c8e0d4
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/README.md.tpl
@@ -0,0 +1,33 @@
+# {{name}}
+
+{{description}}
+
+Built with the DeepSeek Harness SDK using the {{model}} model.
+
+{{#if isAcp}}
+## Run as an ACP server
+
+Run `{{packageManager}} start` and configure your ACP client to launch this project. Standard output is reserved for ACP JSON-RPC.
+{{else}}
+{{#if isStdio}}
+## Run in a terminal
+
+Run `{{packageManager}} start` to start the interactive agent.
+{{else}}
+## Embed the harness
+
+Import and call the exported `main()` from `index.ts` in your host application.
+{{/if}}
+{{/if}}
+
+## Development
+
+Install NPM dependencies with `{{packageManager}} {{installArgs}}`, then use:
+
+- `dev`: `{{packageManager}} run dev`
+- `build`: `{{packageManager}} {{buildArgs}}`
+- `typecheck`: `{{packageManager}} run typecheck`
+- `start`: `{{packageManager}} start`
+- `config`: `{{packageManager}} run config`
+
+Edit `cordis.yml` to change the runtime plugin tree. Add or remove builtin features with `{{packageManager}} exec dsh-sdk config`.
diff --git a/packages/sdk/helper/src/templates/assets/gitignore.tpl b/packages/sdk/helper/src/templates/assets/gitignore.tpl
new file mode 100644
index 0000000000..30b03c75c5
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/gitignore.tpl
@@ -0,0 +1,5 @@
+node_modules/
+lib/
+.env
+.sessions/
+*.tsbuildinfo
diff --git a/packages/sdk/helper/src/templates/assets/index.ts.tpl b/packages/sdk/helper/src/templates/assets/index.ts.tpl
new file mode 100644
index 0000000000..d0315db80c
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/index.ts.tpl
@@ -0,0 +1,45 @@
+{{#if isAcp}}
+import { startSDK, type SdkBootContext } from '@deepseek-ai/dsh-scripts'
+{{else}}
+import { randomUUID } from 'node:crypto'
+import { AgentId } from '@deepseek-ai/dsh-agent'
+import { SessionId } from '@deepseek-ai/dsh-session'
+import { startSDK, type SdkBootContext } from '@deepseek-ai/dsh-scripts'
+{{/if}}
+
+/** Boot this project's cordis.yml when invoked by dsh-scripts. */
+export async function main(boot: SdkBootContext) {
+  const ctx = await startSDK(new URL('./cordis.yml', import.meta.url))
+{{#if isStdio}}
+  const model = boot.args.model
+  if (typeof model !== 'string' || model.length === 0) throw new Error('stdio startup requires --model=<name>')
+  const resume = boot.args.resume
+  if (resume !== undefined && (typeof resume !== 'string' || resume.length === 0)) {
+    throw new Error('stdio startup requires --resume=<session-id>')
+  }
+  if (resume === undefined) {
+    await ctx.agents.create({
+      agentId: AgentId('main'),
+      sessionId: SessionId(`main-session-${randomUUID()}`),
+      meta: { cwd: boot.cwd },
+      agentOptions: { model },
+    })
+  } else {
+    await ctx.agents.resume({
+      agentId: AgentId('main'),
+      resumeSessionId: SessionId(resume),
+      agentOptions: { model },
+    })
+  }
+{{else}}
+{{#if isEmbed}}
+  await ctx.agents.create({
+    agentId: AgentId('main'),
+    sessionId: SessionId(`main-session-${randomUUID()}`),
+    meta: { cwd: boot.cwd },
+    agentOptions: { model: {{modelLiteral}} },
+  })
+{{/if}}
+{{/if}}
+  return ctx
+}
diff --git a/packages/sdk/helper/src/templates/assets/local-plugin-tsdown.config.ts.tpl b/packages/sdk/helper/src/templates/assets/local-plugin-tsdown.config.ts.tpl
new file mode 100644
index 0000000000..d24fb6f870
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/local-plugin-tsdown.config.ts.tpl
@@ -0,0 +1,13 @@
+import { defineConfig } from 'tsdown'
+import { PluginBuild } from '@deepseek-ai/dsh-scripts/dev/tsdown-config'
+
+export default defineConfig(PluginBuild({
+  entry: ['src/index.ts'],
+  outDir: 'lib',
+  format: ['esm'],
+  platform: 'node',
+  target: 'es2024',
+  fixedExtension: false,
+  dts: true,
+  clean: false,
+}))
diff --git a/packages/sdk/helper/src/templates/assets/local-plugin.ts.tpl b/packages/sdk/helper/src/templates/assets/local-plugin.ts.tpl
new file mode 100644
index 0000000000..2f15bdef91
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/local-plugin.ts.tpl
@@ -0,0 +1,9 @@
+/** Local Cordis plugin. */
+import type { Context } from 'cordis'
+
+export const name = '{{pluginName}}'
+
+/** Register this plugin's project-local behavior. */
+export function apply(ctx: Context): void {
+  ctx.effect(() => () => {})
+}
diff --git a/packages/sdk/helper/src/templates/assets/local-tool.ts.tpl b/packages/sdk/helper/src/templates/assets/local-tool.ts.tpl
new file mode 100644
index 0000000000..4333d41c70
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/local-tool.ts.tpl
@@ -0,0 +1,17 @@
+/** Project-local model-facing tool. */
+import type { Context } from 'cordis'
+import { defineTool } from '@deepseek-ai/dsh-tools'
+
+export const name = '{{pluginName}}'
+export const inject = ['tools']
+
+/** Register the {{toolName}} tool. */
+export function apply(ctx: Context): void {
+  ctx.tools.register(defineTool({
+    name: '{{toolName}}',
+    description: 'Project-local {{toolTitle}} tool.',
+    parameters: {},
+    execute: async () => [{ type: 'text', text: '{{toolName}} completed.' }],
+    presentCall: args => ({ card: 'generic', title: '{{toolTitle}}', kind: 'other', rawInput: args }),
+  }))
+}
diff --git a/packages/sdk/helper/src/templates/assets/package.json.tpl b/packages/sdk/helper/src/templates/assets/package.json.tpl
new file mode 100644
index 0000000000..b5aa0fde0d
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/package.json.tpl
@@ -0,0 +1,14 @@
+{
+  "name": {{name}},
+  "version": "0.0.0",
+  "private": true,
+  "description": {{description}},
+  "type": "module",
+  "scripts": {
+    "build": "dsh-sdk build",
+    "typecheck": "tsc -b",
+    "config": "dsh-sdk config"
+  },
+  "dependencies": {{dependencies}},
+  "devDependencies": {{devDependencies}}
+}
diff --git a/packages/sdk/helper/src/templates/assets/persona.txt.tpl b/packages/sdk/helper/src/templates/assets/persona.txt.tpl
new file mode 100644
index 0000000000..b17a6e802a
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/persona.txt.tpl
@@ -0,0 +1,3 @@
+You are a coding assistant powered by the \{{model}} model. Your working directory is \{{cwd}}.
+
+Verify your work by running the code or tests. Keep answers brief and factual.
diff --git a/packages/sdk/helper/src/templates/assets/tsconfig.base.json.tpl b/packages/sdk/helper/src/templates/assets/tsconfig.base.json.tpl
new file mode 100644
index 0000000000..f712277e9a
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/tsconfig.base.json.tpl
@@ -0,0 +1,14 @@
+{
+  "compilerOptions": {
+    "target": "ES2024",
+    "module": "NodeNext",
+    "moduleResolution": "NodeNext",
+    "strict": true,
+    "declaration": true,
+    "composite": true,
+    "outDir": "lib",
+    "rootDir": ".",
+    "types": ["node"],
+    "skipLibCheck": true
+  }
+}
diff --git a/packages/sdk/helper/src/templates/assets/tsdown.config.ts.tpl b/packages/sdk/helper/src/templates/assets/tsdown.config.ts.tpl
new file mode 100644
index 0000000000..c923dd2d5b
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/tsdown.config.ts.tpl
@@ -0,0 +1,13 @@
+import { defineConfig } from 'tsdown'
+import { ProjectBuild } from '@deepseek-ai/dsh-scripts/dev/tsdown-config'
+
+export default defineConfig(ProjectBuild({
+  entry: ['index.ts'],
+  outDir: '.',
+  format: ['esm'],
+  platform: 'node',
+  target: 'es2024',
+  fixedExtension: false,
+  dts: false,
+  clean: false,
+}))
diff --git a/packages/sdk/helper/src/templates/assets/yarnrc.yml.tpl b/packages/sdk/helper/src/templates/assets/yarnrc.yml.tpl
new file mode 100644
index 0000000000..3186f3f079
--- /dev/null
+++ b/packages/sdk/helper/src/templates/assets/yarnrc.yml.tpl
@@ -0,0 +1 @@
+nodeLinker: node-modules
diff --git a/packages/sdk/helper/src/templates/project-template.ts b/packages/sdk/helper/src/templates/project-template.ts
new file mode 100644
index 0000000000..b214126b6e
--- /dev/null
+++ b/packages/sdk/helper/src/templates/project-template.ts
@@ -0,0 +1,113 @@
+/**
+ * Strict Handlebars wrapper and complete-file SDK project artifacts.
+ *
+ * @module @deepseek-ai/dsh-helper/templates/project-template
+ */
+
+import { PackageJsonFile } from '../documents/package-json-file.ts'
+import { TextProjectFile } from '../documents/project-file.ts'
+import type { PackageManagerName } from '../package-managers/package-manager.ts'
+import { baselineNpmDependencies } from '../project/npm-dependency-policy.ts'
+import type { ProjectProfile, RunInterface } from '../project/types.ts'
+import { loadHelperTemplate } from './template-assets.ts'
+import type { TextTemplate } from './text-template.ts'
+
+/** Stable typed view consumed by all generated text artifacts. */
+export interface ProjectTemplateContext {
+  name: string
+  description: string
+  releaseVersion: string
+  model: string
+  modelLiteral: string
+  isAcp: boolean
+  isStdio: boolean
+  isEmbed: boolean
+  packageManager: PackageManagerName
+  installArgs: string
+  buildArgs: string
+}
+
+/** Complete project-file template artifact. */
+export class TemplateArtifact<TModel extends object> extends TextProjectFile {
+  /** Render and own one complete project file. */
+  constructor(relativePath: string, template: TextTemplate<TModel>, model: TModel) {
+    super(relativePath, template.render(model))
+  }
+}
+
+const README_TEMPLATE = loadHelperTemplate<ProjectTemplateContext>('README.md.tpl')
+const PACKAGE_JSON_TEMPLATE = loadHelperTemplate<{
+  name: string
+  description: string
+  dependencies: string
+  devDependencies: string
+}>('package.json.tpl')
+const INDEX_TEMPLATE = loadHelperTemplate<ProjectTemplateContext>('index.ts.tpl')
+const TSDOWN_TEMPLATE = loadHelperTemplate<ProjectTemplateContext>('tsdown.config.ts.tpl')
+const TSCONFIG_BASE_TEMPLATE = loadHelperTemplate<ProjectTemplateContext>('tsconfig.base.json.tpl')
+const GITIGNORE_TEMPLATE = loadHelperTemplate<ProjectTemplateContext>('gitignore.tpl')
+const YARNRC_TEMPLATE = loadHelperTemplate<ProjectTemplateContext>('yarnrc.yml.tpl')
+
+/** Build the template model for one project and selected run interface. */
+export function createProjectTemplateContext(
+  profile: ProjectProfile,
+  runInterface: RunInterface = profile.runInterface,
+): ProjectTemplateContext {
+  return {
+    name: profile.name,
+    description: profile.description,
+    releaseVersion: profile.releaseVersion,
+    model: profile.runtime.model,
+    modelLiteral: JSON.stringify(profile.runtime.model),
+    isAcp: runInterface === 'acp',
+    isStdio: runInterface === 'stdio',
+    isEmbed: runInterface === 'embed',
+    packageManager: profile.packageManager.name,
+    installArgs: profile.packageManager.installCommand().join(' '),
+    buildArgs: profile.packageManager.buildCommand().join(' '),
+  }
+}
+
+/** Render the complete root package defaults before structured contributions merge. */
+export function createPackageJsonDoc(context: ProjectTemplateContext): PackageJsonFile {
+  const npmDependencies = baselineNpmDependencies(context.releaseVersion)
+  return PackageJsonFile.create(PACKAGE_JSON_TEMPLATE.render({
+    name: JSON.stringify(context.name),
+    description: JSON.stringify(context.description),
+    dependencies: JSON.stringify(npmDependencies.dependencies),
+    devDependencies: JSON.stringify(npmDependencies.devDependencies),
+  }))
+}
+
+/** Build interface-independent one-shot project artifacts. */
+export function createBaselineProjectArtifacts(
+  context: ProjectTemplateContext,
+): TemplateArtifact<ProjectTemplateContext>[] {
+  return [
+    new TemplateArtifact('tsdown.config.ts', TSDOWN_TEMPLATE, context),
+    new TemplateArtifact('tsconfig.base.json', TSCONFIG_BASE_TEMPLATE, context),
+    new TemplateArtifact('.gitignore', GITIGNORE_TEMPLATE, context),
+    ...context.packageManager === 'yarn'
+      ? [new TemplateArtifact('.yarnrc.yml', YARNRC_TEMPLATE, context)]
+      : [],
+  ]
+}
+
+/** Build files owned by the selected app feature option. */
+export function createAppProjectArtifacts(
+  context: ProjectTemplateContext,
+): TemplateArtifact<ProjectTemplateContext>[] {
+  return [
+    new TemplateArtifact('README.md', README_TEMPLATE, context),
+    new TemplateArtifact('index.ts', INDEX_TEMPLATE, context),
+  ]
+}
+
+/** Build package scripts owned by the selected app feature option. */
+export function createAppPackageScripts(context: ProjectTemplateContext): Readonly<Record<'dev' | 'start', string>> {
+  const modelArg = context.isStdio ? ` -- --model=${JSON.stringify(context.model)}` : ''
+  return {
+    dev: `dsh-sdk dev index.ts${modelArg}`,
+    start: `dsh-sdk start index.js${modelArg}`,
+  }
+}
diff --git a/packages/sdk/helper/src/templates/template-assets.ts b/packages/sdk/helper/src/templates/template-assets.ts
new file mode 100644
index 0000000000..fbf7fbec45
--- /dev/null
+++ b/packages/sdk/helper/src/templates/template-assets.ts
@@ -0,0 +1,19 @@
+/**
+ * Asset loader for templates owned by dsh-helper.
+ *
+ * @module @deepseek-ai/dsh-helper/templates/template-assets
+ */
+
+import { TextTemplate } from './text-template.ts'
+
+/**
+ * Load one helper-owned template in source and bundled layouts.
+ * @param filename - basename under the helper template asset directory.
+ * @returns compiled typed template.
+ */
+export function loadHelperTemplate<TModel extends object>(filename: string): TextTemplate<TModel> {
+  if (filename.includes('/') || filename.includes('\\')) {
+    throw new Error(`helper template filename must not contain a directory: ${filename}`)
+  }
+  return TextTemplate.fromFile<TModel>(new URL(`./assets/${filename}`, import.meta.url))
+}
diff --git a/packages/sdk/helper/src/templates/text-template.ts b/packages/sdk/helper/src/templates/text-template.ts
new file mode 100644
index 0000000000..b5fa1652cc
--- /dev/null
+++ b/packages/sdk/helper/src/templates/text-template.ts
@@ -0,0 +1,44 @@
+/**
+ * Strict typed rendering for package-owned text templates.
+ *
+ * @module @deepseek-ai/dsh-helper/templates/text-template
+ */
+
+import { readFileSync } from 'node:fs'
+import Handlebars from 'handlebars'
+
+/** Strict Handlebars template with no HTML escaping or custom extensions. */
+export class TextTemplate<TModel extends object> {
+  private readonly renderer: Handlebars.TemplateDelegate<TModel>
+
+  /**
+   * Compile one template under the SDK's fixed rendering policy.
+   * @param source - complete template source.
+   */
+  constructor(source: string) {
+    const handlebars = Handlebars.create()
+    this.renderer = handlebars.compile<TModel>(source, {
+      strict: true,
+      noEscape: true,
+      preventIndent: true,
+    })
+  }
+
+  /**
+   * Load a template asset owned by the calling package.
+   * @param url - source or bundled asset URL.
+   * @returns compiled template.
+   */
+  static fromFile<T extends object>(url: URL): TextTemplate<T> {
+    return new TextTemplate<T>(readFileSync(url, 'utf8'))
+  }
+
+  /**
+   * Render text from one complete typed model.
+   * @param model - values referenced by the template.
+   * @returns rendered text.
+   */
+  render(model: TModel): string {
+    return this.renderer(model)
+  }
+}
diff --git a/packages/sdk/helper/tests/documents.spec.ts b/packages/sdk/helper/tests/documents.spec.ts
new file mode 100644
index 0000000000..1ba8b9bd68
--- /dev/null
+++ b/packages/sdk/helper/tests/documents.spec.ts
@@ -0,0 +1,384 @@
+import { chmod, mkdtemp, mkdir, readFile, rm, writeFile } from 'node:fs/promises'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { afterEach, describe, expect, it } from 'vitest'
+import { CordisYamlFile, JsExpression } from '../src/documents/cordis-yaml-file.ts'
+import { EnvFile } from '../src/documents/env-file.ts'
+import { PackageJsonFile } from '../src/documents/package-json-file.ts'
+import { PnpmWorkspaceFile } from '../src/documents/pnpm-workspace-file.ts'
+import { TsConfigFile } from '../src/documents/tsconfig-file.ts'
+import { TextProjectFile, withTrailingNewline } from '../src/documents/project-file.ts'
+import { featureId, resourceKey } from '../src/ids.ts'
+import { LinkWorkspace } from '../src/package-managers/link-workspace.ts'
+import { LocalPluginBlueprint } from '../src/plugins/local-plugin-blueprint.ts'
+import {
+  NpmPackageManager,
+  NodeCommandRunner,
+  PnpmPackageManager,
+  YarnPackageManager,
+  createPackageManager,
+  inferPackageManagerName,
+  probePackageManagerVersion,
+  scrubEnvironment,
+  type CommandRunner,
+} from '../src/package-managers/package-manager.ts'
+import { createBaselineProjectArtifacts } from '../src/templates/project-template.ts'
+import { loadHelperTemplate } from '../src/templates/template-assets.ts'
+import { TextTemplate } from '../src/templates/text-template.ts'
+import { resolveNpmDependency } from '../src/project/npm-dependency-policy.ts'
+
+const temporary: string[] = []
+
+afterEach(async () => {
+  await Promise.all(temporary.splice(0).map(path => rm(path, { recursive: true, force: true })))
+})
+
+describe('structured project documents', () => {
+  it('normalizes trailing newlines and preserves managed package fields', () => {
+    expect(withTrailingNewline('a\n\n')).toBe('a\n')
+    const manifest = PackageJsonFile.parse('{"name":"demo","custom":1,"dependencies":{"z":"1"}}\n')
+    manifest.setScript('start', 'node index.js')
+    manifest.setNpmDependency('dependencies', 'a', '2')
+    manifest.setNpmDependency('devDependencies', 'typescript', '3')
+    manifest.removeNpmDependency('dependencies', 'z')
+    manifest.addWorkspace('plugins/*')
+    manifest.addWorkspace('plugins/*')
+    manifest.setPackageManager('pnpm@10.0.0')
+    manifest.setResolution('a', 'portal:../a')
+    manifest.validate()
+    expect(manifest.npmDependency('a')).toEqual({ section: 'dependencies', spec: '2' })
+    expect(manifest.npmDependencyNames()).toEqual(['a', 'typescript'])
+    expect(JSON.parse(manifest.serialize())).toMatchObject({
+      name: 'demo',
+      custom: 1,
+      dependencies: { a: '2' },
+      workspaces: ['plugins/*'],
+    })
+    manifest.setPackageManager(undefined)
+    expect(manifest.value().packageManager).toBeUndefined()
+    expect(() => PackageJsonFile.parse('[]')).toThrow('root must be an object')
+    expect(() => PackageJsonFile.parse('{')).toThrow('invalid package.json')
+    for (const [text, diagnostic] of [
+      ['{}', 'name must be'],
+      ['{"name":"x","scripts":null}', 'scripts must be an object'],
+      ['{"name":"x","dependencies":[]}', 'dependencies must be an object'],
+      ['{"name":"x","devDependencies":{"bad":""}}', 'must be a non-empty string'],
+      ['{"name":"x","workspaces":"bad"}', 'workspaces must be an array'],
+    ] as const) expect(() => { PackageJsonFile.parse(text).validate() }).toThrow(diagnostic)
+    const minimal = PackageJsonFile.parse('{"name":"x"}')
+    minimal.validate()
+    expect(minimal.npmDependency('missing')).toBeUndefined()
+    expect(minimal.serialize()).toBe('{\n  "name": "x"\n}\n')
+    expect(PackageJsonFile.parse('{"name":"x","devDependencies":{"a":"1"}}').npmDependencyNames()).toEqual(['a'])
+  })
+
+  it('round-trips Cordis comments and !!js while editing owned fields', () => {
+    const created = CordisYamlFile.create().clone()
+    created.addEntry({ id: 'created', name: 'created-package' }, `Uncomment this example.
+config:
+  value: true`)
+    expect(created.serialize()).toMatch(/^- id: created/m)
+    expect(created.serialize()).toContain('  # Uncomment this example.\n  # config:\n  #   value: true')
+    expect(created.serialize()).not.toMatch(/^\[/)
+    const flow = CordisYamlFile.parse('[{ id: flow, name: flow-package, config: { root: ./flow } }]\n')
+    expect(flow.serialize()).toContain('- id: flow\n  name: flow-package\n  config:\n    root: ./flow')
+    expect(flow.serialize()).not.toContain('{')
+    const document = CordisYamlFile.parse(`# lead
+- id: provider
+  name: '@deepseek-ai/dsh-llm-deepseek'
+  config:
+    apiKey: !!js process.env.DEEPSEEK_API_KEY
+    custom: keep
+`)
+    const apiKey = document.entry('provider')?.config?.apiKey
+    expect(apiKey).toBeInstanceOf(JsExpression)
+    document.updateOwnedConfig('provider', ['apiKey'], { apiKey: new JsExpression('process.env.NEXT_KEY') })
+    document.setDisabled('provider', true)
+    document.addEntry({ id: 'tool', name: 'demo-tool' })
+    document.validate()
+    const text = document.serialize()
+    expect(text).toContain('# lead')
+    expect(text).toContain('!!js process.env.NEXT_KEY')
+    expect(text).toContain('custom: keep')
+    expect(document.removeEntry('tool')).toBe(true)
+    expect(document.removeEntry('tool')).toBe(false)
+    document.setDisabled('provider', false)
+    expect(document.entry('provider')?.disabled).toBeUndefined()
+    expect(() => { document.addEntry({ id: 'provider', name: 'duplicate' }) }).toThrow('already exists')
+    expect(() => CordisYamlFile.parse('{}')).toThrow('root must be a sequence')
+  })
+
+  it('rejects malformed Cordis config entries and missing mutation targets', () => {
+    expect(() => new JsExpression(' ')).toThrow('must not be empty')
+    expect(() => CordisYamlFile.parse('[')).toThrow('invalid cordis.yml')
+    for (const [text, diagnostic] of [
+      ['- nope\n', 'every entry must be a mapping'],
+      ['- name: pkg\n', 'id must be'],
+      ['- id: x\n', 'name must be'],
+      ['- id: x\n  name: pkg\n  config: nope\n', 'config must be'],
+      ['- id: x\n  name: pkg\n  disabled: nope\n', 'disabled must be'],
+    ] as const) expect(() => CordisYamlFile.parse(text).entries()).toThrow(diagnostic)
+    const duplicate = CordisYamlFile.parse('- id: x\n  name: one\n- id: x\n  name: two\n')
+    expect(() => { duplicate.validate() }).toThrow('duplicate Cordis config entry id')
+    const document = CordisYamlFile.create()
+    expect(() => { document.setDisabled('missing', true) }).toThrow('does not exist')
+    expect(() => { document.updateOwnedConfig('missing', [], {}) }).toThrow('does not exist')
+    document.addEntry({ id: 'plain', name: 'pkg' })
+    document.updateOwnedConfig('plain', [], { value: 1 })
+    expect(document.entry('plain')?.config).toEqual({ value: 1 })
+    document.updateOwnedConfig('plain', ['value'], {})
+    expect(document.entry('plain')?.config).toBeUndefined()
+    const scalar = CordisYamlFile.parse('- id: plain\n  name: pkg\n  config: value\n')
+    expect(() => { scalar.updateOwnedConfig('plain', [], {}) }).toThrow('config is not a mapping')
+    expect(() => { CordisYamlFile.parse('- nope\n').setDisabled('missing', true) }).toThrow('does not exist')
+  })
+
+  it('keeps .env append-only while managing .env.example strictly', () => {
+    const document = EnvFile.parse('.env', '# keep\nA=1\nexport B=2\n')
+    expect(document.get('A')).toBe('1')
+    expect(document.get('B')).toBe('2')
+    expect(document.append('A', 'next')).toBe(false)
+    expect(document.append('C', '', 'Required')).toBe(true)
+    expect(document.serialize()).toBe('# keep\nA=1\nexport B=2\n# Required\nC=\n')
+    expect(() => { document.append('bad-name', 'x') }).toThrow('invalid environment variable')
+    expect(() => { document.append('D', '', '') }).toThrow('non-empty line')
+    expect(() => { document.append('D', '', 'bad\ncomment') }).toThrow('non-empty line')
+    expect(() => { document.set('A', 'next') }).toThrow('.env is append-only')
+    expect(() => { document.remove('A') }).toThrow('.env is append-only')
+    const duplicateEnv = EnvFile.parse('.env', 'A=1\nA=2\n')
+    expect(duplicateEnv.get('A')).toBe('2')
+    expect(duplicateEnv.append('A', 'next')).toBe(false)
+    expect(() => { duplicateEnv.validate() }).not.toThrow()
+
+    const example = EnvFile.parse('.env.example', '# keep\nA=1\n')
+    example.set('A', 'next')
+    example.set('C', '')
+    example.remove('C')
+    expect(example.serialize()).toBe('# keep\nA=next\n')
+    expect(() => { example.append('C', '') }).toThrow('.env.example is SDK-managed')
+    expect(() => { example.set('bad-name', 'x') }).toThrow('invalid environment variable')
+    const duplicate = EnvFile.parse('.env.example', 'A=1\nA=2\n')
+    expect(() => { duplicate.validate() }).toThrow('duplicate variable A')
+    expect(() => duplicate.get('A')).toThrow('duplicate variable A')
+    expect(() => { duplicate.set('A', 'next') }).toThrow('duplicate variable A')
+    expect(() => { duplicate.remove('A') }).toThrow('duplicate variable A')
+    example.remove('missing')
+    expect(document.get('missing')).toBeUndefined()
+    expect(EnvFile.parse('.env', '').clone().serialize()).toBe('\n')
+  })
+
+  it('patches JSONC references without erasing comments', () => {
+    const document = TsConfigFile.parse(`{
+  // retained
+  "references": [{ "path": "./plugins/a" }]
+}`)
+    document.addReference('./plugins/a')
+    document.addReference('./plugins/b')
+    document.validate()
+    expect(document.serialize()).toContain('// retained')
+    expect(document.serialize()).toContain('./plugins/b')
+    expect(() => TsConfigFile.parse('{')).toThrow('valid JSONC object')
+    const malformed = TsConfigFile.parse('{"references": {}}')
+    expect(() => { malformed.addReference('./plugins/x') }).toThrow('must be an array')
+    const badItem = TsConfigFile.parse('{"references":[null]}')
+    expect(() => { badItem.addReference('./plugins/x') }).toThrow('must contain')
+    expect(() => { badItem.validate() }).toThrow('must contain')
+    const created = TsConfigFile.create()
+    created.validate()
+    expect(created.clone().serialize()).toContain('"references": []')
+    TsConfigFile.parse('{}').validate()
+    const noReferences = TsConfigFile.parse('{}')
+    noReferences.addReference('./plugin')
+    expect(noReferences.serialize()).toContain('./plugin')
+    expect(() => { TsConfigFile.parse('{"references":{}}').validate() }).toThrow('must be an array')
+  })
+
+  it('creates and parses pnpm workspace policy', () => {
+    const document = PnpmWorkspaceFile.create()
+    document.addPackage('plugins/*')
+    document.addPackage('plugins/*')
+    document.disableAutoInstallPeers()
+    document.validate()
+    expect(document.serialize()).toContain('autoInstallPeers: false')
+    const parsed = PnpmWorkspaceFile.parse(document.serialize())
+    expect(parsed.clone().serialize()).toBe(document.serialize())
+    expect(() => PnpmWorkspaceFile.parse('packages: [')).toThrow('invalid pnpm-workspace.yaml')
+    expect(() => PnpmWorkspaceFile.parse('packages: nope')).toThrow('packages must be an array')
+    expect(() => PnpmWorkspaceFile.parse('packages: [{}]')).toThrow('packages must be an array')
+    expect(() => PnpmWorkspaceFile.parse('packages: [1]')).toThrow('packages must be an array')
+    expect(() => PnpmWorkspaceFile.parse('[]')).toThrow('root must be an object')
+    expect(() => PnpmWorkspaceFile.parse('packages: []\nautoInstallPeers: nope')).toThrow('must be boolean')
+    const invalid = PnpmWorkspaceFile.create()
+    invalid.addPackage(' ')
+    expect(() => { invalid.validate() }).toThrow('must not be empty')
+    expect(PnpmWorkspaceFile.parse('packages: []\n').serialize()).not.toContain('autoInstallPeers')
+    const preserved = PnpmWorkspaceFile.parse(`# keep workspace settings
+packages:
+  - apps/*
+catalog:
+  react: ^19.0.0
+overrides:
+  legacy: modern
+`)
+    preserved.disableAutoInstallPeers()
+    const preservedText = preserved.clone().serialize()
+    expect(preservedText).toContain('# keep workspace settings')
+    expect(preservedText).toContain('catalog:\n  react: ^19.0.0')
+    expect(preservedText).toContain('overrides:\n  legacy: modern')
+    expect(preservedText).toContain('autoInstallPeers: false')
+  })
+
+  it('renders strict complete-file templates without escaping code text', () => {
+    const template = new TextTemplate<{ value: string }>('value={{value}} missing={{missing}}')
+    expect(() => template.render({ value: '<code>' })).toThrow()
+    const valid = new TextTemplate<{ value: string }>('value={{value}}')
+    expect(valid.render({ value: '<code>' })).toBe('value=<code>')
+    expect(new TextTemplate<Record<string, never>>('\\{{model}}').render({})).toBe('{{model}}')
+    expect(() => new TextProjectFile('/absolute', 'x')).toThrow('stay inside')
+    expect(() => new TextProjectFile('../outside', 'x')).toThrow('stay inside')
+    expect(new TextProjectFile('inside', 'x').clone().serialize()).toBe('x\n')
+    expect(() => featureId('Bad Id')).toThrow('invalid feature id')
+    expect(() => resourceKey('')).toThrow('must not be empty')
+    expect(() => loadHelperTemplate('../bad.tpl')).toThrow('must not contain a directory')
+    expect(createBaselineProjectArtifacts({
+      name: 'demo', description: 'demo', releaseVersion: '0.0.1', model: 'model', modelLiteral: '"model"', packageManager: 'yarn',
+      isAcp: false, isStdio: false, isEmbed: true,
+      installArgs: 'install', buildArgs: 'build',
+    }).map(document => document.relativePath)).toContain('.yarnrc.yml')
+    expect(() => new LocalPluginBlueprint('---', 'plugin')).toThrow('invalid local plugin name')
+    expect(new LocalPluginBlueprint('tool', 'tool').packageName('@scope/project')).toBe('@scope/project-tool')
+    expect(new LocalPluginBlueprint('tool', 'tool').packageName('@invalid')).toBe('@invalid-tool')
+  })
+})
+
+describe('package manager strategies', () => {
+  it('owns workspace fields, execution commands, and supported version floors', () => {
+    const npm = new NpmPackageManager('10.1.0')
+    const pnpm = new PnpmPackageManager('10.2.0')
+    const yarn = new YarnPackageManager('4.0.0')
+    for (const manager of [npm, pnpm, yarn]) manager.validateVersion()
+    expect(npm.localPluginSpec()).toBe('*')
+    expect(npm.linkSpec('../x')).toBe('file:../x')
+    expect(npm.configureWorkspace(PackageJsonFile.create('{"name":"demo"}'))).toEqual([])
+    const pnpmManifest = PackageJsonFile.create('{"name":"demo"}')
+    expect(pnpm.configureWorkspace(pnpmManifest)[0]).toBeInstanceOf(PnpmWorkspaceFile)
+    expect(pnpm.localPluginSpec()).toBe('workspace:*')
+    expect(pnpm.linkSpec('../x')).toBe('link:../x')
+    const yarnManifest = PackageJsonFile.create('{"name":"demo"}')
+    expect(yarn.configureWorkspace(yarnManifest)).toEqual([])
+    expect(yarn.localPluginSpec()).toBe('workspace:*')
+    expect(yarn.linkSpec('../x')).toBe('portal:../x')
+    expect(yarn.buildCommand()).toEqual(['build'])
+    expect(npm.installCommand()).toEqual(['install'])
+    expect(npm.buildCommand()).toEqual(['run', 'build'])
+    expect(() => createPackageManager('npm', '9.0.0')).toThrow('npm >=10')
+    expect(() => createPackageManager('pnpm', '9.0.0')).toThrow('pnpm >=10')
+    expect(() => createPackageManager('yarn', '1.22.0')).toThrow('Yarn >=2')
+    expect(inferPackageManagerName(undefined, 'pnpm/10.0.0 node/v24')).toBe('pnpm')
+    expect(inferPackageManagerName(undefined, 'unknown/1')).toBeUndefined()
+    expect(inferPackageManagerName('yarn', undefined)).toBe('yarn')
+    expect(() => createPackageManager('npm', 'invalid')).toThrow('invalid package manager version')
+    expect(resolveNpmDependency('cordis', 'devDependencies', '0.0.1')).toEqual({
+      section: 'devDependencies', spec: '^4.0.0-rc.7',
+    })
+    expect(resolveNpmDependency('@cordisjs/plugin-hmr', 'dependencies', '0.0.1').spec).toBe('^1.0.15')
+    expect(resolveNpmDependency('@deepseek-ai/dsh-tools', 'dependencies', '1.2.3').spec).toBe('^1.2.3')
+    expect(() => resolveNpmDependency('unknown', 'dependencies', '0.0.1')).toThrow('no generated-project')
+  })
+
+  it('checks install/build process outcomes and scrubs credential-shaped names', async () => {
+    const calls: string[][] = []
+    const runner: CommandRunner = {
+      run: async (command, args) => {
+        calls.push([command, ...args])
+        return { exitCode: 0, signal: null }
+      },
+    }
+    const npm = new NpmPackageManager('10.0.0')
+    await npm.install('/tmp', runner)
+    await npm.build('/tmp', runner)
+    expect(calls).toEqual([['npm', 'install'], ['npm', 'run', 'build']])
+    const failed: CommandRunner = { run: async () => ({ exitCode: 2, signal: null }) }
+    await expect(npm.install('/tmp', failed)).rejects.toThrow('exited with code 2')
+    const killed: CommandRunner = { run: async () => ({ exitCode: null, signal: 'SIGTERM' }) }
+    await expect(npm.build('/tmp', killed)).rejects.toThrow('killed by SIGTERM')
+    expect(scrubEnvironment({ PATH: '/bin', API_KEY: 'secret', TOKEN_VALUE: 'secret' })).toEqual({ PATH: '/bin' })
+  })
+
+  it('probes versions and runs real child-process boundaries', async () => {
+    await expect(probePackageManagerVersion('npm', process.cwd())).resolves.toMatch(/^\d+/)
+    await expect(probePackageManagerVersion('npm', '/missing/dsh-cwd')).rejects.toThrow('cannot run npm --version')
+    const root = await mkdtemp(join(tmpdir(), 'dsh-empty-version-'))
+    temporary.push(root)
+    const executable = join(root, 'npm')
+    await writeFile(executable, '#!/bin/sh\nexit 0\n')
+    await chmod(executable, 0o755)
+    const before = process.env.PATH
+    process.env.PATH = root
+    await expect(probePackageManagerVersion('npm', root)).rejects.toThrow('empty version output')
+    process.env.PATH = before
+    const runner = new NodeCommandRunner()
+    await expect(runner.run(process.execPath, ['-e', ''], root)).resolves.toEqual({ exitCode: 0, signal: null })
+    await expect(runner.run('missing-dsh-command', [], root)).rejects.toThrow()
+  })
+
+  it('discovers and rewrites a repository-local NPM dependency closure', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-link-workspace-'))
+    temporary.push(root)
+    await mkdir(join(root, 'vendor', 'cordis'), { recursive: true })
+    await mkdir(join(root, 'packages', 'sdk', 'scripts'), { recursive: true })
+    await mkdir(join(root, 'packages', 'sdk', 'helper'), { recursive: true })
+    await writeFile(join(root, 'vendor', 'cordis', 'package.json'), JSON.stringify({ name: 'cordis' }))
+    await writeFile(join(root, 'packages', 'sdk', 'helper', 'package.json'), JSON.stringify({ name: '@deepseek-ai/dsh-helper' }))
+    await writeFile(join(root, 'packages', 'sdk', 'scripts', 'package.json'), JSON.stringify({
+      name: '@deepseek-ai/dsh-scripts', dependencies: { '@deepseek-ai/dsh-helper': '^0.0.1' }, peerDependencies: { cordis: '^4' },
+    }))
+    const workspace = await LinkWorkspace.open(root)
+    expect(workspace.closure(['@deepseek-ai/dsh-scripts'])).toEqual([
+      '@deepseek-ai/dsh-helper', '@deepseek-ai/dsh-scripts', 'cordis',
+    ])
+    const manifest = PackageJsonFile.create('{"name":"consumer","description":"test"}')
+    manifest.setNpmDependency('dependencies', '@deepseek-ai/dsh-scripts', '^0.0.1')
+    const pnpmWorkspace = PnpmWorkspaceFile.create()
+    workspace.apply(join(root, 'consumer'), manifest, new PnpmPackageManager('10.0.0'), [pnpmWorkspace])
+    expect(manifest.npmDependency('cordis')?.spec).toMatch(/^link:/)
+    expect(pnpmWorkspace.serialize()).toContain('autoInstallPeers: false')
+    expect(workspace.packageDirectory('cordis')).toBe(join(root, 'vendor', 'cordis'))
+    expect(await readFile(join(root, 'vendor', 'cordis', 'package.json'), 'utf8')).toContain('cordis')
+    expect(workspace.packageDirectory('missing')).toBeUndefined()
+    const yarnManifest = PackageJsonFile.create('{"name":"consumer"}')
+    yarnManifest.setNpmDependency('dependencies', '@deepseek-ai/dsh-scripts', '^0.0.1')
+    workspace.apply(join(root, 'consumer-yarn'), yarnManifest, new YarnPackageManager('4.0.0'), [])
+    expect(yarnManifest.value().resolutions).toBeDefined()
+    const pnpmManifest = PackageJsonFile.create('{"name":"consumer"}')
+    pnpmManifest.setNpmDependency('dependencies', '@deepseek-ai/dsh-scripts', '^0.0.1')
+    expect(() => { workspace.apply(join(root, 'consumer-pnpm'), pnpmManifest, new PnpmPackageManager('10.0.0'), []) })
+      .toThrow('requires pnpm-workspace.yaml')
+  })
+
+  it('rejects malformed linked repositories', async () => {
+    const missing = await mkdtemp(join(tmpdir(), 'dsh-link-missing-'))
+    temporary.push(missing)
+    await mkdir(join(missing, 'vendor'), { recursive: true })
+    await mkdir(join(missing, 'packages'), { recursive: true })
+    await expect(LinkWorkspace.open(missing)).rejects.toThrow('not a DeepSeek Harness repository root')
+    const unreadable = await mkdtemp(join(tmpdir(), 'dsh-link-unreadable-'))
+    temporary.push(unreadable)
+    await mkdir(join(unreadable, 'vendor', 'bad'), { recursive: true })
+    await mkdir(join(unreadable, 'packages'), { recursive: true })
+    await expect(LinkWorkspace.open(unreadable)).rejects.toThrow('cannot read linked package')
+    const unnamed = await mkdtemp(join(tmpdir(), 'dsh-link-unnamed-'))
+    temporary.push(unnamed)
+    await mkdir(join(unnamed, 'vendor', 'unnamed'), { recursive: true })
+    await mkdir(join(unnamed, 'packages'), { recursive: true })
+    await writeFile(join(unnamed, 'vendor', 'unnamed', 'package.json'), '{}')
+    await expect(LinkWorkspace.open(unnamed)).rejects.toThrow('not a DeepSeek Harness repository root')
+    const duplicate = await mkdtemp(join(tmpdir(), 'dsh-link-duplicate-'))
+    temporary.push(duplicate)
+    await mkdir(join(duplicate, 'vendor', 'one'), { recursive: true })
+    await mkdir(join(duplicate, 'packages', 'group', 'two'), { recursive: true })
+    await writeFile(join(duplicate, 'vendor', 'one', 'package.json'), '{"name":"duplicate"}')
+    await writeFile(join(duplicate, 'packages', 'group', 'two', 'package.json'), '{"name":"duplicate"}')
+    await expect(LinkWorkspace.open(duplicate)).rejects.toThrow('duplicate linked package name')
+  })
+})
diff --git a/packages/sdk/helper/tests/project.spec.ts b/packages/sdk/helper/tests/project.spec.ts
new file mode 100644
index 0000000000..d55bb16be9
--- /dev/null
+++ b/packages/sdk/helper/tests/project.spec.ts
@@ -0,0 +1,954 @@
+import { chmod, mkdtemp, readFile, rm, stat, writeFile } from 'node:fs/promises'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { fileURLToPath } from 'node:url'
+import { afterEach, describe, expect, it } from 'vitest'
+import {
+  FeatureOption,
+  ExclusiveOptionFeature,
+  MultiOptionFeature,
+  FixedFeature,
+  type FeatureProjectView,
+} from '../src/features/feature.ts'
+import { createBuiltinRegistry } from '../src/features/builtin/index.ts'
+import {
+  npmCordisConfigEntry,
+  cordisConfigEntry,
+  environment as environmentResource,
+  optionalString,
+  ownedTextFile,
+  requiredString,
+  stringArray,
+} from '../src/features/builtin/helpers.ts'
+import { defineFeatures, defineFeature } from '../src/features/define-feature.ts'
+import { FeatureRegistry } from '../src/features/registry.ts'
+import { ProjectContribution } from '../src/features/resources.ts'
+import type { CordisConfigEntryResource, ProjectResource } from '../src/features/resources.ts'
+import type { CordisConfigEntry } from '../src/documents/cordis-yaml-file.ts'
+import { PackageJsonFile } from '../src/documents/package-json-file.ts'
+import { TextProjectFile } from '../src/documents/project-file.ts'
+import { featureId, resourceKey } from '../src/ids.ts'
+import { NpmPackageManager } from '../src/package-managers/package-manager.ts'
+import { LocalPluginBlueprint } from '../src/plugins/local-plugin-blueprint.ts'
+import { SdkProject } from '../src/project/sdk-project.ts'
+import type {
+  FeatureSelection,
+  ProjectCreationRequest,
+  ProjectProfile,
+} from '../src/project/types.ts'
+
+const temporary: string[] = []
+const repoRoot = fileURLToPath(new URL('../../../../', import.meta.url))
+
+afterEach(async () => {
+  await Promise.all(temporary.splice(0).map(path => rm(path, { recursive: true, force: true })))
+})
+
+function selection(id: string, options: readonly string[], secrets?: Record<string, string>): FeatureSelection {
+  return { id: featureId(id), options, ...secrets ? { secrets } : {} }
+}
+
+function request(
+  extra: readonly FeatureSelection[] = [],
+  plugins: readonly LocalPluginBlueprint[] = [],
+  app: 'acp' | 'stdio' | 'embed' = 'stdio',
+  bash: 'local' | 'sandbox' = 'local',
+): ProjectCreationRequest {
+  return {
+    name: 'test-agent',
+    description: 'test project',
+    runtime: { model: 'deepseek-v4-flash' },
+    packageManager: new NpmPackageManager('10.0.0'),
+    releaseVersion: '0.0.1',
+    features: [
+      selection('provider', ['deepseek'], { apiKey: 'test-key' }),
+      selection('bash', [bash]),
+      selection('app', [app]),
+      selection('persistence', ['jsonl']),
+      ...extra,
+    ],
+    localPlugins: plugins,
+  }
+}
+
+async function createCommitted(
+  extra: readonly FeatureSelection[] = [],
+  plugins: readonly LocalPluginBlueprint[] = [],
+): Promise<SdkProject> {
+  const root = await mkdtemp(join(tmpdir(), 'dsh-project-domain-'))
+  temporary.push(root)
+  const creation = request(extra, plugins)
+  const project = SdkProject.create(root, creation)
+  const registry = createBuiltinRegistry(project.profile)
+  const edit = project.edit(registry)
+  for (const item of creation.features) edit.installFeature(registry.get(item.id), item)
+  for (const plugin of plugins) edit.addPlugin(plugin)
+  return (await edit.commit()).project
+}
+
+describe('SdkProject and ProjectEditSession', () => {
+  it('derives existing-project profiles and tolerates malformed optional documents', async () => {
+    const make = async (
+      name: string,
+      manifest: Record<string, unknown>,
+      cordis: string,
+      extras: Record<string, string> = {},
+    ): Promise<SdkProject> => {
+      const root = await mkdtemp(join(tmpdir(), `${name}-`))
+      temporary.push(root)
+      await writeFile(join(root, 'package.json'), JSON.stringify(manifest))
+      await writeFile(join(root, 'cordis.yml'), cordis)
+      for (const [path, text] of Object.entries(extras)) await writeFile(join(root, path), text)
+      return SdkProject.open(root)
+    }
+    const acp = await make('dsh-open-acp', {
+      name: 'acp', description: 'ACP', packageManager: 'pnpm@10.1.0',
+      dependencies: { '@deepseek-ai/dsh-scripts': '^1.2.3' },
+    }, `- id: acp
+  name: '@deepseek-ai/dsh-acp'
+  config: { model: app-model }
+`, { '.env': 'KEY=value\n', 'tsconfig.json': '{bad', 'pnpm-workspace.yaml': 'bad' })
+    expect(acp.profile).toMatchObject({
+      runInterface: 'acp', runtime: { model: 'app-model' }, releaseVersion: '1.2.3', description: 'ACP',
+    })
+    expect(acp.profile.packageManager.name).toBe('pnpm')
+    expect(acp.readEnvironment('.env', 'KEY')).toBe('value')
+    expect(() => acp.readEnvironment('.env.example', 'KEY')).not.toThrow()
+    expect(acp.document('tsconfig.json')).toBeInstanceOf(TextProjectFile)
+    const stdio = await make('dsh-open-stdio', {}, `- id: provider
+  name: '@deepseek-ai/dsh-llm-deepseek'
+  config: { models: [provider-model] }
+- id: stdio
+  name: '@deepseek-ai/dsh-stdio'
+`, { 'yarn.lock': '' })
+    expect(stdio.profile.runInterface).toBe('stdio')
+    expect(stdio.profile.runtime.model).toBe('provider-model')
+    expect(stdio.profile.packageManager.name).toBe('yarn')
+    expect(stdio.profile.name).toBe(stdio.root.split('/').at(-1))
+    const pnpm = await make('dsh-open-pnpm', { name: 'pnpm' }, '[]\n', { 'pnpm-lock.yaml': '' })
+    expect(pnpm.profile.packageManager.name).toBe('pnpm')
+    const defaults = await make('dsh-open-default', { name: 'default', packageManager: 'npm@10.0.0' }, '[]\n')
+    expect(defaults.profile).toMatchObject({
+      runInterface: 'embed', runtime: { model: 'deepseek-v4-flash' }, releaseVersion: '0.0.1',
+    })
+    expect(() => SdkProject.create(defaults.root, { ...request(), features: [] })).toThrow('requires one app')
+    await expect(make('dsh-open-invalid-manager', { name: 'bad', packageManager: 'bad' }, '[]\n'))
+      .rejects.toThrow('invalid packageManager field')
+    const providerFallback = await make('dsh-open-provider-fallback', { name: 'fallback' }, `- id: stdio
+  name: '@deepseek-ai/dsh-stdio'
+  config: { model: '' }
+- id: provider
+  name: '@deepseek-ai/dsh-llm-deepseek'
+  config: { models: [fallback-model] }
+`)
+    expect(providerFallback.profile.runtime.model).toBe('fallback-model')
+    const pnpmRequest = { ...request(), packageManager: new (await import('../src/package-managers/package-manager.ts')).PnpmPackageManager('10.0.0') }
+    expect(SdkProject.create(join(defaults.root, 'pnpm'), pnpmRequest).hasDocument('pnpm-workspace.yaml')).toBe(true)
+  })
+
+  it('commits a complete blueprint and round-trips every installed feature', async () => {
+    const project = await createCommitted([
+      selection('hmr', ['default']),
+      selection('fs', ['local']),
+      selection('todo', ['default']),
+      selection('web', ['exa'], { apiKey: 'exa-key' }),
+      selection('subagent', ['fork']),
+      selection('workflow', ['workerthread']),
+      selection('hooks', ['claude', 'codex']),
+    ], [new LocalPluginBlueprint('sample', 'plugin'), new LocalPluginBlueprint('lookup', 'tool')])
+    const registry = createBuiltinRegistry(project.profile)
+    const inspections = registry.inspect(project)
+    expect(inspections.filter(item => item.state === 'enabled').map(item => item.id)).toEqual([
+      'provider', 'spine', 'bash', 'app', 'persistence', 'hmr', 'fs', 'todo', 'web', 'subagent', 'workflow', 'hooks',
+    ])
+    expect(inspections.find(item => item.id === 'subagent')?.options).toEqual(['spawn', 'fork'])
+    expect(project.cordisConfigEntries().map(entry => entry.id)).toContain('lookup')
+    const index = await readFile(join(project.root, 'index.ts'), 'utf8')
+    expect(index).toContain('SdkBootContext')
+    expect(index).toContain('agents.create')
+    expect(index).toContain('boot.args.resume')
+    expect(project.packageManifest().scripts).toEqual({
+      dev: 'dsh-sdk dev index.ts -- --model="deepseek-v4-flash"',
+      build: 'dsh-sdk build',
+      typecheck: 'tsc -b',
+      start: 'dsh-sdk start index.js -- --model="deepseek-v4-flash"',
+      config: 'dsh-sdk config',
+    })
+    expect(await readFile(join(project.root, '.env.example'), 'utf8')).toContain('EXA_API_KEY=')
+    expect(project.cordis.entry('stdio')?.config).toMatchObject({ agent: 'main' })
+    expect(project.cordis.entry('stdio')?.config).not.toHaveProperty('model')
+    expect(project.cordis.entry('agent-loop')?.config).toEqual({ agents: [] })
+    expect(project.cordis.entry('system-prompt')?.config?.persona).toContain('{{cwd}}')
+    expect(project.packageManifest().dependencies?.['@cordisjs/plugin-timer']).toBe('^1.1.2')
+    expect(project.packageManifest().dependencies?.['@cordisjs/plugin-hmr']).toBe('^1.0.15')
+    expect(project.packageManifest().dependencies).not.toHaveProperty('node-addon-require-builtin')
+    expect(project.cordis.entry('hmr')).toMatchObject({ name: '@cordisjs/plugin-hmr' })
+    expect(project.cordis.entry('llm-deepseek')?.config).not.toHaveProperty('baseURL')
+    expect(project.cordis.entry('llm-deepseek')?.config).not.toHaveProperty('models')
+  })
+
+  it('round-trips embed app projects without a front-door Cordis config entry', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-embed-app-'))
+    temporary.push(root)
+    const creation = request([], [], 'embed')
+    const project = SdkProject.create(root, creation)
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    for (const item of creation.features) edit.installFeature(registry.get(item.id), item)
+    const committed = (await edit.commit()).project
+    const app = createBuiltinRegistry(committed.profile).get(featureId('app')).inspect(committed)
+    expect(app).toMatchObject({ state: 'enabled', options: ['embed'] })
+    expect(app.selection).toEqual(selection('app', ['embed']))
+    expect(committed.cordis.entry('agent-loop')?.config).toEqual({ agents: [] })
+    expect(committed.cordis.entry('acp')).toBeUndefined()
+    expect(committed.cordis.entry('stdio')).toBeUndefined()
+  })
+
+  it('emits the sandbox workspace-write example as inactive Cordis config', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-sandbox-bash-'))
+    temporary.push(root)
+    const creation = request([], [], 'stdio', 'sandbox')
+    const project = SdkProject.create(root, creation)
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    for (const item of creation.features) edit.installFeature(registry.get(item.id), item)
+    await edit.commit()
+    const cordis = await readFile(join(root, 'cordis.yml'), 'utf8')
+    expect(cordis).toContain(`- id: bash
+  name: "@deepseek-ai/dsh-bash-sandbox"
+  # Uncomment to allow writes under the project workspace.
+  # config:
+  #   mode: workspace-write
+  #   workspaceRoot: !!js process.cwd()`)
+  })
+
+  it('round-trips the custom pi-ai provider with explicit endpoint and default model', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-custom-provider-'))
+    temporary.push(root)
+    const base = request()
+    const creation: ProjectCreationRequest = {
+      ...base,
+      features: [
+        {
+          id: featureId('provider'),
+          options: ['custom'],
+          values: { baseURL: 'https://custom.example/v1' },
+          secrets: { apiKey: 'custom-key' },
+        },
+        ...base.features.filter(item => item.id !== 'provider'),
+      ],
+    }
+    const project = SdkProject.create(root, creation)
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    for (const item of creation.features) edit.installFeature(registry.get(item.id), item)
+    const committed = (await edit.commit()).project
+    expect(committed.cordis.entry('llm-pi-ai')).toMatchObject({
+      name: '@deepseek-ai/dsh-llm-pi-ai',
+      config: { baseURL: 'https://custom.example/v1' },
+    })
+    expect(committed.cordis.entry('llm-pi-ai')?.config).not.toHaveProperty('models')
+    expect(createBuiltinRegistry(committed.profile).get(featureId('provider')).inspect(committed)).toMatchObject({
+      state: 'enabled', options: ['custom'],
+    })
+  })
+
+  it('switches exclusive options and refuses disabling a required feature', async () => {
+    const project = await createCommitted([selection('subagent', ['spawn']), selection('workflow', ['workerthread'])])
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    const persistence = registry.get(featureId('persistence'))
+    edit.configureFeature(persistence, selection('persistence', ['sqlite']))
+    expect(() => { edit.disableFeature(registry.get(featureId('subagent'))) }).toThrow('required by workflow')
+    expect(() => { edit.disableFeature(registry.get(featureId('app'))) }).toThrow('required feature')
+    const committed = await edit.commit()
+    expect(committed.project.cordis.entry('session-persistence')?.name).toContain('sqlite')
+    expect(committed.changes.npmDependenciesChanged).toBe(true)
+  })
+
+  it('switches app-owned files and scripts while protecting user edits', async () => {
+    const project = await createCommitted()
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    edit.configureFeature(registry.get(featureId('app')), selection('app', ['acp']))
+    const acp = (await edit.commit()).project
+    expect(acp.profile.runInterface).toBe('acp')
+    expect(acp.packageManifest().scripts).toMatchObject({
+      dev: 'dsh-sdk dev index.ts',
+      start: 'dsh-sdk start index.js',
+    })
+    expect(await readFile(join(acp.root, 'README.md'), 'utf8')).toContain('Run as an ACP server')
+    expect(await readFile(join(acp.root, 'index.ts'), 'utf8')).not.toContain('agents.create')
+
+    const acpRegistry = createBuiltinRegistry(acp.profile)
+    const embedEdit = acp.edit(acpRegistry)
+    embedEdit.configureFeature(acpRegistry.get(featureId('app')), selection('app', ['embed']))
+    const embed = (await embedEdit.commit()).project
+    expect(embed.profile.runInterface).toBe('embed')
+    expect(await readFile(join(embed.root, 'README.md'), 'utf8')).toContain('Embed the harness')
+    expect(await readFile(join(embed.root, 'index.ts'), 'utf8')).toContain('agents.create')
+
+    await writeFile(join(embed.root, 'README.md'), '# Custom README\n')
+    const modified = await SdkProject.open(embed.root)
+    const modifiedRegistry = createBuiltinRegistry(modified.profile)
+    expect(() => { modified.edit(modifiedRegistry).configureFeature(
+      modifiedRegistry.get(featureId('app')),
+      selection('app', ['stdio']),
+    ) }).toThrow('feature-owned file was modified: README.md')
+
+    const manifest = PackageJsonFile.parse(await readFile(join(embed.root, 'package.json'), 'utf8'))
+    manifest.removeScript('dev')
+    await writeFile(join(embed.root, 'package.json'), manifest.serialize())
+    const incomplete = await SdkProject.open(embed.root)
+    expect(createBuiltinRegistry(incomplete.profile).get(featureId('app')).inspect(incomplete).diagnostics)
+      .toContain('missing package.json script dev')
+  })
+
+  it('rejects enabled features that do not apply to the target app interface', async () => {
+    const project = await createCommitted([selection('ask-user', ['default'])])
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    edit.configureFeature(registry.get(featureId('app')), selection('app', ['embed']))
+    await expect(edit.commit()).rejects.toThrow('feature ask-user is not available for embed')
+  })
+
+  it('supports disabled feature reconfiguration and rejects invalid state operations', async () => {
+    const project = await createCommitted([selection('todo', ['default'])])
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    expect(edit.inspections()).not.toHaveLength(0)
+    const todo = registry.get(featureId('todo'))
+    edit.configureFeature(registry.get(featureId('web')), selection('web', ['deepseek']))
+    edit.disableFeature(todo)
+    edit.configureFeature(todo, selection('todo', ['default']))
+    edit.enableFeature(todo)
+    expect(() => { edit.enableFeature(registry.get(featureId('ask-user'))) }).toThrow('not installed')
+    expect(() => { edit.disableFeature(registry.get(featureId('ask-user'))) }).toThrow('not installed')
+    expect(() => { edit.setCustomPluginDisabled('missing', true) }).toThrow('does not exist')
+    const committed = await edit.commit()
+    expect(committed.changes.enabledFeatures).toContain('todo')
+    expect(() => { edit.enableFeature(todo) }).toThrow('already committed')
+  })
+
+  it('preserves custom entries and toggles only their Loader disabled state', async () => {
+    const project = await createCommitted([], [new LocalPluginBlueprint('sample', 'plugin')])
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    edit.setCustomPluginDisabled('sample', true)
+    expect(edit.cordisConfigEntries().find(entry => entry.id === 'sample')?.disabled).toBe(true)
+    expect(() => { edit.setCustomPluginDisabled('stdio', true) }).toThrow('builtin feature')
+    const next = (await edit.commit()).project
+    const enable = next.edit(createBuiltinRegistry(next.profile))
+    enable.setCustomPluginDisabled('sample', false)
+    expect((await enable.commit()).project.cordis.entry('sample')?.disabled).toBeUndefined()
+  })
+
+  it('rejects local plugin collisions and invalid optional document shapes', async () => {
+    const project = await createCommitted([], [new LocalPluginBlueprint('sample', 'plugin')])
+    const registry = createBuiltinRegistry(project.profile)
+    const npmDependencyConflict = project.edit(registry)
+    expect(() => { npmDependencyConflict.addPlugin(new LocalPluginBlueprint('sample', 'plugin')) })
+      .toThrow('NPM dependency already exists')
+    await writeFile(join(project.root, 'tsconfig.json'), 'not-json\n')
+    const malformed = await SdkProject.open(project.root)
+    expect(() => { malformed.edit(createBuiltinRegistry(malformed.profile)).addPlugin(
+      new LocalPluginBlueprint('other', 'plugin'),
+    ) }).toThrow('requires a valid tsconfig')
+    const entryProject = await createCommitted()
+    const entryEdit = entryProject.edit(createBuiltinRegistry(entryProject.profile))
+    ;(entryEdit as unknown as { cordis(): { addEntry(entry: CordisConfigEntry): void } }).cordis()
+      .addEntry({ id: 'sample', name: 'manual' })
+    expect(() => { entryEdit.addPlugin(new LocalPluginBlueprint('sample', 'plugin')) }).toThrow('entry already exists')
+    const fileEdit = entryProject.edit(createBuiltinRegistry(entryProject.profile))
+    ;(fileEdit as unknown as { documents: Map<string, TextProjectFile> }).documents
+      .set('plugins/other/package.json', new TextProjectFile('plugins/other/package.json', '{}'))
+    expect(() => { fileEdit.addPlugin(new LocalPluginBlueprint('other', 'plugin')) }).toThrow('file already exists')
+  })
+
+  it('reinstalls existing/disabled features and detects requirement cycles', async () => {
+    const project = await createCommitted([selection('todo', ['default']), selection('subagent', ['spawn'])])
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    const todo = registry.get(featureId('todo'))
+    edit.installFeature(todo, selection('todo', ['default']))
+    edit.disableFeature(todo)
+    edit.installFeature(todo, selection('todo', ['default']))
+    const subagent = registry.get(featureId('subagent'))
+    edit.disableFeature(subagent)
+    edit.installFeature(registry.get(featureId('workflow')), selection('workflow', ['workerthread']))
+    expect(edit.cordisConfigEntries().find(entry => entry.id === 'subagent-spawn')?.disabled).toBeUndefined()
+
+    class Cyclic extends FixedFeature {
+      override readonly summary = 'cyclic'
+      override readonly options = [new (class extends FeatureOption {
+        override readonly id = 'one'
+        override readonly label = 'One'
+        override contribution(): ProjectContribution { return new ProjectContribution([]) }
+      })()]
+      override readonly id
+      override readonly requires
+      constructor(id: string, required: string) {
+        super()
+        this.id = featureId(id)
+        this.requires = [featureId(required)]
+      }
+    }
+    const one = new Cyclic('cycle-one', 'cycle-two')
+    const two = new Cyclic('cycle-two', 'cycle-one')
+    const cycleRegistry = new FeatureRegistry([one, two], project.profile)
+    expect(() => { project.edit(cycleRegistry).installFeature(one, selection('cycle-one', ['one'])) })
+      .toThrow('cyclic feature requirement')
+  })
+
+  it('removes clean owned files and detects files disappearing before commit', async () => {
+    const project = await createCommitted([selection('hooks', ['claude', 'codex']), selection('todo', ['default'])])
+    const registry = createBuiltinRegistry(project.profile)
+    const remove = project.edit(registry)
+    remove.configureFeature(registry.get(featureId('hooks')), selection('hooks', ['codex']))
+    const committed = await remove.commit()
+    expect(committed.changes.changedFiles).toContain('hooks.json')
+    const edit = committed.project.edit(createBuiltinRegistry(committed.project.profile))
+    edit.disableFeature(createBuiltinRegistry(committed.project.profile).get(featureId('todo')))
+    await rm(join(committed.project.root, 'cordis.yml'))
+    await expect(edit.commit()).rejects.toThrow('cannot verify project file cordis.yml')
+  })
+
+  it('guards internal resource collisions and malformed aggregate documents', async () => {
+    const project = await createCommitted()
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    type Internals = {
+      documents: Map<string, TextProjectFile>
+      states: Map<ReturnType<typeof featureId>, unknown>
+      applyResource(resource: ProjectResource, previous: ProjectResource | undefined): void
+      removeResource(resource: ProjectResource): void
+      replaceContribution(previous: ProjectContribution | undefined, next: ProjectContribution): void
+      finalProfile(): ProjectProfile
+      manifest(): unknown
+      cordis(): unknown
+      environment(path: '.env' | '.env.example'): unknown
+      state(feature: FixedFeature): unknown
+    }
+    const internals = edit as unknown as Internals
+    const collidingEntry: ProjectResource = {
+      kind: 'cordis-config-entry', key: resourceKey('cordis-config-entry:stdio'),
+      entry: { id: 'stdio', name: 'other-package' }, ownedConfigKeys: [],
+    }
+    expect(() => { internals.applyResource(collidingEntry, undefined) }).toThrow('is owned by')
+    const existingFile: ProjectResource = {
+      kind: 'owned-file', key: resourceKey('file:tsconfig.json'),
+      document: new TextProjectFile('tsconfig.json', 'replacement'), removeOnlyWhenUnchanged: true,
+    }
+    expect(() => { internals.applyResource(existingFile, undefined) }).toThrow('already exists')
+    internals.documents.set('owned.txt', new TextProjectFile('owned.txt', 'old'))
+    const previousFile: ProjectResource = {
+      ...existingFile, key: resourceKey('file:owned.txt'), document: new TextProjectFile('owned.txt', 'old'),
+    }
+    const nextFile: ProjectResource = {
+      ...existingFile, key: resourceKey('file:owned.txt'), document: new TextProjectFile('owned.txt', 'replacement'),
+    }
+    internals.applyResource(nextFile, previousFile)
+    expect(internals.documents.get('owned.txt')?.serialize()).toBe('replacement\n')
+    internals.documents.set('owned.txt', new TextProjectFile('owned.txt', 'user edit'))
+    expect(() => { internals.applyResource(nextFile, previousFile) }).toThrow('was modified')
+    const existingScript: ProjectResource = {
+      kind: 'package-script', key: resourceKey('package-script:build'),
+      name: 'build', command: 'other build', removeOnlyWhenUnchanged: true,
+    }
+    expect(() => { internals.applyResource(existingScript, undefined) }).toThrow('script already exists')
+    const transientScript: ProjectResource = {
+      kind: 'package-script', key: resourceKey('package-script:transient'),
+      name: 'transient', command: 'first', removeOnlyWhenUnchanged: true,
+    }
+    internals.applyResource(transientScript, undefined)
+    const nextScript: ProjectResource = { ...transientScript, command: 'second' }
+    internals.applyResource(nextScript, transientScript)
+    internals.applyResource(nextScript, transientScript)
+    ;(internals.manifest() as PackageJsonFile).setScript('transient', 'user edit')
+    expect(() => { internals.applyResource(transientScript, nextScript) }).toThrow('script was modified')
+    expect(() => { internals.removeResource(nextScript) }).toThrow('script was modified')
+    ;(internals.manifest() as PackageJsonFile).setScript('transient', 'second')
+    internals.removeResource(nextScript)
+    expect(() => { internals.removeResource(nextScript) }).toThrow('script is missing')
+    expect(() => { internals.removeResource({
+      ...existingFile, key: resourceKey('file:missing.txt'), document: new TextProjectFile('missing.txt', 'missing'),
+    }) }).toThrow('owned file is missing')
+    expect(() => { internals.removeResource({
+      kind: 'cordis-config-entry', key: resourceKey('cordis-config-entry:missing'),
+      entry: { id: 'missing', name: 'missing' }, ownedConfigKeys: [],
+    }) }).toThrow('cannot confirm old Cordis resource')
+    const transient: ProjectResource = {
+      kind: 'owned-file', key: resourceKey('file:transient.txt'),
+      document: new TextProjectFile('transient.txt', 'transient'), removeOnlyWhenUnchanged: true,
+    }
+    internals.applyResource(transient, undefined)
+    internals.removeResource(transient)
+    internals.replaceContribution(
+      new ProjectContribution([{ kind: 'npm-dependency', key: resourceKey('shared'), name: 'cordis', section: 'dependencies' }]),
+      new ProjectContribution([{
+        kind: 'cordis-config-entry', key: resourceKey('shared'), entry: { id: 'new', name: 'new' }, ownedConfigKeys: [],
+      }]),
+    )
+    internals.replaceContribution(
+      new ProjectContribution([{
+        kind: 'environment', key: resourceKey('environment:SAME'), name: 'SAME', value: 'old', exampleValue: '',
+      }]),
+      new ProjectContribution([{
+        kind: 'environment', key: resourceKey('environment:SAME'), name: 'SAME', value: 'new', exampleValue: '',
+      }]),
+    )
+    internals.documents.set('.env', new TextProjectFile('.env', 'bad'))
+    expect(() => edit.readEnvironment('.env', 'KEY')).toThrow('not an environment document')
+    expect(() => { internals.environment('.env') }).toThrow('not an environment document')
+    internals.documents.delete('package.json')
+    expect(() => { internals.manifest() }).toThrow('package.json is missing')
+    internals.documents.delete('cordis.yml')
+    expect(() => { internals.cordis() }).toThrow('cordis.yml is missing')
+    class Foreign extends FixedFeature {
+      override readonly id = featureId('foreign')
+      override readonly summary = 'foreign'
+      override readonly options = []
+    }
+    expect(() => { internals.state(new Foreign()) }).toThrow('not applicable')
+    internals.states.delete(featureId('app'))
+    expect(internals.finalProfile()).toBe(project.profile)
+    const sourceDocuments = (project as unknown as { documents: Map<string, TextProjectFile> }).documents
+    sourceDocuments.set('.env', new TextProjectFile('.env', 'bad'))
+    expect(() => project.readEnvironment('.env', 'KEY')).toThrow('not an environment document')
+    sourceDocuments.delete('package.json')
+    expect(() => project.packageJson).toThrow('package.json is missing or invalid')
+    sourceDocuments.delete('cordis.yml')
+    expect(() => project.cordis).toThrow('cordis.yml is missing or invalid')
+  })
+
+  it('rejects external edits before writing any affected file', async () => {
+    const project = await createCommitted([selection('todo', ['default'])])
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    edit.disableFeature(registry.get(featureId('todo')))
+    const manifestBefore = await readFile(join(project.root, 'package.json'), 'utf8')
+    await writeFile(join(project.root, 'cordis.yml'), '# external\n[]\n')
+    await expect(edit.commit()).rejects.toThrow('changed outside this edit session')
+    expect(await readFile(join(project.root, 'package.json'), 'utf8')).toBe(manifestBefore)
+  })
+
+  it('rejects a create target file that appeared after the edit session opened', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-create-conflict-'))
+    temporary.push(root)
+    const creation = request()
+    const project = SdkProject.create(root, creation)
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    for (const item of creation.features) edit.installFeature(registry.get(item.id), item)
+    await writeFile(join(root, 'README.md'), 'external\n')
+    await expect(edit.commit()).rejects.toThrow('changed outside this edit session: README.md')
+  })
+
+  it('uses Cordis config entries as the installation anchor and rejects partial resources', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-inconsistent-'))
+    temporary.push(root)
+    await writeFile(join(root, 'package.json'), JSON.stringify({
+      name: 'partial', dependencies: { '@deepseek-ai/dsh-llm-deepseek': '^0.0.1' },
+    }))
+    await writeFile(join(root, 'cordis.yml'), '[]\n')
+    const project = await SdkProject.open(root)
+    const registry = createBuiltinRegistry(project.profile)
+    expect(registry.get(featureId('provider')).inspect(project)).toMatchObject({
+      state: 'absent', diagnostics: [],
+    })
+
+    const partialRoot = await mkdtemp(join(tmpdir(), 'dsh-entry-partial-'))
+    temporary.push(partialRoot)
+    await writeFile(join(partialRoot, 'package.json'), JSON.stringify({ name: 'partial-entry' }))
+    await writeFile(join(partialRoot, 'cordis.yml'), `- id: llm-deepseek
+  name: '@deepseek-ai/dsh-llm-deepseek'
+  config:
+    apiKey: test
+`)
+    const partial = await SdkProject.open(partialRoot)
+    const installation = createBuiltinRegistry(partial.profile)
+      .get(featureId('provider')).inspect(partial)
+    expect(installation.state).toBe('inconsistent')
+    expect(installation.diagnostics).toContain('missing package.json dependencies entry @deepseek-ai/dsh-llm-deepseek')
+    const partialEdit = partial.edit(createBuiltinRegistry(partial.profile))
+    const provider = createBuiltinRegistry(partial.profile).get(featureId('provider'))
+    expect(() => { partialEdit.configureFeature(provider, selection('provider', ['deepseek'])) }).toThrow('inconsistent')
+    expect(() => { partialEdit.enableFeature(provider) }).toThrow('inconsistent')
+    expect(() => { partialEdit.disableFeature(provider) }).toThrow('required feature')
+  })
+
+  it('rejects inconsistent optional features and incompatible requirement options', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-optional-inconsistent-'))
+    temporary.push(root)
+    await writeFile(join(root, 'package.json'), '{"name":"partial"}')
+    await writeFile(join(root, 'cordis.yml'), `- id: web-search-exa
+  name: '@deepseek-ai/dsh-web-search-exa'
+`)
+    const project = await SdkProject.open(root)
+    const builtin = createBuiltinRegistry(project.profile)
+    const edit = project.edit(builtin)
+    const web = builtin.get(featureId('web'))
+    expect(() => { edit.disableFeature(web) }).toThrow('inconsistent')
+    expect(() => { edit.installFeature(web, selection('web', ['deepseek'])) }).toThrow('inconsistent')
+
+    class RequiresWeb extends FixedFeature {
+      override readonly id = featureId('requires-web')
+      override readonly summary = 'requires web'
+      override readonly requires = [featureId('web')]
+      override readonly options = [new (class extends FeatureOption {
+        override readonly id = 'one'
+        override readonly label = 'One'
+        override contribution(): ProjectContribution { return new ProjectContribution([]) }
+      })()]
+    }
+    const requiresWeb = new RequiresWeb()
+    const webRegistry = new FeatureRegistry([web, requiresWeb], project.profile)
+    expect(() => { project.edit(webRegistry).installFeature(requiresWeb, selection('requires-web', ['one'])) })
+      .toThrow('required feature web is inconsistent')
+
+    class RequiresAcp extends FixedFeature {
+      override readonly id = featureId('requires-acp')
+      override readonly summary = 'requires acp'
+      override readonly options = [new (class extends FeatureOption {
+        override readonly id = 'one'
+        override readonly label = 'One'
+        override contribution(): ProjectContribution { return new ProjectContribution([]) }
+      })()]
+      override requirements(): readonly [{ id: ReturnType<typeof featureId>; options: readonly string[] }] {
+        return [{ id: featureId('app'), options: ['acp'] }]
+      }
+    }
+    const requiring = new RequiresAcp()
+    const complete = await createCommitted()
+    const app = createBuiltinRegistry(complete.profile).get(featureId('app'))
+    const registry = new FeatureRegistry([app, requiring], complete.profile)
+    expect(() => { complete.edit(registry).installFeature(requiring, selection('requires-acp', ['one'])) })
+      .toThrow('does not satisfy the option requirement')
+  })
+
+  it('removes obsolete environment resources when switching options', async () => {
+    const project = await createCommitted([selection('web', ['exa'], { apiKey: 'exa' })])
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    edit.configureFeature(registry.get(featureId('web')), selection('web', ['deepseek']))
+    expect(edit.readEnvironment('.env.example', 'EXA_API_KEY')).toBeUndefined()
+  })
+
+  it('preserves duplicate and existing .env values while appending differently named secrets', async () => {
+    const project = await createCommitted()
+    if (process.platform !== 'win32') {
+      expect((await stat(join(project.root, '.env'))).mode & 0o777).toBe(0o600)
+    }
+    const original = '# keep\nDEEPSEEK_API_KEY=first\nDEEPSEEK_API_KEY=second\n'
+    await writeFile(join(project.root, '.env'), original)
+    if (process.platform !== 'win32') await chmod(join(project.root, '.env'), 0o640)
+    const reopened = await SdkProject.open(project.root)
+    const registry = createBuiltinRegistry(reopened.profile)
+    expect(registry.get(featureId('provider')).inspect(reopened)).toMatchObject({
+      state: 'enabled', selection: { secrets: { apiKey: 'second' } },
+    })
+    const edit = reopened.edit(registry)
+    edit.configureFeature(
+      registry.get(featureId('provider')),
+      selection('provider', ['deepseek'], { apiKey: 'replacement' }),
+    )
+    edit.installFeature(registry.get(featureId('web')), selection('web', ['exa'], { apiKey: 'exa-key' }))
+    const withExa = (await edit.commit()).project
+    expect(await readFile(join(withExa.root, '.env'), 'utf8')).toBe(`${original}EXA_API_KEY=exa-key\n`)
+    if (process.platform !== 'win32') {
+      expect((await stat(join(withExa.root, '.env'))).mode & 0o777).toBe(0o640)
+    }
+    const nextRegistry = createBuiltinRegistry(withExa.profile)
+    const remove = withExa.edit(nextRegistry)
+    remove.configureFeature(nextRegistry.get(featureId('web')), selection('web', ['deepseek']))
+    await remove.commit()
+    expect(await readFile(join(withExa.root, '.env'), 'utf8')).toBe(`${original}EXA_API_KEY=exa-key\n`)
+  })
+
+  it('refuses to remove a feature-owned file after user edits', async () => {
+    const project = await createCommitted([selection('hooks', ['claude', 'codex'])])
+    await writeFile(join(project.root, 'hooks.json'), '{"hooks":{}}\n')
+    const reopened = await SdkProject.open(project.root)
+    const registry = createBuiltinRegistry(reopened.profile)
+    const edit = reopened.edit(registry)
+    expect(() => { edit.configureFeature(
+      registry.get(featureId('hooks')),
+      selection('hooks', ['codex']),
+    ) }).toThrow('owned file was modified: hooks.json')
+  })
+
+  it('does not mistake a linked NPM dependency closure for an installed feature', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-link-closure-inspection-'))
+    temporary.push(root)
+    const base = request([selection('hooks', ['claude'])])
+    const creation: ProjectCreationRequest = { ...base, linkWorkspaceRoot: repoRoot }
+    const project = SdkProject.create(root, creation)
+    const registry = createBuiltinRegistry(project.profile)
+    const edit = project.edit(registry)
+    for (const item of creation.features) edit.installFeature(registry.get(item.id), item)
+    const committed = (await edit.commit()).project
+    expect(committed.packageManifest().dependencies?.['@deepseek-ai/dsh-subagent']).toMatch(/^file:/)
+    expect(createBuiltinRegistry(committed.profile).get(featureId('subagent')).inspect(committed).state).toBe('absent')
+  })
+})
+
+describe('extension points', () => {
+  const profile: ProjectProfile = {
+    name: 'test', description: 'test', runtime: { model: 'm' }, runInterface: 'embed',
+    packageManager: new NpmPackageManager('10.0.0'), releaseVersion: '0.0.1',
+  }
+
+  it('rejects cross-feature resource ownership conflicts at registry construction', () => {
+    class TestOption extends FeatureOption {
+      override readonly id = 'default'
+      override readonly label = 'Default'
+      override contribution(): ProjectContribution {
+        return new ProjectContribution([{
+          kind: 'npm-dependency', key: resourceKey('npm-dependency:shared'), name: 'shared', section: 'dependencies',
+        }])
+      }
+    }
+    class TestFeature extends FixedFeature {
+      override readonly id
+      override readonly summary = 'test'
+      override readonly options = [new TestOption()]
+      constructor(id: string) {
+        super()
+        this.id = featureId(id)
+      }
+    }
+    expect(() => new FeatureRegistry([
+      new TestFeature('one'), new TestFeature('two'),
+    ], profile)).toThrow('declared by both one and two')
+    expect(new TestFeature('one').defaultOptions()).toEqual(['default'])
+    expect(() => new FeatureRegistry([
+      new TestFeature('one'), new TestFeature('one'),
+    ], profile)).toThrow('duplicate feature id')
+  })
+
+  it('validates selection modes and declarative feature definitions', () => {
+    const option = { id: 'one', label: 'One', default: true, resources: [] } as const
+    expect(() => defineFeature({ id: 'bad-single', summary: 'bad', mode: 'single', options: [] }))
+      .toThrow('requires one default option')
+    expect(() => defineFeature({
+      id: 'bad-exclusive', summary: 'bad', mode: 'exclusive', options: [{ ...option, default: false }],
+    })).toThrow('exactly one default option')
+    expect(() => defineFeature({
+      id: 'bad-multiple', summary: 'bad', mode: 'multiple', options: [{ ...option, default: false }],
+    })).toThrow('at least one default option')
+    const exclusive = defineFeature({
+      id: 'defined', summary: 'Defined', mode: 'exclusive', supportedInterfaces: ['embed'],
+      requires: [{ id: 'base' }], suggests: ['suggested'],
+      baseResources: [{ kind: 'npm-dependency', name: 'base', section: 'devDependencies' }],
+      options: [
+        {
+          id: 'one', label: 'One', default: true,
+          requires: [{ id: 'option', options: ['required'] }],
+          secrets: [{ id: 'token', environment: 'TOKEN', message: 'Token', required: true }],
+          resources: [
+            {
+              kind: 'npm-cordis-config-entry', id: 'one', package: 'one-package',
+              config: { nested: { value: 1 }, list: ['x'], nullable: null },
+            },
+            { kind: 'owned-file', path: 'one.txt', text: 'one', removeOnlyWhenUnchanged: false },
+          ],
+        },
+        {
+          id: 'two', label: 'Two', resources: [
+            { kind: 'file-cordis-config-entry', id: 'two', path: './two.ts' },
+          ],
+        },
+      ],
+    })
+    expect(exclusive.defaultOptions(profile)).toEqual(['one'])
+    expect(exclusive.isApplicable(profile)).toBe(true)
+    expect(exclusive.isApplicable({ ...profile, runInterface: 'stdio' })).toBe(false)
+    expect(exclusive.requirements(selection('defined', ['one']))).toEqual([
+      { id: 'base' }, { id: 'option', options: ['required'] },
+    ])
+    const contribution = exclusive.contribution({
+      id: featureId('defined'), options: ['one'], secrets: { token: 'secret' },
+    }, profile)
+    expect(contribution.resources.map(resource => resource.kind)).toEqual([
+      'npm-dependency', 'npm-dependency', 'cordis-config-entry', 'owned-file', 'environment',
+    ])
+    const entry = contribution.resources.find(resource => resource.kind === 'cordis-config-entry')
+    expect(entry?.validateConfig?.({ nested: { value: 2 }, list: ['a', 'b'], nullable: null })).toEqual([])
+    expect(entry?.validateConfig?.({ nested: [], list: 'bad' })).toHaveLength(3)
+    expect(() => exclusive.normalizeSelection(selection('other', ['one']), profile)).toThrow('does not belong')
+    expect(() => exclusive.normalizeSelection(selection('defined', ['one']), { ...profile, runInterface: 'stdio' }))
+      .toThrow('not available')
+    expect(() => exclusive.normalizeSelection(selection('defined', ['missing']), profile)).toThrow('unknown')
+    expect(() => exclusive.normalizeSelection(selection('defined', ['one', 'two']), profile)).toThrow('exactly one')
+    expect(defineFeatures([exclusive, {
+      id: 'fixed', summary: 'Fixed', mode: 'single', options: [option],
+    }])).toHaveLength(2)
+    expect(() => new FeatureRegistry([], profile).get(featureId('missing'))).toThrow('unknown feature')
+    expect(new FeatureRegistry([exclusive], profile).ownerOfPackage('one-package', { ...profile, runInterface: 'stdio' }))
+      .toBeUndefined()
+    class Unsupported extends FixedFeature {
+      override readonly id = featureId('unsupported')
+      override readonly summary = 'unsupported'
+      override readonly options = [new (class extends FeatureOption {
+        override readonly id = 'one'
+        override readonly label = 'One'
+        override contribution(): ProjectContribution { return new ProjectContribution([]) }
+      })()]
+      override readonly supportedInterfaces = []
+    }
+    expect(() => new FeatureRegistry([new Unsupported()], profile)).toThrow('supports no run interface')
+  })
+
+  it('covers feature base classes and resource conflict checks', () => {
+    class EmptySimple extends FixedFeature {
+      override readonly id = featureId('empty')
+      override readonly summary = 'empty'
+      override readonly options = []
+    }
+    expect(() => new EmptySimple().defaultOptions()).toThrow('has no option')
+    class BadSimple extends FixedFeature {
+      override readonly id = featureId('bad-simple')
+      override readonly summary = 'bad'
+      override readonly options = [new (class extends FeatureOption {
+        override readonly id = 'one'
+        override readonly label = 'One'
+        override contribution(): ProjectContribution { return new ProjectContribution([]) }
+      })(), new (class extends FeatureOption {
+        override readonly id = 'two'
+        override readonly label = 'Two'
+        override contribution(): ProjectContribution { return new ProjectContribution([]) }
+      })()]
+    }
+    expect(() => new BadSimple().normalizeSelection(selection('bad-simple', ['one']), profile)).toThrow('one fixed option')
+    class EmptyMulti extends MultiOptionFeature {
+      override readonly id = featureId('multi')
+      override readonly summary = 'multi'
+      override readonly options = []
+      override defaultOptions(): readonly string[] { return [] }
+    }
+    expect(() => new EmptyMulti().normalizeSelection(selection('multi', []), profile)).toThrow('at least one')
+    class EmptyExclusive extends ExclusiveOptionFeature {
+      override readonly id = featureId('exclusive')
+      override readonly summary = 'exclusive'
+      override readonly options = []
+      override defaultOptions(): readonly string[] { return [] }
+    }
+    expect(() => new EmptyExclusive().normalizeSelection(selection('exclusive', []), profile)).toThrow('exactly one')
+    const resource = {
+      kind: 'npm-dependency' as const, key: resourceKey('same'), name: 'one', section: 'dependencies' as const,
+    }
+    expect(() => new ProjectContribution([resource, resource])).toThrow('duplicate contribution')
+    expect(() => ProjectContribution.merge(
+      new ProjectContribution([resource]),
+      new ProjectContribution([{ ...resource, name: 'two' }]),
+    )).toThrow('conflicting definitions')
+    expect(ProjectContribution.merge(new ProjectContribution([resource]), new ProjectContribution([resource])).byKey().size)
+      .toBe(1)
+    expect(ownedTextFile('owner', 'file.txt', 'text').document).toBeInstanceOf(TextProjectFile)
+    expect(optionalString({ value: 1 }, 'value')).toHaveLength(1)
+    expect(optionalString({}, 'value')).toEqual([])
+    expect(requiredString({ value: 'x' }, 'value')).toEqual([])
+    expect(stringArray({ value: ['a'] }, 'value')).toEqual([])
+    expect(stringArray({ value: [1] }, 'value')).toHaveLength(1)
+    expect(cordisConfigEntry('owner', { id: 'entry', name: 'pkg' }).ownedConfigKeys).toEqual([])
+    expect(npmCordisConfigEntry('owner', { id: 'entry', name: 'pkg' })[1].ownedConfigKeys).toEqual([])
+    expect(environmentResource('owner', 'EMPTY', undefined)).not.toHaveProperty('value')
+    const builtins = createBuiltinRegistry(profile)
+    expect(builtins.get(featureId('app')).defaultOptions(profile)).toEqual(['embed'])
+    expect(builtins.get(featureId('hmr')).defaultOptions(profile)).toEqual(['default'])
+    const app = builtins.get(featureId('app'))
+    const acpEntry = builtins.get(featureId('app')).contribution(selection('app', ['acp']), profile).resources
+      .find((resource): resource is CordisConfigEntryResource =>
+        resource.kind === 'cordis-config-entry' && resource.entry.id === 'acp')
+    expect(acpEntry?.entry.id).toBe('acp')
+    expect(acpEntry?.validateConfig?.({ model: '' })).toHaveLength(1)
+    const embedOption = app.options.find(option => option.id === 'embed')
+    expect(embedOption?.markerConfigEntries(profile)).toEqual([])
+    expect(embedOption?.contribution(profile, {}).resources.map(resource => resource.kind)).toEqual([
+      'owned-file', 'owned-file', 'package-script', 'package-script',
+    ])
+    expect(embedOption?.matchesConfigEntries([
+      { id: 'agent-loop', name: '@deepseek-ai/dsh-agent-loop' },
+      { id: 'stdio', name: '@deepseek-ai/dsh-stdio' },
+    ], profile)).toBe(false)
+    const spineAgentLoop = builtins.get(featureId('spine')).contribution(selection('spine', ['default']), profile).resources
+      .find((resource): resource is CordisConfigEntryResource =>
+        resource.kind === 'cordis-config-entry' && resource.entry.id === 'agent-loop')
+    expect(spineAgentLoop?.validateConfig?.({ agents: 'main' })).toEqual(['agents must be an array'])
+    expect(spineAgentLoop?.validateConfig?.({ agents: ['main'] })).toEqual(['agents must be empty'])
+    expect(spineAgentLoop?.validateConfig?.({ agents: [] })).toEqual([])
+    expect(builtins.get(featureId('provider')).defaultOptions(profile)).toEqual(['deepseek'])
+    expect(() => builtins.get(featureId('provider')).contribution({
+      id: featureId('provider'), options: ['custom'], values: { baseURL: 1 },
+    }, profile)).toThrow('baseURL must be a string')
+    const alternateModel = builtins.get(featureId('provider')).contribution({
+      id: featureId('provider'), options: ['deepseek'],
+    }, { ...profile, runtime: { model: 'other' } }).resources
+      .find(resource => resource.kind === 'cordis-config-entry')
+    expect(alternateModel?.entry.config?.models).toEqual(['other'])
+    class RequiringSimple extends BadSimple {
+      override readonly requires = [featureId('npm-dependency')]
+    }
+    expect(new RequiringSimple().requirements(selection('bad-simple', ['one']))).toEqual([{ id: 'npm-dependency' }])
+    expect(builtins.get(featureId('bash')).defaultOptions(profile)).toEqual(['local'])
+  })
+
+  it('reports every inconsistent feature resource shape', () => {
+    const feature = defineFeature({
+      id: 'inspectable', summary: 'Inspectable', mode: 'single',
+      baseResources: [{ kind: 'file-cordis-config-entry', id: 'base', path: 'pkg' }],
+      options: [{
+        id: 'one', label: 'One', default: true,
+        secrets: [{ id: 'token', environment: 'TOKEN', message: 'Token', required: true }],
+        resources: [
+          { kind: 'file-cordis-config-entry', id: 'one', path: 'pkg', config: { value: 'x' } },
+          { kind: 'npm-dependency', name: 'dep' },
+          { kind: 'owned-file', path: 'owned.txt', text: 'owned' },
+        ],
+      }],
+    })
+    const view = (entries: readonly CordisConfigEntry[]): FeatureProjectView => ({
+      profile,
+      cordisConfigEntries: () => entries,
+      packageManifest: () => ({}),
+      hasDocument: () => false,
+      readEnvironment: (path) => {
+        if (path === '.env.example') throw new Error('bad env')
+        return 'secret'
+      },
+    })
+    expect(feature.inspect(view([])).state).toBe('absent')
+    const inconsistent = feature.inspect(view([
+      { id: 'one', name: 'pkg', config: { value: 1 } },
+      { id: 'extra', name: 'pkg', disabled: true },
+    ]))
+    expect(inconsistent.state).toBe('inconsistent')
+    expect(inconsistent.diagnostics.join('\n')).toContain('missing Cordis config entry base')
+    expect(inconsistent.diagnostics.join('\n')).toContain('unexpected owned Cordis config entry extra')
+    expect(inconsistent.diagnostics.join('\n')).toContain('missing package.json dependencies entry dep')
+    expect(inconsistent.diagnostics.join('\n')).toContain('missing owned file owned.txt')
+    expect(inconsistent.diagnostics.join('\n')).toContain('bad env')
+    expect(inconsistent.diagnostics.join('\n')).toContain('mixed enabled states')
+    expect(feature.inspect(view([{ id: 'unknown', name: 'pkg' }])).state).toBe('inconsistent')
+    const ambiguous = defineFeature({
+      id: 'ambiguous', summary: 'Ambiguous', mode: 'exclusive',
+      options: [
+        { id: 'one', label: 'One', default: true, resources: [], markers: [{ id: 'one', name: 'pkg' }] },
+        { id: 'two', label: 'Two', resources: [], markers: [{ id: 'two', name: 'pkg' }] },
+      ],
+    })
+    expect(ambiguous.inspect(view([{ id: 'one', name: 'pkg' }, { id: 'two', name: 'pkg' }])).state)
+      .toBe('inconsistent')
+    const noValidator = defineFeature({
+      id: 'no-validator', summary: 'No validator', mode: 'single',
+      options: [{
+        id: 'one', label: 'One', default: true,
+        resources: [{ kind: 'file-cordis-config-entry', id: 'plain', path: 'plain-package' }],
+      }],
+    })
+    expect(noValidator.inspect(view([{ id: 'plain', name: 'plain-package' }])).state).toBe('enabled')
+    const app = createBuiltinRegistry(profile).get(featureId('app'))
+    expect(app.inspect(view([{ id: 'acp', name: '@deepseek-ai/dsh-acp' }])).state)
+      .toBe('inconsistent')
+  })
+})
diff --git a/packages/sdk/helper/tests/questions.spec.ts b/packages/sdk/helper/tests/questions.spec.ts
new file mode 100644
index 0000000000..463f3b979f
--- /dev/null
+++ b/packages/sdk/helper/tests/questions.spec.ts
@@ -0,0 +1,453 @@
+import { PassThrough, Writable } from 'node:stream'
+import { stripVTControlCharacters } from 'node:util'
+import { S_CHECKBOX_SELECTED, S_RADIO_ACTIVE, S_WARN } from '@clack/prompts'
+import { describe, expect, it } from 'vitest'
+import { createBuiltinRegistry } from '../src/features/builtin/index.ts'
+import { FeatureConfigurator } from '../src/features/feature-configurator.ts'
+import { FeatureOption, ExclusiveOptionFeature } from '../src/features/feature.ts'
+import { ProjectContribution } from '../src/features/resources.ts'
+import { featureId } from '../src/ids.ts'
+import { NpmPackageManager } from '../src/package-managers/package-manager.ts'
+import { ClackPromptPort } from '../src/questions/clack-prompt-port.ts'
+import {
+  PromptCancelledError,
+  requireAnswer,
+  type ConfirmPromptRequest,
+  type MultiSelectPromptRequest,
+  type NestedMultiSelectRequest,
+  type NestedMultiSelectValue,
+  type PromptOutcome,
+  type PromptPort,
+  type SecretPromptRequest,
+  type SelectPromptRequest,
+  type TextPromptRequest,
+} from '../src/questions/prompt-port.ts'
+import {
+  ConfirmQuestion,
+  MultiSelectQuestion,
+  SecretQuestion,
+  SelectQuestion,
+  TextQuestion,
+} from '../src/questions/question.ts'
+import type { ProjectProfile } from '../src/project/types.ts'
+import { clackNestedMultiselect } from '../src/questions/clack-nested-multiselect.ts'
+
+function validateString(
+  outcome: PromptOutcome<string>,
+  validate: ((value: string) => string | undefined) | undefined,
+): PromptOutcome<string> {
+  if (outcome.status === 'answered') {
+    const diagnostic = validate?.(outcome.value)
+    if (diagnostic) throw new Error(diagnostic)
+  }
+  return outcome
+}
+
+class QueuePromptPort implements PromptPort {
+  readonly answers: unknown[]
+  readonly requests: string[] = []
+
+  constructor(answers: unknown[]) {
+    this.answers = [...answers]
+  }
+
+  next<T>(message: string): PromptOutcome<T> {
+    this.requests.push(message)
+    const value = this.answers.shift()
+    return value === QueuePromptPort.cancel ? { status: 'cancelled' } : { status: 'answered', value: value as T }
+  }
+
+  async text(request: TextPromptRequest): Promise<PromptOutcome<string>> {
+    return validateString(this.next<string>(request.message), request.validate)
+  }
+
+  async secret(request: SecretPromptRequest): Promise<PromptOutcome<string>> {
+    return validateString(this.next<string>(request.message), request.validate)
+  }
+
+  select<T>(request: SelectPromptRequest<T>): Promise<PromptOutcome<T>> {
+    return Promise.resolve(this.next(request.message))
+  }
+
+  multiselect<T>(request: MultiSelectPromptRequest<T>): Promise<PromptOutcome<readonly T[]>> {
+    return Promise.resolve(this.next(request.message))
+  }
+
+  confirm(request: ConfirmPromptRequest): Promise<PromptOutcome<boolean>> {
+    return Promise.resolve(this.next(request.message))
+  }
+  nestedMultiselect<TValue, TChoice>(
+    request: NestedMultiSelectRequest<TValue, TChoice>,
+  ): Promise<PromptOutcome<readonly NestedMultiSelectValue<TValue, TChoice>[]>> {
+    return Promise.resolve(this.next(request.message))
+  }
+
+  static readonly cancel = Symbol('cancel')
+}
+
+describe('typed questions', () => {
+  it('uses and validates prefilled answers without prompting', async () => {
+    const port = new QueuePromptPort([])
+    const text = new TextQuestion({ id: 'name', message: 'Name', validate: value => value ? undefined : 'required' })
+    await expect(text.resolve(port, 'demo')).resolves.toEqual({ status: 'answered', value: 'demo' })
+    await expect(text.resolve(port, '')).rejects.toThrow('name: required')
+    const select = new SelectQuestion({
+      id: 'choice', message: 'Choice', options: [{ value: 'a', label: 'A' }], initialValue: 'a',
+    })
+    await expect(select.resolve(port, 'b')).rejects.toThrow('unknown or disabled option')
+    await expect(select.resolve(port, 'a')).resolves.toMatchObject({ value: 'a' })
+    const disabled = new SelectQuestion({
+      id: 'disabled', message: 'Disabled', options: [{ value: 'a', label: 'A', disabled: true }],
+    })
+    await expect(disabled.resolve(port, 'a')).rejects.toThrow('disabled option')
+    const multi = new MultiSelectQuestion({
+      id: 'many', message: 'Many', options: [{ value: 'a', label: 'A' }], required: true,
+    })
+    await expect(multi.resolve(port, [])).rejects.toThrow('choose at least one')
+    await expect(multi.resolve(port, ['missing'])).rejects.toThrow('unknown or disabled option')
+    await expect(new MultiSelectQuestion({
+      id: 'disabled-many', message: 'Disabled many', options: [{ value: 'a', label: 'A', disabled: true }],
+    }).resolve(port, ['a'])).rejects.toThrow('disabled option')
+    await expect(new MultiSelectQuestion({
+      id: 'optional', message: 'Optional', options: [{ value: 'a', label: 'A' }],
+    }).resolve(port, [])).resolves.toMatchObject({ value: [] })
+    const secret = new SecretQuestion({ id: 'secret', message: 'Secret', validate: value => value ? undefined : 'required' })
+    await expect(secret.resolve(port, 'value')).resolves.toMatchObject({ value: 'value' })
+    await expect(secret.resolve(port, '')).rejects.toThrow('secret: required')
+    await expect(new ConfirmQuestion({ id: 'confirm', message: 'Confirm' }).resolve(port, false))
+      .resolves.toEqual({ status: 'answered', value: false })
+    expect(port.requests).toEqual([])
+  })
+
+  it('delegates each interaction shape and propagates cancellation', async () => {
+    const port = new QueuePromptPort(['text', 'secret', 'a', ['a'], true, QueuePromptPort.cancel])
+    await expect(new TextQuestion({ id: 't', message: 'Text' }).resolve(port)).resolves.toMatchObject({ value: 'text' })
+    await expect(new SecretQuestion({ id: 's', message: 'Secret' }).resolve(port)).resolves.toMatchObject({ value: 'secret' })
+    await expect(new SelectQuestion({
+      id: 'one', message: 'One', options: [{ value: 'a', label: 'A' }],
+    }).resolve(port)).resolves.toMatchObject({ value: 'a' })
+    await expect(new MultiSelectQuestion({
+      id: 'many', message: 'Many', options: [{ value: 'a', label: 'A' }],
+    }).resolve(port)).resolves.toMatchObject({ value: ['a'] })
+    await expect(new ConfirmQuestion({ id: 'yes', message: 'Yes?' }).resolve(port)).resolves.toMatchObject({ value: true })
+    const cancelled = await new ConfirmQuestion({ id: 'cancel', message: 'Cancel?' }).resolve(port)
+    expect(() => requireAnswer(cancelled)).toThrow(PromptCancelledError)
+    const optionsPort = new QueuePromptPort(['full', 'a', ['a']])
+    await new TextQuestion({
+      id: 'full', message: 'Full', placeholder: 'p', initialValue: 'i', defaultValue: 'd', validate: () => undefined,
+    }).resolve(optionsPort)
+    await new SelectQuestion({
+      id: 'initial', message: 'Initial', options: [{ value: 'a', label: 'A' }], initialValue: 'a',
+    }).resolve(optionsPort)
+    await new MultiSelectQuestion({
+      id: 'initial-many', message: 'Initial many', options: [{ value: 'a', label: 'A' }],
+      initialValues: ['a'], required: true,
+    }).resolve(optionsPort)
+  })
+
+  it('accepts a visible placeholder default before required validation', async () => {
+    const input = new PassThrough()
+    const output = new Writable({ write(_chunk, _encoding, callback) { callback() } })
+    const pending = new ClackPromptPort(input, output).text({
+      message: 'Directory',
+      placeholder: 'my-agent',
+      defaultValue: 'my-agent',
+      validate: value => value ? undefined : 'required',
+    })
+    setTimeout(() => input.write('\r'), 0)
+    await expect(pending).resolves.toEqual({ status: 'answered', value: 'my-agent' })
+  })
+
+  it('renders warning confirmations with a yellow warning marker', async () => {
+    const input = new PassThrough()
+    let screen = ''
+    const output = new Writable({ write(chunk, _encoding, callback) { screen += String(chunk); callback() } })
+    const pending = new ClackPromptPort(input, output).confirm({
+      message: 'Keep empty?',
+      initialValue: true,
+      tone: 'warning',
+    })
+    setTimeout(() => input.write('\r'), 0)
+    await expect(pending).resolves.toEqual({ status: 'answered', value: true })
+    expect(stripVTControlCharacters(screen)).toContain(`${S_WARN} Keep empty?`)
+  })
+
+  it('adapts secret, select, multiselect, nested, and cancellation prompts', async () => {
+    const run = async <T>(
+      start: (port: ClackPromptPort) => Promise<PromptOutcome<T>>,
+      keys: string,
+    ): Promise<PromptOutcome<T>> => {
+      const input = new PassThrough()
+      const output = new Writable({ write(_chunk, _encoding, callback) { callback() } })
+      const pending = start(new ClackPromptPort(input, output))
+      setTimeout(() => input.write(keys), 0)
+      return pending
+    }
+    await expect(run(port => port.secret({ message: 'Secret', validate: value => value ? undefined : 'required' }), 'key\r'))
+      .resolves.toEqual({ status: 'answered', value: 'key' })
+    await expect(run(port => port.secret({ message: 'Secret' }), 'plain\r'))
+      .resolves.toEqual({ status: 'answered', value: 'plain' })
+    await expect(run(port => port.text({ message: 'Text', initialValue: 'seed' }), '\r'))
+      .resolves.toEqual({ status: 'answered', value: 'seed' })
+    let validated = 'unset'
+    await expect(run(port => port.text({
+      message: 'Empty', validate: (value) => { validated = value; return undefined },
+    }), '\r')).resolves.toEqual({ status: 'answered', value: '' })
+    expect(validated).toBe('')
+    await expect(run(port => port.select({
+      message: 'Select', options: [{ value: 'a', label: 'A', hint: 'hint' }, { value: 'b', label: 'B', disabled: true }],
+      initialValue: 'a',
+    }), '\r')).resolves.toEqual({ status: 'answered', value: 'a' })
+    await expect(run(port => port.multiselect({
+      message: 'Many', options: [{ value: 'a', label: 'A' }], initialValues: ['a'], required: true,
+    }), '\r')).resolves.toEqual({ status: 'answered', value: ['a'] })
+    await expect(run(port => port.multiselect({
+      message: 'Many', options: [{ value: 'a', label: 'A' }],
+    }), ' \r')).resolves.toEqual({ status: 'answered', value: ['a'] })
+    await expect(run(port => port.nestedMultiselect({
+      message: 'Nested', options: [{ value: 'a', label: 'A', default: true }],
+    }), '\r')).resolves.toEqual({ status: 'answered', value: [{ value: 'a', choices: [] }] })
+    await expect(run(port => port.confirm({ message: 'Cancel' }), '\u0003')).resolves.toEqual({ status: 'cancelled' })
+    expect(new ClackPromptPort()).toBeInstanceOf(ClackPromptPort)
+  })
+})
+
+describe('nested Clack picker', () => {
+  it('navigates root options, ignores disabled rows, and toggles optional rows', async () => {
+    const input = new PassThrough()
+    const output = new Writable({ write(_chunk, _encoding, callback) { callback() } })
+    const pending = clackNestedMultiselect({
+      message: 'Features', showChanges: true, input, output,
+      options: [
+        { value: 'required', label: 'Required', required: true },
+        { value: 'optional', label: 'Optional', default: true },
+        { value: 'added', label: 'Added' },
+        { value: 'disabled', label: 'Disabled', disabled: true, warning: 'disabled warning' },
+      ],
+    })
+    setTimeout(() => input.write('\x1b[A \x1b[B\x1b[B \x1b[B \x1b[A\r'), 0)
+    await expect(pending).resolves.toEqual({
+      status: 'answered',
+      value: [{ value: 'required', choices: [] }, { value: 'added', choices: [] }],
+    })
+  })
+
+  it('cancels from the root layer', async () => {
+    const input = new PassThrough()
+    const output = new Writable({ write(_chunk, _encoding, callback) { callback() } })
+    const pending = clackNestedMultiselect({
+      message: 'Features', input, output, options: [{ value: 'one', label: 'One' }],
+    })
+    setTimeout(() => input.write('\u0003'), 0)
+    await expect(pending).resolves.toEqual({ status: 'cancelled' })
+  })
+
+  it('enters an exclusive child with Right and commits the selected option', async () => {
+    const input = new PassThrough()
+    let screen = ''
+    const output = new Writable({ write(chunk, _encoding, callback) { screen += String(chunk); callback() } })
+    const pending = clackNestedMultiselect({
+      message: 'Features',
+      showChanges: true,
+      input,
+      output,
+      options: [
+        {
+          value: 'persistence',
+          label: 'Session storage',
+          required: true,
+          default: true,
+          choiceMode: 'exclusive',
+          choices: [
+            { value: 'jsonl', label: 'JSONL', default: true },
+            { value: 'sqlite', label: 'SQLite' },
+          ],
+        },
+        { value: 'fs', label: 'Filesystem', default: true },
+      ],
+    })
+    setTimeout(() => input.write('\x1b[C\x1b[B\x1b[A\x1b[B\x1b[C\r\r'), 0)
+    await expect(pending).resolves.toEqual({
+      status: 'answered',
+      value: [
+        { value: 'persistence', choices: ['sqlite'] },
+        { value: 'fs', choices: [] },
+      ],
+    })
+    const rendered = stripVTControlCharacters(screen)
+    expect(rendered).toContain(`› ${S_CHECKBOX_SELECTED} Session storage`)
+    expect(rendered).toContain(`› ${S_RADIO_ACTIVE} SQLite`)
+    expect(rendered).toContain('● changed')
+  })
+
+  it('highlights and blocks a selected multiple feature with no child option', async () => {
+    const input = new PassThrough()
+    let screen = ''
+    const output = new Writable({ write(chunk, _encoding, callback) { screen += String(chunk); callback() } })
+    const pending = clackNestedMultiselect({
+      message: 'Features',
+      input,
+      output,
+      options: [{
+        value: 'hooks',
+        label: 'Hooks',
+        default: true,
+        choiceMode: 'multiple',
+        choices: [
+          { value: 'claude', label: 'Claude', default: true },
+          { value: 'codex', label: 'Codex' },
+        ],
+      }],
+    })
+    setTimeout(() => input.write('\x1b[C \x1b[D \x1b[D\r'), 0)
+    await expect(pending).resolves.toEqual({
+      status: 'answered',
+      value: [{ value: 'hooks', choices: ['claude'] }],
+    })
+    expect(stripVTControlCharacters(screen)).toContain('▲ choose at least one')
+  })
+
+  it('blocks root submission for an exclusive feature with no selected option', async () => {
+    const input = new PassThrough()
+    let screen = ''
+    const output = new Writable({ write(chunk, _encoding, callback) { screen += String(chunk); callback() } })
+    const pending = clackNestedMultiselect({
+      message: 'Features', input, output,
+      options: [{
+        value: 'provider', label: 'Provider', default: true, choiceMode: 'exclusive',
+        choices: [{ value: 'one', label: 'One' }],
+      }],
+    })
+    setTimeout(() => input.write('\r\x1b[C\x1b[C\r\r'), 0)
+    await expect(pending).resolves.toEqual({
+      status: 'answered', value: [{ value: 'provider', choices: ['one'] }],
+    })
+    expect(stripVTControlCharacters(screen)).toContain('Choose one Provider option')
+  })
+
+  it('blocks root submission for a multiple feature with no selected option', async () => {
+    const input = new PassThrough()
+    const output = new Writable({ write(_chunk, _encoding, callback) { callback() } })
+    const pending = clackNestedMultiselect({
+      message: 'Features', input, output,
+      options: [{
+        value: 'hooks', label: 'Hooks', default: true, choiceMode: 'multiple',
+        choices: [{ value: 'one', label: 'One' }],
+      }],
+    })
+    setTimeout(() => input.write('\r\x1b[C \r\r'), 0)
+    await expect(pending).resolves.toEqual({
+      status: 'answered', value: [{ value: 'hooks', choices: ['one'] }],
+    })
+  })
+
+  it('renders an unchanged checked option while another child is focused', async () => {
+    const input = new PassThrough()
+    const output = new Writable({ write(_chunk, _encoding, callback) { callback() } })
+    const pending = clackNestedMultiselect({
+      message: 'Features', input, output,
+      options: [{
+        value: 'hooks', label: 'Hooks', default: true, choiceMode: 'multiple',
+        choices: [
+          { value: 'one', label: 'One', default: true },
+          { value: 'two', label: 'Two', default: true },
+        ],
+      }],
+    })
+    setTimeout(() => input.write('\x1b[C\x1b[B\x1b[D\r'), 0)
+    await expect(pending).resolves.toEqual({
+      status: 'answered', value: [{ value: 'hooks', choices: ['one', 'two'] }],
+    })
+  })
+
+  it('submits an empty optional selection', async () => {
+    const input = new PassThrough()
+    const output = new Writable({ write(_chunk, _encoding, callback) { callback() } })
+    const pending = clackNestedMultiselect({
+      message: 'Features', input, output, options: [{ value: 'one', label: 'One' }],
+    })
+    setTimeout(() => input.write('\r'), 0)
+    await expect(pending).resolves.toEqual({ status: 'answered', value: [] })
+  })
+})
+
+describe('feature configurator', () => {
+  const profile: ProjectProfile = {
+    name: 'demo',
+    description: 'demo',
+    runtime: { model: 'deepseek-v4-flash' },
+    runInterface: 'stdio',
+    packageManager: new NpmPackageManager('10.0.0'),
+    releaseVersion: '0.0.1',
+  }
+
+  it('shares exclusive, multiple, fixed, and secret behavior', async () => {
+    const registry = createBuiltinRegistry(profile)
+    const port = new QueuePromptPort(['sqlite', ['spawn', 'fork'], 'deepseek', 'new-key'])
+    const configurator = new FeatureConfigurator(port)
+    await expect(configurator.configure(registry.get(featureId('persistence')), profile)).resolves.toMatchObject({
+      options: ['sqlite'],
+    })
+    await expect(configurator.configure(registry.get(featureId('subagent')), profile)).resolves.toMatchObject({
+      options: ['spawn', 'fork'],
+    })
+    await expect(configurator.configure(
+      registry.get(featureId('provider')),
+      profile,
+      { id: featureId('provider'), options: ['deepseek'], secrets: { apiKey: 'old-key' } },
+    )).resolves.toMatchObject({ secrets: { apiKey: 'new-key' } })
+    expect(port.requests).toEqual([
+      'Choose durable session storage',
+      'Choose delegate work to child agents',
+      'Choose model provider',
+      'DeepSeek API key (leave empty to keep current)',
+    ])
+  })
+
+  it('validates feature values, defaults, and retained secrets', async () => {
+    const registry = createBuiltinRegistry(profile)
+    const fixed = new FeatureConfigurator(new QueuePromptPort([]))
+    await expect(fixed.configure(registry.get(featureId('bash')), profile, undefined, ['local'])).resolves.toMatchObject({
+      options: ['local'],
+    })
+    const requiredSecret = new FeatureConfigurator(new QueuePromptPort([]))
+    await expect(requiredSecret.configure(
+      registry.get(featureId('provider')), profile, undefined, ['deepseek'], { apiKey: '' },
+    )).rejects.toThrow('required')
+    const keep = new FeatureConfigurator(new QueuePromptPort(['deepseek', '']))
+    await expect(keep.configure(
+      registry.get(featureId('provider')),
+      profile,
+      { id: featureId('provider'), options: ['deepseek'], secrets: { apiKey: 'old' } },
+    )).resolves.toMatchObject({ secrets: { apiKey: 'old' } })
+    const custom = registry.get(featureId('provider'))
+    await expect(new FeatureConfigurator(new QueuePromptPort(['custom'])).configure(
+      custom,
+      profile,
+      { id: featureId('provider'), options: ['custom'], values: { baseURL: 1 }, secrets: { apiKey: 'old' } },
+    )).rejects.toThrow('current value must be a string')
+    await expect(new FeatureConfigurator(new QueuePromptPort(['custom', ''])).configure(
+      custom, profile, undefined,
+    )).rejects.toThrow('required')
+    await expect(new FeatureConfigurator(new QueuePromptPort(['custom', 'https://next', ''])).configure(
+      custom,
+      profile,
+      {
+        id: featureId('provider'), options: ['custom'],
+        values: { baseURL: 'https://old' }, secrets: { apiKey: 'old' },
+      },
+    )).resolves.toMatchObject({ values: { baseURL: 'https://next' }, secrets: { apiKey: 'old' } })
+    class EmptyExclusive extends ExclusiveOptionFeature {
+      override readonly id = featureId('empty-exclusive')
+      override readonly summary = 'Empty'
+      override readonly options = [new (class extends FeatureOption {
+        override readonly id = 'one'
+        override readonly label = 'One'
+        override contribution(): ProjectContribution { return new ProjectContribution([]) }
+      })()]
+      override defaultOptions(): readonly string[] { return [] }
+    }
+    await expect(new FeatureConfigurator(new QueuePromptPort([])).configure(new EmptyExclusive(), profile))
+      .rejects.toThrow('has no default option')
+  })
+})
diff --git a/packages/sdk/helper/tsconfig.json b/packages/sdk/helper/tsconfig.json
new file mode 100644
index 0000000000..e4a8604575
--- /dev/null
+++ b/packages/sdk/helper/tsconfig.json
@@ -0,0 +1,19 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": ["src"],
+  "references": [
+    { "path": "../../util/brand" },
+    { "path": "../../compact/compact-basic" },
+    { "path": "../../hooks/hooks-claude" },
+    { "path": "../../hooks/hooks-codex" },
+    { "path": "../../session-persistence/session-persistence-jsonl" },
+    { "path": "../../session-persistence/session-persistence-sqlite" },
+    { "path": "../../subagent/tool-subagent" },
+    { "path": "../../web/tool-web" },
+    { "path": "../../../vendor/cordis" }
+  ]
+}
diff --git a/packages/sdk/helper/tsdown.config.ts b/packages/sdk/helper/tsdown.config.ts
new file mode 100644
index 0000000000..cb8b9fafef
--- /dev/null
+++ b/packages/sdk/helper/tsdown.config.ts
@@ -0,0 +1,14 @@
+import { defineConfig } from 'tsdown'
+
+/** Bundle helper runtime and mirror template assets beside the bundle. */
+export default defineConfig({
+  entry: ['lib/types/index.js'],
+  outDir: 'lib',
+  format: ['esm'],
+  platform: 'node',
+  target: 'es2024',
+  fixedExtension: false,
+  dts: false,
+  clean: false,
+  copy: [{ from: 'src/templates/assets/*', to: 'lib/assets' }],
+})
diff --git a/packages/sdk/scripts/README.md b/packages/sdk/scripts/README.md
new file mode 100644
index 0000000000..e87c375a4f
--- /dev/null
+++ b/packages/sdk/scripts/README.md
@@ -0,0 +1,30 @@
+# `@deepseek-ai/dsh-scripts`
+
+The `dsh-sdk` launcher owns SDK project startup and configuration.
+
+| Command | Behavior |
+|---|---|
+| `dsh-sdk start [target] [-- args…]` | Import a module target and invoke `main(bootContext)`, or boot `cordis.yml` when omitted; arguments after `--` are forwarded |
+| `dsh-sdk dev [target] [-- args…]` | Register TypeScript and local-workspace source resolution, then use the start path |
+| `dsh-sdk build [args…]` | Invoke the project's installed tsdown with the project arguments |
+| `dsh-sdk config` | Open one interactive edit session, review accumulated changes, commit once, and install once when NPM dependencies changed |
+
+`ProjectBuild(tsdownConfig)` and `PluginBuild(tsdownConfig)` are exported only from `@deepseek-ai/dsh-scripts/dev/tsdown-config`. Development and production read the same `cordis.yml`.
+
+Generated project scripts invoke `dsh-sdk` for dev, build, start, and config; typecheck runs `tsc -b` directly. HMR remains an explicit `cordis.yml` feature loaded by both dev and start.
+
+The runtime library exports `startSDK(source)` to load `.env` and `cordis.yml` and return the live context, and `runSDK(target)` to import a project module and invoke its `main(bootContext)` (`runSDK()` without a target delegates to `startSDK('./cordis.yml')`). `SdkBootContext` carries the raw forwarded `argv`, generic `args`, the absolute launcher `cwd`, and the `start`/`dev` mode. The launcher declares no project options: Node `parseArgs()` runs with zero schema, so valued flags use `--key=value`, bare flags become booleans, `--no-cache` becomes `args.cache = false`, and option names retain Node's spelling (`--max-depth=3` → `args['max-depth']`).
+
+`start` never builds. `dev` registers the project-installed tsx transform plus an exact package-name map from `plugins/*/package.json` to each `src/index.ts`, then follows the same start path. `build` invokes the project-installed tsdown and forwards its arguments; an absent tsdown config is a successful no-op.
+
+`config` requires a TTY. One feature tree selects the desired enabled set; changed rows are highlighted, Right changes finite feature options, required rows cannot be deselected, inconsistent rows show diagnostics, and custom/manual Cordis config entries support enable/disable. The workflow reconciles that target into one edit session. Review & Apply commits once, then NPM dependency changes trigger one package-manager install. A failed install does not undo committed files.
+
+The root library exports `startSDK`, `runSDK`, and the `SdkBootArgs`/`SdkBootContext` types; command composition remains private to the bin. No `src/*`, bin, or package-manifest subpath is exported.
+
+## Model Experience
+
+Indirectly, through the project `cordis.yml` tree loaded by `start` or `dev`.
+
+## Known Limitations and Deferred Work
+
+- **Launcher arguments are schema-free** — `start` and `dev` preserve Node `parseArgs()` output rather than validating project-specific flags.
diff --git a/packages/sdk/scripts/package.json b/packages/sdk/scripts/package.json
new file mode 100644
index 0000000000..ba441a528c
--- /dev/null
+++ b/packages/sdk/scripts/package.json
@@ -0,0 +1,54 @@
+{
+  "name": "@deepseek-ai/dsh-scripts",
+  "description": "DeepSeek Harness SDK launcher for start, dev, build, and project configuration",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "bin": {
+    "dsh-sdk": "lib/bin.js"
+  },
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./dev/tsdown-config": {
+      "types": "./lib/types/dev/tsdown-config.d.ts",
+      "default": "./lib/dev/tsdown-config.js"
+    }
+  },
+  "files": [
+    "lib/index.js",
+    "lib/bin.js",
+    "lib/dev/tsdown-config.js",
+    "lib/local-plugin-loader-hooks.js",
+    "lib/assets",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "dependencies": {
+    "@deepseek-ai/dsh-helper": "workspace:^",
+    "commander": "^15.0.0",
+    "node-addon-require-builtin": "^0.1.0"
+  },
+  "peerDependencies": {
+    "@deepseek-ai/dsh-app-boot": "workspace:^",
+    "cordis": "^4.0.0-rc.7",
+    "tsdown": "^0.22.2",
+    "tsx": "^4.22.4"
+  },
+  "peerDependenciesMeta": {
+    "tsdown": { "optional": true },
+    "tsx": { "optional": true }
+  },
+  "devDependencies": {
+    "@deepseek-ai/dsh-app-boot": "workspace:^",
+    "cordis": "^4.0.0-rc.7",
+    "tsdown": "^0.22.2",
+    "tsx": "^4.22.4"
+  }
+}
diff --git a/packages/sdk/scripts/src/args.ts b/packages/sdk/scripts/src/args.ts
new file mode 100644
index 0000000000..1b91269592
--- /dev/null
+++ b/packages/sdk/scripts/src/args.ts
@@ -0,0 +1,70 @@
+/**
+ * Commander adapter for the dsh-sdk subcommand surface.
+ *
+ * @module @deepseek-ai/dsh-scripts/args
+ */
+
+import { parseArgs as parseNodeArgs } from 'node:util'
+import { Command } from 'commander'
+
+/** Commands implemented by the dsh-sdk launcher. */
+type DshSdkCommand = 'start' | 'dev' | 'build' | 'config'
+
+/** Parsed dsh-sdk invocation. */
+export interface DshSdkArgs {
+  command?: DshSdkCommand
+  target?: string
+  forwarded: readonly string[]
+  help: boolean
+}
+
+/** Parse arbitrary project flags through Node's zero-schema argument parser. */
+export function parseSdkBootArgs(argv: readonly string[]): Record<string, string | boolean | undefined> {
+  return parseNodeArgs({
+    args: [...argv],
+    strict: false,
+    allowPositionals: true,
+    allowNegative: true,
+  }).values
+}
+
+/** Parse one launcher invocation through real Commander subcommands. */
+export function parseDshSdkArgs(argv: readonly string[]): DshSdkArgs {
+  if (argv.length === 0 || argv[0] === '--help' || argv[0] === '-h') {
+    return { forwarded: [], help: true }
+  }
+  const separator = argv.indexOf('--')
+  const launcherArgv = separator === -1 ? argv : argv.slice(0, separator)
+  const passthrough = separator === -1 ? [] : argv.slice(separator + 1)
+  let parsed: DshSdkArgs | undefined
+  const program = new Command()
+    .name('dsh-sdk')
+    .helpOption(false)
+    .showHelpAfterError(false)
+    .exitOverride()
+    .configureOutput({
+      /* v8 ignore next -- the command wrapper renders the package-owned usage template */
+      writeOut: () => {},
+      /* v8 ignore next -- Commander errors are returned to the command wrapper */
+      writeErr: () => {},
+    })
+  program.command('start [target]').helpOption(false).action((target?: string) => {
+    parsed = { command: 'start', ...target ? { target } : {}, forwarded: [], help: false }
+  })
+  program.command('dev [target]').helpOption(false).action((target?: string) => {
+    parsed = { command: 'dev', ...target ? { target } : {}, forwarded: [], help: false }
+  })
+  program.command('build [args...]').helpOption(false).allowUnknownOption(true).action((args: string[] = []) => {
+    parsed = { command: 'build', forwarded: args, help: false }
+  })
+  program.command('config').helpOption(false).action(() => {
+    parsed = { command: 'config', forwarded: [], help: false }
+  })
+  program.parse([...launcherArgv], { from: 'user' })
+  /* v8 ignore next -- every registered Commander action above assigns parsed or Commander throws */
+  if (!parsed) throw new Error('dsh-sdk command did not resolve')
+  if (parsed.command === 'config' && passthrough.length > 0) {
+    throw new Error('dsh-sdk config does not accept forwarded arguments')
+  }
+  return { ...parsed, forwarded: [...parsed.forwarded, ...passthrough] }
+}
diff --git a/packages/sdk/scripts/src/bin.ts b/packages/sdk/scripts/src/bin.ts
new file mode 100644
index 0000000000..f125f0f4ac
--- /dev/null
+++ b/packages/sdk/scripts/src/bin.ts
@@ -0,0 +1,10 @@
+#!/usr/bin/env node
+/**
+ * Self-executing dsh-sdk launcher.
+ *
+ * @module @deepseek-ai/dsh-scripts/bin
+ */
+
+import { runDshSdkCommand } from './command.ts'
+
+process.exitCode = await runDshSdkCommand()
diff --git a/packages/sdk/scripts/src/build.ts b/packages/sdk/scripts/src/build.ts
new file mode 100644
index 0000000000..025ac4b2ab
--- /dev/null
+++ b/packages/sdk/scripts/src/build.ts
@@ -0,0 +1,94 @@
+/**
+ * User-owned tsdown configuration wrappers and child-process invocation.
+ *
+ * @module @deepseek-ai/dsh-scripts/build
+ */
+
+import { createRequire } from 'node:module'
+import { existsSync, readFileSync, readdirSync } from 'node:fs'
+import { dirname, resolve } from 'node:path'
+import type { UserConfig } from 'tsdown'
+import { NodeCommandRunner, type CommandRunner } from '@deepseek-ai/dsh-helper'
+
+function hasLocalPluginPackages(root: string): boolean {
+  const directory = resolve(root, 'plugins')
+  return existsSync(directory) && readdirSync(directory, { withFileTypes: true }).some(
+    item => item.isDirectory() && existsSync(resolve(directory, item.name, 'package.json')),
+  )
+}
+
+function hasTsdownConfig(root: string): boolean {
+  const hasConfigFile = [
+    'tsdown.config.ts', 'tsdown.config.mts', 'tsdown.config.cts',
+    'tsdown.config.js', 'tsdown.config.mjs', 'tsdown.config.cjs',
+    'tsdown.config.json',
+  ]
+    .some(name => existsSync(resolve(root, name)))
+  if (hasConfigFile) return true
+  let manifestText: string
+  try {
+    manifestText = readFileSync(resolve(root, 'package.json'), 'utf8')
+  } catch (error) {
+    if ((error as NodeJS.ErrnoException).code === 'ENOENT') return false
+    throw error
+  }
+  const manifest: unknown = JSON.parse(manifestText)
+  return manifest !== null && !Array.isArray(manifest) && typeof manifest === 'object'
+    && Object.hasOwn(manifest, 'tsdown')
+}
+
+/**
+ * Preserve the developer's root config and append a separate workspace pass
+ * when generated local plugin packages exist.
+ * @param tsdownConfig - developer-owned root tsdown config.
+ * @returns root tsdown config and optional local-plugin workspace pass.
+ */
+export function ProjectBuild(tsdownConfig: UserConfig): UserConfig[] {
+  if (tsdownConfig.workspace !== undefined) {
+    throw new Error('ProjectBuild owns workspace discovery; remove config.workspace')
+  }
+  const root = resolve(tsdownConfig.cwd ?? process.cwd())
+  return hasLocalPluginPackages(root)
+    ? [{ ...tsdownConfig }, { workspace: { include: ['plugins/*'] } }]
+    : [{ ...tsdownConfig }]
+}
+
+/**
+ * Preserve a local plugin package's developer-owned tsdown config.
+ * @param tsdownConfig - developer-owned plugin tsdown config.
+ * @returns validated tsdown config copy.
+ */
+export function PluginBuild(tsdownConfig: UserConfig): UserConfig {
+  if (tsdownConfig.workspace !== undefined) throw new Error('PluginBuild does not accept nested workspace config')
+  return { ...tsdownConfig }
+}
+
+function resolveTsdownBin(cwd: string): string {
+  const require = createRequire(resolve(cwd, 'package.json'))
+  let manifestPath: string
+  try {
+    manifestPath = require.resolve('tsdown/package.json')
+  } catch (error) {
+    throw new Error(`dsh-sdk build requires tsdown in this project: ${String(error)}`)
+  }
+  const manifest = JSON.parse(readFileSync(manifestPath, 'utf8')) as { bin?: unknown }
+  const bin = typeof manifest.bin === 'string'
+    ? manifest.bin
+    : manifest.bin && typeof manifest.bin === 'object'
+      ? (manifest.bin as Record<string, unknown>).tsdown
+      : undefined
+  if (typeof bin !== 'string') throw new Error('installed tsdown package has no executable')
+  return resolve(dirname(manifestPath), bin)
+}
+
+/** Invoke the project's installed tsdown, forwarding all build arguments. */
+export async function runProjectBuild(
+  args: readonly string[],
+  cwd: string = process.cwd(),
+  runner: CommandRunner = new NodeCommandRunner(),
+): Promise<void> {
+  if (!hasTsdownConfig(cwd)) return
+  const result = await runner.run(process.execPath, [resolveTsdownBin(cwd), ...args], resolve(cwd))
+  if (result.signal) throw new Error(`tsdown was killed by ${result.signal}`)
+  if (result.exitCode !== 0) throw new Error(`tsdown exited with code ${String(result.exitCode)}`)
+}
diff --git a/packages/sdk/scripts/src/command.ts b/packages/sdk/scripts/src/command.ts
new file mode 100644
index 0000000000..9351cfa39c
--- /dev/null
+++ b/packages/sdk/scripts/src/command.ts
@@ -0,0 +1,58 @@
+/**
+ * Internal dsh-sdk command composition used by the package bin.
+ *
+ * @module @deepseek-ai/dsh-scripts/command
+ */
+
+import { parseDshSdkArgs } from './args.ts'
+import { runProjectBuild } from './build.ts'
+import { runConfigCommand, type ConfigCommandContext } from './config.ts'
+import { runSDK } from './runtime.ts'
+import { DSH_SDK_TEMPLATES } from './templates/dsh-sdk-templates.ts'
+
+/** Injectable process and command boundaries used by the dsh-sdk bin. */
+export interface DshSdkCommandContext extends ConfigCommandContext {
+  cwd: string
+  stdin: NodeJS.ReadStream
+  stdout: NodeJS.WriteStream
+  stderr: NodeJS.WriteStream
+  run?: typeof runSDK
+  build?: typeof runProjectBuild
+  config?: typeof runConfigCommand
+}
+
+/** Run one parsed dsh-sdk command and return its process exit code. */
+export async function runDshSdkCommand(
+  argv: readonly string[] = process.argv.slice(2),
+  context: DshSdkCommandContext = {
+    cwd: process.cwd(),
+    stdin: process.stdin,
+    stdout: process.stdout,
+    stderr: process.stderr,
+  },
+): Promise<number> {
+  try {
+    const args = parseDshSdkArgs(argv)
+    if (args.help || !args.command) {
+      context.stdout.write(DSH_SDK_TEMPLATES.usage.render({}))
+      return 0
+    }
+    const run = context.run ?? runSDK
+    const build = context.build ?? runProjectBuild
+    const config = context.config ?? runConfigCommand
+    switch (args.command) {
+      case 'start': await run(args.target, { cwd: context.cwd, argv: args.forwarded }); break
+      case 'dev': await run(args.target, { cwd: context.cwd, dev: true, argv: args.forwarded }); break
+      case 'build': await build(args.forwarded, context.cwd); break
+      case 'config': {
+        const result = await config(context)
+        if (result.installError) return 1
+        break
+      }
+    }
+    return 0
+  } catch (error) {
+    context.stderr.write(`dsh-sdk: ${error instanceof Error ? error.message : String(error)}\n`)
+    return 1
+  }
+}
diff --git a/packages/sdk/scripts/src/config.ts b/packages/sdk/scripts/src/config.ts
new file mode 100644
index 0000000000..88d8cc035a
--- /dev/null
+++ b/packages/sdk/scripts/src/config.ts
@@ -0,0 +1,37 @@
+/**
+ * dsh-sdk config command composition.
+ *
+ * @module @deepseek-ai/dsh-scripts/config
+ */
+
+import {
+  ClackPromptPort,
+  SdkProject,
+  createBuiltinRegistry,
+  type PromptPort,
+} from '@deepseek-ai/dsh-helper'
+import { ConfigWorkflow, type ConfigWorkflowResult } from './config/config-workflow.ts'
+
+/** Process stream slice required by dsh-sdk config. */
+export interface ConfigCommandContext {
+  cwd: string
+  stdin: NodeJS.ReadStream
+  stdout: NodeJS.WriteStream
+  port?: PromptPort
+  install?: (project: SdkProject) => Promise<void>
+}
+
+/** Open and interactively edit one existing SDK project. */
+export async function runConfigCommand(context: ConfigCommandContext): Promise<ConfigWorkflowResult> {
+  if (!context.port && (!context.stdin.isTTY || !context.stdout.isTTY)) {
+    throw new Error('dsh-sdk config requires an interactive TTY')
+  }
+  const project = await SdkProject.open(context.cwd)
+  const registry = createBuiltinRegistry(project.profile)
+  return new ConfigWorkflow(
+    /* v8 ignore next -- production TTY wiring is exercised by the built-bin smoke */
+    context.port ?? new ClackPromptPort(context.stdin, context.stdout),
+    context.stdout,
+    context.install,
+  ).run(project, registry)
+}
diff --git a/packages/sdk/scripts/src/config/config-workflow.ts b/packages/sdk/scripts/src/config/config-workflow.ts
new file mode 100644
index 0000000000..7d3d7ee43a
--- /dev/null
+++ b/packages/sdk/scripts/src/config/config-workflow.ts
@@ -0,0 +1,216 @@
+/**
+ * Tree-shaped existing-project feature workflow and single Apply boundary.
+ *
+ * @module @deepseek-ai/dsh-scripts/config/config-workflow
+ */
+
+import type { Writable } from 'node:stream'
+import {
+  FeatureConfigurator,
+  ConfirmQuestion,
+  requireAnswer,
+  type Feature,
+  type FeatureInstallation,
+  type FeatureRegistry,
+  type FeatureSelection,
+  type ChangeSet,
+  type NestedMultiSelectValue,
+  type ProjectCommitResult,
+  type PromptPort,
+  type RunInterface,
+  type SdkProject,
+} from '@deepseek-ai/dsh-helper'
+import { DSH_SDK_TEMPLATES } from '../templates/dsh-sdk-templates.ts'
+
+/** Config result, including an install failure that happened after commit. */
+export interface ConfigWorkflowResult {
+  commit?: ProjectCommitResult<SdkProject>
+  installError?: Error
+}
+
+function featureTarget(feature: Feature): string {
+  return `feature:${feature.id}`
+}
+
+function pluginTarget(id: string): string {
+  return `plugin:${id}`
+}
+
+function sameOptions(left: readonly string[], right: readonly string[]): boolean {
+  return [...left].sort().join('\0') === [...right].sort().join('\0')
+}
+
+function targetRunInterface(
+  current: RunInterface,
+  desired: ReadonlyMap<string, NestedMultiSelectValue<string, string>>,
+): RunInterface {
+  const selected = desired.get('feature:app')?.choices[0]
+  return selected === 'acp' || selected === 'stdio' || selected === 'embed' ? selected : current
+}
+
+/** Reconcile one tree selection into domain commands, then review and commit once. */
+export class ConfigWorkflow {
+  private readonly port: PromptPort
+  private readonly output: Writable
+  private readonly install: (project: SdkProject) => Promise<void>
+
+  /** Bind terminal prompts and descriptive output. */
+  constructor(
+    port: PromptPort,
+    output: Writable = process.stdout,
+    install: (project: SdkProject) => Promise<void> = project => project.profile.packageManager.install(project.root),
+  ) {
+    this.port = port
+    this.output = output
+    this.install = install
+  }
+
+  /** Select desired state, reconcile the working copy, review, and apply. */
+  async run(project: SdkProject, registry: FeatureRegistry): Promise<ConfigWorkflowResult> {
+    const edit = project.edit(registry)
+    const configurator = new FeatureConfigurator(this.port)
+    const features = registry.all().filter(feature => feature.isApplicable(project.profile))
+    const inspections = new Map(edit.inspections().map(item => [item.id, item]))
+    const custom = edit.cordisConfigEntries().filter(entry => !registry.ownerOfPackage(entry.name, project.profile))
+    const desired = requireAnswer(await this.port.nestedMultiselect<string, string>({
+      message: 'Configure the project',
+      showChanges: true,
+      options: [
+        ...features.map((feature) => {
+          const installation = inspections.get(feature.id)
+          /* v8 ignore next -- inspections() is built from this exact feature registry */
+          if (!installation) throw new Error(`feature inspection is missing: ${feature.id}`)
+          const inconsistent = installation.state === 'inconsistent'
+          const selectedOptions = new Set(installation.options.length > 0
+            ? installation.options
+            : feature.defaultOptions(project.profile))
+          return {
+            value: featureTarget(feature),
+            label: feature.summary,
+            required: feature.required,
+            default: feature.required || installation.state === 'enabled' || inconsistent,
+            disabled: inconsistent,
+            ...inconsistent ? { warning: installation.diagnostics.join('; ') } : {},
+            ...feature.mode === 'single' ? {} : {
+              choiceMode: feature.mode,
+              choices: feature.options.map(option => ({
+                value: option.id,
+                label: option.label,
+                default: selectedOptions.has(option.id),
+              })),
+            },
+          }
+        }),
+        ...custom.map(entry => ({
+          value: pluginTarget(entry.id),
+          label: `${entry.name} [custom]`,
+          default: !entry.disabled,
+        })),
+      ],
+    }))
+    const desiredByTarget = new Map(desired.map(item => [item.value, item]))
+    const targetProfile = {
+      ...project.profile,
+      runInterface: targetRunInterface(project.profile.runInterface, desiredByTarget),
+    }
+    for (const feature of features) {
+      if (!feature.isApplicable(targetProfile)) desiredByTarget.delete(featureTarget(feature))
+    }
+
+    for (const feature of features) {
+      const installation = inspections.get(feature.id)
+      /* v8 ignore next -- inspections() is built from this exact feature registry */
+      if (!installation) throw new Error(`feature inspection is missing: ${feature.id}`)
+      if (installation.state === 'inconsistent') continue
+      const choice = desiredByTarget.get(featureTarget(feature))
+      if (!choice && !feature.required) continue
+      await this.enableOrConfigure(feature, installation, choice, project, edit, configurator)
+    }
+
+    for (const feature of [...features].reverse()) {
+      const installation = inspections.get(feature.id)
+      /* v8 ignore next -- inspections() is built from this exact feature registry */
+      if (!installation) throw new Error(`feature inspection is missing: ${feature.id}`)
+      if (feature.required || installation.state !== 'enabled'
+        || desiredByTarget.has(featureTarget(feature))) continue
+      edit.disableFeature(feature)
+    }
+
+    for (const entry of custom) {
+      const enabled = desiredByTarget.has(pluginTarget(entry.id))
+      if (enabled === !entry.disabled) continue
+      edit.setCustomPluginDisabled(entry.id, !enabled)
+    }
+
+    const changes = edit.changes()
+    if (changes.changedFiles.length === 0) {
+      this.output.write('No changes.\n')
+      return {}
+    }
+    this.renderReview(changes)
+    const apply = requireAnswer(await new ConfirmQuestion({
+      id: 'config.apply', message: 'Apply these changes?', initialValue: true,
+    }).resolve(this.port))
+    if (!apply) return {}
+    const commit = await edit.commit()
+    if (!commit.changes.npmDependenciesChanged) return { commit }
+    try {
+      await this.install(project)
+      return { commit }
+    } catch (error) {
+      const installError = error instanceof Error ? error : new Error(String(error))
+      const manager = project.profile.packageManager
+      this.output.write(DSH_SDK_TEMPLATES.configInstallFailure.render({
+        error: installError.message,
+        packageManager: manager.name,
+        installArgs: manager.installCommand().join(' '),
+      }))
+      return { commit, installError }
+    }
+  }
+
+  private async enableOrConfigure(
+    feature: Feature,
+    installation: FeatureInstallation,
+    choice: NestedMultiSelectValue<string, string> | undefined,
+    project: SdkProject,
+    edit: ReturnType<SdkProject['edit']>,
+    configurator: FeatureConfigurator,
+  ): Promise<void> {
+    const options = choice?.choices.length
+      ? choice.choices
+      : installation.options.length > 0
+        ? installation.options
+        : feature.defaultOptions(project.profile)
+    if (installation.state === 'absent') {
+      const selection = await configurator.configure(feature, project.profile, undefined, options)
+      edit.installFeature(feature, selection)
+      return
+    }
+    /* v8 ignore next -- non-absent/non-inconsistent inspections always carry their normalized selection */
+    if (!installation.selection) throw new Error(`feature ${feature.id} has no readable selection`)
+    if (!sameOptions(installation.options, options)) {
+      const selection: FeatureSelection = await configurator.configure(
+        feature,
+        project.profile,
+        installation.selection,
+        options,
+      )
+      edit.configureFeature(feature, selection)
+    }
+    if (installation.state === 'disabled') edit.enableFeature(feature)
+  }
+
+  private renderReview(changes: ChangeSet): void {
+    const lines = [
+      ...changes.addedFeatures.map(id => `Install feature: ${id}`),
+      ...changes.enabledFeatures.map(id => `Enable feature: ${id}`),
+      ...changes.disabledFeatures.map(id => `Disable feature: ${id}`),
+      ...changes.configuredFeatures.map(id => `Configure feature: ${id}`),
+      ...changes.enabledPlugins.map(id => `Enable custom plugin: ${id}`),
+      ...changes.disabledPlugins.map(id => `Disable custom plugin: ${id}`),
+      ...changes.changedFiles.map(path => `Change file: ${path}`),
+    ]
+    this.output.write(`${lines.join('\n')}\n`)
+  }
+}
diff --git a/packages/sdk/scripts/src/dev/tsdown-config.ts b/packages/sdk/scripts/src/dev/tsdown-config.ts
new file mode 100644
index 0000000000..15e5603cfa
--- /dev/null
+++ b/packages/sdk/scripts/src/dev/tsdown-config.ts
@@ -0,0 +1,7 @@
+/**
+ * Generated-project tsdown config wrappers.
+ *
+ * @module @deepseek-ai/dsh-scripts/dev/tsdown-config
+ */
+
+export { PluginBuild, ProjectBuild } from '../build.ts'
diff --git a/packages/sdk/scripts/src/index.ts b/packages/sdk/scripts/src/index.ts
new file mode 100644
index 0000000000..5180f885c8
--- /dev/null
+++ b/packages/sdk/scripts/src/index.ts
@@ -0,0 +1,7 @@
+/**
+ * Public DeepSeek Harness SDK runtime entry points.
+ *
+ * @module @deepseek-ai/dsh-scripts
+ */
+
+export { runSDK, startSDK, type SdkBootContext } from './runtime.ts'
diff --git a/packages/sdk/scripts/src/local-plugin-loader-hooks.ts b/packages/sdk/scripts/src/local-plugin-loader-hooks.ts
new file mode 100644
index 0000000000..b57a76da08
--- /dev/null
+++ b/packages/sdk/scripts/src/local-plugin-loader-hooks.ts
@@ -0,0 +1,27 @@
+/**
+ * Node module customization hook for project-local plugin package names.
+ *
+ * @module @deepseek-ai/dsh-scripts/local-plugin-loader-hooks
+ */
+
+import type { ResolveHookContext, ResolveFnOutput } from 'node:module'
+
+interface HookData {
+  mappings: Readonly<Record<string, string>>
+}
+
+let mappings: Readonly<Record<string, string>> = {}
+
+/** Receive the package-name to source-URL map from the launcher thread. */
+export function initialize(data: HookData): void {
+  mappings = { ...data.mappings }
+}
+
+/** Resolve exact local workspace package names to their TypeScript entry source. */
+export async function resolve(
+  specifier: string,
+  context: ResolveHookContext,
+  nextResolve: (specifier: string, context: ResolveHookContext) => Promise<ResolveFnOutput>,
+): Promise<ResolveFnOutput> {
+  return nextResolve(mappings[specifier] ?? specifier, context)
+}
diff --git a/packages/sdk/scripts/src/runtime.ts b/packages/sdk/scripts/src/runtime.ts
new file mode 100644
index 0000000000..426ac90e5b
--- /dev/null
+++ b/packages/sdk/scripts/src/runtime.ts
@@ -0,0 +1,137 @@
+/**
+ * Shared start/dev runtime and project-local module resolution.
+ *
+ * @module @deepseek-ai/dsh-scripts/runtime
+ */
+
+import { register as registerHook } from 'node:module'
+import { access, readFile, readdir } from 'node:fs/promises'
+import { dirname, resolve } from 'node:path'
+import { fileURLToPath, pathToFileURL } from 'node:url'
+import type { Context } from 'cordis'
+import { boot, installFailLoud, loadEnv, resolveConfigPath } from '@deepseek-ai/dsh-app-boot'
+import { parseSdkBootArgs } from './args.ts'
+
+/** Options that distinguish dev boot from production boot. */
+interface BootProjectOptions {
+  cwd?: string
+  dev?: boolean
+  argv?: readonly string[]
+}
+
+/** Startup context passed to a generated project's exported `main()`. */
+export interface SdkBootContext {
+  /** Developer arguments forwarded after the launcher's `--` separator. */
+  readonly argv: readonly string[]
+  /** SDK-recognized structured arguments parsed from {@link argv}. */
+  readonly args: Record<string, string | boolean | undefined>
+  /** Absolute project working directory selected by the launcher. */
+  readonly cwd: string
+  /** Whether the launcher is running the built or TypeScript development entry. */
+  readonly mode: 'start' | 'dev'
+}
+
+async function localPluginMappings(cwd: string): Promise<Record<string, string>> {
+  const mappings: Record<string, string> = {}
+  let directories
+  try {
+    directories = await readdir(resolve(cwd, 'plugins'), { withFileTypes: true })
+  } catch (error) {
+    /* v8 ignore else -- the other arm requires a filesystem permission/IO fault from readdir */
+    if ((error as NodeJS.ErrnoException).code === 'ENOENT') return mappings
+    /* v8 ignore next -- paired with the ignored defensive readdir-error arm above */
+    throw error
+  }
+  for (const directory of directories) {
+    if (!directory.isDirectory()) continue
+    const root = resolve(cwd, 'plugins', directory.name)
+    let manifest: { name?: unknown }
+    try {
+      manifest = JSON.parse(await readFile(resolve(root, 'package.json'), 'utf8')) as { name?: unknown }
+      await access(resolve(root, 'src/index.ts'))
+    } catch (error) {
+      throw new Error(`cannot load local plugin metadata from ${root}: ${String(error)}`)
+    }
+    if (typeof manifest.name !== 'string' || manifest.name.length === 0) {
+      throw new Error(`local plugin package has no name: ${root}`)
+    }
+    if (mappings[manifest.name]) throw new Error(`duplicate local plugin package name: ${manifest.name}`)
+    mappings[manifest.name] = pathToFileURL(resolve(root, 'src/index.ts')).href
+  }
+  return mappings
+}
+
+/** Register tsx and exact local-plugin source mappings for the current process. */
+async function registerDevRuntime(cwd: string = process.cwd()): Promise<void> {
+  let registerTsx: typeof import('tsx/esm/api')['register']
+  try {
+    ({ register: registerTsx } = await import('tsx/esm/api'))
+  } catch (error) {
+    /* v8 ignore next -- tsx is a declared project NPM dependency; missing-package behavior is defensive */
+    throw new Error(`dsh-sdk dev requires the project's tsx NPM dependency: ${String(error)}`)
+  }
+  registerTsx()
+  const mappings = await localPluginMappings(resolve(cwd))
+  const hook = new URL(
+    /* v8 ignore next -- the .js arm is exercised by the built-bin smoke rather than source coverage */
+    import.meta.url.endsWith('.ts')
+      ? './local-plugin-loader-hooks.ts'
+      : './local-plugin-loader-hooks.js', import.meta.url)
+  registerHook(hook, { data: { mappings } })
+}
+
+/**
+ * Boot one cordis.yml after loading its sibling .env.
+ * @param source - file path or file URL to cordis.yml.
+ * @param options - working directory and development-runtime options.
+ * @returns live Cordis context.
+ */
+export async function startSDK(
+  source: string | URL = './cordis.yml',
+  options: BootProjectOptions = {},
+): Promise<Context> {
+  const cwd = resolve(options.cwd ?? process.cwd())
+  if (options.dev) await registerDevRuntime(cwd)
+  if (source instanceof URL && source.protocol !== 'file:') {
+    throw new Error(`cordis.yml URL must use file:, got ${source.protocol}`)
+  }
+  const requested = source instanceof URL ? fileURLToPath(source) : source
+  const absolute = resolveConfigPath(requested, undefined, cwd)
+  loadEnv('dsh-sdk', dirname(absolute))
+  installFailLoud('dsh-sdk')
+  return boot('dsh-sdk', absolute)
+}
+
+/**
+ * Import and invoke a module target's main(), or directly boot cordis.yml.
+ * @param target - module path relative to the project, or absent for cordis.yml.
+ * @param options - working directory and development-runtime options.
+ * @returns target main result or live Cordis context.
+ */
+export async function runSDK(
+  target?: string,
+  options: BootProjectOptions = {},
+): Promise<unknown> {
+  /* v8 ignore next -- the bin always supplies cwd; direct consumers normally accept process.cwd() */
+  const cwd = resolve(options.cwd ?? process.cwd())
+  if (options.dev) await registerDevRuntime(cwd)
+  if (!target) return startSDK('./cordis.yml', { cwd })
+  const absolute = resolve(cwd, target)
+  try {
+    await access(absolute)
+  } catch (error) {
+    const hint = options.dev ? '' : ' Run dsh-sdk build first if this is a TypeScript project.'
+    throw new Error(`cannot start missing target ${target}.${hint} ${String(error)}`)
+  }
+  const module = await import(pathToFileURL(absolute).href) as { main?: (context: SdkBootContext) => unknown }
+  if (typeof module.main !== 'function') {
+    throw new Error(`dsh-sdk target ${target} must export function main()`)
+  }
+  const argv = [...options.argv ?? []]
+  return module.main({
+    argv,
+    args: parseSdkBootArgs(argv),
+    cwd,
+    mode: options.dev ? 'dev' : 'start',
+  })
+}
diff --git a/packages/sdk/scripts/src/templates/assets/config-install-failure.txt.tpl b/packages/sdk/scripts/src/templates/assets/config-install-failure.txt.tpl
new file mode 100644
index 0000000000..ee1da80f76
--- /dev/null
+++ b/packages/sdk/scripts/src/templates/assets/config-install-failure.txt.tpl
@@ -0,0 +1,2 @@
+Changes were committed, but install failed: {{error}}
+Retry: {{packageManager}} {{installArgs}}
diff --git a/packages/sdk/scripts/src/templates/assets/usage.txt.tpl b/packages/sdk/scripts/src/templates/assets/usage.txt.tpl
new file mode 100644
index 0000000000..d4122198f5
--- /dev/null
+++ b/packages/sdk/scripts/src/templates/assets/usage.txt.tpl
@@ -0,0 +1,7 @@
+Usage: dsh-sdk <command> [options]
+
+Commands:
+  start [target] [-- args...]  Import a built module, or boot cordis.yml
+  dev [target] [-- args...]    Start with TypeScript and local-plugin source resolution
+  build [args...] Run the project's installed tsdown
+  config          Interactively edit project features
diff --git a/packages/sdk/scripts/src/templates/dsh-sdk-templates.ts b/packages/sdk/scripts/src/templates/dsh-sdk-templates.ts
new file mode 100644
index 0000000000..a7dabc3b85
--- /dev/null
+++ b/packages/sdk/scripts/src/templates/dsh-sdk-templates.ts
@@ -0,0 +1,21 @@
+/**
+ * Package-owned terminal templates for the dsh-sdk launcher.
+ *
+ * @module @deepseek-ai/dsh-scripts/templates/dsh-sdk-templates
+ */
+
+import { TextTemplate, type PackageManagerName } from '@deepseek-ai/dsh-helper'
+
+interface ConfigInstallFailureTemplateModel {
+  error: string
+  packageManager: PackageManagerName
+  installArgs: string
+}
+
+/** Compiled dsh-sdk terminal templates. */
+export const DSH_SDK_TEMPLATES = {
+  usage: TextTemplate.fromFile<Record<string, never>>(new URL('./assets/usage.txt.tpl', import.meta.url)),
+  configInstallFailure: TextTemplate.fromFile<ConfigInstallFailureTemplateModel>(
+    new URL('./assets/config-install-failure.txt.tpl', import.meta.url),
+  ),
+} as const
diff --git a/packages/sdk/scripts/tests/__snapshots__/config.snapshot.ts.snap b/packages/sdk/scripts/tests/__snapshots__/config.snapshot.ts.snap
new file mode 100644
index 0000000000..31e21dd8a9
--- /dev/null
+++ b/packages/sdk/scripts/tests/__snapshots__/config.snapshot.ts.snap
@@ -0,0 +1,303 @@
+// Vitest Snapshot v1, https://vitest.dev/guide/snapshot.html
+
+exports[`dsh-sdk config terminal contract > pins the feature tree and Review & Apply output 1`] = `
+{
+  "committed": {
+    "addedFeatures": [
+      "todo",
+    ],
+    "addedPlugins": [],
+    "changedFiles": [
+      "cordis.yml",
+      "package.json",
+    ],
+    "configuredFeatures": [],
+    "disabledFeatures": [],
+    "disabledPlugins": [],
+    "enabledFeatures": [],
+    "enabledPlugins": [],
+    "npmDependenciesChanged": true,
+  },
+  "installs": 1,
+  "review": "Install feature: todo
+Change file: cordis.yml
+Change file: package.json
+",
+  "transcript": [
+    {
+      "kind": "nested-multiselect",
+      "message": "Configure the project",
+      "options": [
+        {
+          "choiceMode": "exclusive",
+          "choices": [
+            {
+              "default": true,
+              "label": "DeepSeek",
+              "value": "deepseek",
+            },
+            {
+              "default": false,
+              "label": "Custom endpoint (pi-ai)",
+              "value": "custom",
+            },
+          ],
+          "default": true,
+          "disabled": false,
+          "label": "Model provider",
+          "required": true,
+          "value": "feature:provider",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": undefined,
+          "choices": undefined,
+          "default": true,
+          "disabled": false,
+          "label": "Agent runtime spine",
+          "required": true,
+          "value": "feature:spine",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": "exclusive",
+          "choices": [
+            {
+              "default": true,
+              "label": "Local executor",
+              "value": "local",
+            },
+            {
+              "default": false,
+              "label": "Sandboxed executor",
+              "value": "sandbox",
+            },
+          ],
+          "default": true,
+          "disabled": false,
+          "label": "Command execution",
+          "required": true,
+          "value": "feature:bash",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": "exclusive",
+          "choices": [
+            {
+              "default": false,
+              "label": "ACP server",
+              "value": "acp",
+            },
+            {
+              "default": true,
+              "label": "Terminal REPL",
+              "value": "stdio",
+            },
+            {
+              "default": false,
+              "label": "Embedded context",
+              "value": "embed",
+            },
+          ],
+          "default": true,
+          "disabled": false,
+          "label": "Run interface",
+          "required": true,
+          "value": "feature:app",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": "exclusive",
+          "choices": [
+            {
+              "default": true,
+              "label": "JSONL files",
+              "value": "jsonl",
+            },
+            {
+              "default": false,
+              "label": "SQLite database",
+              "value": "sqlite",
+            },
+          ],
+          "default": true,
+          "disabled": false,
+          "label": "Durable session storage",
+          "required": true,
+          "value": "feature:persistence",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": undefined,
+          "choices": undefined,
+          "default": false,
+          "disabled": false,
+          "label": "Hot-module reload",
+          "required": false,
+          "value": "feature:hmr",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": undefined,
+          "choices": undefined,
+          "default": false,
+          "disabled": false,
+          "label": "Read, write, and edit local files",
+          "required": false,
+          "value": "feature:fs",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": undefined,
+          "choices": undefined,
+          "default": false,
+          "disabled": false,
+          "label": "Model-facing task tracking",
+          "required": false,
+          "value": "feature:todo",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": undefined,
+          "choices": undefined,
+          "default": false,
+          "disabled": false,
+          "label": "Local skill discovery",
+          "required": false,
+          "value": "feature:skill",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": "exclusive",
+          "choices": [
+            {
+              "default": true,
+              "label": "DeepSeek search",
+              "value": "deepseek",
+            },
+            {
+              "default": false,
+              "label": "Exa search",
+              "value": "exa",
+            },
+            {
+              "default": false,
+              "label": "Perplexity search",
+              "value": "perplexity",
+            },
+            {
+              "default": false,
+              "label": "Fetch only",
+              "value": "fetch-only",
+            },
+          ],
+          "default": false,
+          "disabled": false,
+          "label": "Web search and fetch tools",
+          "required": false,
+          "value": "feature:web",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": "multiple",
+          "choices": [
+            {
+              "default": true,
+              "label": "Fresh child agent",
+              "value": "spawn",
+            },
+            {
+              "default": false,
+              "label": "Fork parent history",
+              "value": "fork",
+            },
+          ],
+          "default": false,
+          "disabled": false,
+          "label": "Delegate work to child agents",
+          "required": false,
+          "value": "feature:subagent",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": undefined,
+          "choices": undefined,
+          "default": false,
+          "disabled": false,
+          "label": "Scripted multi-agent workflows",
+          "required": false,
+          "value": "feature:workflow",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": undefined,
+          "choices": undefined,
+          "default": false,
+          "disabled": false,
+          "label": "Automatic context compaction",
+          "required": false,
+          "value": "feature:compact",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": "multiple",
+          "choices": [
+            {
+              "default": true,
+              "label": "Claude Code hooks",
+              "value": "claude",
+            },
+            {
+              "default": false,
+              "label": "Codex hooks",
+              "value": "codex",
+            },
+          ],
+          "default": false,
+          "disabled": false,
+          "label": "Run Claude Code or Codex hooks",
+          "required": false,
+          "value": "feature:hooks",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": undefined,
+          "choices": undefined,
+          "default": false,
+          "disabled": false,
+          "label": "Loop-hygiene reminders",
+          "required": false,
+          "value": "feature:guard",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": undefined,
+          "choices": undefined,
+          "default": false,
+          "disabled": false,
+          "label": "Tool timeout policy",
+          "required": false,
+          "value": "feature:timeout-policy",
+          "warning": undefined,
+        },
+        {
+          "choiceMode": undefined,
+          "choices": undefined,
+          "default": false,
+          "disabled": false,
+          "label": "Ask the user from the model loop",
+          "required": false,
+          "value": "feature:ask-user",
+          "warning": undefined,
+        },
+      ],
+      "showChanges": true,
+    },
+    {
+      "initialValue": true,
+      "kind": "confirm",
+      "message": "Apply these changes?",
+    },
+  ],
+}
+`;
diff --git a/packages/sdk/scripts/tests/config.snapshot.ts b/packages/sdk/scripts/tests/config.snapshot.ts
new file mode 100644
index 0000000000..80a79a5f20
--- /dev/null
+++ b/packages/sdk/scripts/tests/config.snapshot.ts
@@ -0,0 +1,128 @@
+import { mkdtemp, rm } from 'node:fs/promises'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { Writable } from 'node:stream'
+import { afterEach, describe, expect, it } from 'vitest'
+import {
+  NpmPackageManager,
+  SdkProject,
+  featureId,
+  createBuiltinRegistry,
+  type NestedMultiSelectValue,
+  type PromptPort,
+} from '@deepseek-ai/dsh-helper'
+import type {
+  ConfirmPromptRequest,
+  MultiSelectPromptRequest,
+  NestedMultiSelectRequest,
+  PromptOutcome,
+  SecretPromptRequest,
+  SelectPromptRequest,
+  TextPromptRequest,
+} from '../../helper/src/questions/prompt-port.ts'
+import { ConfigWorkflow } from '../src/config/config-workflow.ts'
+
+class RecordingPort implements PromptPort {
+  readonly transcript: unknown[] = []
+  readonly #answers: unknown[]
+
+  constructor(answers: unknown[]) { this.#answers = [...answers] }
+
+  answer<T>(record: unknown): Promise<PromptOutcome<T>> {
+    this.transcript.push(record)
+    return Promise.resolve({ status: 'answered', value: this.#answers.shift() as T })
+  }
+
+  text(request: TextPromptRequest): Promise<PromptOutcome<string>> {
+    return this.answer({ kind: 'text', message: request.message })
+  }
+  secret(request: SecretPromptRequest): Promise<PromptOutcome<string>> {
+    return this.answer({ kind: 'secret', message: request.message })
+  }
+  select<T>(request: SelectPromptRequest<T>): Promise<PromptOutcome<T>> {
+    return this.answer({
+      kind: 'select', message: request.message,
+      options: request.options.map(option => ({ value: option.value, label: option.label })),
+    })
+  }
+  multiselect<T>(request: MultiSelectPromptRequest<T>): Promise<PromptOutcome<readonly T[]>> {
+    return this.answer({ kind: 'multiselect', message: request.message })
+  }
+  confirm(request: ConfirmPromptRequest): Promise<PromptOutcome<boolean>> {
+    return this.answer({ kind: 'confirm', message: request.message, initialValue: request.initialValue })
+  }
+  nestedMultiselect<TValue, TChoice>(
+    request: NestedMultiSelectRequest<TValue, TChoice>,
+  ): Promise<PromptOutcome<readonly NestedMultiSelectValue<TValue, TChoice>[]>> {
+    return this.answer({
+      kind: 'nested-multiselect',
+      message: request.message,
+      showChanges: request.showChanges,
+      options: request.options.map(option => ({
+        value: option.value,
+        label: option.label,
+        required: option.required,
+        default: option.default,
+        disabled: option.disabled,
+        warning: option.warning,
+        choiceMode: option.choiceMode,
+        choices: option.choices?.map(choice => ({
+          value: choice.value,
+          label: choice.label,
+          default: choice.default,
+        })),
+      })),
+    })
+  }
+}
+
+const temporary: string[] = []
+
+afterEach(async () => {
+  await Promise.all(temporary.splice(0).map(path => rm(path, { recursive: true, force: true })))
+})
+
+async function baseProject(): Promise<SdkProject> {
+  const root = await mkdtemp(join(tmpdir(), 'dsh-config-snapshot-'))
+  temporary.push(root)
+  const request = {
+    name: 'snapshot-agent',
+    description: 'snapshot',
+    runtime: { model: 'deepseek-v4-flash' },
+    packageManager: new NpmPackageManager('10.0.0'),
+    releaseVersion: '0.0.1',
+    features: [
+      { id: featureId('provider'), options: ['deepseek'], secrets: { apiKey: 'key' } },
+      { id: featureId('bash'), options: ['local'] },
+      { id: featureId('app'), options: ['stdio'] },
+      { id: featureId('persistence'), options: ['jsonl'] },
+    ],
+    localPlugins: [],
+  }
+  const project = SdkProject.create(root, request)
+  const registry = createBuiltinRegistry(project.profile)
+  const edit = project.edit(registry)
+  for (const item of request.features) edit.installFeature(registry.get(item.id), item)
+  return (await edit.commit()).project
+}
+
+describe('dsh-sdk config terminal contract', () => {
+  it('pins the feature tree and Review & Apply output', async () => {
+    const project = await baseProject()
+    const registry = createBuiltinRegistry(project.profile)
+    const port = new RecordingPort([
+      [{ value: 'feature:todo', choices: [] }],
+      true,
+    ])
+    let output = ''
+    const stream = new Writable({ write(chunk, _encoding, callback) { output += String(chunk); callback() } })
+    let installs = 0
+    const result = await new ConfigWorkflow(port, stream, async () => { installs += 1 }).run(project, registry)
+    expect({
+      transcript: port.transcript,
+      review: output,
+      installs,
+      committed: result.commit?.changes,
+    }).toMatchSnapshot()
+  })
+})
diff --git a/packages/sdk/scripts/tests/scripts.spec.ts b/packages/sdk/scripts/tests/scripts.spec.ts
new file mode 100644
index 0000000000..8d110799ea
--- /dev/null
+++ b/packages/sdk/scripts/tests/scripts.spec.ts
@@ -0,0 +1,538 @@
+import { mkdtemp, mkdir, readFile, rm, symlink, writeFile } from 'node:fs/promises'
+import { tmpdir } from 'node:os'
+import { dirname, join } from 'node:path'
+import { PassThrough, Writable } from 'node:stream'
+import { fileURLToPath, pathToFileURL } from 'node:url'
+import { afterEach, describe, expect, expectTypeOf, it, vi } from 'vitest'
+import {
+  LocalPluginBlueprint,
+  NpmPackageManager,
+  SdkProject,
+  featureId,
+  createBuiltinRegistry,
+  type CommandRunner,
+  type NestedMultiSelectValue,
+  type ProjectCreationRequest,
+  type PromptPort,
+} from '@deepseek-ai/dsh-helper'
+import type {
+  ConfirmPromptRequest,
+  MultiSelectPromptRequest,
+  NestedMultiSelectRequest,
+  PromptOutcome,
+  SecretPromptRequest,
+  SelectPromptRequest,
+  TextPromptRequest,
+} from '../../helper/src/questions/prompt-port.ts'
+import { runSDK, startSDK } from '@deepseek-ai/dsh-scripts'
+import { parseDshSdkArgs, parseSdkBootArgs } from '../src/args.ts'
+import { PluginBuild, ProjectBuild, runProjectBuild } from '../src/build.ts'
+import { runDshSdkCommand, type DshSdkCommandContext } from '../src/command.ts'
+import { runConfigCommand } from '../src/config.ts'
+import { ConfigWorkflow } from '../src/config/config-workflow.ts'
+import { initialize, resolve as resolveLocalPlugin } from '../src/local-plugin-loader-hooks.ts'
+
+const temporary: string[] = []
+
+afterEach(async () => {
+  await Promise.all(temporary.splice(0).map(path => rm(path, { recursive: true, force: true })))
+})
+
+class QueuePort implements PromptPort {
+  readonly #answers: unknown[]
+  constructor(answers: unknown[]) { this.#answers = [...answers] }
+  next<T>(): Promise<PromptOutcome<T>> {
+    return Promise.resolve({ status: 'answered', value: this.#answers.shift() as T })
+  }
+  text(_request: TextPromptRequest): Promise<PromptOutcome<string>> { return this.next() }
+  secret(_request: SecretPromptRequest): Promise<PromptOutcome<string>> { return this.next() }
+  select<T>(_request: SelectPromptRequest<T>): Promise<PromptOutcome<T>> { return this.next() }
+  multiselect<T>(_request: MultiSelectPromptRequest<T>): Promise<PromptOutcome<readonly T[]>> { return this.next() }
+  confirm(_request: ConfirmPromptRequest): Promise<PromptOutcome<boolean>> { return this.next() }
+  nestedMultiselect<TValue, TChoice>(
+    _request: NestedMultiSelectRequest<TValue, TChoice>,
+  ): Promise<PromptOutcome<readonly NestedMultiSelectValue<TValue, TChoice>[]>> { return this.next() }
+}
+
+function outputBuffer(): { stream: Writable; read: () => string } {
+  let text = ''
+  return {
+    stream: new Writable({ write(chunk, _encoding, callback) { text += String(chunk); callback() } }),
+    read: () => text,
+  }
+}
+
+function commandContext(cwd: string): DshSdkCommandContext & { readStdout: () => string; readStderr: () => string } {
+  let stdout = ''
+  let stderr = ''
+  const stdin = Object.assign(new PassThrough(), { isTTY: true }) as unknown as NodeJS.ReadStream
+  const output = Object.assign(new Writable({
+    write(chunk, _encoding, callback) { stdout += String(chunk); callback() },
+  }), { isTTY: true }) as unknown as NodeJS.WriteStream
+  const error = new Writable({
+    write(chunk, _encoding, callback) { stderr += String(chunk); callback() },
+  }) as unknown as NodeJS.WriteStream
+  return {
+    cwd, stdin, stdout: output, stderr: error,
+    readStdout: () => stdout,
+    readStderr: () => stderr,
+  }
+}
+
+function creation(
+  extra: ProjectCreationRequest['features'] = [],
+  localPlugins: readonly LocalPluginBlueprint[] = [],
+  app: 'acp' | 'stdio' | 'embed' = 'embed',
+): ProjectCreationRequest {
+  return {
+    name: 'config-agent',
+    description: 'config test',
+    runtime: { model: 'deepseek-v4-flash' },
+    packageManager: new NpmPackageManager('10.0.0'),
+    releaseVersion: '0.0.1',
+    features: [
+      { id: featureId('provider'), options: ['deepseek'], secrets: { apiKey: 'key' } },
+      { id: featureId('bash'), options: ['local'] },
+      { id: featureId('app'), options: [app] },
+      { id: featureId('persistence'), options: ['jsonl'] },
+      ...extra,
+    ],
+    localPlugins,
+  }
+}
+
+async function committedProject(
+  extra: ProjectCreationRequest['features'] = [],
+  localPlugins: readonly LocalPluginBlueprint[] = [],
+  app: 'acp' | 'stdio' | 'embed' = 'embed',
+): Promise<SdkProject> {
+  const root = await mkdtemp(join(tmpdir(), 'dsh-config-workflow-'))
+  temporary.push(root)
+  const request = creation(extra, localPlugins, app)
+  const project = SdkProject.create(root, request)
+  const registry = createBuiltinRegistry(project.profile)
+  const edit = project.edit(registry)
+  for (const item of request.features) edit.installFeature(registry.get(item.id), item)
+  for (const plugin of localPlugins) edit.addPlugin(plugin)
+  return (await edit.commit()).project
+}
+
+describe('Commander launcher arguments', () => {
+  it('parses real subcommands and forwards arbitrary build options', () => {
+    expect(parseDshSdkArgs([])).toMatchObject({ help: true })
+    expect(parseDshSdkArgs(['start', 'index.js'])).toMatchObject({ command: 'start', target: 'index.js' })
+    expect(parseDshSdkArgs(['dev'])).toEqual({ command: 'dev', forwarded: [], help: false })
+    expect(parseDshSdkArgs(['build', '--watch', '--minify'])).toMatchObject({
+      command: 'build', forwarded: ['--watch', '--minify'],
+    })
+    expect(parseDshSdkArgs(['start', 'index.js', '--', '--resume', 'session-1'])).toMatchObject({
+      command: 'start', target: 'index.js', forwarded: ['--resume', 'session-1'],
+    })
+    expect(parseDshSdkArgs(['config'])).toMatchObject({ command: 'config' })
+    expect(parseDshSdkArgs(['start'])).toEqual({ command: 'start', forwarded: [], help: false })
+    expect(parseDshSdkArgs(['dev', 'index.ts'])).toMatchObject({ command: 'dev', target: 'index.ts' })
+    expect(parseDshSdkArgs(['-h'])).toMatchObject({ help: true })
+    expect(() => parseDshSdkArgs(['unknown'])).toThrow()
+    expect(() => parseDshSdkArgs(['config', 'extra'])).toThrow()
+    expect(() => parseDshSdkArgs(['config', '--', 'extra'])).toThrow('does not accept forwarded')
+    expect(parseSdkBootArgs([
+      '--model=mock', '--resume=session-1', '--custom=value', '--verbose', '--no-cache', '--max-depth=-1',
+    ])).toEqual({
+      model: 'mock', resume: 'session-1', custom: 'value', verbose: true, cache: false, 'max-depth': '-1',
+    })
+  })
+
+  it('dispatches every command and maps failures to exit codes', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-command-'))
+    temporary.push(root)
+    const context = commandContext(root)
+    const calls: unknown[] = []
+    context.run = async (target, options) => { calls.push(['run', target, options]); return undefined }
+    context.build = async (args, cwd) => { calls.push(['build', args, cwd]) }
+    context.config = async () => { calls.push(['config']); return {} }
+    await expect(runDshSdkCommand(['start', 'index.js', '--', '--resume', 'session-1'], context)).resolves.toBe(0)
+    await expect(runDshSdkCommand(['dev', 'index.ts'], context)).resolves.toBe(0)
+    await expect(runDshSdkCommand(['build', '--watch'], context)).resolves.toBe(0)
+    await expect(runDshSdkCommand(['config'], context)).resolves.toBe(0)
+    expect(calls).toHaveLength(4)
+    expect(calls[0]).toEqual(['run', 'index.js', { cwd: root, argv: ['--resume', 'session-1'] }])
+    expect(calls[1]).toEqual(['run', 'index.ts', { cwd: root, dev: true, argv: [] }])
+    context.config = async () => ({ installError: new Error('offline') })
+    await expect(runDshSdkCommand(['config'], context)).resolves.toBe(1)
+    context.config = async () => { throw 'broken' }
+    await expect(runDshSdkCommand(['config'], context)).resolves.toBe(1)
+    expect(context.readStderr()).toContain('broken')
+    await expect(runDshSdkCommand(['unknown'], context)).resolves.toBe(1)
+    await expect(runDshSdkCommand([], context)).resolves.toBe(0)
+    expect(context.readStdout()).toContain('Usage: dsh-sdk')
+
+    const defaults = commandContext(root)
+    await writeFile(join(root, 'main.mjs'), 'export function main() { return "ok" }\n')
+    await expect(runDshSdkCommand(['start', 'main.mjs'], defaults)).resolves.toBe(0)
+    await expect(runDshSdkCommand(['build'], defaults)).resolves.toBe(0)
+    defaults.port = new QueuePort([[]])
+    await expect(runDshSdkCommand(['config'], defaults)).resolves.toBe(1)
+  })
+})
+
+describe('build profiles and invocation', () => {
+  it('discovers root and plugin targets and creates independent profiles', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-build-profile-'))
+    temporary.push(root)
+    await mkdir(join(root, 'plugins', 'one', 'src'), { recursive: true })
+    await writeFile(join(root, 'index.ts'), 'export {}\n')
+    await writeFile(join(root, 'plugins', 'one', 'package.json'), '{"name":"one"}\n')
+    await writeFile(join(root, 'plugins', 'one', 'src', 'index.ts'), 'export {}\n')
+    expect(ProjectBuild({ cwd: root, entry: ['index.ts'] })).toEqual([
+      { cwd: root, entry: ['index.ts'] },
+      { workspace: { include: ['plugins/*'] } },
+    ])
+    expect(PluginBuild({ entry: ['src/index.ts'], dts: true })).toEqual({ entry: ['src/index.ts'], dts: true })
+    expect(() => ProjectBuild({ workspace: true })).toThrow('owns workspace discovery')
+    expect(() => PluginBuild({ workspace: true })).toThrow('does not accept nested workspace')
+    expect(ProjectBuild({ cwd: join(root, 'empty'), entry: ['index.ts'] })).toEqual([
+      { cwd: join(root, 'empty'), entry: ['index.ts'] },
+    ])
+    expect(ProjectBuild({ entry: ['index.ts'] })[0]).toMatchObject({ entry: ['index.ts'] })
+  })
+
+  it('runs the project-installed tsdown and reports child failure', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-build-run-'))
+    temporary.push(root)
+    await writeFile(join(root, 'package.json'), '{"type":"module"}\n')
+    await writeFile(join(root, 'index.ts'), 'export {}\n')
+    await writeFile(join(root, 'tsdown.config.ts'), 'export default {}\n')
+    await mkdir(join(root, 'node_modules'), { recursive: true })
+    const manifest = fileURLToPath(import.meta.resolve('tsdown/package.json'))
+    await symlink(dirname(manifest), join(root, 'node_modules', 'tsdown'))
+    const calls: string[][] = []
+    const runner: CommandRunner = {
+      run: async (command, args) => {
+        calls.push([command, ...args])
+        return { exitCode: 0, signal: null }
+      },
+    }
+    await runProjectBuild(['--watch'], root, runner)
+    expect(calls[0]?.[0]).toBe(process.execPath)
+    expect(calls[0]?.at(-1)).toBe('--watch')
+    const failed: CommandRunner = { run: async () => ({ exitCode: 2, signal: null }) }
+    await expect(runProjectBuild([], root, failed)).rejects.toThrow('exited with code 2')
+    const killed: CommandRunner = { run: async () => ({ exitCode: null, signal: 'SIGTERM' }) }
+    await expect(runProjectBuild([], root, killed)).rejects.toThrow('killed by SIGTERM')
+  })
+
+  it('recognizes every tsdown config source', async () => {
+    const manifest = fileURLToPath(import.meta.resolve('tsdown/package.json'))
+    for (const extension of ['cts', 'cjs', 'json']) {
+      const root = await mkdtemp(join(tmpdir(), `dsh-build-${extension}-`))
+      temporary.push(root)
+      await writeFile(join(root, 'package.json'), '{"type":"module"}\n')
+      await writeFile(join(root, `tsdown.config.${extension}`), '{}\n')
+      await mkdir(join(root, 'node_modules'), { recursive: true })
+      await symlink(dirname(manifest), join(root, 'node_modules', 'tsdown'))
+      let called = false
+      await runProjectBuild([], root, {
+        run: async () => { called = true; return { exitCode: 0, signal: null } },
+      })
+      expect(called).toBe(true)
+    }
+    const root = await mkdtemp(join(tmpdir(), 'dsh-build-package-json-'))
+    temporary.push(root)
+    await writeFile(join(root, 'package.json'), '{"type":"module","tsdown":{}}\n')
+    await mkdir(join(root, 'node_modules'), { recursive: true })
+    await symlink(dirname(manifest), join(root, 'node_modules', 'tsdown'))
+    let called = false
+    await runProjectBuild([], root, {
+      run: async () => { called = true; return { exitCode: 0, signal: null } },
+    })
+    expect(called).toBe(true)
+  })
+
+  it('reports missing and malformed project tsdown executables', async () => {
+    const missing = await mkdtemp(join(tmpdir(), 'dsh-build-missing-'))
+    temporary.push(missing)
+    await writeFile(join(missing, 'package.json'), '{"type":"module"}')
+    await writeFile(join(missing, 'tsdown.config.ts'), 'export default {}\n')
+    await expect(runProjectBuild([], missing)).rejects.toThrow('requires tsdown')
+    const malformed = await mkdtemp(join(tmpdir(), 'dsh-build-malformed-'))
+    temporary.push(malformed)
+    await writeFile(join(malformed, 'package.json'), '{"type":"module"}')
+    await writeFile(join(malformed, 'tsdown.config.ts'), 'export default {}\n')
+    await mkdir(join(malformed, 'node_modules', 'tsdown'), { recursive: true })
+    await writeFile(join(malformed, 'node_modules', 'tsdown', 'package.json'), JSON.stringify({
+      name: 'tsdown', version: '0.0.0', exports: { './package.json': './package.json' }, bin: {},
+    }))
+    await expect(runProjectBuild([], malformed)).rejects.toThrow('has no executable')
+    await writeFile(join(malformed, 'node_modules', 'tsdown', 'package.json'), JSON.stringify({
+      name: 'tsdown', version: '0.0.0', exports: { './package.json': './package.json' },
+    }))
+    await expect(runProjectBuild([], malformed)).rejects.toThrow('has no executable')
+    const stringBin = await mkdtemp(join(tmpdir(), 'dsh-build-string-bin-'))
+    temporary.push(stringBin)
+    await writeFile(join(stringBin, 'package.json'), '{"type":"module"}')
+    await writeFile(join(stringBin, 'tsdown.config.js'), 'export default {}\n')
+    await mkdir(join(stringBin, 'node_modules', 'tsdown'), { recursive: true })
+    await writeFile(join(stringBin, 'node_modules', 'tsdown', 'package.json'), JSON.stringify({
+      name: 'tsdown', version: '0.0.0', exports: { './package.json': './package.json' }, bin: 'cli.js',
+    }))
+    await writeFile(join(stringBin, 'node_modules', 'tsdown', 'cli.js'), '')
+    let command = ''
+    await runProjectBuild([], stringBin, {
+      run: async (_node, args) => { command = args[0] ?? ''; return { exitCode: 0, signal: null } },
+    })
+    expect(command).toContain('cli.js')
+  })
+
+  it('returns a no-op for a project with no build targets and hints on a missing start target', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-no-build-'))
+    temporary.push(root)
+    let called = false
+    await runProjectBuild([], root, { run: async () => { called = true; return { exitCode: 0, signal: null } } })
+    expect(called).toBe(false)
+    const unreadableManifest = await mkdtemp(join(tmpdir(), 'dsh-build-unreadable-manifest-'))
+    temporary.push(unreadableManifest)
+    await mkdir(join(unreadableManifest, 'package.json'))
+    await expect(runProjectBuild([], unreadableManifest)).rejects.toThrow()
+    await expect(runSDK('index.js', { cwd: root })).rejects.toThrow('Run dsh-sdk build first')
+  })
+
+  it('invokes the target module main export and rejects passive modules', async () => {
+    const root = await mkdtemp(join(tmpdir(), 'dsh-module-main-'))
+    temporary.push(root)
+    await writeFile(join(root, 'main.mjs'), 'export function main(context) { return context }\n')
+    await writeFile(join(root, 'passive.mjs'), 'export const value = 1\n')
+    await expect(runSDK('main.mjs', {
+      cwd: root,
+      argv: ['--model=mock', '--resume=session-1', 'custom'],
+    })).resolves.toEqual({
+      argv: ['--model=mock', '--resume=session-1', 'custom'],
+      args: { model: 'mock', resume: 'session-1' }, cwd: root, mode: 'start',
+    })
+    await expect(runSDK('passive.mjs', { cwd: root })).rejects.toThrow('must export function main()')
+  })
+
+  it('boots empty Cordis configs and delegates targetless runs', async () => {
+    expectTypeOf(runSDK).toBeCallableWith()
+    const root = await mkdtemp(join(tmpdir(), 'dsh-start-sdk-'))
+    temporary.push(root)
+    await writeFile(join(root, 'cordis.yml'), '[]\n')
+    const byUrl = await startSDK(pathToFileURL(join(root, 'cordis.yml')))
+    await byUrl.fiber.dispose()
+    const byRun = await runSDK(undefined, { cwd: root }) as import('cordis').Context
+    await byRun.fiber.dispose()
+    const dev = await startSDK('./cordis.yml', { cwd: root, dev: true })
+    await dev.fiber.dispose()
+    await expect(startSDK(new URL('https://example.invalid/cordis.yml'), { cwd: root })).rejects.toThrow()
+  })
+
+  it('validates local plugin metadata in dev mode', async () => {
+    const malformed = await mkdtemp(join(tmpdir(), 'dsh-dev-malformed-'))
+    temporary.push(malformed)
+    await mkdir(join(malformed, 'plugins', 'bad'), { recursive: true })
+    await expect(runSDK('missing.ts', { cwd: malformed, dev: true })).rejects.toThrow('cannot load local plugin metadata')
+    const absent = await mkdtemp(join(tmpdir(), 'dsh-dev-absent-'))
+    temporary.push(absent)
+    await expect(runSDK('missing.ts', { cwd: absent, dev: true })).rejects.toThrow('cannot start missing target')
+
+    const unnamed = await mkdtemp(join(tmpdir(), 'dsh-dev-unnamed-'))
+    temporary.push(unnamed)
+    await mkdir(join(unnamed, 'plugins', 'bad', 'src'), { recursive: true })
+    await writeFile(join(unnamed, 'plugins', 'bad', 'package.json'), '{}')
+    await writeFile(join(unnamed, 'plugins', 'bad', 'src/index.ts'), 'export {}\n')
+    await expect(runSDK('missing.ts', { cwd: unnamed, dev: true })).rejects.toThrow('has no name')
+
+    const duplicate = await mkdtemp(join(tmpdir(), 'dsh-dev-duplicate-'))
+    temporary.push(duplicate)
+    for (const name of ['one', 'two']) {
+      await mkdir(join(duplicate, 'plugins', name, 'src'), { recursive: true })
+      await writeFile(join(duplicate, 'plugins', name, 'package.json'), '{"name":"same"}')
+      await writeFile(join(duplicate, 'plugins', name, 'src/index.ts'), 'export {}\n')
+    }
+    await expect(runSDK('missing.ts', { cwd: duplicate, dev: true })).rejects.toThrow('duplicate local plugin')
+
+    const valid = await mkdtemp(join(tmpdir(), 'dsh-dev-valid-'))
+    temporary.push(valid)
+    await mkdir(join(valid, 'plugins', 'one', 'src'), { recursive: true })
+    await writeFile(join(valid, 'plugins', 'README.md'), 'skip\n')
+    await writeFile(join(valid, 'plugins', 'one', 'package.json'), '{"name":"local"}')
+    await writeFile(join(valid, 'plugins', 'one', 'src/index.ts'), 'export {}\n')
+    await writeFile(join(valid, 'main.ts'), 'export function main() { return "dev" }\n')
+    await expect(runSDK('main.ts', { cwd: valid, dev: true })).resolves.toBe('dev')
+    await expect(runSDK('missing.ts', { cwd: valid, dev: true })).rejects.toThrow('cannot start missing target')
+  })
+
+  it('maps only exact local package names through the loader hook', async () => {
+    initialize({ mappings: { local: 'file:///tmp/local.ts' } })
+    const next = async (specifier: string) => ({ url: specifier, format: 'module' as const })
+    const context: import('node:module').ResolveHookContext = {
+      conditions: [], importAttributes: {}, parentURL: undefined,
+    }
+    await expect(resolveLocalPlugin('local', context, next)).resolves.toMatchObject({ url: 'file:///tmp/local.ts' })
+    await expect(resolveLocalPlugin('other', context, next)).resolves.toMatchObject({ url: 'other' })
+  })
+})
+
+describe('ConfigWorkflow', () => {
+  it('opens a project through the config command prompt seam', async () => {
+    const project = await committedProject()
+    const context = commandContext(project.root)
+    context.port = new QueuePort([[]])
+    context.install = async () => { throw new Error('install should not run') }
+    await expect(runConfigCommand(context)).resolves.toEqual({})
+    delete context.port
+    delete context.install
+    context.stdin.isTTY = false
+    await expect(runConfigCommand(context)).rejects.toThrow('interactive TTY')
+    context.stdin.isTTY = true
+    context.stdout.isTTY = false
+    await expect(runConfigCommand(context)).rejects.toThrow('interactive TTY')
+  })
+  it('accumulates a disable and commits only after Review & Apply', async () => {
+    const project = await committedProject([{ id: featureId('todo'), options: ['default'] }])
+    const registry = createBuiltinRegistry(project.profile)
+    const output = outputBuffer()
+    const workflow = new ConfigWorkflow(new QueuePort([
+      [], true,
+    ]), output.stream, async () => { throw new Error('install should not run') })
+    const result = await workflow.run(project, registry)
+    expect(result.commit?.project.cordis.entry('tool-todo')?.disabled).toBe(true)
+    expect(output.read()).toContain('Disable feature: todo')
+  })
+
+  it('installs once after NPM dependency changes and keeps committed files on install failure', async () => {
+    const project = await committedProject()
+    const registry = createBuiltinRegistry(project.profile)
+    const output = outputBuffer()
+    let installs = 0
+    const workflow = new ConfigWorkflow(new QueuePort([
+      [{ value: 'feature:todo', choices: [] }], true,
+    ]), output.stream, async () => {
+      installs += 1
+      throw new Error('offline')
+    })
+    const result = await workflow.run(project, registry)
+    expect(installs).toBe(1)
+    expect(result.installError?.message).toBe('offline')
+    expect(result.commit?.project.cordis.entry('tool-todo')).toBeDefined()
+    expect(output.read()).toContain('Changes were committed, but install failed')
+  })
+
+  it('cancels apply and enables a disabled feature without reinstalling', async () => {
+    const project = await committedProject([{ id: featureId('todo'), options: ['default'] }])
+    const registry = createBuiltinRegistry(project.profile)
+    const cancelled = await new ConfigWorkflow(new QueuePort([[], false]), outputBuffer().stream).run(project, registry)
+    expect(cancelled).toEqual({})
+    const disable = project.edit(registry)
+    disable.disableFeature(registry.get(featureId('todo')))
+    const disabled = (await disable.commit()).project
+    let installs = 0
+    const enabled = await new ConfigWorkflow(new QueuePort([
+      [{ value: 'feature:todo', choices: [] }], true,
+    ]), outputBuffer().stream, async () => { installs += 1 }).run(disabled, createBuiltinRegistry(disabled.profile))
+    expect(enabled.commit?.project.cordis.entry('tool-todo')?.disabled).toBeUndefined()
+    expect(installs).toBe(0)
+  })
+
+  it('toggles custom Cordis config entries without changing NPM dependencies', async () => {
+    const project = await committedProject([], [new LocalPluginBlueprint('sample', 'plugin')])
+    await expect(new ConfigWorkflow(new QueuePort([
+      [{ value: 'plugin:sample', choices: [] }],
+    ]), outputBuffer().stream).run(project, createBuiltinRegistry(project.profile))).resolves.toEqual({})
+    const output = outputBuffer()
+    const disabled = await new ConfigWorkflow(new QueuePort([[], true]), output.stream).run(
+      project, createBuiltinRegistry(project.profile),
+    )
+    expect(disabled.commit?.project.cordis.entry('sample')?.disabled).toBe(true)
+    expect(output.read()).toContain('Disable custom plugin: sample')
+    const next = disabled.commit?.project
+    if (!next) throw new Error('custom toggle did not commit')
+    const enabled = await new ConfigWorkflow(new QueuePort([
+      [{ value: 'plugin:sample', choices: [] }], true,
+    ]), outputBuffer().stream).run(next, createBuiltinRegistry(next.profile))
+    expect(enabled.commit?.project.cordis.entry('sample')?.disabled).toBeUndefined()
+  })
+
+  it('shows inconsistent features as diagnostic-only rows', async () => {
+    const complete = await committedProject()
+    await writeFile(join(complete.root, 'cordis.yml'), `${await readFile(join(complete.root, 'cordis.yml'), 'utf8')}- id: web-search-exa
+  name: '@deepseek-ai/dsh-web-search-exa'
+`)
+    const project = await SdkProject.open(complete.root)
+    const port = new QueuePort([[]])
+    await expect(new ConfigWorkflow(port, outputBuffer().stream).run(project, createBuiltinRegistry(project.profile)))
+      .resolves.toEqual({})
+  })
+
+  it('uses the default installer and normalizes non-Error install failures', async () => {
+    const project = await committedProject()
+    const install = vi.spyOn(NpmPackageManager.prototype, 'install').mockResolvedValue()
+    await new ConfigWorkflow(new QueuePort([
+      [{ value: 'feature:todo', choices: [] }], true,
+    ])).run(project, createBuiltinRegistry(project.profile))
+    expect(install).toHaveBeenCalledOnce()
+    install.mockRestore()
+    const next = await committedProject()
+    const failed = await new ConfigWorkflow(new QueuePort([
+      [{ value: 'feature:todo', choices: [] }], true,
+    ]), outputBuffer().stream, async () => { throw 'offline-string' }).run(next, createBuiltinRegistry(next.profile))
+    expect(failed.installError?.message).toBe('offline-string')
+  })
+
+  it('reconciles a child option selected in the feature tree', async () => {
+    const project = await committedProject()
+    const registry = createBuiltinRegistry(project.profile)
+    let installs = 0
+    const workflow = new ConfigWorkflow(new QueuePort([
+      [{ value: 'feature:persistence', choices: ['sqlite'] }], true,
+    ]), outputBuffer().stream, async () => { installs += 1 })
+    const result = await workflow.run(project, registry)
+    expect(result.commit?.project.cordis.entry('session-persistence')).toMatchObject({
+      name: '@deepseek-ai/dsh-session-persistence-sqlite',
+      config: { path: './.sessions/sessions.sqlite' },
+    })
+    expect(installs).toBe(1)
+  })
+
+  it('switches required provider and interface options', async () => {
+    const project = await committedProject()
+    const registry = createBuiltinRegistry(project.profile)
+    const workflow = new ConfigWorkflow(new QueuePort([
+      [
+        { value: 'feature:provider', choices: ['custom'] },
+        { value: 'feature:app', choices: ['stdio'] },
+        { value: 'feature:persistence', choices: ['jsonl'] },
+      ],
+      'https://provider.example/v1',
+      'custom-key',
+      true,
+    ]), outputBuffer().stream, async () => {})
+    const result = await workflow.run(project, registry)
+    const provider = result.commit?.project.cordis.entry('llm-pi-ai')
+    expect(provider?.config?.apiKey).toBeDefined()
+    expect(provider?.config?.baseURL).toBe('https://provider.example/v1')
+    expect(result.commit?.project.cordis.entry('stdio')).toBeDefined()
+    expect(result.commit?.project.cordis.entry('agent-loop')).toBeDefined()
+    expect(result.commit?.project.cordis.entry('agent-core')).toBeUndefined()
+  })
+
+  it('disables ask-user when switching its app interface to embed', async () => {
+    const project = await committedProject([
+      { id: featureId('ask-user'), options: ['default'] },
+    ], [], 'acp')
+    const registry = createBuiltinRegistry(project.profile)
+    const output = outputBuffer()
+    const workflow = new ConfigWorkflow(new QueuePort([
+      [
+        { value: 'feature:provider', choices: ['deepseek'] },
+        { value: 'feature:app', choices: ['embed'] },
+        { value: 'feature:persistence', choices: ['jsonl'] },
+        { value: 'feature:ask-user', choices: ['default'] },
+      ],
+      true,
+    ]), output.stream, async () => {})
+    const result = await workflow.run(project, registry)
+    expect(result.commit?.project.profile.runInterface).toBe('embed')
+    expect(result.commit?.project.cordis.entry('tool-ask-user')?.disabled).toBe(true)
+    expect(output.read()).toContain('Disable feature: ask-user')
+  })
+})
diff --git a/packages/sdk/scripts/tsconfig.json b/packages/sdk/scripts/tsconfig.json
new file mode 100644
index 0000000000..848de9a314
--- /dev/null
+++ b/packages/sdk/scripts/tsconfig.json
@@ -0,0 +1,13 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": ["src"],
+  "references": [
+    { "path": "../helper" },
+    { "path": "../../ui/app-boot" },
+    { "path": "../../../vendor/cordis" }
+  ]
+}
diff --git a/packages/sdk/scripts/tsdown.config.ts b/packages/sdk/scripts/tsdown.config.ts
new file mode 100644
index 0000000000..2218faa7be
--- /dev/null
+++ b/packages/sdk/scripts/tsdown.config.ts
@@ -0,0 +1,22 @@
+import { defineConfig } from 'tsdown'
+
+/** Bundle each public or runtime entry and mirror package-owned terminal templates. */
+export default defineConfig([
+  {
+    entry: ['lib/types/index.js'], outDir: 'lib', format: ['esm'], platform: 'node', target: 'es2024',
+    fixedExtension: false, outputOptions: { codeSplitting: false }, dts: false, clean: false,
+    copy: [{ from: 'src/templates/assets/*', to: 'lib/assets' }],
+  },
+  {
+    entry: ['lib/types/bin.js'], outDir: 'lib', format: ['esm'], platform: 'node', target: 'es2024',
+    fixedExtension: false, outputOptions: { codeSplitting: false }, dts: false, clean: false,
+  },
+  {
+    entry: ['lib/types/dev/tsdown-config.js'], outDir: 'lib/dev', format: ['esm'], platform: 'node',
+    target: 'es2024', fixedExtension: false, outputOptions: { codeSplitting: false }, dts: false, clean: false,
+  },
+  {
+    entry: ['lib/types/local-plugin-loader-hooks.js'], outDir: 'lib', format: ['esm'], platform: 'node',
+    target: 'es2024', fixedExtension: false, outputOptions: { codeSplitting: false }, dts: false, clean: false,
+  },
+])
diff --git a/packages/session-persistence/session-persistence-jsonl/README.md b/packages/session-persistence/session-persistence-jsonl/README.md
index 343fb4a70a..43f4443107 100644
--- a/packages/session-persistence/session-persistence-jsonl/README.md
+++ b/packages/session-persistence/session-persistence-jsonl/README.md
@@ -19,6 +19,8 @@ The JSONL durable session-persistence backend — a concrete `SessionPersistence
 |---|---|---|
 | `root` | `string` (required) | Root directory for all session files. **No default** — a `process.cwd()` default would scatter files as the process's cwd changes (bash calls, subprocesses). |
 
+`locate(meta)` returns `{ kind: 'jsonl', path }` using the resolved absolute root and the same cwd-bucket/id encoding as materialization. It performs no filesystem I/O: the target can be returned before the file exists, and an existing file contains only the last flushed prefix.
+
 ## Durability and crash semantics
 
 - **Lazy materialization.** `create(meta)` writes nothing; on the first `append`, the backend writes and `fsync`s a temporary file, publishes it without overwrite via a hard link, then `fsync`s the directory. A created-but-never-appended session leaves nothing on disk and is absent from `list`.
diff --git a/packages/session-persistence/session-persistence-jsonl/src/index.ts b/packages/session-persistence/session-persistence-jsonl/src/index.ts
index 1d13ff424e..0c8a2ee3ef 100644
--- a/packages/session-persistence/session-persistence-jsonl/src/index.ts
+++ b/packages/session-persistence/session-persistence-jsonl/src/index.ts
@@ -1,7 +1,8 @@
 /**
  * JSONL durable session-persistence backend. It stores a header and contiguous
  * events in one append-only file per session, and delegates orchestration to
- * {@link PersistenceCoordinator}.
+ * {@link PersistenceCoordinator}. Its side-effect-free locator returns the
+ * absolute per-session log target before materialization.
  * @module @deepseek-ai/dsh-session-persistence-jsonl
  */
 
@@ -12,7 +13,7 @@ import { dirname, resolve } from 'node:path'
 import { randomBytes } from 'node:crypto'
 import {
   SessionPersistence, PersistenceCoordinator,
-  type PersistenceBackend, type StoredPrefix,
+  type PersistenceBackend, type SessionLocation, type StoredPrefix,
 } from '@deepseek-ai/dsh-session-persistence'
 import type { SessionEvent, SessionId, SessionHeader } from '@deepseek-ai/dsh-session'
 import {
@@ -68,6 +69,11 @@ export class SessionPersistenceJsonl extends SessionPersistence implements Persi
   /* jscpd:ignore-start */
   // --- SessionPersistence service surface (delegated to the coordinator) ---
 
+  /** Resolve the absolute target path without touching the filesystem. */
+  locate(meta: SessionHeader): SessionLocation {
+    return { kind: 'jsonl', path: logPath(this.root, meta.cwd, meta.id) }
+  }
+
   create(meta: SessionHeader): Promise<void> {
     return this.coordinator.create(meta)
   }
diff --git a/packages/session-persistence/session-persistence-jsonl/tests/jsonl.spec.ts b/packages/session-persistence/session-persistence-jsonl/tests/jsonl.spec.ts
index 75fcb48732..4b279e3c6e 100644
--- a/packages/session-persistence/session-persistence-jsonl/tests/jsonl.spec.ts
+++ b/packages/session-persistence/session-persistence-jsonl/tests/jsonl.spec.ts
@@ -2,11 +2,11 @@ import { afterEach, beforeEach, describe, expect, it, vi } from 'vitest'
 import { Context } from 'cordis'
 import { appendFile, mkdtemp, mkdir, rm, readFile, writeFile, readdir, stat } from 'node:fs/promises'
 import { tmpdir } from 'node:os'
-import { join } from 'node:path'
+import { isAbsolute, join, relative, resolve } from 'node:path'
 import SessionStore, { SessionId } from '@deepseek-ai/dsh-session'
 import type { Session, SessionEvent, SessionHeader } from '@deepseek-ai/dsh-session'
 import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
-import { encodeSegment, logPath, scanLog, sessionDir } from '../src/format.ts'
+import { encodeSegment, logPath, scanLog, sessionDir, toHeaderLine } from '../src/format.ts'
 import { runPersistenceContract, meta, oneTurnLog, appendLog } from '../../session-persistence/tests/contract.ts'
 import { runCoordinatorContract, type CoordinatorFixture } from '../../session-persistence/tests/coordinator-contract.ts'
 
@@ -112,6 +112,19 @@ describe('SessionPersistenceJsonl: format helpers', () => {
   it('encodeSegment rejects an empty id', () => {
     expect(() => encodeSegment('')).toThrow(/empty/)
   })
+
+  it('resolves a relative custom root before locating a session', async () => {
+    const absoluteRoot = await freshRoot()
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const fiber = await ctx.plugin(SessionPersistenceJsonl, { root: relative(process.cwd(), absoluteRoot) })
+    const m = meta('relative-location', '/work')
+    expect(ctx.sessionPersistence.locate(m)).toEqual({
+      kind: 'jsonl',
+      path: logPath(resolve(absoluteRoot), '/work', m.id),
+    })
+    await fiber.dispose()
+  })
 })
 
 describe('SessionPersistenceJsonl: durability and crash semantics', () => {
@@ -126,8 +139,13 @@ describe('SessionPersistenceJsonl: durability and crash semantics', () => {
 
   it('lazy materialization: create() writes no file until the first append', async () => {
     const m = meta('lazy', '/work')
+    const location = ctx.sessionPersistence.locate(m)
+    expect(location).toEqual({ kind: 'jsonl', path: logPath(root, '/work', m.id) })
+    expect(isAbsolute(location!.path)).toBe(true)
+
     await ctx.sessionPersistence.create(m)
-    // nothing on disk yet
+    // locate() is a pure target-path calculation: neither it nor create()
+    // materializes a file before the first append.
     const dir = sessionDir(root, '/work')
     await expect(stat(logPath(root, '/work', m.id))).rejects.toThrow()
     expect((await ctx.sessionPersistence.list()).map(h => h.id)).not.toContain(m.id)
@@ -139,6 +157,26 @@ describe('SessionPersistenceJsonl: durability and crash semantics', () => {
     void dir
   })
 
+  it('keeps the same location on resume and gives a fork its own location', async () => {
+    const parent = meta('location-parent', '/work')
+    const parentLocation = ctx.sessionPersistence.locate(parent)
+    await ctx.sessionPersistence.create(parent)
+    await ctx.sessionPersistence.append(parent.id, oneTurnLog())
+
+    const loaded = await ctx.sessionPersistence.load(parent.id)
+    expect(ctx.sessionPersistence.locate(loaded.meta)).toEqual(parentLocation)
+
+    const child = {
+      ...loaded.meta,
+      id: SessionId('location-child'),
+      parentSession: parent.id,
+      seedLength: loaded.events.length,
+    }
+    const childLocation = ctx.sessionPersistence.locate(child)
+    expect(childLocation?.path).not.toBe(parentLocation?.path)
+    expect(childLocation).toEqual({ kind: 'jsonl', path: logPath(root, '/work', child.id) })
+  })
+
   it('round-trip is byte-identical (incl. assistant/chunk verbatim)', async () => {
     const m = meta('chunks')
     const log: SessionEvent[] = [
@@ -146,7 +184,7 @@ describe('SessionPersistenceJsonl: durability and crash semantics', () => {
       { type: 'step/start', seq: 1, time: 2, data: { turn: 1, step: 1 } },
       { type: 'assistant/chunk', seq: 2, time: 3, data: { turn: 1, step: 1, chunk: { type: 'text-delta', index: 0, text: 'he' } } },
       { type: 'assistant/chunk', seq: 3, time: 4, data: { turn: 1, step: 1, chunk: { type: 'text-delta', index: 0, text: 'llo' } } },
-      { type: 'assistant/message', seq: 4, time: 5, data: { turn: 1, step: 1, content: [{ type: 'text', text: 'hello' }] }, surfaceOp: 'append', sourceEventSeqs: [2, 3] },
+      { type: 'assistant/message', seq: 4, time: 5, data: { turn: 1, step: 1, content: [{ type: 'text', text: 'hello' }], provenance: { provider: 'mock', model: 'mock' } }, surfaceOp: 'append', sourceEventSeqs: [2, 3] },
       { type: 'step/end', seq: 5, time: 6, data: { turn: 1, step: 1 } },
       { type: 'turn/end', seq: 6, time: 7, data: { turn: 1, reason: { kind: 'completed' } } },
     ]
@@ -156,6 +194,40 @@ describe('SessionPersistenceJsonl: durability and crash semantics', () => {
     expect(loaded.events).toEqual(log) // chunks preserved, contiguous seqs
   })
 
+  it('rejects a stored v0 log containing a legacy request/header-delta event', async () => {
+    const m = meta('legacy-header-delta', '/legacy')
+    const path = logPath(root, m.cwd, m.id)
+    await mkdir(sessionDir(root, m.cwd), { recursive: true })
+    await writeFile(path, [
+      JSON.stringify(toHeaderLine(m)),
+      JSON.stringify({ type: 'turn/start', seq: 0, time: 1, data: { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } } }),
+      JSON.stringify({ type: 'request/header-delta', seq: 1, time: 2, data: { config: { model: 'legacy' } } }),
+      JSON.stringify({ type: 'turn/end', seq: 2, time: 3, data: { turn: 1, reason: { kind: 'completed' } } }),
+      '',
+    ].join('\n'))
+
+    await expect(ctx.sessionPersistence.load(m.id)).rejects.toThrow(/unsupported legacy request\/header-delta event at seq 1/)
+  })
+
+  it('rejects a stored v0 full header carrying the legacy fallback reason', async () => {
+    const m = meta('legacy-header-fallback', '/legacy')
+    const path = logPath(root, m.cwd, m.id)
+    await mkdir(sessionDir(root, m.cwd), { recursive: true })
+    await writeFile(path, [
+      JSON.stringify(toHeaderLine(m)),
+      JSON.stringify({
+        type: 'request/header',
+        seq: 0,
+        time: 1,
+        data: { header: { config: { model: 'legacy' } }, reason: 'fallback' },
+      }),
+      '',
+    ].join('\n'))
+
+    await expect(ctx.sessionPersistence.load(m.id))
+      .rejects.toThrow(/unsupported legacy request\/header reason "fallback" at seq 0/)
+  })
+
   it('persists a forked child seed through the existing session write path', async () => {
     const source = ctx.sessions.create(SessionId('persist-parent'), { meta: { cwd: '/workspace' } })
     appendClosedTurn(source)
diff --git a/packages/session-persistence/session-persistence-sqlite/README.md b/packages/session-persistence/session-persistence-sqlite/README.md
index 1411e177be..241e7ceb38 100644
--- a/packages/session-persistence/session-persistence-sqlite/README.md
+++ b/packages/session-persistence/session-persistence-sqlite/README.md
@@ -2,6 +2,8 @@
 
 A SQLite durable session-persistence backend — a second `SessionPersistence` implementation ([session persistence](../../../docs/rfc/implemented/architecture/2026-06-14-session-persistence.md)), built to validate that the abstract seam and the shared `runPersistenceContract` suite are genuinely backend-agnostic. It satisfies the SAME contract as `dsh-session-persistence-jsonl` (append-only, contiguous-seq, lazy materialization, interrupted-turn close on load), expressed over `node:sqlite` rows instead of file bytes.
 
+`locate(meta)` returns `undefined`: all sessions share one database, so there is no honest independent per-session transcript path.
+
 > **TODO:** this backend talks to `node:sqlite` directly. If a cordis database service (`cordis/db` / a `@cordisjs` SQL driver plugin) is adopted, route through that instead of holding a raw `DatabaseSync` here — the contract surface (`SessionPersistence`) would not change, only the storage driver.
 
 ## Storage model
diff --git a/packages/session-persistence/session-persistence-sqlite/src/index.ts b/packages/session-persistence/session-persistence-sqlite/src/index.ts
index 7d23292e9d..d886124961 100644
--- a/packages/session-persistence/session-persistence-sqlite/src/index.ts
+++ b/packages/session-persistence/session-persistence-sqlite/src/index.ts
@@ -1,7 +1,8 @@
 /**
  * SQLite durable session-persistence backend. It maps each session header and
  * event to rows, and delegates write-path orchestration to
- * {@link PersistenceCoordinator}.
+ * {@link PersistenceCoordinator}. It has no independent per-session artifact,
+ * so its locator returns `undefined`.
  * @module @deepseek-ai/dsh-session-persistence-sqlite
  */
 
@@ -12,7 +13,7 @@ import { mkdir } from 'node:fs/promises'
 import { dirname, resolve } from 'node:path'
 import {
   SessionPersistence, PersistenceCoordinator,
-  type PersistenceBackend, type StoredPrefix,
+  type PersistenceBackend, type SessionLocation, type StoredPrefix,
 } from '@deepseek-ai/dsh-session-persistence'
 import type { SessionEvent, SurfaceEventType, SessionId, SessionHeader } from '@deepseek-ai/dsh-session'
 import {
@@ -95,6 +96,11 @@ export class SessionPersistenceSqlite extends SessionPersistence implements Pers
 
   // --- SessionPersistence service surface (delegated to the coordinator) ---
 
+  /** SQLite has one database, not an independent local artifact per session. */
+  locate(_meta: SessionHeader): SessionLocation | undefined {
+    return undefined
+  }
+
   create(meta: SessionHeader): Promise<void> {
     return this.coordinator.create(meta)
   }
diff --git a/packages/session-persistence/session-persistence-sqlite/tests/sqlite.spec.ts b/packages/session-persistence/session-persistence-sqlite/tests/sqlite.spec.ts
index bae79b0a53..a728cc0b71 100644
--- a/packages/session-persistence/session-persistence-sqlite/tests/sqlite.spec.ts
+++ b/packages/session-persistence/session-persistence-sqlite/tests/sqlite.spec.ts
@@ -153,6 +153,48 @@ describe('scanRows', () => {
 })
 
 describe('SessionPersistenceSqlite: durability and crash semantics', () => {
+  it('rejects a stored v0 log containing a legacy request/header-delta event', async () => {
+    const path = await freshDbPath()
+    const m = meta('legacy-header-delta', '/legacy')
+    const db = openDatabase(path, 'wal')
+    db.prepare('INSERT INTO sessions (id, version, created_at, cwd, parent_session, seed_length) VALUES (?, ?, ?, ?, NULL, NULL)')
+      .run(m.id, m.version, m.createdAt, m.cwd ?? null)
+    const insert = db.prepare('INSERT INTO events (session_id, seq, type, time, data) VALUES (?, ?, ?, ?, ?)')
+    insert.run(m.id, 0, 'turn/start', 1, JSON.stringify({ turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } }))
+    insert.run(m.id, 1, 'request/header-delta', 2, JSON.stringify({ config: { model: 'legacy' } }))
+    insert.run(m.id, 2, 'turn/end', 3, JSON.stringify({ turn: 1, reason: { kind: 'completed' } }))
+    db.close()
+
+    const mounted = await backend(path)
+    await expect(mounted.ctx.sessionPersistence.load(m.id)).rejects.toThrow(/unsupported legacy request\/header-delta event at seq 1/)
+    await mounted.dispose()
+  })
+
+  it('rejects a stored v0 full header carrying the legacy fallback reason', async () => {
+    const path = await freshDbPath()
+    const m = meta('legacy-header-fallback', '/legacy')
+    const db = openDatabase(path, 'wal')
+    db.prepare('INSERT INTO sessions (id, version, created_at, cwd, parent_session, seed_length) VALUES (?, ?, ?, ?, NULL, NULL)')
+      .run(m.id, m.version, m.createdAt, m.cwd ?? null)
+    db.prepare('INSERT INTO events (session_id, seq, type, time, data) VALUES (?, ?, ?, ?, ?)')
+      .run(m.id, 0, 'request/header', 1, JSON.stringify({
+        header: { config: { model: 'legacy' } },
+        reason: 'fallback',
+      }))
+    db.close()
+
+    const mounted = await backend(path)
+    await expect(mounted.ctx.sessionPersistence.load(m.id))
+      .rejects.toThrow(/unsupported legacy request\/header reason "fallback" at seq 0/)
+    await mounted.dispose()
+  })
+
+  it('has no independent per-session log location', async () => {
+    const { ctx, dispose } = await backend()
+    expect(ctx.sessionPersistence.locate(meta('sqlite-location'))).toBeUndefined()
+    await dispose()
+  })
+
   it('an interrupted turn (rows after the last turn/end) is PRESERVED and closed during load', async () => {
     const path = await freshDbPath()
     const m = meta('crash')
@@ -471,7 +513,7 @@ describe('surface field round-trip', () => {
     const session = ctx.sessions.create(SessionId('roundtrip-surface'))
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
     session.append('user/message', { content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [0] })
+    session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [0] })
     session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
     await ctx.parallel('session/flush', session)
     const loaded = await ctx.sessionPersistence.load(SessionId('roundtrip-surface'))
diff --git a/packages/session-persistence/session-persistence/README.md b/packages/session-persistence/session-persistence/README.md
index f09a21251b..3390b2cc30 100644
--- a/packages/session-persistence/session-persistence/README.md
+++ b/packages/session-persistence/session-persistence/README.md
@@ -8,6 +8,7 @@ The persisted unit IS the existing `SessionEvent` (event-sourced model — the l
 
 | Method | Contract |
 |---|---|
+| `locate(meta): SessionLocation \| undefined` | Resolve an absolute per-session artifact target without I/O or materialization. Backends without an independent local artifact return `undefined`. |
 | `create(meta): Promise<void>` | Register a new session's metadata. MAY defer the physical write until the first `append` (lazy materialization). |
 | `append(id, events): Promise<void>` | Durably persist a batch (from the `session/flush` drain). Append-only; first event `seq` == stored next-seq after any repair; rejects non-JSON-serializable data naming the offending type. |
 | `load(id): Promise<{ meta; events }>` | Reload meta + log. Preserves an interrupted (unclosed) final turn and closes it with synthetic closers — an error `tool/result` per unanswered `tool-call`, then `step/end?`+`turn/end {interrupted}` (a turn can be huge — never truncated); only a torn tail fragment is dropped. Events contiguous (`events[i].seq === i`); rejects a committed-region gap/parse error or unknown `version`. |
@@ -24,6 +25,10 @@ The persisted unit IS the existing `SessionEvent` (event-sourced model — the l
 
 `PersistenceCoordinator` owns per-id state, write-behind buffers and serialization, the `session/event` → `session/flush` drain, lazy materialization, crash-tail repair, session adoption, and quiescent disposal. A first-party backend composes one, implements the small `PersistenceBackend` storage hook interface, and delegates its four public service methods. JSONL and SQLite therefore share lifecycle correctness while retaining different storage primitives; see the [coordinator RFC](../../../docs/rfc/implemented/architecture/2026-06-18-shared-persistence-write-coordinator.md).
 
+When a live session emits `session/disposed`, the coordinator waits for its initialization, serializes a final buffer drain, then releases every map entry owned by that exact `Session` object. A failed final drain keeps the pending buffer for backend teardown to retry. Backend teardown stops event admission first, awaits all in-flight session retirements and remaining per-id operations, drains any retained buffers, and only then closes the storage handle.
+
+The side-effect-free `locate` query remains backend-owned because it describes storage topology rather than write orchestration.
+
 The `PersistenceBackend<TornMarker>` hooks (the only seam between the coordinator and storage):
 
 | Hook | Role |
@@ -44,9 +49,9 @@ Import `runPersistenceContract` from `tests/contract.ts` (the public-API contrac
 
 Three backends run these suites: an in-memory reference (in `tests/`), `dsh-session-persistence-jsonl` (append-only file log) and `dsh-session-persistence-sqlite` (`node:sqlite`, each `SessionEvent` one row `(session_id, seq, type, time, data, source_event_seqs, surface_op)`). All passing the same contract + coordinator suite is the proof that the seam is genuinely backend-agnostic — lazy materialization, crash-tail-on-load, and contiguous-seq hold identically over file bytes and over a transactional store.
 
-## Metadata types
+## Metadata and location types
 
-Re-exported from `dsh-session`: `SessionHeader` (immutable session metadata: `version`, `id`, `createdAt`, `cwd?`, `parentSession?`, `seedLength?`).
+Re-exported from `dsh-session`: `SessionHeader` (immutable session metadata: `version`, `id`, `createdAt`, `cwd?`, `parentSession?`, `seedLength?`). `SessionLocation` is `{ readonly kind: string; readonly path: string }`; its path is an absolute backend target, not proof that the artifact exists or contains an unflushed turn.
 
 ## Model Experience
 
diff --git a/packages/session-persistence/session-persistence/src/coordinator.ts b/packages/session-persistence/session-persistence/src/coordinator.ts
index 4a2fed8d34..beb5fca483 100644
--- a/packages/session-persistence/session-persistence/src/coordinator.ts
+++ b/packages/session-persistence/session-persistence/src/coordinator.ts
@@ -118,6 +118,20 @@ function seedCoversPrefix(seed: readonly SessionEvent[], prefix: readonly Sessio
     })
 }
 
+/** Reject events from an obsolete v0 vocabulary that this build cannot replay. */
+function assertSupportedEvents(events: readonly SessionEvent[], id: SessionId): void {
+  const legacyType: string = 'request/header-delta'
+  const legacy = events.find(event => event.type === legacyType)
+  if (legacy !== undefined) {
+    throw new Error(`session "${id}" contains unsupported legacy request/header-delta event at seq ${legacy.seq}`)
+  }
+  const fallback = events.find(event => event.type === 'request/header'
+    && (event.data as { reason?: string }).reason === 'fallback')
+  if (fallback !== undefined) {
+    throw new Error(`session "${id}" contains unsupported legacy request/header reason "fallback" at seq ${fallback.seq}`)
+  }
+}
+
 /**
  * Owns the backend-agnostic session write-path orchestration. A backend
  * constructs one (`new PersistenceCoordinator(ctx, this)`), implements
@@ -126,7 +140,8 @@ function seedCoversPrefix(seed: readonly SessionEvent[], prefix: readonly Sessio
  *
  * All per-id operations are serialized (a per-id promise chain) so concurrent
  * flushes / a flush racing a load never interleave storage writes. The
- * constructor installs the write-path listeners and the dispose effect.
+ * constructor installs the write-path listeners, per-session retirement, and
+ * the backend dispose effect.
  *
  * @typeParam TornMarker - the backend's opaque torn-tail repair token.
  */
@@ -146,6 +161,8 @@ export class PersistenceCoordinator<TornMarker = unknown> {
    * observation boundary; callers do not inspect this bookkeeping directly.
    */
   private inits = new Map<Session, Promise<void>>()
+  /** Final drains started by fire-and-forget session disposal notifications. */
+  private retirements = new Set<Promise<void>>()
 
   constructor(private ctx: Context, private backend: PersistenceBackend<TornMarker>) {
     this.installWritePath()
@@ -204,6 +221,11 @@ export class PersistenceCoordinator<TornMarker = unknown> {
   }
 
   private async appendCore(id: SessionId, events: readonly SessionEvent[]): Promise<void> {
+    // Every append route converges here: the public service, live write-behind
+    // drains, and HMR seed/suffix adoption. Keep vocabulary rejection at that
+    // shared boundary so a stale JavaScript plugin cannot persist an event that
+    // this same backend will refuse to load.
+    assertSupportedEvents(events, id)
     if (events.length === 0) return
     let state = this.states.get(id)
     if (state === undefined) state = await this.adopt(id) // calls loadCore, not load
@@ -238,6 +260,7 @@ export class PersistenceCoordinator<TornMarker = unknown> {
     if (stored === undefined) throw new Error(`session "${id}" not found`)
     const { meta, events, tornMarker } = stored
     this.assertVersion(meta)
+    assertSupportedEvents(events, id)
 
     // Preserve complete interrupted events and synthesize only missing closers.
     const closers = interruptedTurnClosers(events)
@@ -267,7 +290,13 @@ export class PersistenceCoordinator<TornMarker = unknown> {
     const next = prior.then(op, op)
     // Keep the chain alive but swallow this op's rejection for the NEXT waiter
     // (the caller still sees the real rejection via `next`).
-    this.chains.set(id, next.then(() => undefined, () => undefined))
+    const tail = next.then(() => undefined, () => undefined)
+    this.chains.set(id, tail)
+    // Settled tails carry no serialization value. Delete only the exact tail
+    // installed above: a later operation may already have replaced it.
+    void tail.then(() => {
+      if (this.chains.get(id) === tail) this.chains.delete(id)
+    })
     return next
   }
 
@@ -293,27 +322,12 @@ export class PersistenceCoordinator<TornMarker = unknown> {
   private installWritePath(): void {
     const ctx = this.ctx
 
-    // Capture the header on creation; persist a fork's seed once. Record the init
-    // promise so flush/dispose can await it (onCreated is async).
-    ctx.on('session/created', (session) => { void this.initFor(session) })
-
-    // Session emits an owned frozen event. Keep a persistence-owned copy anyway
-    // so the write-behind queue owns exactly the record it will flush rather than
-    // retaining a product-layer record by identity. Serializability is guaranteed
-    // at the source, so structuredClone is safe.
-    ctx.on('session/event', (session, event) => {
-      let buffer = this.buffers.get(session)
-      if (!buffer) this.buffers.set(session, buffer = [])
-      buffer.push(structuredClone(event))
-    })
-
-    // Drain to the backend at the durability checkpoint.
-    ctx.on('session/flush', session => this.flush(session))
-
-    // Dispose must reach quiescence: await every init + final drain BEFORE
-    // returning, then close the backend's own resources (AFTER the drain), so no
-    // write lands after teardown and a close failure never MASKS a drain error.
+    // Register the disposer BEFORE the listeners. Cordis tears effects down in
+    // reverse registration order, so event admission closes before this final
+    // drain reaches quiescence and closes the backend.
     ctx.effect(() => async () => {
+      await this.awaitRetirements()
+
       let disposeError: unknown
       try {
         const errors = [
@@ -341,11 +355,63 @@ export class PersistenceCoordinator<TornMarker = unknown> {
       }
     }, `${this.backend.name} write path`)
 
+    // Capture the header on creation; persist a fork's seed once. Record the init
+    // promise so flush/dispose can await it (onCreated is async).
+    ctx.on('session/created', (session) => { void this.initFor(session) })
+
+    // Session emits an owned frozen event. Keep a persistence-owned copy anyway
+    // so the write-behind queue owns exactly the record it will flush rather than
+    // retaining a product-layer record by identity. Serializability is guaranteed
+    // at the source, so structuredClone is safe.
+    ctx.on('session/event', (session, event) => {
+      let buffer = this.buffers.get(session)
+      if (!buffer) this.buffers.set(session, buffer = [])
+      buffer.push(structuredClone(event))
+    })
+
+    // Drain to the backend at the durability checkpoint.
+    ctx.on('session/flush', session => this.flush(session))
+
+    // Session disposal is observe-only, so the coordinator observes the
+    // detached task itself and backend teardown awaits quiescence.
+    ctx.on('session/disposed', (session) => { this.retire(session) })
+
     // HMR: a hot reload does not replay session/created, so seed existing live
     // sessions (mirrors dsh-invariants).
     for (const session of ctx.sessions.list()) void this.initFor(session)
   }
 
+  /** Start, observe, and track one disposed session's final drain. */
+  private retire(session: Session): void {
+    const task = this.retireCore(session)
+    this.retirements.add(task)
+    const settled = (): void => { this.retirements.delete(task) }
+    void task.then(settled, (error: unknown) => {
+      settled()
+      this.ctx.logger.warn(`${this.backend.name}: session "${session.id}" retirement failed: ${String(error)}`)
+    })
+  }
+
+  /** Drain and release state owned by one exact disposed Session lifecycle. */
+  private async retireCore(session: Session): Promise<void> {
+    await this.inits.get(session)
+
+    const id = session.header.id
+    await this.serialize(id, async () => {
+      await this.drain(session)
+      this.buffers.delete(session)
+      this.inits.delete(session)
+      if (this.states.get(id)?.owner === session) this.states.delete(id)
+    })
+  }
+
+  /** Await every retirement admitted before listener teardown. */
+  private async awaitRetirements(): Promise<void> {
+    while (this.retirements.size > 0) {
+      await Promise.allSettled([...this.retirements])
+    }
+  }
+
   /** Start (once) the async init for a session and remember its promise. */
   private initFor(session: Session): Promise<void> {
     const existing = this.inits.get(session)
@@ -463,6 +529,7 @@ export class PersistenceCoordinator<TornMarker = unknown> {
   private async adoptLivePrefix(session: Session, seed: readonly SessionEvent[], stored: StoredPrefix<TornMarker>): Promise<void> {
     const { meta, events, tornMarker } = stored
     this.assertVersion(meta)
+    assertSupportedEvents(events, session.header.id)
     if (!seedCoversPrefix(seed, events)) {
       throw new Error(`session "${session.header.id}" already has a persisted log on disk that does not match this live session (id collision)`)
     }
diff --git a/packages/session-persistence/session-persistence/src/index.ts b/packages/session-persistence/session-persistence/src/index.ts
index 3e6c9c21ed..3c102e6ede 100644
--- a/packages/session-persistence/session-persistence/src/index.ts
+++ b/packages/session-persistence/session-persistence/src/index.ts
@@ -21,6 +21,18 @@ declare module 'cordis' {
   }
 }
 
+/**
+ * A backend-resolved, per-session local artifact location. The path is an
+ * absolute target path and can name an artifact that has not materialized yet.
+ * Consumers must treat it as a location hint, never as an authorization token.
+ */
+export interface SessionLocation {
+  /** Backend-specific artifact kind, for example `jsonl`. */
+  readonly kind: string
+  /** Absolute path to this session's backend-owned artifact. */
+  readonly path: string
+}
+
 /**
  * Durable append-only session storage. Implementations preserve contiguous,
  * losslessly JSON-serializable events; {@link append} resolves only after
@@ -32,6 +44,15 @@ export abstract class SessionPersistence extends Service {
     super(ctx, 'sessionPersistence')
   }
 
+  /**
+   * Resolve this backend's independent local artifact for a session without
+   * reading, creating, flushing, or otherwise materializing it. Backends such
+   * as SQLite that do not own one artifact per session return `undefined`.
+   * @param meta - the immutable session header whose artifact is requested.
+   * @returns the backend-specific absolute location, when one exists.
+   */
+  abstract locate(meta: SessionHeader): SessionLocation | undefined
+
   /**
    * Register a new session's metadata. A backend MAY defer the physical write
    * until the first {@link append} (lazy materialization), in which case a
diff --git a/packages/session-persistence/session-persistence/tests/contract.ts b/packages/session-persistence/session-persistence/tests/contract.ts
index d8cd9fc230..84386c016b 100644
--- a/packages/session-persistence/session-persistence/tests/contract.ts
+++ b/packages/session-persistence/session-persistence/tests/contract.ts
@@ -36,7 +36,7 @@ export function oneTurnLog(): SessionEvent[] {
     { type: 'turn/start', seq: 0, time: 1, data: { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } } },
     { type: 'user/message', seq: 1, time: 2, data: { content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } }, surfaceOp: 'append' },
     { type: 'step/start', seq: 2, time: 3, data: { turn: 1, step: 1 } },
-    { type: 'assistant/message', seq: 3, time: 4, data: { turn: 1, step: 1, content: [{ type: 'text', text: 'hello' }] }, surfaceOp: 'append' },
+    { type: 'assistant/message', seq: 3, time: 4, data: { turn: 1, step: 1, content: [{ type: 'text', text: 'hello' }], provenance: { provider: 'mock', model: 'mock' } }, surfaceOp: 'append' },
     { type: 'step/end', seq: 4, time: 5, data: { turn: 1, step: 1 } },
     { type: 'turn/end', seq: 5, time: 6, data: { turn: 1, reason: { kind: 'completed' } } },
   ]
@@ -136,7 +136,7 @@ export function runPersistenceContract(name: string, make: () => Promise<Contrac
           { type: 'step/start', seq: 7, time: 8, data: { turn: 2, step: 1 } },
           { type: 'assistant/message', seq: 8, time: 9, data: { turn: 2, step: 1, content: [
             { type: 'tool-call', id: CallId('call-x'), name: 'bash', arguments: '{}' },
-          ] } },
+          ], provenance: { provider: 'mock', model: 'mock' } } },
         ])
 
         const loaded = await persistence.load(m.id)
diff --git a/packages/session-persistence/session-persistence/tests/coordinator-contract.ts b/packages/session-persistence/session-persistence/tests/coordinator-contract.ts
index f6692b9175..defcaefb6d 100644
--- a/packages/session-persistence/session-persistence/tests/coordinator-contract.ts
+++ b/packages/session-persistence/session-persistence/tests/coordinator-contract.ts
@@ -9,7 +9,7 @@
  * @module @deepseek-ai/dsh-session-persistence/tests/coordinator-contract
  */
 
-import { describe, expect, it } from 'vitest'
+import { describe, expect, it, vi } from 'vitest'
 import { Context, type Fiber } from 'cordis'
 import SessionStore, { SESSION_FORMAT_VERSION, SessionId } from '@deepseek-ai/dsh-session'
 import type { Session, SessionEvent } from '@deepseek-ai/dsh-session'
@@ -401,7 +401,7 @@ export function runCoordinatorContract(name: string, makeFixture: () => Promise<
       }
     })
 
-    it('does NOT reclaim an id whose abandoned owner still has buffered (unflushed) events', async () => {
+    it('session disposal drains buffered events before retiring ownership', async () => {
       const fix = await makeFixture()
       const { ctx, fiber } = await freshCtx(fix)
       try {
@@ -413,13 +413,20 @@ export function runCoordinatorContract(name: string, makeFixture: () => Promise<
         // Append a turn but do NOT flush — events sit in the write-behind buffer.
         first.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
         first.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
-        await firstFiber.dispose() // disposed before flush; not materialized, buffer pending
+        await firstFiber.dispose()
+
+        // Disposal is an observe-only notification. Poll storage rather than
+        // assuming the owning fiber awaits the coordinator's detached drain.
+        await vi.waitFor(async () => {
+          expect((await ctx.sessionPersistence.list()).map(meta => meta.id)).toContain(SessionId('buffered'))
+        })
+        expect((await ctx.sessionPersistence.load(SessionId('buffered'))).events.map(event => event.seq)).toEqual([0, 1])
 
         let reuse!: Session
         await ctx.plugin(Object.assign((inner: Context) => {
           reuse = inner.sessions.create(SessionId('buffered'), { meta: { cwd: WORK } })
         }, { inject: ['sessions'] }))
-        await expect(ctx.sessions.flush(reuse)).rejects.toThrow(/already bound to a different live session/)
+        await expect(ctx.sessions.flush(reuse)).rejects.toThrow(/persisted log|id collision/)
       } finally {
         await fiber.dispose()
         await fix.cleanup()
diff --git a/packages/session-persistence/session-persistence/tests/persistence.spec.ts b/packages/session-persistence/session-persistence/tests/persistence.spec.ts
index a937a39da7..e083ac3543 100644
--- a/packages/session-persistence/session-persistence/tests/persistence.spec.ts
+++ b/packages/session-persistence/session-persistence/tests/persistence.spec.ts
@@ -1,7 +1,7 @@
-import { describe, expect, it } from 'vitest'
+import { describe, expect, it, vi } from 'vitest'
 import { Context } from 'cordis'
 import SessionStore, { SessionId, isJsonValue } from '@deepseek-ai/dsh-session'
-import type { SessionEvent, SessionHeader } from '@deepseek-ai/dsh-session'
+import type { Session, SessionEvent, SessionHeader } from '@deepseek-ai/dsh-session'
 import {
   SessionPersistence, PersistenceCoordinator,
   type PersistenceBackend, type StoredPrefix,
@@ -12,9 +12,38 @@ import { runCoordinatorContract, type CoordinatorFixture } from './coordinator-c
 /** The durable store shape: materialized sessions only (no lazy entries). */
 type MemoryStore = Map<string, { meta: SessionHeader; events: SessionEvent[] }>
 
+/** An obsolete event fixture that emulates an untyped pre-change producer. */
+function legacyHeaderDelta(seq = 0): SessionEvent {
+  return {
+    type: 'request/header-delta',
+    seq,
+    time: 1,
+    data: { config: { model: 'legacy' } },
+  } as unknown as SessionEvent
+}
+
+/** An obsolete full-header reason fixture from the removed delta codec. */
+function legacyFallbackHeader(seq = 0): SessionEvent {
+  return {
+    type: 'request/header',
+    seq,
+    time: 1,
+    data: { header: { config: { model: 'legacy' } }, reason: 'fallback' },
+  } as unknown as SessionEvent
+}
+
 /** Optional plugin config: an EXTERNAL store shared across backend instances. */
 interface MemoryConfig { store?: MemoryStore }
 
+/** Test-only view of the coordinator containers whose retirement is the contract under test. */
+interface CoordinatorInternals {
+  states: Map<unknown, unknown>
+  buffers: Map<unknown, unknown>
+  chains: Map<unknown, unknown>
+  inits: Map<unknown, unknown>
+  retirements: Set<Promise<void>>
+}
+
 /**
  * Reference {@link PersistenceCoordinator} vehicle and abstract-service coverage, backed by a
  * dependency-free map with atomic writes and no torn-tail marker. Supplying the map lets multiple
@@ -41,6 +70,10 @@ class MemoryPersistence extends SessionPersistence implements PersistenceBackend
 
   // --- service surface (delegated to the coordinator) ---
 
+  locate(_meta: SessionHeader): undefined {
+    return undefined
+  }
+
   create(m: SessionHeader): Promise<void> {
     return this.coordinator.create(m)
   }
@@ -97,6 +130,49 @@ class MemoryPersistence extends SessionPersistence implements PersistenceBackend
   }
 }
 
+/** Controllable storage primitive for serialization and retirement failure tests. */
+class ControlledBackend implements PersistenceBackend<never> {
+  readonly name = 'session-persistence-controlled'
+  readonly store: MemoryStore = new Map()
+  readonly lifecycle: string[] = []
+  appendAttempts = 0
+  loadAttempts = 0
+  beforeAppend?: (attempt: number) => Promise<void>
+  beforeLoadStored?: (attempt: number) => Promise<void>
+
+  async loadStored(id: SessionId): Promise<StoredPrefix<never> | undefined> {
+    await this.beforeLoadStored?.(++this.loadAttempts)
+    const entry = this.store.get(id)
+    if (entry === undefined) return undefined
+    return { meta: structuredClone(entry.meta), events: structuredClone(entry.events) }
+  }
+
+  loadLive(id: SessionId, _cwd: string | undefined): Promise<StoredPrefix<never> | undefined> {
+    return this.loadStored(id)
+  }
+
+  async appendBatch(m: SessionHeader, events: readonly SessionEvent[], _isMaterialized: boolean): Promise<void> {
+    const attempt = ++this.appendAttempts
+    await this.beforeAppend?.(attempt)
+    const entry = this.store.get(m.id)
+    if (entry === undefined) {
+      this.store.set(m.id, { meta: structuredClone(m), events: structuredClone(events) as SessionEvent[] })
+    } else {
+      entry.events.push(...structuredClone(events) as SessionEvent[])
+    }
+  }
+
+  async commitRepair(_m: SessionHeader, _tornMarker: undefined, _closers: readonly SessionEvent[]): Promise<void> {}
+
+  async list(): Promise<SessionHeader[]> {
+    return [...this.store.values()].map(entry => structuredClone(entry.meta))
+  }
+
+  async close(): Promise<void> {
+    this.lifecycle.push('close')
+  }
+}
+
 // Run the shared contract against the in-memory backend.
 runPersistenceContract('memory', async () => {
   const ctx = new Context()
@@ -118,6 +194,230 @@ runCoordinatorContract('memory', async (): Promise<CoordinatorFixture> => {
   }
 })
 
+describe('PersistenceCoordinator retirement', () => {
+  it('a retiring unmaterialized owner without buffered events releases its id', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const backend = new ControlledBackend()
+    let coordinator!: PersistenceCoordinator<never>
+    const backendFiber = await ctx.plugin(Object.assign((inner: Context) => {
+      coordinator = new PersistenceCoordinator(inner, backend)
+    }, { inject: ['sessions'] }))
+    const loadGate = Promise.withResolvers<boolean>()
+
+    try {
+      const id = SessionId('retiring-lazy-owner')
+      let first!: Session
+      const firstFiber = await ctx.plugin(Object.assign((inner: Context) => {
+        first = inner.sessions.create(id)
+      }, { inject: ['sessions'] }))
+      await ctx.sessions.flush(first)
+
+      const baselineLoads = backend.loadAttempts
+      backend.beforeLoadStored = async () => { await loadGate.promise }
+      const blockingLoad = coordinator.load(id)
+      await vi.waitFor(() => { expect(backend.loadAttempts).toBe(baselineLoads + 1) })
+      await firstFiber.dispose()
+
+      let reuse!: Session
+      await ctx.plugin(Object.assign((inner: Context) => {
+        reuse = inner.sessions.create(id)
+      }, { inject: ['sessions'] }))
+      await vi.waitFor(() => { expect(backend.loadAttempts).toBe(baselineLoads + 2) })
+
+      loadGate.resolve(true)
+      await expect(blockingLoad).rejects.toThrow(/not found/)
+      await expect(ctx.sessions.flush(reuse)).resolves.toBeUndefined()
+    } finally {
+      loadGate.resolve(true)
+      await backendFiber.dispose()
+      await ctx.fiber.dispose()
+    }
+  })
+
+  it('a retiring owner with buffered events still rejects same-id reuse', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const backend = new ControlledBackend()
+    let coordinator!: PersistenceCoordinator<never>
+    const backendFiber = await ctx.plugin(Object.assign((inner: Context) => {
+      coordinator = new PersistenceCoordinator(inner, backend)
+    }, { inject: ['sessions'] }))
+    const loadGate = Promise.withResolvers<boolean>()
+
+    try {
+      const id = SessionId('retiring-buffered-owner')
+      let first!: Session
+      const firstFiber = await ctx.plugin(Object.assign((inner: Context) => {
+        first = inner.sessions.create(id)
+      }, { inject: ['sessions'] }))
+      await ctx.sessions.flush(first)
+      first.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
+      first.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
+
+      const baselineLoads = backend.loadAttempts
+      backend.beforeLoadStored = async () => { await loadGate.promise }
+      const blockingLoad = coordinator.load(id)
+      await vi.waitFor(() => { expect(backend.loadAttempts).toBe(baselineLoads + 1) })
+      await firstFiber.dispose()
+
+      let reuse!: Session
+      await ctx.plugin(Object.assign((inner: Context) => {
+        reuse = inner.sessions.create(id)
+      }, { inject: ['sessions'] }))
+      await expect(ctx.sessions.flush(reuse)).rejects.toThrow(/bound to a different live session/)
+
+      loadGate.resolve(true)
+      await expect(blockingLoad).rejects.toThrow(/not found/)
+      await vi.waitFor(() => {
+        expect(backend.store.get(id)?.events.map(event => event.seq)).toEqual([0, 1])
+      })
+    } finally {
+      loadGate.resolve(true)
+      await backendFiber.dispose()
+      await ctx.fiber.dispose()
+    }
+  })
+
+  it('a settled chain tail cannot delete a newer operation for the same id', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const backend = new ControlledBackend()
+    let coordinator!: PersistenceCoordinator<never>
+    const fiber = await ctx.plugin(Object.assign((inner: Context) => {
+      coordinator = new PersistenceCoordinator(inner, backend)
+    }, { inject: ['sessions'] }))
+    const internals = coordinator as unknown as CoordinatorInternals
+    const first = Promise.withResolvers<boolean>()
+    const second = Promise.withResolvers<boolean>()
+    backend.beforeAppend = async (attempt) => {
+      if (attempt === 1) await first.promise
+      if (attempt === 2) await second.promise
+    }
+
+    try {
+      const id = SessionId('chain-tail')
+      await coordinator.create(meta(id))
+      const firstAppend = coordinator.append(id, [{
+        type: 'turn/start',
+        seq: 0,
+        time: 1,
+        data: { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } },
+      }])
+      const secondAppend = coordinator.append(id, [{
+        type: 'turn/end',
+        seq: 1,
+        time: 2,
+        data: { turn: 1, reason: { kind: 'completed' } },
+      }])
+
+      await vi.waitFor(() => { expect(backend.appendAttempts).toBe(1) })
+      first.resolve(true)
+      await vi.waitFor(() => { expect(backend.appendAttempts).toBe(2) })
+      expect(internals.chains.size).toBe(1)
+      second.resolve(true)
+      await Promise.all([firstAppend, secondAppend])
+      await vi.waitFor(() => { expect(internals.chains.size).toBe(0) })
+      expect(backend.store.get(id)?.events.map(event => event.seq)).toEqual([0, 1])
+    } finally {
+      first.resolve(true)
+      second.resolve(true)
+      await fiber.dispose()
+      await ctx.fiber.dispose()
+    }
+  })
+
+  it('backend teardown retries a failed session retirement before close', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const backend = new ControlledBackend()
+    let coordinator!: PersistenceCoordinator<never>
+    const backendFiber = await ctx.plugin(Object.assign((inner: Context) => {
+      coordinator = new PersistenceCoordinator(inner, backend)
+    }, { inject: ['sessions'] }))
+    const internals = coordinator as unknown as CoordinatorInternals
+    backend.beforeAppend = async (attempt) => {
+      if (attempt === 1) {
+        backend.lifecycle.push('append-failed')
+        throw new Error('transient append failure')
+      }
+      backend.lifecycle.push('append-committed')
+    }
+
+    try {
+      let session!: Session
+      const sessionFiber = await ctx.plugin(Object.assign((inner: Context) => {
+        session = inner.sessions.create(SessionId('retry-retirement'))
+      }, { inject: ['sessions'] }))
+      session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
+      session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
+      await sessionFiber.dispose()
+
+      await vi.waitFor(() => {
+        expect(backend.appendAttempts).toBe(1)
+        expect(internals.retirements.size).toBe(0)
+      })
+      expect([...internals.buffers.values()]).toEqual([expect.arrayContaining([
+        expect.objectContaining({ seq: 0 }),
+        expect.objectContaining({ seq: 1 }),
+      ])])
+
+      await backendFiber.dispose()
+      expect(backend.store.get(SessionId('retry-retirement'))?.events.map(event => event.seq)).toEqual([0, 1])
+      expect(backend.lifecycle).toEqual(['append-failed', 'append-committed', 'close'])
+    } finally {
+      await backendFiber.dispose()
+      await ctx.fiber.dispose()
+    }
+  })
+
+  it('backend teardown waits for an in-flight session retirement before close', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const backend = new ControlledBackend()
+    let coordinator!: PersistenceCoordinator<never>
+    const backendFiber = await ctx.plugin(Object.assign((inner: Context) => {
+      coordinator = new PersistenceCoordinator(inner, backend)
+    }, { inject: ['sessions'] }))
+    const internals = coordinator as unknown as CoordinatorInternals
+    const appendGate = Promise.withResolvers<boolean>()
+    backend.beforeAppend = async () => {
+      backend.lifecycle.push('append-started')
+      await appendGate.promise
+      backend.lifecycle.push('append-committed')
+    }
+
+    try {
+      let session!: Session
+      const sessionFiber = await ctx.plugin(Object.assign((inner: Context) => {
+        session = inner.sessions.create(SessionId('inflight-retirement'))
+      }, { inject: ['sessions'] }))
+      session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
+      session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
+      await sessionFiber.dispose()
+      await vi.waitFor(() => {
+        expect(backend.appendAttempts).toBe(1)
+        expect(internals.retirements.size).toBe(1)
+      })
+
+      let disposed = false
+      const teardown = backendFiber.dispose().then(() => { disposed = true })
+      await Promise.resolve()
+      expect(disposed).toBe(false)
+      expect(backend.lifecycle).toEqual(['append-started'])
+
+      appendGate.resolve(true)
+      await teardown
+      expect(backend.store.get(SessionId('inflight-retirement'))?.events.map(event => event.seq)).toEqual([0, 1])
+      expect(backend.lifecycle).toEqual(['append-started', 'append-committed', 'close'])
+    } finally {
+      appendGate.resolve(true)
+      await backendFiber.dispose()
+      await ctx.fiber.dispose()
+    }
+  })
+})
+
 describe('SessionPersistence service registration', () => {
   it('registers as ctx.sessionPersistence and is removed on fiber dispose (HMR safety)', async () => {
     const ctx = new Context()
@@ -151,4 +451,95 @@ describe('SessionPersistence service registration', () => {
       .rejects.toThrow('session metadata must be losslessly JSON-serializable')
     await fiber.dispose()
   })
+
+  it('rejects a legacy header delta from a pre-change live producer', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const fiber = await ctx.plugin(MemoryPersistence)
+    const session = ctx.sessions.create(SessionId('legacy-live'), { meta: { cwd: '/legacy' } })
+    // Model the runtime shape available to JavaScript or a hot-loaded plugin
+    // compiled against the obsolete event vocabulary.
+    const appendLegacy = session.append.bind(session) as (type: string, data: unknown) => SessionEvent
+    expect(() => appendLegacy('request/header-delta', { config: { model: 'legacy' } }))
+      .toThrow(/unsupported legacy request\/header-delta format/)
+    expect(session.events).toHaveLength(0)
+    await fiber.dispose()
+  })
+
+  it('rejects a legacy fallback header buffered by a pre-change live producer', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const fiber = await ctx.plugin(MemoryPersistence)
+    const session = ctx.sessions.create(SessionId('legacy-fallback-live'), { meta: { cwd: '/legacy' } })
+    const appendLegacy = session.append.bind(session) as (type: string, data: unknown) => SessionEvent
+
+    expect(() => appendLegacy('request/header', legacyFallbackHeader().data))
+      .toThrow('unsupported legacy request/header reason "fallback"')
+    expect(session.events).toHaveLength(0)
+    await fiber.dispose()
+  })
+
+  it('rejects a legacy stored prefix during live HMR adoption', async () => {
+    const id = SessionId('legacy-hmr')
+    const m = meta(id, '/legacy')
+    const legacy = legacyHeaderDelta()
+    const store: MemoryStore = new Map([[id, { meta: m, events: [legacy] }]])
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    // A current live session cannot carry the obsolete event in its seed, but
+    // HMR still has to identify the persisted prefix as unsupported rather than
+    // treating it as an ordinary live-prefix collision.
+    const session = ctx.sessions.create(id, { meta: { cwd: '/legacy' } })
+    const fiber = await ctx.plugin(MemoryPersistence, { store })
+
+    await expect(ctx.sessions.flush(session))
+      .rejects.toThrow(/unsupported legacy request\/header-delta event at seq 0/)
+    await Promise.allSettled([fiber.dispose()])
+  })
+
+  it('rejects a stored legacy fallback header during load', async () => {
+    const id = SessionId('legacy-fallback-load')
+    const m = meta(id, '/legacy')
+    const store: MemoryStore = new Map([[id, { meta: m, events: [legacyFallbackHeader()] }]])
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const fiber = await ctx.plugin(MemoryPersistence, { store })
+
+    await expect(ctx.sessionPersistence.load(id))
+      .rejects.toThrow('unsupported legacy request/header reason "fallback" at seq 0')
+    await fiber.dispose()
+  })
+
+  it('retires all coordinator bookkeeping for disposed sessions', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SessionStore)
+    const fiber = await ctx.plugin(MemoryPersistence)
+    const { coordinator } = ctx.sessionPersistence as unknown as { coordinator: CoordinatorInternals }
+
+    try {
+      for (let index = 0; index < 3; index += 1) {
+        let session!: Session
+        const sessionFiber = await ctx.plugin(Object.assign((inner: Context) => {
+          session = inner.sessions.create(SessionId(`disposed-${index}`))
+        }, { inject: ['sessions'] }))
+        session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
+        session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
+        await ctx.sessions.flush(session)
+        await sessionFiber.dispose()
+      }
+
+      await vi.waitFor(() => {
+        expect(ctx.sessions.list()).toHaveLength(0)
+        expect({
+          states: coordinator.states.size,
+          buffers: coordinator.buffers.size,
+          chains: coordinator.chains.size,
+          inits: coordinator.inits.size,
+          retirements: coordinator.retirements.size,
+        }).toEqual({ states: 0, buffers: 0, chains: 0, inits: 0, retirements: 0 })
+      })
+    } finally {
+      await fiber.dispose()
+    }
+  })
 })
diff --git a/packages/session-query/README.md b/packages/session-query/README.md
index 8b0c06a30c..4c4b1c75c4 100644
--- a/packages/session-query/README.md
+++ b/packages/session-query/README.md
@@ -1,9 +1,9 @@
 # session-query/ — session retrieval capability family
 
-Trusted exact reads over live and durable session logs. Phase one contains one interface package that owns `ctx.sessionQuery`, logical-corpus precedence, surface classification, and bounded event reads.
+Trusted exact reads and relationship traces over live and durable session logs. The family contains one interface package that owns `ctx.sessionQuery`, logical-corpus precedence, surface classification, bounded event reads, lineage, and direct event relationships.
 
 | Package | Role | ctx key |
 |---|---|---|
-| [`session-query/`](session-query/README.md) | Logical-corpus and exact-event read service | `ctx.sessionQuery` |
+| [`session-query/`](session-query/README.md) | Logical-corpus exact-read and relationship-tracing service | `ctx.sessionQuery` |
 
-The family is independent of compaction: it reads the canonical session log but does not participate in compaction policy or execution. Full-text search remains proposed as a phase-two SQLite package rather than a speculative provider seam in this interface package.
+The family is independent of compaction: it reads canonical lineage, surface operations, and logged provenance but does not participate in compaction policy or execution. Full-text search remains a proposed SQLite package rather than a speculative provider seam in this interface package.
diff --git a/packages/session-query/session-query/README.md b/packages/session-query/session-query/README.md
index f2293cbbee..269ba51e3e 100644
--- a/packages/session-query/session-query/README.md
+++ b/packages/session-query/session-query/README.md
@@ -1,16 +1,20 @@
 # @deepseek-ai/dsh-session-query
 
-Exact session-history retrieval through `ctx.sessionQuery`. The service presents live `ctx.sessions` and an optional, dynamically mounted `ctx.sessionPersistence` as one logical corpus. Matching ids produce one record: live events win, while `live` and `persisted` report both source availabilities. Conflicting immutable headers fail with `SESSION_QUERY_SOURCE_CONFLICT`.
+Exact session-history retrieval and relationship tracing through `ctx.sessionQuery`. The service presents live `ctx.sessions` and an optional, dynamically mounted `ctx.sessionPersistence` as one logical corpus. Matching ids produce one record: live events win, while `live` and `persisted` report both source availabilities. Conflicting immutable headers fail with `SESSION_QUERY_SOURCE_CONFLICT`.
 
 ## Reads
 
 - `listSessions()` reads current persistence metadata, merges live records with live precedence, and returns cloned records in deterministic newest-first order.
 - `listEvents(sessionId)` loads the live-preferred raw log and classifies each event as `current`, `shadowed`, or `log-only` with the shared `dsh-session` surface fold.
 - `readEvent(request)` returns a cloned header, the full target event, and a bounded raw-seq window. `before` and `after` default to zero and may not exceed `readWindowMax`.
+- `traceSession(sessionId)` reads the corpus once and returns immediate-to-outward ancestors plus deterministic recursive descendant trees. `complete: false` identifies the first missing parent; a target-connected cycle fails with `SESSION_QUERY_INVALID_LINEAGE`.
+- `traceEvent(request)` loads the logical log once and returns direct positional replacements and direct logged provenance. `replacementChain` follows positional replacers to the final replacement; provenance links remain non-transitive.
 
-Persistence is optional and may mount or unmount dynamically. A cross-corpus list fails with `SESSION_QUERY_PERSISTENCE_FAILED` while mounted persistence is unreadable. A read targeting a known live session does not consult persistence, so durable backend health cannot make current in-memory history unreadable. Persisted exact reads list before loading, and reject a metadata mismatch rather than combining inconsistent observations.
+Persistence is optional and may mount or unmount dynamically. Cross-corpus listing and lineage tracing fail with `SESSION_QUERY_PERSISTENCE_FAILED` while mounted persistence is unreadable. An event read or trace targeting a known live session does not consult persistence, so durable backend health cannot make current in-memory history unreadable. Persisted event operations list before loading and reject a metadata mismatch rather than combining inconsistent observations.
 
-`SessionQueryError.code` is a closed union: `SESSION_QUERY_EVENT_NOT_FOUND`, `SESSION_QUERY_INVALID_CONFIG`, `SESSION_QUERY_INVALID_SURFACE`, `SESSION_QUERY_INVALID_WINDOW`, `SESSION_QUERY_PERSISTENCE_FAILED`, `SESSION_QUERY_SESSION_NOT_FOUND`, and `SESSION_QUERY_SOURCE_CONFLICT`.
+`listEvents()` and `traceEvent()` run the same one-pass `dsh-session` surface fold. A loaded log is valid only when event seqs are zero-based and contiguous, surface markers obey event-type eligibility, provenance arrays are nonempty and duplicate-free, references name earlier events, and each positional replacement names and cites every surface node it removes; every violation fails with `SESSION_QUERY_INVALID_SURFACE`.
+
+`SessionQueryError.code` is a closed union: `SESSION_QUERY_EVENT_NOT_FOUND`, `SESSION_QUERY_INVALID_CONFIG`, `SESSION_QUERY_INVALID_LINEAGE`, `SESSION_QUERY_INVALID_SURFACE`, `SESSION_QUERY_INVALID_WINDOW`, `SESSION_QUERY_PERSISTENCE_FAILED`, `SESSION_QUERY_SESSION_NOT_FOUND`, and `SESSION_QUERY_SOURCE_CONFLICT`.
 
 ## Configuration
 
@@ -25,4 +29,4 @@ None, as this trusted query service returns cloned session records only to its c
 ## Known Limitations and Deferred Work
 
 - **No caller authorization** — this is trusted context-wide infrastructure; a future model tool or UI must constrain which sessions its caller may inspect.
-- **Exact retrieval only** — filters, lineage/provenance traversal, extraction, search-provider protocol, index synchronization, and a model-facing tool are absent. Full-text search belongs beside its first implementation; the proposed SQLite package and its single transaction/reconciliation owner are described in the [phase-two RFC](../../../docs/rfc/proposed/feature/2026-07-10-sqlite-session-query-provider.md).
+- **No search or extraction** — filters, extraction registry, search-provider protocol, index synchronization, and a model-facing tool are absent. The [tracing decision](../../../docs/rfc/implemented/feature/2026-07-13-session-query-tracing.md) owns relationship semantics; content-bearing full-text-search results and their chainable filters belong beside their first implementation in the proposed [SQLite package](../../../docs/rfc/proposed/feature/2026-07-10-sqlite-session-query-provider.md).
diff --git a/packages/session-query/session-query/package.json b/packages/session-query/session-query/package.json
index e87327de13..ae2767598a 100644
--- a/packages/session-query/session-query/package.json
+++ b/packages/session-query/session-query/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@deepseek-ai/dsh-session-query",
-  "description": "Live-preferred exact session-history retrieval service (ctx.sessionQuery)",
+  "description": "Live-preferred exact session-history retrieval and tracing service (ctx.sessionQuery)",
   "version": "0.0.1",
   "private": true,
   "type": "module",
diff --git a/packages/session-query/session-query/src/config.ts b/packages/session-query/session-query/src/config.ts
index 2736f68cbd..4a15366c2c 100644
--- a/packages/session-query/session-query/src/config.ts
+++ b/packages/session-query/session-query/src/config.ts
@@ -5,16 +5,17 @@ import { HarnessError } from '@deepseek-ai/dsh-llm'
 /** Default maximum `before`/`after` raw-event window. */
 export const SESSION_QUERY_READ_WINDOW_MAX = 50
 
-/** Configuration for exact session-query reads. */
+/** Configuration for exact session-query reads and traces. */
 export interface Config {
   /** Maximum accepted raw read context on either side. Defaults to 50. */
   readWindowMax?: number
 }
 
-/** Stable machine-routable failure taxonomy for exact session reads. */
+/** Stable machine-routable failure taxonomy for exact session reads and traces. */
 export type SessionQueryErrorCode =
   | 'SESSION_QUERY_EVENT_NOT_FOUND'
   | 'SESSION_QUERY_INVALID_CONFIG'
+  | 'SESSION_QUERY_INVALID_LINEAGE'
   | 'SESSION_QUERY_INVALID_SURFACE'
   | 'SESSION_QUERY_INVALID_WINDOW'
   | 'SESSION_QUERY_PERSISTENCE_FAILED'
diff --git a/packages/session-query/session-query/src/index.ts b/packages/session-query/session-query/src/index.ts
index 828fe2ec88..bd35b51442 100644
--- a/packages/session-query/session-query/src/index.ts
+++ b/packages/session-query/session-query/src/index.ts
@@ -1,17 +1,19 @@
 /**
- * Exact session-history reads over live and optionally persisted logs.
+ * Exact session-history reads and traces over live and optionally persisted logs.
  *
  * @module @deepseek-ai/dsh-session-query
  */
 
 import { Context, Service } from 'cordis'
 import z from 'schemastery'
-import { foldSurface } from '@deepseek-ai/dsh-session'
-import type { SessionEvent, SessionId } from '@deepseek-ai/dsh-session'
+import type { SessionId } from '@deepseek-ai/dsh-session'
 import type {
   SessionEventReadRequest,
   SessionEventRecord,
+  SessionEventTrace,
+  SessionEventTraceRequest,
   SessionEventWindow,
+  SessionLineageTrace,
   SessionRecord,
 } from './types.ts'
 import {
@@ -20,6 +22,7 @@ import {
   type Config,
 } from './config.ts'
 import { SessionCorpus } from './corpus.ts'
+import * as tracing from './tracing.ts'
 
 export type * from './types.ts'
 export type { Config, SessionQueryErrorCode } from './config.ts'
@@ -31,7 +34,7 @@ declare module 'cordis' {
   }
 }
 
-/** Live-preferred logical-corpus and exact-event read service. */
+/** Live-preferred logical-corpus exact-read and relationship-tracing service. */
 export class SessionQueryService extends Service {
   static inject = ['sessions']
   static Config: z<Config> = z.object({
@@ -68,7 +71,29 @@ export class SessionQueryService extends Service {
    */
   async listEvents(sessionId: SessionId): Promise<SessionEventRecord[]> {
     const loaded = await this._corpus.load(sessionId)
-    return eventRecords(sessionId, loaded.events)
+    return tracing.eventRecords(sessionId, loaded.events)
+  }
+
+  /**
+   * Trace known ancestry and descendants from one corpus observation.
+   * @param sessionId - logical session id to trace.
+   * @returns a complete lineage or an explicit unresolved parent boundary.
+   * @throws when corpus resolution fails, the target is absent, or its known ancestry cycles.
+   */
+  async traceSession(sessionId: SessionId): Promise<SessionLineageTrace> {
+    const records = await this._corpus.listSessions()
+    return tracing.traceSession(records, sessionId)
+  }
+
+  /**
+   * Trace one event's direct positional and provenance relationships.
+   * @param request - target session id and event seq.
+   * @returns direct links plus the target's positional replacement chain.
+   * @throws when source resolution fails, the target is absent, or surface/provenance validation fails.
+   */
+  async traceEvent(request: SessionEventTraceRequest): Promise<SessionEventTrace> {
+    const loaded = await this._corpus.load(request.sessionId)
+    return tracing.traceEvent(request.sessionId, loaded.events, request.seq)
   }
 
   /**
@@ -110,27 +135,4 @@ export class SessionQueryService extends Service {
   }
 }
 
-function eventRecords(sessionId: SessionId, events: readonly SessionEvent[]): SessionEventRecord[] {
-  let folded: ReturnType<typeof foldSurface>
-  try {
-    folded = foldSurface(events)
-  } catch (error: unknown) {
-    throw new SessionQueryError(
-      /* v8 ignore next -- foldSurface throws Error instances */
-      `invalid session surface: ${error instanceof Error ? error.message : 'unknown error'}`,
-      'SESSION_QUERY_INVALID_SURFACE',
-      { cause: error },
-    )
-  }
-  const current = new Set(folded.nodes.map(node => node.seq))
-  const shadowed = new Set(folded.replacements.flatMap(replacement => replacement.shadowedSeqs))
-  return events.map(event => ({
-    sessionId,
-    seq: event.seq,
-    type: event.type,
-    time: event.time,
-    surface: current.has(event.seq) ? 'current' : shadowed.has(event.seq) ? 'shadowed' : 'log-only',
-  }))
-}
-
 export default SessionQueryService
diff --git a/packages/session-query/session-query/src/tracing.ts b/packages/session-query/session-query/src/tracing.ts
new file mode 100644
index 0000000000..82d9f12852
--- /dev/null
+++ b/packages/session-query/session-query/src/tracing.ts
@@ -0,0 +1,222 @@
+/** One-shot session-lineage and event-relationship tracing helpers. */
+
+import { foldSurface } from '@deepseek-ai/dsh-session'
+import type { SessionEvent, SessionId, SurfaceEventType } from '@deepseek-ai/dsh-session'
+import { SessionQueryError } from './config.ts'
+import type {
+  SessionEventRecord,
+  SessionEventTrace,
+  SessionLineageNode,
+  SessionLineageTrace,
+  SessionRecord,
+} from './types.ts'
+
+interface EventLogAnalysis {
+  records: SessionEventRecord[]
+  replacedBy: Map<number, number>
+  replacedEventSeqs: Map<number, number[]>
+}
+
+/**
+ * Classify a raw event log with one canonical surface fold.
+ * @param sessionId - owner of the event log.
+ * @param events - detached raw event log.
+ * @returns lightweight records in ascending log order.
+ */
+export function eventRecords(
+  sessionId: SessionId,
+  events: readonly SessionEvent[],
+): SessionEventRecord[] {
+  return analyzeEventLog(sessionId, events).records
+}
+
+/**
+ * Trace one target after one canonical surface fold and whole-log validation.
+ * @param sessionId - owner of the event log.
+ * @param events - detached raw event log.
+ * @param seq - target event seq.
+ * @returns direct surface and provenance relationships.
+ */
+export function traceEvent(
+  sessionId: SessionId,
+  events: readonly SessionEvent[],
+  seq: number,
+): SessionEventTrace {
+  const target = events[seq]
+  if (target === undefined || target.seq !== seq) {
+    throw new SessionQueryError(
+      `session "${sessionId}" has no event at seq ${seq}`,
+      'SESSION_QUERY_EVENT_NOT_FOUND',
+    )
+  }
+
+  const analysis = analyzeEventLog(sessionId, events)
+
+  const replacementChain: number[] = []
+  let replacement = analysis.replacedBy.get(seq)
+  while (replacement !== undefined) {
+    replacementChain.push(replacement)
+    replacement = analysis.replacedBy.get(replacement)
+  }
+
+  const derivedEventSeqs: number[] = []
+  for (const event of events) {
+    if (event.seq <= seq) continue
+    if (eventSources(event).includes(seq)) derivedEventSeqs.push(event.seq)
+  }
+
+  // The target check above proves the parallel record exists at this index.
+  // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+  const targetRecord = analysis.records[seq]!
+  const replacedBy = analysis.replacedBy.get(seq)
+  return {
+    target: targetRecord,
+    ...replacedBy === undefined ? {} : { replacedBy },
+    replacementChain,
+    replacedEventSeqs: analysis.replacedEventSeqs.get(seq) ?? [],
+    sourceEventSeqs: [...eventSources(target)],
+    derivedEventSeqs,
+  }
+}
+
+/**
+ * Trace one target's known ancestry and recursively known descendants.
+ * @param records - complete logical corpus from one observation.
+ * @param sessionId - target session id.
+ * @returns complete or explicitly partial lineage.
+ */
+export function traceSession(
+  records: readonly SessionRecord[],
+  sessionId: SessionId,
+): SessionLineageTrace {
+  const byId = new Map(records.map(record => [record.header.id, record]))
+  const target = byId.get(sessionId)
+  if (target === undefined) {
+    throw new SessionQueryError(
+      `session "${sessionId}" not found`,
+      'SESSION_QUERY_SESSION_NOT_FOUND',
+    )
+  }
+
+  const ancestors: SessionRecord[] = []
+  const ancestrySeen = new Set<SessionId>([sessionId])
+  let unresolvedParentId: SessionId | undefined
+  let parentId = target.header.parentSession
+  while (parentId !== undefined) {
+    if (ancestrySeen.has(parentId)) {
+      throw new SessionQueryError(
+        `session lineage contains a cycle at "${parentId}"`,
+        'SESSION_QUERY_INVALID_LINEAGE',
+      )
+    }
+    ancestrySeen.add(parentId)
+    const parent = byId.get(parentId)
+    if (parent === undefined) {
+      unresolvedParentId = parentId
+      break
+    }
+    ancestors.push(parent)
+    parentId = parent.header.parentSession
+  }
+
+  const childrenByParent = new Map<SessionId, SessionRecord[]>()
+  for (const record of records) {
+    const parent = record.header.parentSession
+    if (parent === undefined) continue
+    const children = childrenByParent.get(parent) ?? []
+    children.push(record)
+    childrenByParent.set(parent, children)
+  }
+  for (const children of childrenByParent.values()) {
+    children.sort((a, b) => a.header.createdAt - b.header.createdAt || a.header.id.localeCompare(b.header.id))
+  }
+
+  const descendants = buildDescendants(childrenByParent, sessionId)
+  const common = {
+    target: cloneRecord(target),
+    ancestors: ancestors.map(cloneRecord),
+    descendants,
+  }
+  if (unresolvedParentId !== undefined) {
+    return { ...common, complete: false, unresolvedParentId }
+  }
+  return {
+    ...common,
+    complete: true,
+    root: cloneRecord(ancestors.at(-1) ?? target),
+  }
+}
+
+function analyzeEventLog(
+  sessionId: SessionId,
+  events: readonly SessionEvent[],
+): EventLogAnalysis {
+  let folded: ReturnType<typeof foldSurface>
+  try {
+    folded = foldSurface(events)
+  } catch (error: unknown) {
+    throw new SessionQueryError(
+      /* v8 ignore next -- foldSurface throws Error instances */
+      `invalid session surface: ${error instanceof Error ? error.message : 'unknown error'}`,
+      'SESSION_QUERY_INVALID_SURFACE',
+      { cause: error },
+    )
+  }
+  const current = new Set(folded.nodes)
+  const replacedBy = new Map<number, number>()
+  const replacedEventSeqs = new Map<number, number[]>()
+  for (const replacement of folded.replacements) {
+    const removed = replacement.shadowedSeqs
+    replacedEventSeqs.set(replacement.seq, removed)
+    for (const removedSeq of removed) {
+      replacedBy.set(removedSeq, replacement.seq)
+    }
+  }
+  return {
+    records: events.map(event => ({
+      sessionId,
+      seq: event.seq,
+      type: event.type,
+      time: event.time,
+      surface: current.has(event.seq)
+        ? 'current'
+        : replacedBy.has(event.seq) ? 'shadowed' : 'log-only',
+    })),
+    replacedBy,
+    replacedEventSeqs,
+  }
+}
+
+function eventSources(event: SessionEvent): readonly number[] {
+  return (event as SessionEvent<SurfaceEventType>).sourceEventSeqs ?? []
+}
+
+function buildDescendants(
+  childrenByParent: ReadonlyMap<SessionId, readonly SessionRecord[]>,
+  sessionId: SessionId,
+): SessionLineageNode[] {
+  const descendants: SessionLineageNode[] = []
+  const stack = [{ sessionId, descendants }]
+  while (stack.length > 0) {
+    // The length guard proves a frame exists.
+    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+    const frame = stack.pop()!
+    const nodes: SessionLineageNode[] = []
+    for (const child of childrenByParent.get(frame.sessionId) ?? []) {
+      const node = { session: cloneRecord(child), descendants: [] }
+      nodes.push(node)
+      frame.descendants.push(node)
+    }
+    for (let index = nodes.length - 1; index >= 0; index -= 1) {
+      // The loop bounds prove this indexed node exists.
+      // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+      const node = nodes[index]!
+      stack.push({ sessionId: node.session.header.id, descendants: node.descendants })
+    }
+  }
+  return descendants
+}
+
+function cloneRecord(record: SessionRecord): SessionRecord {
+  return { ...record, header: structuredClone(record.header) }
+}
diff --git a/packages/session-query/session-query/src/types.ts b/packages/session-query/session-query/src/types.ts
index 5c49695dda..38f0225ee4 100644
--- a/packages/session-query/session-query/src/types.ts
+++ b/packages/session-query/session-query/src/types.ts
@@ -1,5 +1,6 @@
 /**
- * Public records for exact reads over the live-preferred logical session corpus.
+ * Public records for exact reads and relationship traces over the
+ * live-preferred logical session corpus.
  *
  * @module @deepseek-ai/dsh-session-query/types
  */
@@ -33,6 +34,61 @@ export interface SessionEventRecord {
   surface: SessionEventSurface
 }
 
+/** Recursive descendant node in a session-lineage trace. */
+export interface SessionLineageNode {
+  /** Detached logical-corpus record for this descendant. */
+  session: SessionRecord
+  /** Direct children, each carrying its own recursive descendants. */
+  descendants: SessionLineageNode[]
+}
+
+/** Known ancestry and descendants for one logical session. */
+export type SessionLineageTrace = {
+  /** Detached record for the session that was traced. */
+  target: SessionRecord
+  /** Known parents from the immediate parent outward. */
+  ancestors: SessionRecord[]
+  /** Complete known descendant trees rooted at the target's direct children. */
+  descendants: SessionLineageNode[]
+} & (
+  | {
+    /** The complete parent chain is present in the logical corpus. */
+    complete: true
+    /** Detached record at the top of the complete lineage. */
+    root: SessionRecord
+  }
+  | {
+    /** The parent chain leaves the visible logical corpus. */
+    complete: false
+    /** First parent id that is not present in the logical corpus. */
+    unresolvedParentId: SessionId
+  }
+)
+
+/** Request for direct surface and provenance relationships around one event. */
+export interface SessionEventTraceRequest {
+  /** Session that owns the target event. */
+  sessionId: SessionId
+  /** Target event seq. */
+  seq: number
+}
+
+/** Direct surface and provenance relationships for one event. */
+export interface SessionEventTrace {
+  /** Lightweight target record. */
+  target: SessionEventRecord
+  /** Immediate positional replacement event, when the target was shadowed. */
+  replacedBy?: number
+  /** Positional replacers from the immediate replacement to the final replacement. */
+  replacementChain: number[]
+  /** Surface nodes directly removed when the target itself performed a replacement. */
+  replacedEventSeqs: number[]
+  /** Direct logged provenance sources in their recorded order. */
+  sourceEventSeqs: number[]
+  /** Later events that directly name the target as a provenance source, in log order. */
+  derivedEventSeqs: number[]
+}
+
 /** Request for one event plus raw neighboring log context. */
 export interface SessionEventReadRequest {
   /** Session that owns the target event. */
diff --git a/packages/session-query/session-query/tests/session-query.spec.ts b/packages/session-query/session-query/tests/session-query.spec.ts
index 3b50feee45..f532edf168 100644
--- a/packages/session-query/session-query/tests/session-query.spec.ts
+++ b/packages/session-query/session-query/tests/session-query.spec.ts
@@ -34,6 +34,10 @@ class TestPersistence extends SessionPersistence {
     this.afterList = undefined
   }
 
+  locate(_meta: SessionHeader): undefined {
+    return undefined
+  }
+
   create(meta: SessionHeader): Promise<void> {
     TestPersistence.entries.set(meta.id, { meta: structuredClone(meta), events: [] })
     return Promise.resolve()
@@ -109,8 +113,8 @@ describe('session-query exact reads', () => {
     })
     session.append(
       'assistant/message',
-      { turn: 1, step: 1, content: [{ type: 'text', text: 'replacement' }] },
-      { surfaceOp: { op: 'replace', start: first.seq, end: first.seq } },
+      { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'replacement' }] },
+      { surfaceOp: { op: 'replace', start: first.seq, end: first.seq }, sourceEventSeqs: [first.seq] },
     )
 
     expect((await ctx.sessionQuery.listEvents(session.id)).map(record => record.surface))
@@ -227,11 +231,13 @@ describe('session-query exact reads', () => {
   it('turns malformed surfaces and direct invalid config into typed errors', async () => {
     const ctx = await liveContext()
     const session = ctx.sessions.create(SessionId('bad-surface'))
-    session.append(
-      'assistant/message',
-      { turn: 1, step: 1, content: [] },
-      { surfaceOp: { op: 'replace', start: 9, end: 9 } },
-    )
+    ;(session as unknown as { log: SessionEvent[] }).log.push({
+      type: 'assistant/message',
+      seq: 0,
+      time: 1,
+      data: { turn: 1, step: 1, content: [], provenance: { provider: 'mock', model: 'mock' } },
+      surfaceOp: { op: 'replace', start: 9, end: 9 },
+    })
     await expect(ctx.sessionQuery.listEvents(session.id))
       .rejects.toThrow(expectCode('SESSION_QUERY_INVALID_SURFACE'))
 
diff --git a/packages/session-query/session-query/tests/tracing.spec.ts b/packages/session-query/session-query/tests/tracing.spec.ts
new file mode 100644
index 0000000000..03e1adad89
--- /dev/null
+++ b/packages/session-query/session-query/tests/tracing.spec.ts
@@ -0,0 +1,426 @@
+import { describe, expect, it } from 'vitest'
+import { Context } from 'cordis'
+import SessionStore, { SESSION_FORMAT_VERSION, SessionId } from '@deepseek-ai/dsh-session'
+import type { Session, SessionEvent, SessionHeader, SessionId as SessionIdType } from '@deepseek-ai/dsh-session'
+import SessionPersistence from '@deepseek-ai/dsh-session-persistence'
+import SessionQueryService, { type SessionQueryErrorCode } from '@deepseek-ai/dsh-session-query'
+
+type MutableSessionHeader = { -readonly [K in keyof SessionHeader]: SessionHeader[K] }
+
+/** Test-only mutable view used to verify detached returned metadata. */
+function mutableHeader(value: SessionHeader): MutableSessionHeader {
+  return value
+}
+
+function header(id: string, createdAt = 1, extra: Partial<SessionHeader> = {}): SessionHeader {
+  return { version: SESSION_FORMAT_VERSION, id: SessionId(id), createdAt, ...extra }
+}
+
+function appendEvent(seq: number, sources?: number[]): SessionEvent {
+  return {
+    type: 'user/message',
+    seq,
+    time: seq + 1,
+    data: { content: [{ type: 'text', text: `event ${seq}` }], source: { kind: 'user' } },
+    surfaceOp: 'append',
+    ...sources === undefined ? {} : { sourceEventSeqs: sources },
+  }
+}
+
+class TracePersistence extends SessionPersistence {
+  static entries = new Map<SessionIdType, { meta: SessionHeader; events: SessionEvent[] }>()
+  static listCalls = 0
+  static loadCalls = 0
+  static listFailure: Error | undefined
+  static loadFailure: Error | undefined
+  static afterList: (() => void) | undefined
+
+  static reset(entries: readonly { meta: SessionHeader; events: SessionEvent[] }[] = []): void {
+    this.entries = new Map(entries.map(entry => [entry.meta.id, structuredClone(entry)]))
+    this.listCalls = 0
+    this.loadCalls = 0
+    this.listFailure = undefined
+    this.loadFailure = undefined
+    this.afterList = undefined
+  }
+
+  locate(_meta: SessionHeader): undefined {
+    return undefined
+  }
+
+  create(meta: SessionHeader): Promise<void> {
+    TracePersistence.entries.set(meta.id, { meta: structuredClone(meta), events: [] })
+    return Promise.resolve()
+  }
+
+  append(id: SessionIdType, events: readonly SessionEvent[]): Promise<void> {
+    const entry = TracePersistence.entries.get(id)
+    if (entry === undefined) return Promise.reject(new Error('missing test session'))
+    entry.events.push(...structuredClone(events))
+    return Promise.resolve()
+  }
+
+  load(id: SessionIdType): Promise<{ meta: SessionHeader; events: SessionEvent[] }> {
+    TracePersistence.loadCalls += 1
+    if (TracePersistence.loadFailure !== undefined) return Promise.reject(TracePersistence.loadFailure)
+    const entry = TracePersistence.entries.get(id)
+    if (entry === undefined) return Promise.reject(new Error('missing test session'))
+    return Promise.resolve(structuredClone(entry))
+  }
+
+  list(): Promise<SessionHeader[]> {
+    TracePersistence.listCalls += 1
+    if (TracePersistence.listFailure !== undefined) return Promise.reject(TracePersistence.listFailure)
+    const result = [...TracePersistence.entries.values()].map(entry => structuredClone(entry.meta))
+    TracePersistence.afterList?.()
+    return Promise.resolve(result)
+  }
+}
+
+async function queryContext(): Promise<Context> {
+  const ctx = new Context()
+  await ctx.plugin(SessionStore)
+  await ctx.plugin(SessionQueryService)
+  return ctx
+}
+
+function expectCode(code: SessionQueryErrorCode): Error {
+  return expect.objectContaining({ code }) as Error
+}
+
+function appendTraceEvents(session: Session): void {
+  session.append('assistant/chunk', {
+    turn: 1,
+    step: 1,
+    chunk: { type: 'text-delta', index: 0, text: 'draft' },
+  })
+  session.append(
+    'user/message',
+    { content: [{ type: 'text', text: 'original' }], source: { kind: 'user' } },
+    { surfaceOp: 'append', sourceEventSeqs: [0] },
+  )
+  session.append(
+    'assistant/message',
+    { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'summary one' }] },
+    { surfaceOp: { op: 'replace', start: 1, end: 1 }, sourceEventSeqs: [1, 0] },
+  )
+  session.append(
+    'context/message',
+    { content: [{ type: 'text', text: 'context' }], source: { kind: 'plugin', plugin: 'test' } },
+    { surfaceOp: 'append' },
+  )
+  session.append(
+    'assistant/message',
+    { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 2, content: [{ type: 'text', text: 'summary two' }] },
+    { surfaceOp: { op: 'replace', start: 2, end: 2 }, sourceEventSeqs: [0, 2] },
+  )
+}
+
+describe('session lineage tracing', () => {
+  it('returns complete ancestry, deterministic descendant trees, and detached records', async () => {
+    const ctx = await queryContext()
+    const root = ctx.sessions.create(SessionId('root'), { meta: { createdAt: 0 } })
+    const parent = ctx.sessions.create(SessionId('parent'), {
+      meta: { createdAt: 1, parentSession: root.id },
+    })
+    const target = ctx.sessions.create(SessionId('target'), {
+      meta: { createdAt: 2, parentSession: parent.id },
+    })
+    ctx.sessions.create(SessionId('b'), { meta: { createdAt: 4, parentSession: target.id } })
+    const childA = ctx.sessions.create(SessionId('a'), {
+      meta: { createdAt: 4, parentSession: target.id },
+    })
+    ctx.sessions.create(SessionId('older'), { meta: { createdAt: 3, parentSession: target.id } })
+    ctx.sessions.create(SessionId('grandchild'), {
+      meta: { createdAt: 5, parentSession: childA.id },
+    })
+
+    const trace = await ctx.sessionQuery.traceSession(target.id)
+    expect(trace.complete).toBe(true)
+    if (!trace.complete) throw new Error('expected complete lineage')
+    expect(trace.ancestors.map(record => record.header.id)).toEqual([parent.id, root.id])
+    expect(trace.root.header.id).toBe(root.id)
+    expect(trace.descendants.map(node => node.session.header.id))
+      .toEqual([SessionId('older'), SessionId('a'), SessionId('b')])
+    expect(trace.descendants[1]?.descendants.map(node => node.session.header.id))
+      .toEqual([SessionId('grandchild')])
+
+    mutableHeader(trace.target.header).createdAt = 99
+    mutableHeader(trace.ancestors[0]!.header).createdAt = 99
+    mutableHeader(trace.root.header).createdAt = 99
+    mutableHeader(trace.descendants[0]!.session.header).createdAt = 99
+    const repeated = await ctx.sessionQuery.traceSession(target.id)
+    expect(repeated.target.header.createdAt).toBe(2)
+    expect(repeated.ancestors[0]?.header.createdAt).toBe(1)
+    expect(repeated.descendants[0]?.session.header.createdAt).toBe(3)
+  })
+
+  it('represents root and unresolved-parent traces explicitly', async () => {
+    const ctx = await queryContext()
+    const root = ctx.sessions.create(SessionId('root'), { meta: { createdAt: 1 } })
+    const partial = ctx.sessions.create(SessionId('partial'), {
+      meta: { createdAt: 2, parentSession: SessionId('outside') },
+    })
+
+    await expect(ctx.sessionQuery.traceSession(root.id)).resolves.toMatchObject({
+      complete: true,
+      root: { header: { id: root.id } },
+      ancestors: [],
+    })
+    await expect(ctx.sessionQuery.traceSession(partial.id)).resolves.toMatchObject({
+      complete: false,
+      unresolvedParentId: SessionId('outside'),
+      ancestors: [],
+    })
+  })
+
+  it('rejects target-connected cycles and missing targets', async () => {
+    const ctx = await queryContext()
+    ctx.sessions.create(SessionId('a'), {
+      meta: { createdAt: 1, parentSession: SessionId('b') },
+    })
+    ctx.sessions.create(SessionId('b'), {
+      meta: { createdAt: 2, parentSession: SessionId('a') },
+    })
+
+    await expect(ctx.sessionQuery.traceSession(SessionId('a')))
+      .rejects.toThrow(expectCode('SESSION_QUERY_INVALID_LINEAGE'))
+    await expect(ctx.sessionQuery.traceSession(SessionId('missing')))
+      .rejects.toThrow(expectCode('SESSION_QUERY_SESSION_NOT_FOUND'))
+  })
+
+  it('uses one cross-corpus observation and preserves persistence failure semantics', async () => {
+    const durable = header('durable')
+    TracePersistence.reset([{ meta: durable, events: [appendEvent(0)] }])
+    const ctx = await queryContext()
+    await ctx.plugin(TracePersistence)
+
+    await expect(ctx.sessionQuery.traceSession(durable.id)).resolves.toMatchObject({
+      target: { live: false, persisted: true },
+      complete: true,
+    })
+    expect(TracePersistence.listCalls).toBe(1)
+    expect(TracePersistence.loadCalls).toBe(0)
+
+    TracePersistence.listFailure = new Error('unavailable')
+    await expect(ctx.sessionQuery.traceSession(durable.id))
+      .rejects.toThrow(expectCode('SESSION_QUERY_PERSISTENCE_FAILED'))
+  })
+
+  it('constructs deeply nested descendants without consuming the JavaScript call stack', async () => {
+    const ctx = await queryContext()
+    const root = ctx.sessions.create(SessionId('deep-0'), { meta: { createdAt: 0 } })
+    let parent = root
+    for (let depth = 1; depth < 3_000; depth += 1) {
+      parent = ctx.sessions.create(SessionId(`deep-${depth}`), {
+        meta: { createdAt: depth, parentSession: parent.id },
+      })
+    }
+
+    const trace = await ctx.sessionQuery.traceSession(root.id)
+    expect(trace.complete).toBe(true)
+    let node = trace.descendants[0]
+    for (let depth = 1; depth < 3_000; depth += 1) {
+      if (node === undefined) throw new Error(`lineage ended before depth ${depth}`)
+      if (depth === 2_999) expect(node.session.header.id).toBe(SessionId('deep-2999'))
+      node = node.descendants[0]
+    }
+    expect(node).toBeUndefined()
+  })
+})
+
+describe('session event tracing', () => {
+  it('returns direct replacement and provenance links in their contract order', async () => {
+    const ctx = await queryContext()
+    const session = ctx.sessions.create(SessionId('trace'))
+    appendTraceEvents(session)
+
+    const original = await ctx.sessionQuery.traceEvent({ sessionId: session.id, seq: 1 })
+    expect(original.target).toMatchObject({
+      sessionId: session.id,
+      seq: 1,
+      type: 'user/message',
+      surface: 'shadowed',
+    })
+    expect(original).toMatchObject({
+      replacedBy: 2,
+      replacementChain: [2, 4],
+      replacedEventSeqs: [],
+      sourceEventSeqs: [0],
+      derivedEventSeqs: [2],
+    })
+    await expect(ctx.sessionQuery.traceEvent({ sessionId: session.id, seq: 2 }))
+      .resolves.toMatchObject({
+        replacedBy: 4,
+        replacementChain: [4],
+        replacedEventSeqs: [1],
+        sourceEventSeqs: [1, 0],
+        derivedEventSeqs: [4],
+      })
+    await expect(ctx.sessionQuery.traceEvent({ sessionId: session.id, seq: 0 }))
+      .resolves.toMatchObject({
+        target: { surface: 'log-only' },
+        replacementChain: [],
+        sourceEventSeqs: [],
+        derivedEventSeqs: [1, 2, 4],
+      })
+    await expect(ctx.sessionQuery.traceEvent({ sessionId: session.id, seq: 4 }))
+      .resolves.toMatchObject({
+        replacementChain: [],
+        replacedEventSeqs: [2],
+        sourceEventSeqs: [0, 2],
+        derivedEventSeqs: [],
+      })
+  })
+
+  it('returns fresh trace arrays and target records', async () => {
+    const ctx = await queryContext()
+    const session = ctx.sessions.create(SessionId('detached'))
+    appendTraceEvents(session)
+
+    const first = await ctx.sessionQuery.traceEvent({ sessionId: session.id, seq: 2 })
+    first.target.time = -1
+    first.replacementChain.push(99)
+    first.replacedEventSeqs.push(99)
+    first.sourceEventSeqs.push(99)
+    first.derivedEventSeqs.push(99)
+    const repeated = await ctx.sessionQuery.traceEvent({ sessionId: session.id, seq: 2 })
+    expect(repeated.target.time).not.toBe(-1)
+    expect(repeated.replacementChain).toEqual([4])
+    expect(repeated.replacedEventSeqs).toEqual([1])
+    expect(repeated.sourceEventSeqs).toEqual([1, 0])
+    expect(repeated.derivedEventSeqs).toEqual([4])
+  })
+
+  it('loads persisted logs once, prefers live logs, and preserves failures and conflicts', async () => {
+    const durable = header('shared', 1, { cwd: '/same' })
+    TracePersistence.reset([{ meta: durable, events: [appendEvent(0)] }])
+    const ctx = await queryContext()
+    await ctx.plugin(TracePersistence)
+
+    await expect(ctx.sessionQuery.traceEvent({ sessionId: durable.id, seq: 0 }))
+      .resolves.toMatchObject({ target: { type: 'user/message', surface: 'current' } })
+    expect([TracePersistence.listCalls, TracePersistence.loadCalls]).toEqual([1, 1])
+
+    const live = ctx.sessions.create(durable.id, { meta: { createdAt: 1, cwd: '/same' } })
+    live.append(
+      'context/message',
+      { content: [{ type: 'text', text: 'live' }], source: { kind: 'plugin', plugin: 'test' } },
+      { surfaceOp: 'append' },
+    )
+    TracePersistence.listFailure = new Error('list unavailable')
+    TracePersistence.loadFailure = new Error('load unavailable')
+    await expect(ctx.sessionQuery.traceEvent({ sessionId: durable.id, seq: 0 }))
+      .resolves.toMatchObject({ target: { type: 'context/message' } })
+    expect([TracePersistence.listCalls, TracePersistence.loadCalls]).toEqual([1, 1])
+
+    TracePersistence.reset([{ meta: durable, events: [appendEvent(0)] }])
+    const failedCtx = await queryContext()
+    await failedCtx.plugin(TracePersistence)
+    TracePersistence.listFailure = new Error('list unavailable')
+    await expect(failedCtx.sessionQuery.traceEvent({ sessionId: durable.id, seq: 0 }))
+      .rejects.toThrow(expectCode('SESSION_QUERY_PERSISTENCE_FAILED'))
+    TracePersistence.listFailure = undefined
+    TracePersistence.loadFailure = new Error('load unavailable')
+    await expect(failedCtx.sessionQuery.traceEvent({ sessionId: durable.id, seq: 0 }))
+      .rejects.toThrow(expectCode('SESSION_QUERY_PERSISTENCE_FAILED'))
+    TracePersistence.loadFailure = undefined
+    TracePersistence.afterList = () => {
+      mutableHeader(TracePersistence.entries.get(durable.id)!.meta).cwd = '/changed'
+    }
+    await expect(failedCtx.sessionQuery.traceEvent({ sessionId: durable.id, seq: 0 }))
+      .rejects.toThrow(expectCode('SESSION_QUERY_SOURCE_CONFLICT'))
+  })
+
+  it('checks target existence before surface or provenance analysis', async () => {
+    const bad = header('bad-target')
+    const malformed: SessionEvent[] = [appendEvent(0), {
+      type: 'assistant/message',
+      seq: 1,
+      time: 2,
+      data: { turn: 1, step: 1, content: [], provenance: { provider: 'mock', model: 'mock' } },
+      surfaceOp: { op: 'replace', start: 9, end: 9 },
+      sourceEventSeqs: [],
+    }]
+    TracePersistence.reset([{ meta: bad, events: malformed }])
+    const ctx = await queryContext()
+    await ctx.plugin(TracePersistence)
+
+    await expect(ctx.sessionQuery.traceEvent({ sessionId: bad.id, seq: 9 }))
+      .rejects.toThrow(expectCode('SESSION_QUERY_EVENT_NOT_FOUND'))
+    await expect(ctx.sessionQuery.traceEvent({ sessionId: bad.id, seq: 0 }))
+      .rejects.toThrow(expectCode('SESSION_QUERY_INVALID_SURFACE'))
+  })
+
+  it.each([
+    ['non-surface sources', [
+      { type: 'turn/start', seq: 0, time: 1, data: { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } }, sourceEventSeqs: [0] },
+    ]],
+    ['invalid source array', [
+      { ...appendEvent(0), sourceEventSeqs: 'invalid' },
+    ]],
+    ['empty sources', [
+      appendEvent(0, []),
+    ]],
+    ['sparse sources', [
+      appendEvent(0, Array<number>(1)),
+    ]],
+    ['duplicate sources', [
+      appendEvent(0),
+      appendEvent(1, [0, 0]),
+    ]],
+    ['missing earlier source', [
+      appendEvent(0),
+      appendEvent(1, [-1]),
+    ]],
+    ['future source', [
+      appendEvent(0, [1]),
+      appendEvent(1),
+    ]],
+    ['replacement without sources', [
+      appendEvent(0),
+      { ...appendEvent(1), surfaceOp: { op: 'replace', start: 0, end: 0 } },
+    ]],
+    ['replacement missing a shadowed source', [
+      { type: 'assistant/chunk', seq: 0, time: 1, data: { turn: 1, step: 1, chunk: { type: 'text-delta', index: 0, text: 'draft' } } },
+      appendEvent(1),
+      { ...appendEvent(2, [0]), surfaceOp: { op: 'replace', start: 1, end: 1 } },
+    ]],
+  ] as const)('rejects an invalid surface log: %s', async (_name, rawEvents) => {
+    const durable = header('invalid-provenance')
+    const events = structuredClone(rawEvents) as unknown as SessionEvent[]
+    TracePersistence.reset([{ meta: durable, events }])
+    const ctx = await queryContext()
+    await ctx.plugin(TracePersistence)
+
+    await expect(ctx.sessionQuery.traceEvent({ sessionId: durable.id, seq: 0 }))
+      .rejects.toThrow(expectCode('SESSION_QUERY_INVALID_SURFACE'))
+  })
+
+  it('rejects surfaceOp on a non-surface event as an invalid surface', async () => {
+    const durable = header('invalid-non-surface-op')
+    const events = [{
+      type: 'turn/start',
+      seq: 0,
+      time: 1,
+      data: { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } },
+      surfaceOp: 'append',
+    }] as unknown as SessionEvent[]
+    TracePersistence.reset([{ meta: durable, events }])
+    const ctx = await queryContext()
+    await ctx.plugin(TracePersistence)
+
+    await expect(ctx.sessionQuery.traceEvent({ sessionId: durable.id, seq: 0 }))
+      .rejects.toThrow(expectCode('SESSION_QUERY_INVALID_SURFACE'))
+  })
+
+  it('applies the same surface contract to listEvents', async () => {
+    const durable = header('list-regression')
+    TracePersistence.reset([{ meta: durable, events: [appendEvent(0), appendEvent(1, [0, 0])] }])
+    const ctx = await queryContext()
+    await ctx.plugin(TracePersistence)
+
+    await expect(ctx.sessionQuery.listEvents(durable.id))
+      .rejects.toThrow(expectCode('SESSION_QUERY_INVALID_SURFACE'))
+  })
+})
diff --git a/packages/skill/skill-local/README.md b/packages/skill/skill-local/README.md
index 9449a40ec5..8034441272 100644
--- a/packages/skill/skill-local/README.md
+++ b/packages/skill/skill-local/README.md
@@ -12,7 +12,7 @@ Requires `ctx.skills` (`inject: ['skills']`).
 
 | Field | Default | Meaning |
 |---|---|---|
-| `dshHome` | `$DSH_HOME` or `~/.dsh` | DeepSeek Harness config root; scans `skills` under this directory. |
+| `dshHome` | `$DSH_HOME` or `~/.dsh` | DeepSeek Harness config root resolved by [`@deepseek-ai/dsh-home`](../../util/home/README.md); scans `skills` under this directory. |
 | `agentsHome` | `$DSH_AGENTS_HOME` or `~/.agents` | Shared agent config root scanned for compatible skills. |
 | `customSkillDirs` | `[]` | Additional local skill roots scanned after project roots and before user roots. |
 
diff --git a/packages/skill/skill-local/package.json b/packages/skill/skill-local/package.json
index d1ca775a26..d490438c51 100644
--- a/packages/skill/skill-local/package.json
+++ b/packages/skill/skill-local/package.json
@@ -23,6 +23,7 @@
   "license": "BSD-3-Clause",
   "peerDependencies": {
     "@deepseek-ai/dsh-fs": "^0.0.1",
+    "@deepseek-ai/dsh-home": "^0.0.1",
     "@deepseek-ai/dsh-skill": "^0.0.1",
     "cordis": "^4.0.0-rc.7"
   },
@@ -32,6 +33,7 @@
   },
   "devDependencies": {
     "@deepseek-ai/dsh-fs": "workspace:^",
+    "@deepseek-ai/dsh-home": "workspace:^",
     "@deepseek-ai/dsh-skill": "workspace:^",
     "cordis": "^4.0.0-rc.7"
   }
diff --git a/packages/skill/skill-local/src/index.ts b/packages/skill/skill-local/src/index.ts
index 19a15f1de8..ee109fbb16 100644
--- a/packages/skill/skill-local/src/index.ts
+++ b/packages/skill/skill-local/src/index.ts
@@ -17,6 +17,7 @@ import z from 'schemastery'
 import type Schema from 'schemastery'
 import { parse as parseYaml } from 'yaml'
 import type { FileSystem, FsDirEntry, FsTarget } from '@deepseek-ai/dsh-fs'
+import { resolveDshHome } from '@deepseek-ai/dsh-home'
 import {
   isSkillName,
   type SkillCandidate,
@@ -92,7 +93,7 @@ export class LocalSkillProvider implements SkillProvider {
   private readonly customSkillDirs: string[]
 
   constructor(private readonly ctx: Context, config: Config = {}) {
-    this.dshHome = resolve(config.dshHome ?? process.env.DSH_HOME ?? join(homedir(), '.dsh'))
+    this.dshHome = resolveDshHome(config.dshHome)
     this.agentsHome = resolve(config.agentsHome ?? process.env.DSH_AGENTS_HOME ?? join(homedir(), '.agents'))
     this.customSkillDirs = (config.customSkillDirs ?? []).map(root => resolve(root))
   }
diff --git a/packages/skill/skill-local/tests/skill-local.spec.ts b/packages/skill/skill-local/tests/skill-local.spec.ts
index 94a48ce53f..43cc70c7ad 100644
--- a/packages/skill/skill-local/tests/skill-local.spec.ts
+++ b/packages/skill/skill-local/tests/skill-local.spec.ts
@@ -4,7 +4,7 @@ import { dirname, join } from 'node:path'
 import { tmpdir } from 'node:os'
 import { Context } from 'cordis'
 import SkillService from '@deepseek-ai/dsh-skill'
-import { FileSystem, FsVersion, type FsDirEntry, type FsEditOutcome, type FsEditRequest, type FsInfo, type FsTarget, type FsWriteOutcome } from '@deepseek-ai/dsh-fs'
+import { FileSystem, FsVersion, type FsDirEntry, type FsEditOutcome, type FsEditRequest, type FsInfo, type FsPathInfo, type FsTarget, type FsWriteOutcome } from '@deepseek-ai/dsh-fs'
 import * as SkillLocal from '../src/index.ts'
 
 async function tempDir(name: string): Promise<string> {
@@ -53,6 +53,20 @@ class TestFileSystem extends FileSystem {
     }
   }
 
+  override async lstat(path: string): Promise<FsPathInfo | undefined> {
+    try {
+      const fs = await import('node:fs/promises')
+      const info = await fs.lstat(path)
+      return {
+        version: FsVersion(String(info.mtimeMs)),
+        type: info.isSymbolicLink() ? 'symlink' : info.isFile() ? 'file' : info.isDirectory() ? 'directory' : 'other',
+        size: info.size,
+      }
+    } catch {
+      return undefined
+    }
+  }
+
   override async readText(target: FsTarget, signal?: AbortSignal): Promise<string> {
     this.readTextSignals.push(signal)
     if (this.readTextOverride !== undefined) return await this.readTextOverride(target, signal)
diff --git a/packages/skill/skill-local/tsconfig.json b/packages/skill/skill-local/tsconfig.json
index 018f0a4a50..f51147abce 100644
--- a/packages/skill/skill-local/tsconfig.json
+++ b/packages/skill/skill-local/tsconfig.json
@@ -9,6 +9,7 @@
     { "path": "../../../vendor/cosmokit" },
     { "path": "../../../vendor/cordis" },
     { "path": "../../../vendor/schemastery" },
+    { "path": "../../util/home" },
     { "path": "../../fs/fs" },
     { "path": "../skill" }
   ]
diff --git a/packages/spill/README.md b/packages/spill/README.md
new file mode 100644
index 0000000000..7d54c91eb5
--- /dev/null
+++ b/packages/spill/README.md
@@ -0,0 +1,13 @@
+# spill/ - spill storage capability family
+
+The tool-output spill capability seam: an abstract storage interface, a local filesystem implementation, and the tool-result policy that uses it. All **product** packages.
+
+| Package | Role | ctx key |
+|---|---|---|
+| `spill/` | Abstract spill storage seam (`saveText` — persist oversized tool text and return a locator + retrieval hint) | `ctx.spillStore` |
+| `spill-local/` | Local-filesystem backend: private, session-scoped files with traversal-safe names | (registers on `ctx.spillStore`) |
+| `spill-policy/` | `tools/post-execute` policy: replaces oversized plain-text results with a preview + spill locator | (no service surface) |
+
+The interface lives at `spill/spill/`. The split mirrors bash/fs: the seam owns storage only, `spill-local` owns the filesystem mechanics, and `spill-policy` owns WHEN to spill and the model-facing notice. Preview mechanics stay in [`util/retention`](../util/README.md) — the policy composes the two without either owning the other's job.
+
+See the [tool output spill RFC](../../docs/rfc/implemented/architecture/2026-07-08-tool-output-spill-files.md) for the design rationale, including why final-result spill is separate from tool-owned early spill (bash streams, subagent rollouts) and why creation belongs to the runtime spill seam rather than the model-facing `write` tool.
diff --git a/packages/spill/spill-local/README.md b/packages/spill/spill-local/README.md
new file mode 100644
index 0000000000..59d23b8a46
--- /dev/null
+++ b/packages/spill/spill-local/README.md
@@ -0,0 +1,28 @@
+# @deepseek-ai/dsh-spill-local
+
+The **local-filesystem** implementation of the [`@deepseek-ai/dsh-spill`](../spill) storage seam. Registers as `ctx.spillStore` and persists a tool's oversized text to a private, session-scoped file; its locator is the file path and its retrieval hint tells the model to use `read` or `grep` on that path.
+
+## Storage layout
+
+Files land at `<root>/session-<hash>/​<random>-<safeName>`:
+
+- **`root`** — the config `root` (resolved to absolute), or a lazily-created private (0700) per-process directory under the OS temp dir when omitted. A predictable, world-readable root would let other local users read spilled tool output or plant symlinks.
+- **`session-<hash>`** — a short `sha256(sessionId)` prefix, so a session's spill files group together and a future cleanup can drop them per session.
+- **`<random>-<safeName>`** — an unpredictable hex prefix (defeats symlink planting in a shared root) plus the caller's `suggestedName` sanitized to one safe path segment (traversal-proof; mirrors the JSONL persistence backend's `encodeSegment`). The write is exclusive + owner-only (`open(path, 'wx', 0o600)`): it fails on any pre-existing path, symlink or not, so a planted target cannot redirect it.
+
+## Config
+
+| Key | Default | Meaning |
+|---|---|---|
+| `root` | private 0700 temp dir | Root directory for spill files. Set to keep them under a known location. |
+
+`saveText` rejects on a real storage failure (permissions, ENOSPC); the spill policy treats a rejection as best-effort and keeps the inline result. See the seam README for the vocabulary and the [tool output spill RFC](../../../docs/rfc/implemented/architecture/2026-07-08-tool-output-spill-files.md) for the design.
+
+## Model Experience
+
+Indirectly, through spill consumers that render the local path and `read`/`grep` retrieval guidance.
+
+## Known Limitations and Deferred Work
+
+- **Local spill files persist until external cleanup** — the backend has no session-lifecycle deletion or age-based retention policy, because persisted, resumed, and forked sessions may still reference a path.
+- **Locators require a co-located filesystem consumer** — a remote or virtual deployment needs another `SpillStore` backend whose locator and retrieval hint are meaningful there.
diff --git a/packages/spill/spill-local/package.json b/packages/spill/spill-local/package.json
new file mode 100644
index 0000000000..a75c4bfc0b
--- /dev/null
+++ b/packages/spill/spill-local/package.json
@@ -0,0 +1,38 @@
+{
+  "name": "@deepseek-ai/dsh-spill-local",
+  "description": "Local-filesystem implementation of the DeepSeek Harness spill storage seam (private session-scoped files)",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "@deepseek-ai/dsh-spill": "^0.0.1",
+    "cordis": "^4.0.0-rc.6"
+  },
+  "dependencies": {
+    "schemastery": "^3.18.0"
+  },
+  "devDependencies": {
+    "@deepseek-ai/dsh-brand": "workspace:^",
+    "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-session": "workspace:^",
+    "@deepseek-ai/dsh-spill": "workspace:^",
+    "cordis": "^4.0.0-rc.6"
+  }
+}
diff --git a/packages/spill/spill-local/src/index.ts b/packages/spill/spill-local/src/index.ts
new file mode 100644
index 0000000000..73e2cad851
--- /dev/null
+++ b/packages/spill/spill-local/src/index.ts
@@ -0,0 +1,65 @@
+/**
+ * `LocalSpillStore`: the host-filesystem implementation of the
+ * `@deepseek-ai/dsh-spill` storage seam. Persists a tool's oversized text to a
+ * private, session-scoped file (see `./store.ts` for the traversal-safe naming
+ * and exclusive owner-only write) and returns a path locator plus local
+ * read/grep retrieval guidance.
+ *
+ * @module @deepseek-ai/dsh-spill-local
+ */
+
+import { Context } from 'cordis'
+import { resolve } from 'node:path'
+import z from 'schemastery'
+import { SpillLocator, SpillStore } from '@deepseek-ai/dsh-spill'
+import type { SaveTextSpill, SpillRef } from '@deepseek-ai/dsh-spill'
+import { privateRoot, saveTextFile } from './store.ts'
+
+export { encodeSegment, privateRoot, saveTextFile, sessionDir } from './store.ts'
+export type { SavedText, SaveTextOptions } from './store.ts'
+
+/** Plugin config (all optional — `static Config` supplies the defaults). */
+export interface Config {
+  /**
+   * Root directory for spill files. Omitted uses a lazily-created private
+   * (0700) per-process directory under the OS temp dir — the safe default for
+   * a local deployment. Set it to keep spill files under a known location.
+   */
+  root?: string
+}
+
+/**
+ * Local-filesystem spill backend. Files land under `<root>/session-<hash>/…`
+ * with unpredictable names, an exclusive owner-only (0600) write, and a private
+ * (0700) root — a spilled tool result must not be readable by other local users
+ * or redirectable via a planted symlink.
+ */
+export class LocalSpillStore extends SpillStore {
+  static Config: z<Config> = z.object({
+    root: z.string(),
+  })
+
+  /** Resolved absolute spill root (config `root`, else the private default), fixed at construction. */
+  readonly root: string
+
+  constructor(ctx: Context, config: Config) {
+    super(ctx)
+    this.root = config.root !== undefined ? resolve(config.root) : privateRoot()
+  }
+
+  async saveText(input: SaveTextSpill): Promise<SpillRef> {
+    const saved = await saveTextFile({
+      root: this.root,
+      sessionId: input.owner.sessionId,
+      suggestedName: input.suggestedName,
+      content: input.content,
+    })
+    return {
+      locator: SpillLocator(saved.path),
+      bytes: saved.bytes,
+      retrievalHint: 'Use read with offset/limit, or grep this path to search within it.',
+    }
+  }
+}
+
+export default LocalSpillStore
diff --git a/packages/spill/spill-local/src/store.ts b/packages/spill/spill-local/src/store.ts
new file mode 100644
index 0000000000..e44418767a
--- /dev/null
+++ b/packages/spill/spill-local/src/store.ts
@@ -0,0 +1,120 @@
+/**
+ * Cordis-free storage mechanics for the local spill backend: private
+ * session-scoped directory selection, safe-name derivation, path-traversal
+ * protection, and the exclusive owner-only write. Kept out of the service class
+ * (like `dsh-bash-local`'s `run.ts`) so the filesystem behavior is unit-testable
+ * without a `ctx` and without the OS temp dir.
+ *
+ * @module @deepseek-ai/dsh-spill-local/store
+ */
+
+import { createHash, randomBytes } from 'node:crypto'
+import { mkdtempSync } from 'node:fs'
+import { mkdir, open } from 'node:fs/promises'
+import { join } from 'node:path'
+import { tmpdir } from 'node:os'
+
+let defaultRoot: string | undefined
+
+/**
+ * The default spill root: a private (0700) per-process directory under the OS
+ * tmpdir, created lazily. Predictable world-readable paths would let other
+ * local users read spilled tool output or pre-create symlinks; `mkdtemp` gives
+ * an unpredictable suffix and 0700 semantics.
+ *
+ * @returns The lazily-created private spill root.
+ */
+export function privateRoot(): string {
+  defaultRoot ??= mkdtempSync(join(tmpdir(), 'dsh-spill-'))
+  return defaultRoot
+}
+
+// Deliberately mirrors the JSONL path encoder, but keeps spill's empty-name
+// policy (`""` -> `"~"`) local so storage backends stay decoupled.
+/* jscpd:ignore-start */
+/**
+ * Encode an arbitrary string as one safe path segment, injectively over ALL JS
+ * (UTF-16) strings. A session id / suggested name is untrusted input, so this
+ * neutralizes `../`, absolute paths, NUL, and separators before any filesystem
+ * use. Each code unit is kept literal (`[A-Za-z0-9._-]`, minus `~`) or escaped
+ * as `~XXXX`; `~` is itself escaped, so the mapping is reversible and distinct
+ * inputs never collide. The whole-segment tokens `.`/`..` are escaped so they
+ * can never traverse. An empty string encodes to `~` (never an empty segment).
+ * (Mirrors the JSONL persistence backend's `encodeSegment`.)
+ *
+ * @param raw The untrusted string to encode as one safe path segment.
+ * @returns An injective, filesystem-safe single path segment.
+ */
+export function encodeSegment(raw: string): string {
+  if (raw.length === 0) return '~'
+  if (raw === '.') return '~002E'
+  if (raw === '..') return '~002E~002E'
+  let out = ''
+  for (let i = 0; i < raw.length; i++) {
+    const code = raw.charCodeAt(i)
+    const ch = String.fromCharCode(code)
+    if (ch !== '~' && /^[A-Za-z0-9._-]$/.test(ch)) {
+      out += ch
+    } else {
+      out += '~' + code.toString(16).toUpperCase().padStart(4, '0')
+    }
+  }
+  return out
+}
+/* jscpd:ignore-end */
+
+/**
+ * The session-scoped directory: `<root>/session-<hash(sessionId)>`, a short stable hash.
+ *
+ * @param root The spill root directory.
+ * @param sessionId The owning session id to hash into a stable directory name.
+ * @returns The absolute session-scoped spill directory path.
+ */
+export function sessionDir(root: string, sessionId: string): string {
+  const hash = createHash('sha256').update(sessionId).digest('hex').slice(0, 12)
+  return join(root, `session-${hash}`)
+}
+
+/** Options for {@link saveTextFile} — the resolved root and the request fields the store needs. */
+export interface SaveTextOptions {
+  /** The spill root directory (configured or the lazy private default). */
+  root: string
+  /** The owning session id (scopes the directory). */
+  sessionId: string
+  /** Caller-suggested base name; sanitized to one safe segment before use. */
+  suggestedName: string
+  /** The full text to persist. */
+  content: string
+}
+
+/** A written spill file. */
+export interface SavedText {
+  path: string
+  bytes: number
+}
+
+/**
+ * Write `content` to a fresh file under the session-scoped directory and return
+ * its path + byte length. The filename is a random hex prefix plus the
+ * sanitized `suggestedName`, so it is unpredictable (defeats symlink planting in
+ * a shared root) AND stays readable. The open is exclusive + owner-only
+ * (`'wx', 0o600`): it fails on any existing path — symlink or not — so a
+ * pre-planted target cannot redirect the write.
+ *
+ * @param options The resolved root and request fields required to save the file.
+ * @returns The written file path and UTF-8 byte length.
+ */
+export async function saveTextFile(options: SaveTextOptions): Promise<SavedText> {
+  const dir = sessionDir(options.root, options.sessionId)
+  await mkdir(dir, { recursive: true, mode: 0o700 })
+  const safeName = encodeSegment(options.suggestedName)
+  const path = join(dir, `${randomBytes(6).toString('hex')}-${safeName}`)
+  const bytes = Buffer.byteLength(options.content, 'utf8')
+  const handle = await open(path, 'wx', 0o600)
+  try {
+    await handle.writeFile(options.content)
+  } finally {
+    await handle.close()
+  }
+  return { path, bytes }
+}
diff --git a/packages/spill/spill-local/tests/spill-local.spec.ts b/packages/spill/spill-local/tests/spill-local.spec.ts
new file mode 100644
index 0000000000..d73fca9fe3
--- /dev/null
+++ b/packages/spill/spill-local/tests/spill-local.spec.ts
@@ -0,0 +1,139 @@
+/**
+ * Tests for the LOCAL spill backend: `saveText` writes a session-scoped file and
+ * returns a locator + byte length + retrieval hint, filename sanitization
+ * neutralizes traversal, the configured `root` is honored (and the private
+ * default when omitted), and a storage failure rejects. The Cordis-free
+ * `store.ts` helpers are exercised directly for the naming/encoding edge cases.
+ */
+
+import { describe, expect, it, beforeEach, afterEach } from 'vitest'
+import { Context } from 'cordis'
+import { mkdtempSync, readFileSync, rmSync, statSync } from 'node:fs'
+import { tmpdir } from 'node:os'
+import { dirname, isAbsolute, join } from 'node:path'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import { SessionId } from '@deepseek-ai/dsh-session'
+import type { SaveTextSpill } from '@deepseek-ai/dsh-spill'
+import LocalSpillStore, { encodeSegment, privateRoot, saveTextFile, sessionDir } from '@deepseek-ai/dsh-spill-local'
+
+let root: string
+
+beforeEach(() => {
+  root = mkdtempSync(join(tmpdir(), 'dsh-spill-test-'))
+})
+afterEach(() => {
+  rmSync(root, { recursive: true, force: true })
+})
+
+function request(overrides: Partial<SaveTextSpill> = {}): SaveTextSpill {
+  return {
+    owner: { sessionId: SessionId('sess-1') },
+    source: { toolName: 'web_fetch', callId: CallId('call-1'), label: 'result' },
+    suggestedName: 'web_fetch.txt',
+    content: 'the full body',
+    ...overrides,
+  }
+}
+
+describe('encodeSegment', () => {
+  it('keeps the safe set literal', () => {
+    expect(encodeSegment('web_fetch.txt')).toBe('web_fetch.txt')
+    expect(encodeSegment('a-B_9.z')).toBe('a-B_9.z')
+  })
+
+  it('escapes separators and tilde (dots are literal except as whole-segment tokens)', () => {
+    // `.` is in the safe set, so `..` inside a longer string stays literal; the
+    // traversal defense is that separators escape, keeping the result ONE segment.
+    expect(encodeSegment('../etc/passwd')).toBe('..~002Fetc~002Fpasswd')
+    expect(encodeSegment('a/b')).toBe('a~002Fb')
+    expect(encodeSegment('~')).toBe('~007E')
+  })
+
+  it('escapes the whole-segment dot tokens', () => {
+    expect(encodeSegment('.')).toBe('~002E')
+    expect(encodeSegment('..')).toBe('~002E~002E')
+  })
+
+  it('encodes the empty string to a non-empty segment', () => {
+    expect(encodeSegment('')).toBe('~')
+  })
+})
+
+describe('sessionDir', () => {
+  it('is a stable per-session hash under the root', () => {
+    const dir = sessionDir('/spill', 'sess-1')
+    expect(dir).toBe(sessionDir('/spill', 'sess-1'))
+    expect(dir).toMatch(/\/spill\/session-[0-9a-f]{12}$/)
+    expect(sessionDir('/spill', 'sess-2')).not.toBe(dir)
+  })
+})
+
+describe('saveTextFile', () => {
+  it('writes the content under the session dir and reports bytes', async () => {
+    const saved = await saveTextFile({ root, sessionId: 'sess-1', suggestedName: 'r.txt', content: 'héllo' })
+    expect(readFileSync(saved.path, 'utf8')).toBe('héllo')
+    expect(saved.bytes).toBe(Buffer.byteLength('héllo', 'utf8'))
+    expect(dirname(saved.path)).toBe(sessionDir(root, 'sess-1'))
+    expect(saved.path).toMatch(/\/[0-9a-f]{12}-r\.txt$/)
+  })
+
+  it('sanitizes a traversal-shaped suggested name into one segment', async () => {
+    const saved = await saveTextFile({ root, sessionId: 'sess-1', suggestedName: '../../evil', content: 'x' })
+    // The separators escaped, so the whole name is one leaf under the session dir.
+    expect(dirname(saved.path)).toBe(sessionDir(root, 'sess-1'))
+    expect(saved.path.includes('/..')).toBe(false)
+  })
+
+  it('creates the session dir with owner-only permissions', async () => {
+    const saved = await saveTextFile({ root, sessionId: 'sess-1', suggestedName: 'r.txt', content: 'x' })
+    // 0o700 dir, 0o600 file (masked by umask, but the owner bits must hold).
+    expect(statSync(dirname(saved.path)).mode & 0o700).toBe(0o700)
+    expect(statSync(saved.path).mode & 0o600).toBe(0o600)
+  })
+
+  it('gives distinct paths to two saves of the same name', async () => {
+    const a = await saveTextFile({ root, sessionId: 'sess-1', suggestedName: 'r.txt', content: 'a' })
+    const b = await saveTextFile({ root, sessionId: 'sess-1', suggestedName: 'r.txt', content: 'b' })
+    expect(a.path).not.toBe(b.path)
+  })
+})
+
+describe('privateRoot', () => {
+  it('is a stable absolute directory under the temp dir', () => {
+    const first = privateRoot()
+    expect(isAbsolute(first)).toBe(true)
+    expect(privateRoot()).toBe(first)
+  })
+})
+
+describe('LocalSpillStore service', () => {
+  it('registers as ctx.spillStore and saves under the configured root', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LocalSpillStore, { root })
+    const ref = await ctx.spillStore.saveText(request())
+    expect(dirname(ref.locator)).toBe(sessionDir(root, 'sess-1'))
+    expect(readFileSync(ref.locator, 'utf8')).toBe('the full body')
+    expect(ref.bytes).toBe(Buffer.byteLength('the full body', 'utf8'))
+    expect(ref.retrievalHint).toBe('Use read with offset/limit, or grep this path to search within it.')
+  })
+
+  it('resolves a relative configured root to absolute', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LocalSpillStore, { root: '.' })
+    expect(isAbsolute((ctx.spillStore as LocalSpillStore).root)).toBe(true)
+  })
+
+  it('falls back to the private root when none is configured', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LocalSpillStore, {})
+    expect((ctx.spillStore as LocalSpillStore).root).toBe(privateRoot())
+  })
+
+  it('rejects when the root is not writable (missing parent, exclusive open)', async () => {
+    const ctx = new Context()
+    // A file (not a dir) as the root makes mkdir under it fail — a real storage error.
+    const filePath = (await saveTextFile({ root, sessionId: 's', suggestedName: 'f', content: 'x' })).path
+    await ctx.plugin(LocalSpillStore, { root: filePath })
+    await expect(ctx.spillStore.saveText(request())).rejects.toThrow()
+  })
+})
diff --git a/packages/spill/spill-local/tsconfig.json b/packages/spill/spill-local/tsconfig.json
new file mode 100644
index 0000000000..8e818212f5
--- /dev/null
+++ b/packages/spill/spill-local/tsconfig.json
@@ -0,0 +1,14 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": ["src"],
+  "references": [
+    { "path": "../../../vendor/cosmokit" },
+    { "path": "../../../vendor/cordis" },
+    { "path": "../../../vendor/schemastery" },
+    { "path": "../spill" }
+  ]
+}
diff --git a/packages/spill/spill-policy/README.md b/packages/spill/spill-policy/README.md
new file mode 100644
index 0000000000..bb20ac9dfc
--- /dev/null
+++ b/packages/spill/spill-policy/README.md
@@ -0,0 +1,46 @@
+# @deepseek-ai/dsh-spill-policy
+
+The **tool-result spill policy**: a `tools/post-execute` transformer that keeps oversized plain-text tool results out of the model's context. When a final result exceeds `maxInlineBytes`, it saves the FULL text through [`ctx.spillStore`](../spill) and replaces the model-facing result with a bounded head/tail preview plus the backend's locator and retrieval hint.
+
+This plugin registers **no service** and owns no storage or preview mechanics: preview is [`@deepseek-ai/dsh-retention`](../../util/retention) (`TextRetainer`), storage is `ctx.spillStore`. It only decides WHEN to spill and composes the notice.
+
+## Config
+
+| Key | Default | Meaning |
+|---|---|---|
+| `maxInlineBytes` | *(omitted)* | Model-facing context cap for a plain-text result, in UTF-8 bytes (a non-negative integer; validated at load). **Omitted disables the policy entirely** (the plugin registers nothing). When set, a larger result is spilled and replaced with a preview derived from the same budget (head/tail split). |
+
+## Behavior
+
+1. Let the tool run (delegates via `next()`, so it bounds whatever a downstream hook accepted).
+2. Skip `read` (avoids a `read → spill → read again` loop) and any non-`accept` decision (a `block`'s corrective feedback passes through).
+3. Flatten the accepted content only when it is **plain text** (all `text` blocks); a result with any non-text block is left untouched.
+4. If its UTF-8 size is `≤ maxInlineBytes`, leave it unchanged.
+5. Otherwise save the full text and replace the result with a preview + this notice, sized so the whole replacement (preview + blank line + notice) stays within `maxInlineBytes` — the notice's byte cost is reserved out of the budget, so the preview shrinks to fit and the model-facing result never exceeds the cap:
+
+   ```text
+   <retained head/tail preview>
+
+   (Omitted N bytes. Full formatted result stored at: /…/session-…/…-web_fetch.txt. Use read with offset/limit, or grep this path to search within it.)
+   ```
+
+   When the notice alone fills the budget (a tiny cap or a long locator) the preview is empty and only the notice is returned. If even that notice-only replacement would exceed `maxInlineBytes`, the policy keeps the inline result — it never emits a replacement over the cap (and a within-cap replacement is always smaller than the original, so this also means spilling never adds bytes).
+
+**Best-effort:** no session owner, no `ctx.spillStore` backend, or a `saveText` rejection ⇒ the policy logs a warning and returns the original result. A spill failure never turns a successful call into an `isError` or hides the inline result.
+
+## Scope
+
+The policy sees only the FINAL formatted tool result — not a tool's internal resource. If a provider already truncated (e.g. `web-fetch-local.maxBodyChars`), the spill artifact holds the full formatted result the tool returned, not the full original source. Provider/resource caps stay mandatory and separate. Tool-owned early spill (bash streams, subagent rollouts) is future work — see the [tool output spill RFC](../../../docs/rfc/implemented/architecture/2026-07-08-tool-output-spill-files.md).
+
+## Model Experience
+
+### Oversized plain-text result
+
+**What the model sees**: Results at or below `maxInlineBytes`, `read` results, blocked decisions, and results containing non-text blocks are unchanged. An oversized plain-text result becomes a bounded head/tail preview followed by `(Omitted <bytes> bytes. Full formatted result stored at: <locator>. <retrievalHint>)`; storage or ownership failures leave the original result visible.
+
+**Token effect**: A successful replacement is at most `maxInlineBytes` UTF-8 bytes and remains in history until compaction; the full spill text is not resent to the model.
+
+## Known Limitations and Deferred Work
+
+- **Only final plain-text results are spillable** — mixed-content results, blocked feedback, and `read` pass through; provider truncation or tool-owned retention that happened earlier cannot be recovered here.
+- **A notice that cannot fit disables replacement for that call** — a tiny cap or long locator leaves the oversized original inline after the backend has already saved an unreferenced spill.
diff --git a/packages/spill/spill-policy/package.json b/packages/spill/spill-policy/package.json
new file mode 100644
index 0000000000..9c28ea5382
--- /dev/null
+++ b/packages/spill/spill-policy/package.json
@@ -0,0 +1,44 @@
+{
+  "name": "@deepseek-ai/dsh-spill-policy",
+  "description": "Tool-result spill policy for the DeepSeek Harness — replaces oversized plain-text tool results with a retained preview plus a spill-file path (no service surface)",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "@deepseek-ai/dsh-llm": "^0.0.1",
+    "@deepseek-ai/dsh-retention": "^0.0.1",
+    "@deepseek-ai/dsh-session": "^0.0.1",
+    "@deepseek-ai/dsh-spill": "^0.0.1",
+    "@deepseek-ai/dsh-tools": "^0.0.1",
+    "cordis": "^4.0.0-rc.6"
+  },
+  "dependencies": {
+    "schemastery": "^3.18.0"
+  },
+  "devDependencies": {
+    "@deepseek-ai/dsh-agent": "workspace:^",
+    "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-retention": "workspace:^",
+    "@deepseek-ai/dsh-session": "workspace:^",
+    "@deepseek-ai/dsh-spill": "workspace:^",
+    "@deepseek-ai/dsh-tools": "workspace:^",
+    "cordis": "^4.0.0-rc.6"
+  }
+}
diff --git a/packages/spill/spill-policy/src/index.ts b/packages/spill/spill-policy/src/index.ts
new file mode 100644
index 0000000000..b7ac4a31fc
--- /dev/null
+++ b/packages/spill/spill-policy/src/index.ts
@@ -0,0 +1,174 @@
+/**
+ * The spill-policy PLUGIN: a `tools/post-execute` result transformer that keeps
+ * oversized plain-text tool results out of the model's context. When a final
+ * result's UTF-8 size exceeds `maxInlineBytes`, it saves the FULL text to a
+ * session-scoped spill artifact (`ctx.spillStore`) and replaces the
+ * model-facing result with a bounded head/tail preview plus the backend's
+ * locator and retrieval guidance.
+ *
+ * It registers NO service and owns NO storage or preview mechanics: preview is
+ * `@deepseek-ai/dsh-retention` (`TextRetainer`), storage is `ctx.spillStore`.
+ * The policy only decides WHEN to spill and composes the notice.
+ *
+ * ## Deliberately narrow
+ *
+ * - Omitted `maxInlineBytes` ⇒ the plugin registers nothing (a true no-op).
+ * - Plain-text results only: a result carrying any non-text block is left
+ *   untouched (the policy knows only the final formatted text, not tool
+ *   internals).
+ * - `read` is skipped to avoid a `read → spill → read again` loop.
+ * - Best-effort: no session owner, no `ctx.spillStore` backend, or a save
+ *   failure ⇒ log and return the original result. A spill failure must NEVER
+ *   turn a successful tool call into an `isError` or hide the inline result.
+ *
+ * It COMPOSES with other post-execute listeners: it delegates via `next()` and
+ * bounds the resulting `accept` content, so a hook that replaced the content
+ * still has its replacement bounded, and a `block` decision passes through
+ * unchanged.
+ *
+ * @module @deepseek-ai/dsh-spill-policy
+ */
+
+import type { Context } from 'cordis'
+import z from 'schemastery'
+import type { ContentBlock } from '@deepseek-ai/dsh-llm'
+import { TextRetainer, describeOmitted } from '@deepseek-ai/dsh-retention'
+import type { Omitted } from '@deepseek-ai/dsh-retention'
+import type { SaveTextSpill, SpillRef } from '@deepseek-ai/dsh-spill'
+import type { SessionId } from '@deepseek-ai/dsh-session'
+import type { PostToolDecision, ToolExecution } from '@deepseek-ai/dsh-tools'
+import type { SpillPolicyExec } from './types.ts'
+
+export type { SpillPolicyExec } from './types.ts'
+
+/** Plugin config. */
+export interface Config {
+  /**
+   * The model-facing context cap for a plain-text tool result, in UTF-8 bytes.
+   * Omitted disables the policy entirely (no-op). When set, a result larger than
+   * this is spilled and replaced with a preview derived from this same budget.
+   */
+  maxInlineBytes?: number
+}
+
+/** Cordis plugin name used by loader diagnostics. */
+export const name = 'spill-policy'
+
+/** Require the tool registry (its `tools/post-execute` waterfall is the seam we transform). */
+export const inject = ['tools']
+
+export const Config: z<Config> = z.object({
+  maxInlineBytes: z.number(),
+})
+
+/** All-text content flattened to one UTF-8 string, or `undefined` if any block is non-text. */
+function flattenPlainText(content: ContentBlock[]): string | undefined {
+  let text = ''
+  for (const block of content) {
+    if (block.type !== 'text') return undefined
+    text += block.text
+  }
+  return text
+}
+
+/** The owning session id, or `undefined` for a call with no agent (a direct/test call). */
+function ownerSessionId(exec: ToolExecution): SessionId | undefined {
+  return (exec as SpillPolicyExec).agent?.session.header.id
+}
+
+/** Build the bounded head/tail preview for `text`, splitting `budget` bytes across the two ends. */
+function preview(text: string, budget: number): { text: string; omitted: Omitted } {
+  const headBytes = Math.ceil(budget / 2)
+  const tailBytes = Math.floor(budget / 2)
+  const retainer = new TextRetainer({ kind: 'headTail', headBytes, tailBytes })
+  retainer.push(text)
+  const kept = retainer.finish()
+  return { text: kept.text, omitted: kept.omittedBytes }
+}
+
+/** The spill-notice line for a given omission + saved reference (no preview, no leading blank line). */
+function spillNotice(omitted: Omitted, ref: SpillRef): string {
+  const omission = describeOmitted(omitted, 'bytes')
+  return `(${omission} Full formatted result stored at: ${ref.locator}. ${ref.retrievalHint})`
+}
+
+export function apply(ctx: Context, config: Config): void {
+  const maxInlineBytes = config.maxInlineBytes
+  // Omitted ⇒ no automatic spill policy: register nothing at all.
+  if (maxInlineBytes === undefined) return
+  // Validate at LOAD, not per call: a negative/fractional cap would reach
+  // TextRetainer's assertBudget and throw, turning every oversized-result call
+  // into an isError. A bad config must fail the deployment, not the tool.
+  if (!Number.isInteger(maxInlineBytes) || maxInlineBytes < 0) {
+    throw new Error(`spill-policy: maxInlineBytes must be a non-negative integer (got ${maxInlineBytes})`)
+  }
+
+  ctx.on('tools/post-execute', async (exec, result, next): Promise<PostToolDecision> => {
+    // Delegate first so a downstream listener (e.g. a hook) settles the result;
+    // we bound whatever it accepted. A block passes through — spill only shapes
+    // accepted plain-text results, never corrective feedback.
+    const decision = await next()
+    // Skip `read` to avoid a read → spill → read again loop.
+    if (decision.kind !== 'accept' || exec.name === 'read') return decision
+
+    const content = decision.content ?? result.content
+    const text = flattenPlainText(content)
+    if (text === undefined) return decision
+    const totalBytes = Buffer.byteLength(text, 'utf8')
+    if (totalBytes <= maxInlineBytes) return decision
+
+    const sessionId = ownerSessionId(exec)
+    if (sessionId === undefined) {
+      ctx.logger.warn(`spill-policy: no session owner for ${exec.name} result; keeping the inline result`)
+      return decision
+    }
+    const spillStore = ctx.get('spillStore')
+    if (!spillStore) {
+      ctx.logger.warn('spill-policy: no ctx.spillStore backend loaded; keeping the inline result')
+      return decision
+    }
+
+    const save: SaveTextSpill = {
+      owner: { sessionId },
+      source: { toolName: exec.name, callId: exec.callId, label: 'result' },
+      suggestedName: `${exec.name}.txt`,
+      content: text,
+    }
+    let ref: SpillRef
+    try {
+      ref = await spillStore.saveText(save)
+    } catch (error: unknown) {
+      // Best-effort: a storage failure (permissions, ENOSPC, backend down) must
+      // never fail the call or hide the result — keep the original inline.
+      ctx.logger.warn(`spill-policy: saveText failed for ${exec.name}: ${String(error)}; keeping the inline result`)
+      return decision
+    }
+
+    // Reserve the notice's byte cost INSIDE maxInlineBytes so the replacement
+    // (preview + blank line + notice) never exceeds the documented cap — a naive
+    // preview that spent the whole budget then appended the notice could be
+    // larger than the cap, and for a marginally-over result even larger than the
+    // original. The reservation uses a notice priced at the worst-case omission
+    // count (the full byte total): its digit count bounds the real count's, so
+    // the reserved size is a safe upper bound and the final notice is never
+    // longer than what we reserved. `\n\n` is the 2-byte join.
+    const reserve = Buffer.byteLength(spillNotice({ kind: 'exact', count: totalBytes }, ref), 'utf8') + 2
+    const previewBudget = Math.max(0, maxInlineBytes - reserve)
+    const { text: previewText, omitted } = preview(text, previewBudget)
+    const notice = spillNotice(omitted, ref)
+    const replacedText = previewText.length > 0 ? `${previewText}\n\n${notice}` : notice
+    // Invariant: the policy NEVER emits a replacement larger than the cap. When
+    // the notice alone exceeds maxInlineBytes (a tiny cap or a long spill root),
+    // there is no within-cap replacement, so keep the inline result — spilling
+    // would break the advertised context cap. (A within-cap replacement is
+    // always smaller than the original, which is > cap by the entry condition,
+    // so this one check subsumes "not smaller than the original" too. The spill
+    // file already written is a harmless orphan; cleanup is deferred.)
+    if (Buffer.byteLength(replacedText, 'utf8') > maxInlineBytes) {
+      ctx.logger.warn(`spill-policy: spill notice for ${exec.name} exceeds maxInlineBytes; keeping the inline result`)
+      return decision
+    }
+    const replaced: ContentBlock[] = [{ type: 'text', text: replacedText }]
+    return { kind: 'accept', content: replaced, ...decision.additionalContexts ? { additionalContexts: decision.additionalContexts } : {} }
+  })
+}
diff --git a/packages/spill/spill-policy/src/types.ts b/packages/spill/spill-policy/src/types.ts
new file mode 100644
index 0000000000..3046e3efe5
--- /dev/null
+++ b/packages/spill/spill-policy/src/types.ts
@@ -0,0 +1,26 @@
+/**
+ * Vocabulary for the spill-policy plugin: the minimal structural view of a tool
+ * execution the policy needs to derive the owning session for a spill artifact.
+ *
+ * `@deepseek-ai/dsh-tools`' `ToolExecution` satisfies this shape, so the policy
+ * reads `exec` straight through without importing `dsh-tools` or `dsh-agent`.
+ * Only the session HEADER id is read — the same identity every other subsystem
+ * keys off (see `dsh-tool-bash`'s owner derivation).
+ *
+ * @module @deepseek-ai/dsh-spill-policy/types
+ */
+
+import type { SessionId } from '@deepseek-ai/dsh-session'
+
+/** Minimal structural view of a tool execution: the owning session's header id, when present. */
+export interface SpillPolicyExec {
+  /** The agent on whose behalf the call runs, when there is one. */
+  agent?: {
+    session: {
+      header: {
+        /** The canonical session identity — the spill owner. */
+        id: SessionId
+      }
+    }
+  }
+}
diff --git a/packages/spill/spill-policy/tests/spill-policy.spec.ts b/packages/spill/spill-policy/tests/spill-policy.spec.ts
new file mode 100644
index 0000000000..2449f26a8c
--- /dev/null
+++ b/packages/spill/spill-policy/tests/spill-policy.spec.ts
@@ -0,0 +1,276 @@
+/**
+ * Tests for the spill-policy PLUGIN. It registers no service, only the
+ * `tools/post-execute` transformer. We drive real tools through
+ * `ctx.tools.execute(...)` and assert: disabled mode is a true no-op, an
+ * oversized plain-text result is spilled and replaced with a preview + locator,
+ * a small result and a non-text result pass through, `read` is skipped, and a
+ * `saveText` failure / missing backend / missing owner all preserve the original
+ * result without an `isError`.
+ */
+
+import { describe, expect, it, vi } from 'vitest'
+import { Context } from 'cordis'
+import Loader from '@cordisjs/plugin-loader'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import type { ContentBlock } from '@deepseek-ai/dsh-llm'
+import { SessionId } from '@deepseek-ai/dsh-session'
+import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import ToolRegistry, { defineTool } from '@deepseek-ai/dsh-tools'
+import type { ToolExecution } from '@deepseek-ai/dsh-tools'
+import { SpillLocator, SpillStore } from '@deepseek-ai/dsh-spill'
+import type { SaveTextSpill, SpillRef } from '@deepseek-ai/dsh-spill'
+import * as SpillPolicy from '@deepseek-ai/dsh-spill-policy'
+
+/** A stub spill backend recording its saves; `fail` exercises the best-effort fallback. */
+class StubStore extends SpillStore {
+  saves: SaveTextSpill[] = []
+  fail = false
+
+  async saveText(input: SaveTextSpill): Promise<SpillRef> {
+    if (this.fail) throw new Error('disk full')
+    this.saves.push(input)
+    return {
+      locator: SpillLocator(`/spill/${input.suggestedName}`),
+      bytes: Buffer.byteLength(input.content, 'utf8'),
+      retrievalHint: 'Use the stub retrieval path.',
+    }
+  }
+}
+
+/** A tool returning `text` verbatim (name configurable so we can register `read`). */
+function textTool(name: string, text: string) {
+  return defineTool({
+    name,
+    description: name,
+    parameters: {},
+    async execute(): Promise<ContentBlock[]> { return [{ type: 'text', text }] },
+  })
+}
+
+/** A minimal exec carrying a session header id (the spill owner). */
+function exec(name: string, session = 's1'): ToolExecution {
+  // Only agent.session.header.id is read by the policy; a structural stub suffices.
+  const agent = { session: { header: { id: SessionId(session) } } }
+  return { callId: CallId(`call-${name}`), name, arguments: {}, agent } as unknown as ToolExecution
+}
+
+/**
+ * Build a context with tools + the policy, and optionally a spill backend.
+ * Returns the context and the backend handle (undefined when `withSpill` false).
+ */
+async function setup(config: SpillPolicy.Config, withSpill = true): Promise<{ ctx: Context; spill?: StubStore; fiber: Awaited<ReturnType<Context['plugin']>> }> {
+  const ctx = new Context()
+  await ctx.plugin(SystemPrompt)
+  await ctx.plugin(ToolRegistry)
+  let spill: StubStore | undefined
+  if (withSpill) {
+    await ctx.plugin(StubStore)
+    spill = ctx.spillStore as StubStore
+  }
+  const fiber = await ctx.plugin(SpillPolicy, config)
+  return { ctx, fiber, ...spill ? { spill } : {} }
+}
+
+/** Flatten a result's text blocks. */
+function textOf(content: ContentBlock[]): string {
+  return content.filter((b): b is Extract<ContentBlock, { type: 'text' }> => b.type === 'text').map(b => b.text).join('')
+}
+
+describe('disabled mode', () => {
+  it('registers no post-execute listener when maxInlineBytes is omitted', async () => {
+    const { ctx, spill } = await setup({})
+    ctx.tools.register(textTool('big', 'x'.repeat(1000)))
+    const result = await ctx.tools.execute(exec('big'))
+    expect(textOf(result.content)).toBe('x'.repeat(1000))
+    expect(result.isError).toBe(false)
+    expect(spill?.saves).toHaveLength(0)
+  })
+})
+
+describe('loader export shape', () => {
+  it('has no default export and keeps name/inject/Config through unwrapExports', () => {
+    expect('default' in SpillPolicy).toBe(false)
+
+    const loader = Object.create(Loader.prototype) as Loader
+    const unwrapped = loader.unwrapExports(SpillPolicy) as Record<string, unknown>
+    expect(unwrapped).toBe(SpillPolicy)
+    expect(unwrapped.name).toBe('spill-policy')
+    expect(unwrapped.inject).toEqual(['tools'])
+    expect(unwrapped.Config).toBeDefined()
+    expect(typeof unwrapped.apply).toBe('function')
+  })
+})
+
+describe('config validation', () => {
+  it('rejects a negative maxInlineBytes at load', async () => {
+    await expect(setup({ maxInlineBytes: -1 })).rejects.toThrow(/non-negative integer/)
+  })
+
+  it('rejects a fractional maxInlineBytes at load', async () => {
+    await expect(setup({ maxInlineBytes: 1.5 })).rejects.toThrow(/non-negative integer/)
+  })
+})
+
+describe('oversized plain-text replacement', () => {
+  it('spills the full text and replaces the result with a preview + locator within the cap', async () => {
+    const { ctx, spill } = await setup({ maxInlineBytes: 200 })
+    const body = 'HEAD'.repeat(200) + 'TAIL'.repeat(200) // 1600 bytes > 200
+    ctx.tools.register(textTool('big', body))
+    const result = await ctx.tools.execute(exec('big'))
+
+    expect(result.isError).toBe(false)
+    expect(spill?.saves).toHaveLength(1)
+    expect(spill?.saves[0]?.content).toBe(body)
+    expect(spill?.saves[0]?.source.toolName).toBe('big')
+    expect(spill?.saves[0]?.suggestedName).toBe('big.txt')
+    expect(spill?.saves[0]?.owner.sessionId).toBe('s1')
+
+    const text = textOf(result.content)
+    expect(text).not.toBe(body)
+    expect(text.startsWith('HEAD')).toBe(true)
+    expect(text).toContain('Full formatted result stored at: /spill/big.txt')
+    expect(text).toContain('Use the stub retrieval path.')
+    expect(text).toContain('Omitted')
+    // The replacement (preview + blank line + notice) stays within the cap and
+    // is smaller than the original — the whole point of spilling.
+    expect(Buffer.byteLength(text, 'utf8')).toBeLessThanOrEqual(200)
+    expect(Buffer.byteLength(text, 'utf8')).toBeLessThan(body.length)
+  })
+
+  it('keeps the inline result when the notice-only replacement would exceed the cap', async () => {
+    // A body just over a tiny cap: the notice alone is larger than the cap, so
+    // there is no within-cap replacement — the policy keeps the inline result.
+    const { ctx } = await setup({ maxInlineBytes: 4 })
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    const body = 'xxxxx' // 5 bytes > 4, but far shorter than the notice
+    ctx.tools.register(textTool('big', body))
+    const result = await ctx.tools.execute(exec('big'))
+    expect(textOf(result.content)).toBe(body)
+    expect(warn).toHaveBeenCalled()
+  })
+
+  it('leaves a small plain-text result unchanged', async () => {
+    const { ctx, spill } = await setup({ maxInlineBytes: 1000 })
+    ctx.tools.register(textTool('small', 'tiny'))
+    const result = await ctx.tools.execute(exec('small'))
+    expect(textOf(result.content)).toBe('tiny')
+    expect(spill?.saves).toHaveLength(0)
+  })
+
+  it('leaves a result with a non-text block unchanged', async () => {
+    const { ctx, spill } = await setup({ maxInlineBytes: 5 })
+    ctx.tools.register(defineTool({
+      name: 'mixed',
+      description: 'mixed',
+      parameters: {},
+      async execute(): Promise<ContentBlock[]> {
+        return [{ type: 'text', text: 'x'.repeat(100) }, { type: 'reasoning', text: 'why' }]
+      },
+    }))
+    const result = await ctx.tools.execute(exec('mixed'))
+    expect(spill?.saves).toHaveLength(0)
+    expect(result.content).toHaveLength(2)
+  })
+})
+
+describe('read skip', () => {
+  it('never spills the read tool result (avoids a read → spill → read loop)', async () => {
+    const { ctx, spill } = await setup({ maxInlineBytes: 10 })
+    ctx.tools.register(textTool('read', 'x'.repeat(1000)))
+    const result = await ctx.tools.execute(exec('read'))
+    expect(textOf(result.content)).toBe('x'.repeat(1000))
+    expect(spill?.saves).toHaveLength(0)
+  })
+})
+
+describe('best-effort fallback', () => {
+  it('keeps the original result when saveText fails', async () => {
+    const { ctx, spill } = await setup({ maxInlineBytes: 10 })
+    spill!.fail = true
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    ctx.tools.register(textTool('big', 'x'.repeat(1000)))
+    const result = await ctx.tools.execute(exec('big'))
+    expect(textOf(result.content)).toBe('x'.repeat(1000))
+    expect(result.isError).toBe(false)
+    expect(warn).toHaveBeenCalled()
+  })
+
+  it('keeps the original result when no spill backend is loaded', async () => {
+    const { ctx } = await setup({ maxInlineBytes: 10 }, false)
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    ctx.tools.register(textTool('big', 'x'.repeat(1000)))
+    const result = await ctx.tools.execute(exec('big'))
+    expect(textOf(result.content)).toBe('x'.repeat(1000))
+    expect(warn).toHaveBeenCalled()
+  })
+
+  it('keeps the original result when the call has no session owner', async () => {
+    const { ctx, spill } = await setup({ maxInlineBytes: 10 })
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    ctx.tools.register(textTool('big', 'x'.repeat(1000)))
+    const result = await ctx.tools.execute({ callId: CallId('c'), name: 'big', arguments: {} })
+    expect(textOf(result.content)).toBe('x'.repeat(1000))
+    expect(spill?.saves).toHaveLength(0)
+    expect(warn).toHaveBeenCalled()
+  })
+})
+
+describe('composition', () => {
+  it('bounds content a downstream post-execute listener replaced', async () => {
+    const { ctx, spill } = await setup({ maxInlineBytes: 200 })
+    // A later-registered listener replaces the (small) tool result with a big one;
+    // the policy delegated via next(), so it bounds the replacement.
+    ctx.on('tools/post-execute', async (_e, _r, _next) =>
+      ({ kind: 'accept', content: [{ type: 'text', text: 'z'.repeat(500) }] }))
+    ctx.tools.register(textTool('small', 'tiny'))
+    const result = await ctx.tools.execute(exec('small'))
+    expect(spill?.saves[0]?.content).toBe('z'.repeat(500))
+    expect(textOf(result.content)).toContain('Full formatted result stored at')
+  })
+
+  it('preserves downstream accept-decision contexts when spilling', async () => {
+    const { ctx } = await setup({ maxInlineBytes: 200 })
+    const context = { content: [{ type: 'text' as const, text: 'note' }], source: { kind: 'plugin' as const, plugin: 'test' } }
+    ctx.on('tools/post-execute', async (_e, _r, _next) =>
+      ({ kind: 'accept', additionalContexts: [context] }))
+    ctx.tools.register(textTool('big', 'x'.repeat(1000)))
+    const result = await ctx.tools.execute(exec('big'))
+    expect(textOf(result.content)).toContain('Full formatted result stored at')
+    expect(result.additionalContexts).toEqual([context])
+  })
+})
+
+describe('cap invariant', () => {
+  it('keeps the inline result when the notice alone exceeds the cap, even for a large original', async () => {
+    // A large body (so it is well over the cap) but a cap smaller than the
+    // notice itself: there is no within-cap replacement, so the policy must keep
+    // the inline result rather than emit content over maxInlineBytes.
+    const { ctx } = await setup({ maxInlineBytes: 8 })
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    const body = 'x'.repeat(5000)
+    ctx.tools.register(textTool('big', body))
+    const result = await ctx.tools.execute(exec('big'))
+    expect(textOf(result.content)).toBe(body)
+    expect(warn).toHaveBeenCalled()
+  })
+})
+
+describe('disposal (HMR safety)', () => {
+  it('stops transforming oversized results after the plugin fiber is disposed', async () => {
+    const { ctx, spill, fiber } = await setup({ maxInlineBytes: 200 })
+    const body = 'HEAD'.repeat(200) + 'TAIL'.repeat(200)
+    ctx.tools.register(textTool('big', body))
+
+    // Live: the listener spills and replaces.
+    const before = await ctx.tools.execute(exec('big'))
+    expect(textOf(before.content)).toContain('Full formatted result stored at')
+    expect(spill?.saves).toHaveLength(1)
+
+    // After disposal the listener is gone — the result passes through untouched
+    // and nothing more is spilled (no leaked registration across reload).
+    await fiber.dispose()
+    const after = await ctx.tools.execute(exec('big'))
+    expect(textOf(after.content)).toBe(body)
+    expect(spill?.saves).toHaveLength(1)
+  })
+})
diff --git a/packages/spill/spill-policy/tsconfig.json b/packages/spill/spill-policy/tsconfig.json
new file mode 100644
index 0000000000..6a81ab2f3c
--- /dev/null
+++ b/packages/spill/spill-policy/tsconfig.json
@@ -0,0 +1,18 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": ["src"],
+  "references": [
+    { "path": "../../../vendor/cosmokit" },
+    { "path": "../../../vendor/cordis" },
+    { "path": "../../../vendor/schemastery" },
+    { "path": "../../util/retention" },
+    { "path": "../../llm/llm" },
+    { "path": "../../core/session" },
+    { "path": "../spill" },
+    { "path": "../../core/tools" }
+  ]
+}
diff --git a/packages/spill/spill/README.md b/packages/spill/spill/README.md
new file mode 100644
index 0000000000..c80277339b
--- /dev/null
+++ b/packages/spill/spill/README.md
@@ -0,0 +1,36 @@
+# @deepseek-ai/dsh-spill
+
+The **spill storage seam**: an abstract `SpillStore` service (`ctx.spillStore`) defining WHAT a spill backend does — persist a tool's oversized text and return a model-facing locator plus retrieval guidance — without saying HOW.
+
+This package is one third of the spill capability, split so each concern evolves (and swaps) independently:
+
+| Package | Role |
+|---|---|
+| `@deepseek-ai/dsh-spill` (this) | the interface: abstract service + vocabulary types |
+| `@deepseek-ai/dsh-spill-local` | an implementation: private session-scoped files on the host filesystem |
+| `@deepseek-ai/dsh-spill-policy` | the tool-result policy that spills oversized final results |
+
+The split mirrors the bash/fs seams. A future remote or virtual backend (e.g. a `spill://…` URI, a database key, or a backend-specific retrieval tool) implements this interface without touching the policy plugin.
+
+## Service API (`ctx.spillStore`)
+
+| Member | Semantics |
+|---|---|
+| `saveText(input)` | Persist `input.content` verbatim; resolves with a `SpillRef` (opaque locator, exact bytes written, and retrieval hint). **Rejects on a real storage failure** (permissions, ENOSPC, backend unavailable) — the caller decides how to degrade. |
+
+Storage is grouped by the request's `owner` session as a save-time namespace; the backend chooses its own private representation and may derive names from — never trust as a path — the caller's `suggestedName`. The seam owns storage only: NO retention policy (that is [`@deepseek-ai/dsh-retention`](../../util/retention)), NO tool-result replacement (that is `@deepseek-ai/dsh-spill-policy`), NO retrieval/search API (the backend's `retrievalHint` tells the model what to do with the locator).
+
+## Vocabulary
+
+`SaveTextSpill` (owner, source, suggestedName, content) is the request; `SpillRef` (locator, bytes, retrievalHint) is the result. `SpillLocator` is [branded](../../util/brand) and rendered to the model as an opaque string — a local path for `dsh-spill-local`, but a future backend may return a URI, key, or command token without changing policy/tool consumers. `SpillOwner.sessionId` is the save-time storage namespace: forked sessions inherit existing locators from the seeded log without copying or re-owning them, and new spills after the fork use the child session id. `SpillSource` (toolName, callId, label) is descriptive provenance for backend naming and inspection, not access control. See `src/types.ts` for the full contracts.
+
+See the [tool output spill RFC](../../../docs/rfc/implemented/architecture/2026-07-08-tool-output-spill-files.md) for the design rationale, including why creation belongs to the runtime spill seam rather than the model-facing `write` tool.
+
+## Model Experience
+
+Indirectly, through spill consumers that render a backend locator and retrieval guidance.
+
+## Known Limitations and Deferred Work
+
+- **The seam has no retrieval or deletion API** — consumers can only render the backend's locator and guidance; lifecycle and access semantics remain backend-specific.
+- **Storage is not access control** — `SpillOwner` namespaces writes but does not authorize reads of a locator; each backend and retrieval consumer must enforce its own boundary.
diff --git a/packages/spill/spill/package.json b/packages/spill/spill/package.json
new file mode 100644
index 0000000000..3103c9cd11
--- /dev/null
+++ b/packages/spill/spill/package.json
@@ -0,0 +1,36 @@
+{
+  "name": "@deepseek-ai/dsh-spill",
+  "description": "Abstract spill storage seam (ctx.spillStore) for the DeepSeek Harness — save oversized tool text and return a retrieval locator",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "@deepseek-ai/dsh-brand": "^0.0.1",
+    "@deepseek-ai/dsh-llm": "^0.0.1",
+    "@deepseek-ai/dsh-session": "^0.0.1",
+    "cordis": "^4.0.0-rc.6"
+  },
+  "devDependencies": {
+    "@deepseek-ai/dsh-brand": "workspace:^",
+    "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-session": "workspace:^",
+    "cordis": "^4.0.0-rc.6"
+  }
+}
diff --git a/packages/spill/spill/src/index.ts b/packages/spill/spill/src/index.ts
new file mode 100644
index 0000000000..4c8826defb
--- /dev/null
+++ b/packages/spill/spill/src/index.ts
@@ -0,0 +1,58 @@
+/**
+ * The spill storage seam (`ctx.spillStore`): an abstract service defining WHAT a
+ * spill backend does — persist a tool's oversized text and return a model-facing
+ * locator plus retrieval guidance — without saying HOW. Implementations
+ * subclass {@link SpillStore} and register as the `spillStore` service;
+ * `@deepseek-ai/dsh-spill-local` (host filesystem) is the first.
+ *
+ * The seam is deliberately minimal: `saveText` and nothing else. It owns NO
+ * retention policy (that is `@deepseek-ai/dsh-retention`), NO tool-result
+ * replacement (that is `@deepseek-ai/dsh-spill-policy`), and NO retrieval or
+ * search API. The backend supplies the locator and retrieval hint appropriate
+ * for its storage substrate.
+ *
+ * @module @deepseek-ai/dsh-spill
+ */
+
+import { Context, Service } from 'cordis'
+import type { SaveTextSpill, SpillRef } from './types.ts'
+
+export { SpillLocator } from './types.ts'
+export type { SaveTextSpill, SpillOwner, SpillRef, SpillSource } from './types.ts'
+
+declare module 'cordis' {
+  interface Context {
+    spillStore: SpillStore
+  }
+}
+
+/**
+ * Abstract spill storage service. Subclass, implement {@link saveText}, and load
+ * the subclass as a plugin — it registers as `ctx.spillStore` (one
+ * implementation per context; loading a second throws, cordis' standard
+ * duplicate-service behavior).
+ *
+ * Semantics every implementation must honor:
+ * - {@link saveText} persists the FULL `content` verbatim and returns an opaque
+ *   locator, exact byte length, and model-facing retrieval guidance.
+ * - Storage is scoped by the request's {@link SaveTextSpill.owner} session; the
+ *   backend chooses a private (not world-readable) location and a collision-free
+ *   name derived from — never equal to — the caller's `suggestedName`.
+ * - `saveText` REJECTS on a real storage failure (permissions, ENOSPC, backend
+ *   unavailable); the caller decides how to degrade (the spill policy treats a
+ *   rejection as best-effort and keeps the inline result).
+ */
+export abstract class SpillStore extends Service {
+  constructor(ctx: Context) {
+    super(ctx, 'spillStore')
+  }
+
+  /**
+   * Persist `input.content` to a session-scoped spill artifact.
+   * @param input - the owner, provenance, suggested name, and full text to save.
+   * @returns the saved artifact's {@link SpillRef}; rejects on a storage failure.
+   */
+  abstract saveText(input: SaveTextSpill): Promise<SpillRef>
+}
+
+export default SpillStore
diff --git a/packages/spill/spill/src/types.ts b/packages/spill/spill/src/types.ts
new file mode 100644
index 0000000000..96376bb268
--- /dev/null
+++ b/packages/spill/spill/src/types.ts
@@ -0,0 +1,73 @@
+/**
+ * Vocabulary for the spill storage seam. Types only — the abstract service
+ * lives in `./index.ts`, implementations in sibling packages
+ * (`@deepseek-ai/dsh-spill-local` first).
+ *
+ * @module @deepseek-ai/dsh-spill/types
+ */
+
+import type { Branded } from '@deepseek-ai/dsh-brand'
+import type { CallId } from '@deepseek-ai/dsh-llm'
+import type { SessionId } from '@deepseek-ai/dsh-session'
+
+/**
+ * Opaque model-facing handle for one spilled artifact. A local backend may use a
+ * filesystem path; a remote or database backend may use a URI or key. Consumers
+ * render it with {@link SpillRef.retrievalHint}, but do not parse it.
+ */
+export type SpillLocator = Branded<'SpillLocator'>
+
+/**
+ * Brand a string as a {@link SpillLocator}.
+ *
+ * @param locator The backend-produced locator string to brand.
+ * @returns The branded spill locator.
+ */
+export function SpillLocator(locator: string): SpillLocator {
+  return locator as SpillLocator
+}
+
+/**
+ * Save-time storage namespace for a spilled artifact. The session id lets a
+ * backend group storage under the producing session, but the returned
+ * {@link SpillLocator} is the model-facing handle. Forked sessions inherit
+ * locators already present in the seeded log; those artifacts are not copied or
+ * re-owned, and spills produced after the fork use the child session id.
+ */
+export interface SpillOwner {
+  sessionId: SessionId
+}
+
+/**
+ * Provenance of one spilled artifact — recorded by the backend for a readable
+ * filename and inspection. Not interpreted for access control; purely
+ * descriptive.
+ */
+export interface SpillSource {
+  /** The tool whose result was spilled (e.g. `web_fetch`). */
+  toolName: string
+  /** The model-issued call id the result belongs to. */
+  callId: CallId
+  /** A short human label for the artifact (e.g. `result`). */
+  label: string
+}
+
+/** One request to persist text to a spill artifact. */
+export interface SaveTextSpill {
+  owner: SpillOwner
+  source: SpillSource
+  /**
+   * A caller-suggested base name (e.g. `web_fetch.txt`). The backend sanitizes
+   * it to a single safe path segment before use — it is a hint, never a path.
+   */
+  suggestedName: string
+  /** The full text to persist (UTF-8). */
+  content: string
+}
+
+/** A saved spill artifact: its locator, byte length, and backend-specific retrieval guidance. */
+export interface SpillRef {
+  locator: SpillLocator
+  bytes: number
+  retrievalHint: string
+}
diff --git a/packages/spill/spill/tests/service.spec.ts b/packages/spill/spill/tests/service.spec.ts
new file mode 100644
index 0000000000..ddbc4086e1
--- /dev/null
+++ b/packages/spill/spill/tests/service.spec.ts
@@ -0,0 +1,60 @@
+/**
+ * Tests for the spill seam INTERFACE: a minimal concrete subclass registers as
+ * `ctx.spillStore`, a second load throws (duplicate service), and disposal
+ * releases the service. The storage behavior is the implementation's concern
+ * (`@deepseek-ai/dsh-spill-local`); here we only pin the seam contract.
+ */
+
+import { describe, expect, it } from 'vitest'
+import { Context } from 'cordis'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import { SessionId } from '@deepseek-ai/dsh-session'
+import { SpillLocator, SpillStore } from '@deepseek-ai/dsh-spill'
+import type { SaveTextSpill, SpillRef } from '@deepseek-ai/dsh-spill'
+
+/** Minimal concrete backend: records the last request, returns a fixed ref. */
+class StubStore extends SpillStore {
+  last: SaveTextSpill | undefined
+
+  async saveText(input: SaveTextSpill): Promise<SpillRef> {
+    this.last = input
+    return {
+      locator: SpillLocator(`/stub/${input.suggestedName}`),
+      bytes: Buffer.byteLength(input.content, 'utf8'),
+      retrievalHint: 'Use the stub reader.',
+    }
+  }
+}
+
+function request(content: string): SaveTextSpill {
+  return {
+    owner: { sessionId: SessionId('s1') },
+    source: { toolName: 'web_fetch', callId: CallId('c1'), label: 'result' },
+    suggestedName: 'web_fetch.txt',
+    content,
+  }
+}
+
+describe('spill seam', () => {
+  it('registers as ctx.spillStore and saves text', async () => {
+    const ctx = new Context()
+    await ctx.plugin(StubStore)
+    const ref = await ctx.spillStore.saveText(request('hello'))
+    expect(ref).toEqual({ locator: '/stub/web_fetch.txt', bytes: 5, retrievalHint: 'Use the stub reader.' })
+    expect((ctx.spillStore as StubStore).last?.content).toBe('hello')
+  })
+
+  it('rejects a second implementation (one per context)', async () => {
+    const ctx = new Context()
+    await ctx.plugin(StubStore)
+    await expect(ctx.plugin(StubStore)).rejects.toThrow()
+  })
+
+  it('releases the service on disposal', async () => {
+    const ctx = new Context()
+    const fiber = await ctx.plugin(StubStore)
+    expect(ctx.spillStore).toBeInstanceOf(StubStore)
+    await fiber.dispose()
+    expect((ctx as Context & { spillStore?: unknown }).spillStore).toBeUndefined()
+  })
+})
diff --git a/packages/spill/spill/tsconfig.json b/packages/spill/spill/tsconfig.json
new file mode 100644
index 0000000000..0c2fd5c57f
--- /dev/null
+++ b/packages/spill/spill/tsconfig.json
@@ -0,0 +1,15 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": ["src"],
+  "references": [
+    { "path": "../../../vendor/cosmokit" },
+    { "path": "../../../vendor/cordis" },
+    { "path": "../../util/brand" },
+    { "path": "../../llm/llm" },
+    { "path": "../../core/session" }
+  ]
+}
diff --git a/packages/subagent/subagent-acp/package.json b/packages/subagent/subagent-acp/package.json
index 5093e3df40..ec55845f91 100644
--- a/packages/subagent/subagent-acp/package.json
+++ b/packages/subagent/subagent-acp/package.json
@@ -35,6 +35,7 @@
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-loader-smoke": "workspace:^",
     "@deepseek-ai/dsh-subagent": "workspace:^",
     "@deepseek-ai/dsh-subagent-subprocess": "workspace:^",
     "@cordisjs/plugin-loader": "^1.0.0-rc.5",
diff --git a/packages/subagent/subagent-acp/tests/mock-acp-server.ts b/packages/subagent/subagent-acp/tests/mock-acp-server.ts
index 000f6d49f0..1496f882aa 100644
--- a/packages/subagent/subagent-acp/tests/mock-acp-server.ts
+++ b/packages/subagent/subagent-acp/tests/mock-acp-server.ts
@@ -2,8 +2,8 @@
  * Minimal no-network ACP child process for keyless backend tests. Environment variables script its
  * text and stop reason, a cancel-cooperative or cancel-ignoring hang, permission requests, and a
  * readiness marker. Disposal fixtures can delay an EOF flush, ignore EOF but exit and mark
- * SIGTERM, or trap SIGTERM to require SIGKILL. The specs spawn this non-test module under tsx with
- * an explicit tsconfig, mirroring real example boot.
+ * SIGTERM, or trap SIGTERM to require SIGKILL. The specs run this protocol-only fixture directly
+ * with Node's type stripping; it imports no harness code or workspace paths.
  * @module @deepseek-ai/dsh-subagent-acp/tests/mock-acp-server
  */
 
diff --git a/packages/subagent/subagent-acp/tests/subagent-acp.e2e.ts b/packages/subagent/subagent-acp/tests/subagent-acp.e2e.ts
index 0c20b27fa1..688b1c44aa 100644
--- a/packages/subagent/subagent-acp/tests/subagent-acp.e2e.ts
+++ b/packages/subagent/subagent-acp/tests/subagent-acp.e2e.ts
@@ -6,6 +6,7 @@ import { afterEach, describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
 import type { Agent } from '@deepseek-ai/dsh-agent'
 import SubagentService from '@deepseek-ai/dsh-subagent'
+import { resolveExampleLaunch } from '@deepseek-ai/dsh-loader-smoke'
 import * as acp from '../src/index.ts'
 
 /**
@@ -17,9 +18,22 @@ import * as acp from '../src/index.ts'
 // The real acp-agent example: its bin + cordis.yml (the live DeepSeek config).
 const binScript = fileURLToPath(new URL('../../../examples/acp-demo/src/bin.ts', import.meta.url))
 const exampleConfig = fileURLToPath(new URL('../../../../examples/acp-agent/cordis.yml', import.meta.url))
-const tsxLoader = fileURLToPath(import.meta.resolve('tsx'))
 const repoTsconfig = fileURLToPath(new URL('../../../../tsconfig.json', import.meta.url))
 
+// How to launch the child acp-agent (src via tsx / lib via plain node, per DSH_EXAMPLE_MODE).
+// buildChildEnv scrubs ambient creds but keeps these extras, so the model key is
+// forwarded explicitly; TSX_TSCONFIG_PATH is added by the resolver in src mode only.
+const childLaunch = resolveExampleLaunch({
+  srcBin: binScript,
+  configArgs: ['--config', exampleConfig],
+  tsconfigPath: repoTsconfig,
+  env: {
+    ...process.env.DEEPSEEK_API_KEY !== undefined ? { DEEPSEEK_API_KEY: process.env.DEEPSEEK_API_KEY } : {},
+    ...process.env.DEEPSEEK_BASE_URL !== undefined ? { DEEPSEEK_BASE_URL: process.env.DEEPSEEK_BASE_URL } : {},
+    DSH_PERMISSION_MODE: 'danger-full-access',
+  },
+})
+
 /** The ACP backend ignores the parent, but the seam requires one. */
 const fakeParent = { id: 'parent', session: { header: {} } } as unknown as Agent
 
@@ -40,18 +54,11 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('ACP backend with-key e2e (drive
     await ctx.plugin(SubagentService)
     await ctx.plugin(acp, {
       providerName: 'acp',
-      command: process.execPath,
-      args: ['--import', tsxLoader, binScript, '--config', exampleConfig],
+      command: childLaunch.command,
+      args: childLaunch.args,
       cwd: workdir,
       permission: 'reject',
-      // The child harness needs the key to reach the model; forward it
-      // explicitly (buildChildEnv scrubs ambient creds but keeps these extras).
-      env: {
-        ...process.env.DEEPSEEK_API_KEY !== undefined ? { DEEPSEEK_API_KEY: process.env.DEEPSEEK_API_KEY } : {},
-        ...process.env.DEEPSEEK_BASE_URL !== undefined ? { DEEPSEEK_BASE_URL: process.env.DEEPSEEK_BASE_URL } : {},
-        TSX_TSCONFIG_PATH: repoTsconfig,
-        DSH_PERMISSION_MODE: 'danger-full-access',
-      },
+      env: childLaunch.env as Record<string, string>,
     })
 
     const run = await ctx.subagents.start('acp', {
@@ -76,17 +83,12 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('ACP backend with-key e2e (drive
     await ctx.plugin(SubagentService)
     await ctx.plugin(acp, {
       providerName: 'acp',
-      command: process.execPath,
-      args: ['--import', tsxLoader, binScript, '--config', exampleConfig],
+      command: childLaunch.command,
+      args: childLaunch.args,
       cwd: workdir,
       // The child needs to act (run bash), so approve its permission prompts.
       permission: 'allow',
-      env: {
-        ...process.env.DEEPSEEK_API_KEY !== undefined ? { DEEPSEEK_API_KEY: process.env.DEEPSEEK_API_KEY } : {},
-        ...process.env.DEEPSEEK_BASE_URL !== undefined ? { DEEPSEEK_BASE_URL: process.env.DEEPSEEK_BASE_URL } : {},
-        TSX_TSCONFIG_PATH: repoTsconfig,
-        DSH_PERMISSION_MODE: 'danger-full-access',
-      },
+      env: childLaunch.env as Record<string, string>,
     })
 
     const run = await ctx.subagents.start('acp', {
diff --git a/packages/subagent/subagent-acp/tests/subagent-acp.spec.ts b/packages/subagent/subagent-acp/tests/subagent-acp.spec.ts
index eed3cfe1ed..bfc8476bae 100644
--- a/packages/subagent/subagent-acp/tests/subagent-acp.spec.ts
+++ b/packages/subagent/subagent-acp/tests/subagent-acp.spec.ts
@@ -21,8 +21,6 @@ import { acpStopReason, acpContentText, DEFAULT_DISPOSE_EOF_GRACE_MS, DEFAULT_DI
  */
 
 const mockServer = fileURLToPath(new URL('./mock-acp-server.ts', import.meta.url))
-const tsxLoader = fileURLToPath(import.meta.resolve('tsx'))
-const repoTsconfig = fileURLToPath(new URL('../../../../tsconfig.json', import.meta.url))
 
 /** A throwaway parent Agent — the ACP backend ignores it, but the seam requires one. */
 const fakeParent = { id: 'parent', session: { header: {} } } as unknown as Agent
@@ -46,11 +44,9 @@ async function setup(mockEnv: SetupEnv = {}, permission: 'allow' | 'reject' = 'r
   await ctx.plugin(acp, {
     providerName: 'acp',
     command: process.execPath,
-    args: ['--import', tsxLoader, mockServer],
+    args: [mockServer],
     permission,
-    // The mock-server scripting vars must reach the child; TSX_TSCONFIG_PATH lets
-    // tsx resolve @deepseek-ai/* from a child cwd outside the repo.
-    env: { ...mockEnv, TSX_TSCONFIG_PATH: repoTsconfig },
+    env: mockEnv,
   })
   return ctx
 }
@@ -197,10 +193,10 @@ describe('dsh-subagent-acp', () => {
     try {
       const spec: AcpRunSpec = {
         command: process.execPath,
-        args: ['--import', tsxLoader, mockServer],
+        args: [mockServer],
         cwd: process.cwd(),
         permission: 'reject',
-        env: { MOCK_TRAP_SIGTERM: '1', MOCK_TEXT: 'x', MOCK_READY_FILE: ready, TSX_TSCONFIG_PATH: repoTsconfig },
+        env: { MOCK_TRAP_SIGTERM: '1', MOCK_TEXT: 'x', MOCK_READY_FILE: ready },
         // Short on BOTH tiers: the trap ignores EOF and SIGTERM, so dispose must
         // burn the EOF window, then the SIGTERM window, then SIGKILL — keep each
         // small so the whole ladder finishes well within the 4000ms bound.
@@ -240,7 +236,7 @@ describe('dsh-subagent-acp', () => {
     try {
       const spec: AcpRunSpec = {
         command: process.execPath,
-        args: ['--import', tsxLoader, mockServer],
+        args: [mockServer],
         cwd: process.cwd(),
         permission: 'reject',
         // MOCK_HANG so the prompt never resolves on its own — we tear down a live
@@ -249,7 +245,7 @@ describe('dsh-subagent-acp', () => {
         // wider grace.
         env: {
           MOCK_HANG: '1', MOCK_TEXT: 'x', MOCK_READY_FILE: ready,
-          MOCK_FLUSH_ON_EOF: flushed, MOCK_FLUSH_DELAY_MS: '400', TSX_TSCONFIG_PATH: repoTsconfig,
+          MOCK_FLUSH_ON_EOF: flushed, MOCK_FLUSH_DELAY_MS: '400',
         },
         disposeEofGraceMs: 2000,
         disposeGraceMs: 50,
@@ -280,12 +276,12 @@ describe('dsh-subagent-acp', () => {
     try {
       const spec: AcpRunSpec = {
         command: process.execPath,
-        args: ['--import', tsxLoader, mockServer],
+        args: [mockServer],
         cwd: process.cwd(),
         permission: 'reject',
         env: {
           MOCK_HANG: '1', MOCK_IGNORE_EOF: '1', MOCK_TEXT: 'x',
-          MOCK_READY_FILE: ready, MOCK_SIGTERM_FILE: sigterm, TSX_TSCONFIG_PATH: repoTsconfig,
+          MOCK_READY_FILE: ready, MOCK_SIGTERM_FILE: sigterm,
         },
         // Tiny EOF grace so the ignored-EOF window elapses fast, then SIGTERM.
         disposeEofGraceMs: 150,
@@ -404,9 +400,9 @@ describe('dsh-subagent-acp', () => {
       await ctx.plugin(acp, {
         providerName: 'acp',
         command: process.execPath,
-        args: ['--import', tsxLoader, mockServer],
+        args: [mockServer],
         permission: 'reject',
-        env: { MOCK_TRAP_SIGTERM: '1', MOCK_TEXT: 'x', MOCK_READY_FILE: ready, TSX_TSCONFIG_PATH: repoTsconfig },
+        env: { MOCK_TRAP_SIGTERM: '1', MOCK_TEXT: 'x', MOCK_READY_FILE: ready },
         disposeEofGraceMs: 150,
         disposeGraceMs: 150,
       })
@@ -455,10 +451,10 @@ describe('dsh-subagent-acp', () => {
       request(),
       {
         command: process.execPath,
-        args: ['--import', tsxLoader, mockServer],
+        args: [mockServer],
         cwd: process.cwd(),
         permission: 'reject',
-        env: { MOCK_CRASH_ON_PROMPT: '1', TSX_TSCONFIG_PATH: repoTsconfig },
+        env: { MOCK_CRASH_ON_PROMPT: '1' },
         disposeEofGraceMs: DEFAULT_DISPOSE_EOF_GRACE_MS,
         disposeGraceMs: DEFAULT_DISPOSE_GRACE_MS,
         onError: (error, stopReason) => { errors.push({ message: error.message, stopReason }) },
@@ -493,10 +489,10 @@ describe('dsh-subagent-acp', () => {
       request(),
       {
         command: process.execPath,
-        args: ['--import', tsxLoader, mockServer],
+        args: [mockServer],
         cwd: process.cwd(),
         permission: 'reject',
-        env: { MOCK_CRASH_ON_PROMPT: '1', TSX_TSCONFIG_PATH: repoTsconfig },
+        env: { MOCK_CRASH_ON_PROMPT: '1' },
         disposeEofGraceMs: DEFAULT_DISPOSE_EOF_GRACE_MS,
         disposeGraceMs: DEFAULT_DISPOSE_GRACE_MS,
         onError: () => { throw new Error('sink boom') },
diff --git a/packages/subagent/subagent-acp/tsconfig.json b/packages/subagent/subagent-acp/tsconfig.json
index e415ace1de..5aa28528ac 100644
--- a/packages/subagent/subagent-acp/tsconfig.json
+++ b/packages/subagent/subagent-acp/tsconfig.json
@@ -28,6 +28,9 @@
     },
     {
       "path": "../subagent-subprocess"
+    },
+    {
+      "path": "../../support/loader-smoke"
     }
   ]
 }
diff --git a/packages/subagent/subagent-fork/package.json b/packages/subagent/subagent-fork/package.json
index 8794884518..63f548d217 100644
--- a/packages/subagent/subagent-fork/package.json
+++ b/packages/subagent/subagent-fork/package.json
@@ -34,14 +34,13 @@
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-invariants": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
     "@deepseek-ai/dsh-subagent": "workspace:^",
     "@deepseek-ai/dsh-subagent-inprocess": "workspace:^",
     "@deepseek-ai/dsh-subagent-spawn": "workspace:^",
-    "@deepseek-ai/dsh-system-prompt": "workspace:^",
-    "@deepseek-ai/dsh-tools": "workspace:^",
     "@cordisjs/plugin-loader": "^1.0.0-rc.5",
     "cordis": "^4.0.0-rc.7"
   }
diff --git a/packages/subagent/subagent-fork/tests/multi-subagent.spec.ts b/packages/subagent/subagent-fork/tests/multi-subagent.spec.ts
index 8060a77fd4..4a2c6b9133 100644
--- a/packages/subagent/subagent-fork/tests/multi-subagent.spec.ts
+++ b/packages/subagent/subagent-fork/tests/multi-subagent.spec.ts
@@ -1,11 +1,8 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import { AgentId } from '@deepseek-ai/dsh-agent'
 import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as Invariants from '@deepseek-ai/dsh-invariants'
 import SubagentService, { type SubagentStartRequest } from '@deepseek-ai/dsh-subagent'
 import * as Spawn from '@deepseek-ai/dsh-subagent-spawn'
@@ -26,18 +23,14 @@ function start(ctx: Context, provider: string, request: Omit<SubagentStartReques
  */
 async function setup(script: Script) {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(Invariants)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(SubagentService)
   await ctx.plugin(Spawn, { providerName: 'spawn' })
   await ctx.plugin(fork, { providerName: 'fork' })
   ctx.llm.registerAdapter(['mock'], new MockAdapter(script))
-  const parent = ctx.agentLoop.create(AgentId('parent'), { model: 'mock' })
+  const parent = ctx.agentLoop.create(AgentId('parent'), { provider: 'mock', model: 'mock' })
   return { ctx, parent }
 }
 
diff --git a/packages/subagent/subagent-fork/tests/subagent-fork.spec.ts b/packages/subagent/subagent-fork/tests/subagent-fork.spec.ts
index b5e129e54f..f1ca8f0b75 100644
--- a/packages/subagent/subagent-fork/tests/subagent-fork.spec.ts
+++ b/packages/subagent/subagent-fork/tests/subagent-fork.spec.ts
@@ -1,12 +1,9 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
 import Loader from '@cordisjs/plugin-loader'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
 import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
 import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as Invariants from '@deepseek-ai/dsh-invariants'
 import SubagentService, { type SubagentStartRequest } from '@deepseek-ai/dsh-subagent'
 import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
@@ -33,17 +30,13 @@ const emptyStop: StreamChunk[] = [{ type: 'finish', reason: { kind: 'stop' } }]
  */
 async function setup(script: Script) {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(Invariants)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(SubagentService)
   await ctx.plugin(fork, { providerName: 'fork' })
   ctx.llm.registerAdapter(['mock'], new MockAdapter(script))
-  const parent = ctx.agentLoop.create(AgentId('parent'), { model: 'mock' })
+  const parent = ctx.agentLoop.create(AgentId('parent'), { provider: 'mock', model: 'mock' })
   return { ctx, parent }
 }
 
diff --git a/packages/subagent/subagent-inprocess/package.json b/packages/subagent/subagent-inprocess/package.json
index aa80dcac3e..b9397e300c 100644
--- a/packages/subagent/subagent-inprocess/package.json
+++ b/packages/subagent/subagent-inprocess/package.json
@@ -33,6 +33,7 @@
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-invariants": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
diff --git a/packages/subagent/subagent-inprocess/src/index.ts b/packages/subagent/subagent-inprocess/src/index.ts
index f6de7200cf..8ec8a6aac2 100644
--- a/packages/subagent/subagent-inprocess/src/index.ts
+++ b/packages/subagent/subagent-inprocess/src/index.ts
@@ -107,8 +107,10 @@ export async function startInProcessRun(
   const childId = AgentId(randomUUID())
   const seedLength = options.seed?.length ?? 0
   const parentHeader = parent.session.header
+  const parentProvider = parent.options.provider
   const parentModel = parent.options.model
   const agentOptions: AgentOptions = {
+    ...parentProvider !== undefined ? { provider: parentProvider } : {},
     ...parentModel !== undefined ? { model: parentModel } : {},
     ...request.agentOptions,
     subagentDepth: childDepth,
diff --git a/packages/subagent/subagent-inprocess/tests/structured.spec.ts b/packages/subagent/subagent-inprocess/tests/structured.spec.ts
index a2c55995c2..4222dde760 100644
--- a/packages/subagent/subagent-inprocess/tests/structured.spec.ts
+++ b/packages/subagent/subagent-inprocess/tests/structured.spec.ts
@@ -1,12 +1,10 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import LlmService, { CallId, type ContentBlock, type GenerateOptions } from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import { CallId, type ContentBlock, type GenerateOptions } from '@deepseek-ai/dsh-llm'
+import { AgentId } from '@deepseek-ai/dsh-agent'
 import type { ContinuationDecision } from '@deepseek-ai/dsh-agent'
 import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as Invariants from '@deepseek-ai/dsh-invariants'
 import SubagentService, { type SubagentStartRequest } from '@deepseek-ai/dsh-subagent'
 import type { Config as ToolConfig, StructuredOutputSchema } from '@deepseek-ai/dsh-tools'
@@ -43,10 +41,9 @@ const SCHEMA: StructuredOutputSchema = {
 async function setup(script: Script, options: SetupOptions = {}) {
   const ctx = new Context()
   const adapter = new MockAdapter(script)
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry, { mode: options.toolMode ?? 'native' })
+  await mountAgentLoopTestDependencies(ctx, {
+    tools: { mode: options.toolMode ?? 'native' },
+  })
   if (options.toolMode === 'code' || options.toolMode === 'both') {
     ctx.provide('codeRuntime', {
       language: 'typescript',
@@ -54,7 +51,6 @@ async function setup(script: Script, options: SetupOptions = {}) {
       run: options.codeRun ?? (() => Promise.resolve({ logs: [] })),
     } as never)
   }
-  await ctx.plugin(AgentRegistry)
   await ctx.plugin(Invariants)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(SubagentService)
@@ -65,7 +61,7 @@ async function setup(script: Script, options: SetupOptions = {}) {
     start: (request: SubagentStartRequest) => startInProcessRun(request, {}),
   })
   ctx.llm.registerAdapter(['mock'], adapter)
-  const parent = ctx.agentLoop.create(AgentId('parent'), { model: 'mock' })
+  const parent = ctx.agentLoop.create(AgentId('parent'), { provider: 'mock', model: 'mock' })
   return { ctx, parent, adapter, disposeProvider }
 }
 
diff --git a/packages/subagent/subagent-inprocess/tests/subagent-inprocess.spec.ts b/packages/subagent/subagent-inprocess/tests/subagent-inprocess.spec.ts
index 0005246523..eeab453e40 100644
--- a/packages/subagent/subagent-inprocess/tests/subagent-inprocess.spec.ts
+++ b/packages/subagent/subagent-inprocess/tests/subagent-inprocess.spec.ts
@@ -1,11 +1,8 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId, type Agent } from '@deepseek-ai/dsh-agent'
+import { AgentId, type Agent } from '@deepseek-ai/dsh-agent'
 import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as Invariants from '@deepseek-ai/dsh-invariants'
 import SubagentService from '@deepseek-ai/dsh-subagent'
 import { MockAdapter, textResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
@@ -15,16 +12,12 @@ type Script = ConstructorParameters<typeof MockAdapter>[0]
 
 async function setup(script: Script) {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(Invariants)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(SubagentService)
   ctx.llm.registerAdapter(['mock'], new MockAdapter(script))
-  const parent = ctx.agentLoop.create(AgentId('parent'), { model: 'mock' })
+  const parent = ctx.agentLoop.create(AgentId('parent'), { provider: 'mock', model: 'mock' })
   return { ctx, parent }
 }
 
diff --git a/packages/subagent/subagent-spawn/package.json b/packages/subagent/subagent-spawn/package.json
index f2500a4a56..1a986e69aa 100644
--- a/packages/subagent/subagent-spawn/package.json
+++ b/packages/subagent/subagent-spawn/package.json
@@ -32,6 +32,7 @@
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-bash-local": "workspace:^",
     "@deepseek-ai/dsh-invariants": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
@@ -39,10 +40,8 @@
     "@deepseek-ai/dsh-session": "workspace:^",
     "@deepseek-ai/dsh-subagent": "workspace:^",
     "@deepseek-ai/dsh-subagent-inprocess": "workspace:^",
-    "@deepseek-ai/dsh-system-prompt": "workspace:^",
     "@deepseek-ai/dsh-tool-bash": "workspace:^",
     "@deepseek-ai/dsh-tool-subagent": "workspace:^",
-    "@deepseek-ai/dsh-tools": "workspace:^",
     "@cordisjs/plugin-loader": "^1.0.0-rc.5",
     "cordis": "^4.0.0-rc.7"
   }
diff --git a/packages/subagent/subagent-spawn/tests/harness.ts b/packages/subagent/subagent-spawn/tests/harness.ts
index e4ee2e6ab7..3b9342a90e 100644
--- a/packages/subagent/subagent-spawn/tests/harness.ts
+++ b/packages/subagent/subagent-spawn/tests/harness.ts
@@ -1,10 +1,7 @@
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { type Agent } from '@deepseek-ai/dsh-agent'
+import type { Agent } from '@deepseek-ai/dsh-agent'
 import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
 import * as ToolBash from '@deepseek-ai/dsh-tool-bash'
 import * as LlmDeepSeek from '@deepseek-ai/dsh-llm-deepseek'
@@ -21,17 +18,15 @@ import * as ToolSubagent from '@deepseek-ai/dsh-tool-subagent'
  */
 export async function spawnHarness(workdir: string): Promise<Context> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
   // This harness installs only the global default persona, so both parent and
   // spawned children render it. It stays neutral for both roles; the
   // delegation nudge lives in the e2e's user prompt and the subagent tool's
   // own description.
-  await ctx.plugin(SystemPrompt, { persona: 'You are a coding agent. Report only when the requested work is done.' })
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx, {
+    systemPrompt: { persona: 'You are a coding agent. Report only when the requested work is done.' },
+  })
   await ctx.plugin(AgentLoop, { agents: [] })
-  await ctx.plugin(LlmDeepSeek, { models: ['deepseek-v4-flash'] })
+  await ctx.plugin(LlmDeepSeek)
   await ctx.plugin(LocalBashExecutor, { cwd: workdir, timeoutMs: 30_000 })
   await ctx.plugin(ToolBash)
   await ctx.plugin(SubagentService)
diff --git a/packages/subagent/subagent-spawn/tests/spawn.e2e.ts b/packages/subagent/subagent-spawn/tests/spawn.e2e.ts
index daa032199e..1b278c9748 100644
--- a/packages/subagent/subagent-spawn/tests/spawn.e2e.ts
+++ b/packages/subagent/subagent-spawn/tests/spawn.e2e.ts
@@ -29,7 +29,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('spawn backend with-key smoke', (
   it('a parent delegates to a child that writes a file on disk', async () => {
     workdir = await mkdtemp(join(tmpdir(), 'dsh-subagent-spawn-e2e-'))
     ctx = await spawnHarness(workdir)
-    const parent = ctx.agentLoop.create(AgentId('e2e-parent'), { model: 'deepseek-v4-flash' })
+    const parent = ctx.agentLoop.create(AgentId('e2e-parent'), { provider: 'deepseek', model: 'deepseek-v4-flash' })
 
     parent.send([{ type: 'text', text:
       'Use the subagent tool to delegate this exact task: "Use the bash tool to write the text '
diff --git a/packages/subagent/subagent-spawn/tests/subagent-spawn.spec.ts b/packages/subagent/subagent-spawn/tests/subagent-spawn.spec.ts
index 43ec8ee1ac..147b83dd9a 100644
--- a/packages/subagent/subagent-spawn/tests/subagent-spawn.spec.ts
+++ b/packages/subagent/subagent-spawn/tests/subagent-spawn.spec.ts
@@ -1,13 +1,10 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
 import Loader from '@cordisjs/plugin-loader'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
 import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
 import { SessionId } from '@deepseek-ai/dsh-session'
 import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as Invariants from '@deepseek-ai/dsh-invariants'
 import SubagentService, { type SubagentStartRequest } from '@deepseek-ai/dsh-subagent'
 import { MockAdapter, maxTokensResponse, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
@@ -26,17 +23,13 @@ type Script = ConstructorParameters<typeof MockAdapter>[0]
 async function setup(script: Script) {
   const ctx = new Context()
   const adapter = new MockAdapter(script)
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(Invariants)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(SubagentService)
   await ctx.plugin(spawn, { providerName: 'spawn' })
   ctx.llm.registerAdapter(['mock'], adapter)
-  const parent = ctx.agentLoop.create(AgentId('parent'), { model: 'mock' })
+  const parent = ctx.agentLoop.create(AgentId('parent'), { provider: 'mock', model: 'mock' })
   return { ctx, parent, adapter }
 }
 
@@ -235,7 +228,7 @@ describe('dsh-subagent-spawn', () => {
       agentId: AgentId('cwd-parent'),
       sessionId: SessionId('cwd-parent-session'),
       meta: { cwd: '/tmp/parent-workspace' },
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     const run = await start(ctx, 'spawn', { prompt: [{ type: 'text', text: 'p' }], parent: parentHandle.agent })
     await run.result
@@ -257,7 +250,7 @@ describe('dsh-subagent-spawn', () => {
     const run = await start(ctx, 'spawn', {
       prompt: [{ type: 'text', text: 'p' }],
       parent: parentHandle.agent,
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     const result = await run.result
     expect(result.stopReason).toBe('completed')
@@ -303,17 +296,13 @@ describe('dsh-subagent-spawn', () => {
     // Rebuild the stack by hand so we hold the backend's fiber.
     const ctx = new Context()
     const adapter = new MockAdapter(['hang'])
-    await ctx.plugin(LlmService)
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
+    await mountAgentLoopTestDependencies(ctx)
     await ctx.plugin(Invariants)
     await ctx.plugin(AgentLoop, { agents: [] })
     await ctx.plugin(SubagentService)
     const fiber = await ctx.plugin(spawn, { providerName: 'spawn' })
     ctx.llm.registerAdapter(['mock'], adapter)
-    const parent = ctx.agentLoop.create(AgentId('parent'), { model: 'mock' })
+    const parent = ctx.agentLoop.create(AgentId('parent'), { provider: 'mock', model: 'mock' })
     const controller = new AbortController()
     const run = await start(ctx, 'spawn', {
       prompt: [{ type: 'text', text: 'q' }],
@@ -336,15 +325,11 @@ describe('dsh-subagent-spawn', () => {
 
   it('a start racing an already-unloading backend cannot begin child creation', async () => {
     const ctx = new Context()
-    await ctx.plugin(LlmService)
-    await ctx.plugin(SessionStore)
-    await ctx.plugin(SystemPrompt)
-    await ctx.plugin(ToolRegistry)
-    await ctx.plugin(AgentRegistry)
+    await mountAgentLoopTestDependencies(ctx)
     await ctx.plugin(AgentLoop, { agents: [] })
     await ctx.plugin(SubagentService)
     const fiber = await ctx.plugin(spawn, { providerName: 'spawn' })
-    const parent = ctx.agentLoop.create(AgentId('parent'), { model: 'mock' })
+    const parent = ctx.agentLoop.create(AgentId('parent'), { provider: 'mock', model: 'mock' })
     const parentEffects = parent.ctx.fiber.getEffects().length
     const published: string[] = []
     ctx.on('session/created', () => void published.push('session/created'))
@@ -439,7 +424,7 @@ describe('dsh-subagent-spawn', () => {
     const parentHandle = await ctx.agents.create({
       agentId: AgentId('doomed-parent'),
       sessionId: SessionId('doomed-s'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     await parentHandle.dispose()
     const before = ctx.agents.list().length
@@ -462,7 +447,7 @@ describe('dsh-subagent-spawn', () => {
     const parentHandle = await ctx.agents.create({
       agentId: AgentId('setup-race-parent'),
       sessionId: SessionId('setup-race-parent-session'),
-      agentOptions: { model: 'mock' },
+      agentOptions: { provider: 'mock', model: 'mock' },
     })
     const published: string[] = []
     ctx.on('session/created', () => void published.push('session/created'))
diff --git a/packages/subagent/subagent/README.md b/packages/subagent/subagent/README.md
index 261effb450..6f1f5fa40a 100644
--- a/packages/subagent/subagent/README.md
+++ b/packages/subagent/subagent/README.md
@@ -58,13 +58,13 @@ Provider additions and removals also emit `subagent/provider-added` and `subagen
 
 ## Collection model
 
-The current model-facing tool collects synchronously: it awaits the child result and disposes the run before returning. Background collection and polling remain outside this seam. See the [capability-seam RFC](../../../docs/rfc/implemented/feature/2026-06-21-subagent-capability-seam.md) and `src/types.ts` for the complete contracts.
+The model-facing tool collects synchronously by default: it awaits the child result and disposes the run before returning. Background delegation does not change this seam; the consumer registers startup and the eventual run with the generic `ctx.tasks` runtime, then collection and cancellation use the shared task tools. See the [background subagent tasks RFC](../../../docs/rfc/implemented/feature/2026-07-08-background-subagent-tasks.md), the [capability-seam RFC](../../../docs/rfc/implemented/feature/2026-06-21-subagent-capability-seam.md), and `src/types.ts` for the complete contracts.
 
 ## Model Experience
 
-Indirectly, through `dsh-tool-subagent`, which retains only a provider's data-dependent final output or exact `Error: no subagent provider registered for "<name>"`, `Error: subagent provider "<name>" does not support the "<capability>" capability`, and `Error: <message>` start failures in the parent while child working tokens remain child-only.
+Indirectly, through `dsh-tool-subagent`, which renders provider-specific schemas and foreground or generic-background results while child working context remains child-only.
 
 ## Known Limitations and Deferred Work
 
-- **The current consumer collects synchronously** — the model-facing tool starts a run and awaits `result`; steering (`sendMessage`) is part of the seam but intentionally unused, and background/poll/spill semantics are deferred to a future long-running-runtime design.
-- **The lifecycle events are observe-only** — a run-affecting `subagent/end` continuation or decision surface is deferred until a consumer needs one.
+- **Runtime steering and continuation are seam-only capabilities** — `sendMessage` and `resume` have no model-facing consumer in the current tool.
+- **Lifecycle events are observe-only** — a run-affecting `subagent/end` continuation or decision surface waits for a concrete consumer.
diff --git a/packages/subagent/subagent/src/index.ts b/packages/subagent/subagent/src/index.ts
index 8657e7d06e..0d156ae725 100644
--- a/packages/subagent/subagent/src/index.ts
+++ b/packages/subagent/subagent/src/index.ts
@@ -4,6 +4,22 @@
  * child before returning its run, so fulfillment is the single publication and
  * ownership-transfer boundary.
  *
+ * Unlike the bash seam (one executor per context, second load throws), MULTIPLE
+ * providers coexist here: each registers under a unique name and a caller picks
+ * one by name. The shape mirrors the LLM adapter registry
+ * (`LlmService.registerAdapter`), not the single-service bash executor.
+ *
+ * This package is the INTERFACE third of the capability seam. Implementations
+ * (`@deepseek-ai/dsh-subagent-spawn`, `-fork`, `-acp`) and the model-facing
+ * consumer (`@deepseek-ai/dsh-tool-subagent`) are separate packages.
+ *
+ * Scope: the seam stays collection-agnostic — a run is started and its
+ * `result` awaited, whether the consumer blocks on it (foreground) or
+ * registers it as a `ctx.tasks` background task (the generic runtime owns
+ * ids/polling/stop; this seam gains nothing task-shaped). Steering
+ * ({@link SubagentRun.sendMessage}) is part of the contract but intentionally
+ * unused.
+ *
  * Same-process providers are trusted typed collaborators. Requests, provider
  * descriptors, results, and lifecycle payloads are borrowed immutable values;
  * serialization and hostile-input validation belong at real process, worker,
diff --git a/packages/subagent/tool-subagent/README.md b/packages/subagent/tool-subagent/README.md
index e981788275..825e252923 100644
--- a/packages/subagent/tool-subagent/README.md
+++ b/packages/subagent/tool-subagent/README.md
@@ -1,66 +1,55 @@
 # @deepseek-ai/dsh-tool-subagent
 
-The `subagent` tool lets the model delegate one self-contained task and collect the child's final output. It is a thin consumer of `ctx.subagents`; changing the configured provider changes the transport without changing the model-facing execution contract.
+The model-facing delegation tool over one configured `ctx.subagents` provider. Changing the provider changes transport without changing the execution contract.
 
-## Provider selection
+## Provider selection and lifecycle
 
-Each plugin instance binds to exactly one provider. The model sees `{ description, prompt }`, not a provider selector. To expose multiple transports, load the plugin multiple times with distinct `toolName` values.
+Each plugin instance binds one `provider` to one `toolName`; the model receives no provider selector. Load another distinctly named instance to expose another transport. The tool registers only while its provider exists, avoiding sibling load-order and provider-reload dependencies. Its description follows `provider.inheritsParentContext`: fresh children require standalone prompts, while forked children already see completed parent turns.
 
-The description is derived from `provider.inheritsParentContext`: spawn and ACP tell the model to provide a standalone prompt, while fork says the child already sees completed conversation turns. The plugin follows `subagent/provider-added` and `subagent/provider-removed`, so concurrent Cordis plugin loading does not create a registration-order dependency.
+A foreground call passes the execution signal through startup and execution, awaits `run.result`, and always awaits `run.dispose()` before returning. Only `completed` returns final text; abort, refusal, token limit, and other failures become errored tool results without partial output.
 
-## Lifecycle
+With `run_in_background: true`, the tool registers the parent-owned task before starting the provider. A task-owned signal covers pending startup and the child after the starting call returns. `task_kill` and owner disposal abort it. Settlement awaits startup rollback or child disposal, then maps completed final text, abort to `killed`, and other failures to `failed`. The task has no incremental read; generic task tools own later status, collection, cancellation, and notices. See the [background subagent RFC](../../../docs/rfc/implemented/feature/2026-07-08-background-subagent-tasks.md).
 
-`execute` passes the tool execution's abort signal when present, otherwise supplies an inert signal to satisfy the required `SubagentStartRequest.signal`. It awaits `ctx.subagents.start(...)`, then awaits `run.result` inside a `try/finally` that always calls `run.dispose()`. The selected signal therefore covers startup and live execution, while disposal guarantees quiescence on success, failure, and abort.
-
-A non-`completed` stop reason becomes an `isError` tool result; partial child output is never reported as success. The current tool blocks the parent turn until collection finishes; background and polling modes are deferred.
+`toolFilter` changes the child's global tool layer but is not a parent-derived authority ceiling. See the [agent-scope security non-goal](../../../docs/rfc/implemented/architecture/2026-07-08-agent-scope-contexts.md#security-and-authority-are-non-goals).
 
 ## Config
 
 | Key | Meaning |
 |---|---|
-| `provider` | Required `ctx.subagents` provider name. |
-| `toolName` | Model-facing tool name (default `subagent`). Must be unique per plugin instance. |
-| `agentOptions` | Default child agent options, currently including `model`. |
+| `provider` (required) | Provider name (`spawn`, `fork`, `acp`, ...). |
+| `toolName` | Model-facing name, default `subagent`; distinct for every loaded instance. |
+| `enableRunInBackground` | Exposes background mode, default `true`; disabling also rejects forced background calls. |
+| `agentOptions` | Default child options, currently including `model`. |
 | `persona` | Per-child persona; requires provider `persona` capability. |
-| `toolFilter` | Per-child global-tool restriction; requires provider `toolFilter` capability. |
-| `maxDepth` | Absolute delegation-depth cap; requires provider `depthLimit` capability. |
+| `toolFilter` | Per-child global-tool restriction; requires `toolFilter` capability. |
+| `maxDepth` | Absolute delegation-depth cap; requires `depthLimit` capability. |
 
-`toolFilter` changes the child's visible global tool layer; it is not a parent-derived authority ceiling. See the [agent-scope security non-goal](../../../docs/rfc/implemented/architecture/2026-07-08-agent-scope-contexts.md#security-and-authority-are-explicit-non-goals).
+## Concurrency
+
+Foreground and background calls are exclusive. Children may share the parent's workspace or external resources, and a unary classifier cannot prove that sibling delegations have disjoint effects. See the [parallel tool-call RFC](../../../docs/rfc/implemented/feature/2026-07-10-parallel-tool-call-execution.md).
 
 ## Model Experience
 
-### Standalone-provider schema
+### Tool schema
 
-**What the model sees**: While a fresh-context provider exists, the configured tool uses the generated default [`subagent` schema](../../../docs/tool-catalog.md#deepseek-aidsh-tool-subagent); the catalog also records how `toolName` changes the visible name.
+**What the model sees**: The generated default [`subagent` schema](../../../docs/tool-catalog.md#deepseek-aidsh-tool-subagent) under this instance's configured name while its provider exists. Provider context inheritance changes the tool and prompt descriptions; enabled background mode adds `run_in_background`.
 
-**Token effect**: Fixed schema cost per parent request while mounted. Removing the provider removes the whole schema.
+**Token effect**: Fixed schema cost per parent request; each provider instance adds one schema.
 
-### Inherited-context-provider schema
+### Foreground result
 
-**What the model sees**: Relative to the generated default [`subagent` schema](../../../docs/tool-catalog.md#deepseek-aidsh-tool-subagent), a provider that seeds completed turns replaces only the tool and `prompt` parameter descriptions with the text below; the shape and `description` parameter stay unchanged.
+**What the model sees**: The call retains the description and prompt. Success contains only the child's final text; other outcomes become `Error: <message>`. Intermediate child steps stay out of the parent.
 
-**Token effect**: Fixed schema cost per parent request while mounted. Exposing multiple providers adds one independently named schema per load.
+**Token effect**: The prompt and result remain in parent history until compaction; child working context remains in the child.
 
-#### Inherited-context-provider tool description
+### Background task result
 
-```markdown
-Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all completed turns so far (it does not see the current in-flight turn), returning only its final result. Use this when the subtask builds on this conversation's context — a follow-up analysis, a review, a continuation — without consuming this conversation's context for the work itself. You receive only its final answer, not its intermediate steps.
-```
+**What the model sees**: Start returns exactly `started background subagent task <id>`. The generic task surface provides later status, final output, cancellation responses, and notices.
 
-#### Inherited-context-provider prompt description
-
-```markdown
-The task for the subagent. It already sees this conversation's completed turns, so build on them freely and state only what is new.
-```
-
-### Tool-call history and result
-
-**What the model sees**: The task description and full prompt remain in the parent assistant tool call. Success contains only the child's data-dependent final text. Other stop reasons become exactly `Error: subagent run was cancelled`, `Error: subagent run failed`, `Error: subagent run hit its token limit before finishing`, `Error: subagent declined the task`, or `Error: subagent run ended abnormally (<reason>)`; a call without an owning agent becomes `Error: subagent tool requires a calling agent (exec.agent was undefined)`. Intermediate child steps never enter the parent.
-
-**Token effect**: Prompt and final output are data-dependent retained tokens. All child working context is paid in the child and omitted from the parent.
+**Token effect**: The acknowledgement is retained; final output enters parent history only when collected or injected.
 
 ## Known Limitations and Deferred Work
 
-- **Delegation blocks the parent turn** — synchronous collect only; background start and poll collection are deferred to the long-running-runtime redesign.
-- **Duplicate `toolName` across waiting loads is detected late** (`TODO(subagent-dup-toolname)`) — two loads waiting on providers collide only when a provider arrives, and the throw rolls back the provider's fiber rather than the misconfigured tool's; config-time detection needs a cross-fiber registry of intended names.
-- **Child policy is fixed per tool registration** — `model`, persona, tool filter, and depth cap come from this plugin load's config, not model-call arguments; exposing another policy requires another distinctly named tool.
+- **Background runs expose final output only** — intermediate child steps stay in the child session.
+- **Duplicate names across waiting instances are detected late** (`TODO(subagent-dup-toolname)`) — preventing provider-registration rollback requires a registry of intended names.
+- **Child policy is fixed per instance** — another model, persona, tool filter, or depth cap requires another distinctly named tool.
diff --git a/packages/subagent/tool-subagent/package.json b/packages/subagent/tool-subagent/package.json
index 88d4f8e4f3..c52f5a74dc 100644
--- a/packages/subagent/tool-subagent/package.json
+++ b/packages/subagent/tool-subagent/package.json
@@ -25,6 +25,7 @@
     "@deepseek-ai/dsh-agent": "^0.0.1",
     "@deepseek-ai/dsh-llm": "^0.0.1",
     "@deepseek-ai/dsh-subagent": "^0.0.1",
+    "@deepseek-ai/dsh-tasks": "^0.0.1",
     "@deepseek-ai/dsh-tools": "^0.0.1",
     "cordis": "^4.0.0-rc.7"
   },
@@ -32,13 +33,15 @@
     "schemastery": "^3.18.0"
   },
   "devDependencies": {
+    "@cordisjs/plugin-loader": "^1.0.0-rc.5",
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-subagent": "workspace:^",
     "@deepseek-ai/dsh-subagent-mock": "workspace:^",
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-tasks": "workspace:^",
+    "@deepseek-ai/dsh-tool-tasks": "workspace:^",
     "@deepseek-ai/dsh-tools": "workspace:^",
-    "@cordisjs/plugin-loader": "^1.0.0-rc.5",
     "cordis": "^4.0.0-rc.7"
   }
 }
diff --git a/packages/subagent/tool-subagent/src/index.ts b/packages/subagent/tool-subagent/src/index.ts
index ef75d1fe97..fdfaacef08 100644
--- a/packages/subagent/tool-subagent/src/index.ts
+++ b/packages/subagent/tool-subagent/src/index.ts
@@ -1,21 +1,20 @@
 /**
- * Model-facing delegation tool bound by configuration to one provider; transport selection is not
- * exposed in its `{ description, prompt }` schema. Provider lifecycle controls registration and
- * re-derives conversation-history wording after reload, so load order is irrelevant.
- *
- * Execution synchronously awaits the child result and always disposes the run. Non-completed stop
- * reasons become error results, while transport details remain behind `ctx.subagents`. Load this
- * plugin more than once to expose multiple configured providers.
+ * Model-facing delegation through one configured `ctx.subagents` provider.
+ * Provider lifecycle controls tool registration and context-sensitive schema
+ * wording. Foreground calls always dispose the run after collection; background
+ * calls use an independent cancellation signal and settle a final-output task
+ * only after child disposal.
  * @module @deepseek-ai/dsh-tool-subagent
  */
 
 import type { Context } from 'cordis'
 import z from 'schemastery'
 import { defineTool } from '@deepseek-ai/dsh-tools'
-import type { AgentOptions } from '@deepseek-ai/dsh-agent'
+import type { Agent, AgentOptions } from '@deepseek-ai/dsh-agent'
 import type { ContentBlock } from '@deepseek-ai/dsh-llm'
 import { assertSubagentMaxDepth } from '@deepseek-ai/dsh-subagent'
 import type { SubagentProvider, SubagentResult, SubagentRun, SubagentStartRequest } from '@deepseek-ai/dsh-subagent'
+import type { TaskOutcome } from '@deepseek-ai/dsh-tasks'
 
 export const name = 'tool-subagent'
 export const inject = ['tools', 'subagents']
@@ -25,34 +24,29 @@ export interface Config {
   /** The `ctx.subagents` provider name to start runs on (e.g. `spawn`, `acp`). */
   provider: string
   /**
-   * The model-facing tool name to register (default `subagent`). To expose more
-   * than one transport, load this plugin once per provider — each load MUST set
-   * a distinct `toolName` (the tool registry rejects a duplicate name), e.g.
-   * `{ provider: 'spawn', toolName: 'subagent' }` and
-   * `{ provider: 'acp', toolName: 'subagent_acp' }`.
+   * Model-facing tool name (default `subagent`). Each loaded instance must use
+   * a distinct name.
    */
   toolName?: string
   /**
-   * Default per-child agent options (model) applied to every spawned child.
-   * Omitted fields fall back to the child loop's own defaults.
+   * Expose `run_in_background` (default true). Disabled instances omit the
+   * parameter and reject forced background calls.
+   */
+  enableRunInBackground?: boolean
+  /**
+   * Agent options applied to every child; omitted fields use child-loop defaults.
    */
   agentOptions?: AgentOptions
   /**
-   * Per-child persona applied to every child this tool spawns: a scoped
-   * `deployment:persona` section shadowing the deployment's persona for the
-   * child alone. Requires the bound provider's `persona` capability
-   * (in-process backends support it; a request against one that doesn't is
-   * rejected at start). Omitted ⇒ the child renders the deployment persona.
+   * Per-child persona that shadows `deployment:persona`. Requires the
+   * provider's `persona` capability; omission preserves the deployment persona.
    */
   persona?: string
   /**
-   * Tool scoping applied to every child this tool spawns (see
-   * `SubagentStartRequest.toolFilter`): the named global tools vanish from
-   * the child's prompt AND refuse to execute. Requires the provider's
-   * `toolFilter` capability. Unknown names fail the spawn loudly. Note the
-   * child otherwise sees every global tool — including this delegation tool
-   * itself; `deny`-listing it (or setting `maxDepth`) is how a deployment
-   * bounds recursion.
+   * Tool filter applied to every child. Filtered tools disappear from its
+   * prompt and reject execution. Requires the provider's `toolFilter`
+   * capability; unknown names fail startup. Children otherwise see this tool,
+   * so deny it or set `maxDepth` to bound recursion.
    */
   toolFilter?: {
     /** Global tool names the child keeps; everything else is removed. */
@@ -61,12 +55,8 @@ export interface Config {
     deny?: string[]
   }
   /**
-   * Recursion cap applied to every child this tool spawns (see
-   * `SubagentStartRequest.maxDepth`): a spawn whose child would sit deeper
-   * than this in the delegation tree is rejected. Requires the provider's
-   * `depthLimit` capability. Must be a non-negative safe integer and is
-   * validated when the plugin loads. Omitted ⇒ unbounded (bound it in
-   * deployments that expose this tool to children).
+   * Maximum child depth. Requires the provider's `depthLimit` capability and a
+   * non-negative safe integer. Omission is unbounded.
    */
   maxDepth?: number
 }
@@ -74,16 +64,14 @@ export interface Config {
 export const Config: z<Config> = z.object({
   provider: z.string().required(),
   toolName: z.string().default('subagent'),
-  // Omitted-object discipline (see the toolFilter note below): without the
-  // forced default an omitted `agentOptions` materializes `{}`, which reads as
-  // present — the request would carry `agentOptions: {}` and the presence
-  // check in execute() could never be false through config.
+  enableRunInBackground: z.boolean().default(true),
+  // Prevent Schemastery from materializing omitted agentOptions as `{}`.
   agentOptions: z.object({
+    provider: z.string(),
     model: z.string(),
-  }).default(undefined as unknown as { model: string }),
+  }).default(undefined as unknown as { provider: string; model: string }),
   persona: z.string(),
-  // Schemastery otherwise materializes omitted objects and nested arrays as `{ allow: [] }`, which
-  // silently means deny all. Preserve omission while retaining an explicit empty allow-list.
+  // Preserve omission; Schemastery's `{ allow: [] }` default would deny every tool.
   toolFilter: z.object({
     allow: z.array(z.string()).default(undefined as unknown as string[]),
     deny: z.array(z.string()).default(undefined as unknown as string[]),
@@ -93,9 +81,8 @@ export const Config: z<Config> = z.object({
 
 /**
  * Flatten a child's final output blocks to text for the tool result. The child
- * may return non-text blocks; this cut surfaces the text content (the common
- * case) and drops the rest, which is acceptable for a synchronous summary —
- * the structured path (`outputSchema`) is the channel for non-text results.
+ * may return non-text blocks; this path returns only text. Structured results
+ * use `outputSchema`.
  */
 function outputText(blocks: ContentBlock[]): string {
   return blocks
@@ -124,6 +111,50 @@ function stopReasonError(result: SubagentResult): string | undefined {
   }
 }
 
+/**
+ * Map a child result to the task outcome: completed carries final text,
+ * aborted is killed, and every other reason is failed without partial output.
+ * @param result - child terminal result.
+ * @returns outcome for the `ctx.tasks` registration.
+ */
+export function runOutcome(result: SubagentResult): TaskOutcome {
+  switch (result.stopReason) {
+    case 'completed':
+      return { status: 'completed', output: outputText(result.output) }
+    case 'aborted':
+      return { status: 'killed' }
+    case 'error':
+    case 'max-tokens':
+    case 'refusal':
+      return { status: 'failed', detail: result.stopReason }
+    // Merge-extensible reasons remain failures with their raw detail.
+    default:
+      return { status: 'failed', detail: String(result.stopReason) }
+  }
+}
+
+/**
+ * Await the child result, dispose the run, then return its task outcome. Result
+ * and disposal failures become `failed`; when both fail, both details survive.
+ * @param run - live run to settle and release.
+ * @returns outcome after child resources are released.
+ */
+export async function settleRun(run: SubagentRun): Promise<TaskOutcome> {
+  let outcome: TaskOutcome
+  try {
+    outcome = runOutcome(await run.result)
+  } catch (error: unknown) {
+    outcome = { status: 'failed', detail: String(error) }
+  }
+  try {
+    await run.dispose()
+  } catch (error: unknown) {
+    const prefix = outcome.detail === undefined ? '' : `${outcome.detail}; `
+    return { status: 'failed', detail: `${prefix}dispose failed: ${String(error)}` }
+  }
+  return outcome
+}
+
 /**
  * Model-facing wording from the provider's conversation-history descriptor
  * ({@link SubagentProvider.inheritsParentContext}).
@@ -140,7 +171,7 @@ export function providerWording(inheritsConversation: boolean): { description: s
   if (inheritsConversation) {
     return {
       description:
-        'Delegate a task to a subagent that INHERITS this conversation: a child agent seeded with all '
+        'Delegate a task to a subagent that inherits this conversation: a child agent seeded with all '
         + 'completed turns so far (it does not see the current in-flight turn), returning only its final '
         + 'result. Use this when the subtask builds on this conversation\'s context — a follow-up analysis, '
         + 'a review, a continuation — without consuming this conversation\'s context for the work itself. '
@@ -163,30 +194,47 @@ export function providerWording(inheritsConversation: boolean): { description: s
   }
 }
 
+function startRequest(config: Config, prompt: string, parent: Agent, signal: AbortSignal): SubagentStartRequest {
+  return {
+    prompt: [{ type: 'text', text: prompt }],
+    parent,
+    signal,
+    ...config.agentOptions !== undefined ? { agentOptions: config.agentOptions } : {},
+    ...config.persona !== undefined ? { persona: config.persona } : {},
+    ...config.toolFilter !== undefined ? { toolFilter: config.toolFilter } : {},
+    ...config.maxDepth !== undefined ? { maxDepth: config.maxDepth } : {},
+  }
+}
+
+/** Settle pending startup without rejecting the task producer contract. */
+async function settleStart(start: Promise<SubagentRun>, signal: AbortSignal): Promise<TaskOutcome> {
+  try {
+    return await settleRun(await start)
+  } catch (error: unknown) {
+    return signal.aborted
+      ? { status: 'killed' }
+      : { status: 'failed', detail: String(error) }
+  }
+}
+
 export function apply(ctx: Context, config: Config): void {
-  // Keep misconfiguration at plugin load even when a caller invokes apply()
-  // directly and bypasses Schemastery's natural/max metadata.
+  // Direct apply() bypasses Schemastery's numeric constraints.
   assertSubagentMaxDepth(config.maxDepth)
-  // Misconfiguration fails loud AT LOAD (the check is self-contained): an
-  // explicit `toolFilter: {}` would otherwise pass the capability gate and
-  // kill every delegation later, in the child-setup `restrict({})` throw.
+  // Reject an empty explicit filter at load instead of failing every delegation.
   if (config.toolFilter !== undefined && config.toolFilter.allow === undefined && config.toolFilter.deny === undefined) {
     throw new Error('tool-subagent: `toolFilter` is configured but names neither `allow` nor `deny` — remove the key or fill the filter')
   }
-  // The tool MIRRORS its provider's lifecycle instead of assuming load order:
-  // the cordis Loader starts sibling entries concurrently, so "backend listed
-  // first in cordis.yml" does not guarantee "provider registered first", and
-  // an HMR reload of the backend replaces the provider while this fiber stays
-  // loaded. Register the tool when the bound provider is (or becomes)
-  // available — deriving the wording from THAT provider — and unregister it
-  // when the provider goes away, so the description can never outlive or
-  // predate the provider it describes.
+  // Mirror provider lifecycle because sibling load order and HMR replacement
+  // can change provider availability while this fiber remains active.
   let disposeTool: (() => void) | undefined
   const mount = (provider: SubagentProvider): void => {
     const wording = providerWording(provider.inheritsParentContext)
+    const backgroundEnabled = config.enableRunInBackground !== false
     disposeTool = ctx.tools.register(defineTool({
       name: config.toolName ?? 'subagent',
-      description: wording.description,
+      description: wording.description + (backgroundEnabled
+        ? ' Set `run_in_background: true` to return a task id; collect with `task_output` and stop with `task_kill`.'
+        : ''),
       parameters: {
         description: {
           type: 'string',
@@ -198,55 +246,85 @@ export function apply(ctx: Context, config: Config): void {
           required: true,
           description: wording.promptDescription,
         },
+        ...backgroundEnabled ? {
+          run_in_background: {
+            type: 'boolean' as const,
+            description: 'Run as a background task and return its id; collect with task_output or stop with task_kill.',
+          },
+        } : {},
       },
       async execute(args, exec): Promise<ContentBlock[]> {
         const parent = exec.agent
         if (!parent) {
-          // The loop sets `exec.agent` for every model-driven call; its absence
-          // means a non-agent caller invoked the tool directly, which has no
-          // parent to attribute the child to. Fail loud rather than guess.
+          // Non-agent callers provide no parent for delegation ownership.
           throw new Error('subagent tool requires a calling agent (exec.agent was undefined)')
         }
 
-        const request: SubagentStartRequest = {
-          prompt: [{ type: 'text', text: args.prompt }],
-          parent,
-          signal: exec.signal ?? new AbortController().signal,
-          ...config.agentOptions !== undefined ? { agentOptions: config.agentOptions } : {},
-          ...config.persona !== undefined ? { persona: config.persona } : {},
-          ...config.toolFilter !== undefined ? { toolFilter: config.toolFilter } : {},
-          ...config.maxDepth !== undefined ? { maxDepth: config.maxDepth } : {},
+        if (args.run_in_background === true) {
+          // The validator permits undeclared keys, so schema omission also needs
+          // execution-time enforcement.
+          if (!backgroundEnabled) {
+            throw new Error('run_in_background is disabled for this tool instance (enableRunInBackground: false)')
+          }
+          const tasks = ctx.get('tasks')
+          if (tasks === undefined) {
+            throw new Error('background tasks unavailable: load @deepseek-ai/dsh-tasks and @deepseek-ai/dsh-tool-tasks')
+          }
+          // Reject cancellation before spawning; after return, the task-owned
+          // signal covers both pending startup and the ready child.
+          if (exec.signal?.aborted) throw new Error('subagent delegation aborted')
+          // Task preflight finishes before the starter can spawn a child.
+          const id = tasks.start({
+            kind: 'subagent',
+            label: args.description,
+            owner: parent,
+            run: () => {
+              const controller = new AbortController()
+              const start = ctx.subagents.start(
+                config.provider,
+                startRequest(config, args.prompt, parent, controller.signal),
+              )
+              return {
+                cancel: (reason?: string) => {
+                  controller.abort(reason ?? 'background subagent task killed')
+                },
+                done: settleStart(start, controller.signal),
+                // No readOutput: the child session owns intermediate detail.
+              }
+            },
+          })
+          return [{ type: 'text', text: `started background subagent task ${id}` }]
         }
 
+        const request = startRequest(
+          config,
+          args.prompt,
+          parent,
+          exec.signal ?? new AbortController().signal,
+        )
+
         const run: SubagentRun = await ctx.subagents.start(config.provider, request)
 
         try {
           const result = await run.result
           const error = stopReasonError(result)
           if (error !== undefined) {
-            // Map a non-clean finish to an isError result (the registry turns a
-            // throw into an isError). Report the reason, not partial output.
+            // The registry converts this throw to isError; partial output is not success.
             throw new Error(error)
           }
           return [{ type: 'text', text: outputText(result.output) }]
         } finally {
-          // Always reach child quiescence — never leak a live idle child/session.
+          // Dispose before returning so no child session outlives the call.
           await run.dispose()
         }
       },
     }))
   }
 
-  // Listeners first, then the presence check: both run synchronously, so no
-  // registration can slip between them; the `disposeTool === undefined` guard
-  // makes a same-tick added-event after a successful mount a no-op.
+  // Register listeners before checking presence so no synchronous change is missed.
   // TODO(subagent-dup-toolname): two WAITING fibers configured with the same
-  // toolName collide only when their provider finally arrives — the duplicate
-  // tool-name throw then propagates through `subagent/provider-added` and
-  // rolls back the PROVIDER registration, so an invalid config blasts the
-  // backend's fiber instead of the misconfigured tool's. Config-time detection
-  // would need a cross-fiber registry of intended tool names; revisit if a
-  // real deployment ever hits it.
+  // toolName collide when their provider appears, and the duplicate-name throw
+  // rolls back the provider registration. Add an intent registry if this occurs.
   ctx.on('subagent/provider-added', (provider) => {
     if (provider.name === config.provider && disposeTool === undefined) mount(provider)
   })
@@ -259,9 +337,7 @@ export function apply(ctx: Context, config: Config): void {
   if (present !== undefined) {
     mount(present)
   } else {
-    // Not an error: the backend's fiber may activate after this one.
-    // The tool appears the moment the provider registers; a typo'd provider
-    // name shows up as this note plus a tool that never materializes.
+    // A backend fiber may activate later; a misspelled provider remains visible in this log.
     ctx.logger.info(`subagent provider "${config.provider}" not registered yet; the "${config.toolName ?? 'subagent'}" tool will register when it appears`)
   }
 }
diff --git a/packages/subagent/tool-subagent/tests/tool-subagent.spec.ts b/packages/subagent/tool-subagent/tests/tool-subagent.spec.ts
index 5830513cf4..90f84bfceb 100644
--- a/packages/subagent/tool-subagent/tests/tool-subagent.spec.ts
+++ b/packages/subagent/tool-subagent/tests/tool-subagent.spec.ts
@@ -5,9 +5,13 @@ import { CallId } from '@deepseek-ai/dsh-llm'
 import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
 import ToolRegistry from '@deepseek-ai/dsh-tools'
 import { AgentId, type Agent } from '@deepseek-ai/dsh-agent'
+import AgentRegistry from '@deepseek-ai/dsh-agent'
 import SubagentService from '@deepseek-ai/dsh-subagent'
+import TaskService from '@deepseek-ai/dsh-tasks'
+import * as ToolTasks from '@deepseek-ai/dsh-tool-tasks'
 import * as mock from '@deepseek-ai/dsh-subagent-mock'
 import * as tool from '../src/index.ts'
+import { runOutcome, settleRun } from '../src/index.ts'
 
 /**
  * Drives the REAL plugin body: mounts `dsh-tool-subagent` on a real
@@ -60,12 +64,50 @@ describe('dsh-tool-subagent', () => {
     expect(text(result)).toBe('child says hi')
   })
 
-  it('exposes only description + prompt to the model (no provider/type parameter)', async () => {
+  it('exposes description + prompt + run_in_background to the model (no provider/type parameter)', async () => {
     const ctx = await setup({ provider: 'mock' })
     const schema = ctx.tools.schemas().find(s => s.name === 'subagent')
     expect(schema).toBeDefined()
     const props = (schema!.parameters as { properties?: Record<string, unknown> }).properties ?? {}
+    expect(Object.keys(props).sort()).toEqual(['description', 'prompt', 'run_in_background'])
+    expect(schema!.description).toContain('task_output')
+  })
+
+  it('omits run_in_background entirely when the instance disables it (schema and capability never disagree)', async () => {
+    const ctx = await setup({ provider: 'mock', enableRunInBackground: false })
+    const schema = ctx.tools.schemas().find(s => s.name === 'subagent')
+    const props = (schema!.parameters as { properties?: Record<string, unknown> }).properties ?? {}
     expect(Object.keys(props).sort()).toEqual(['description', 'prompt'])
+    expect(schema!.description).not.toContain('task_output')
+  })
+
+  it('refuses a forced run_in_background at execution time when the instance disables it', async () => {
+    // Schema omission is advertising, not enforcement: the arg validator
+    // allows undeclared keys, so the opt-out must also hold in execute().
+    const ctx = await setup({ provider: 'mock', enableRunInBackground: false })
+    const parent = { id: AgentId('agent-sess-off'), inject: () => {}, session: { header: { version: 0, id: 'sess-off', createdAt: 0 } } } as unknown as Agent
+
+    const forced = await callSubagent(ctx, { description: 'd', prompt: 'p', run_in_background: true }, { agent: parent })
+    expect(forced.isError).toBe(true)
+    expect(text(forced)).toContain('run_in_background is disabled for this tool instance')
+    // The provider was never asked to start a child.
+    expect(ctx.subagents.getProvider('mock')).toBeDefined()
+    const foreground = await callSubagent(ctx, { description: 'd', prompt: 'p' }, { agent: parent })
+    expect(foreground.isError).toBe(false)
+  })
+
+  it('keeps foreground and background calls exclusive', async () => {
+    const ctx = await setup({ provider: 'mock' })
+    expect(ctx.tools.executionMode({
+      callId: CallId('subagent-foreground'),
+      name: 'subagent',
+      arguments: { description: 'do work', prompt: 'Reply OK' },
+    })).toEqual({ kind: 'exclusive' })
+    expect(ctx.tools.executionMode({
+      callId: CallId('subagent-background'),
+      name: 'subagent',
+      arguments: { description: 'do work', prompt: 'Reply OK', run_in_background: true },
+    })).toEqual({ kind: 'exclusive' })
   })
 
   it.each([
@@ -225,7 +267,7 @@ describe('dsh-tool-subagent', () => {
     // Backend reloads with a DIFFERENT conversation-history descriptor: the wording is re-derived
     // from the fresh provider, not served stale from the first mount.
     await ctx.plugin(mock, { name: 'mock', inheritsParentContext: true })
-    expect(ctx.tools.schemas().find(s => s.name === 'subagent')!.description).toContain('INHERITS this conversation')
+    expect(ctx.tools.schemas().find(s => s.name === 'subagent')!.description).toContain('inherits this conversation')
   })
 
   it('the tool PLUGIN fiber owns its lifecycle listeners: disposal unmounts, and a disposed fiber never zombie-mounts', async () => {
@@ -275,10 +317,10 @@ describe('dsh-tool-subagent', () => {
     expect(props['prompt']!.description).toContain('include everything it needs')
   })
 
-  it('derives fork-shaped wording from a seeded-conversation provider (the description stops lying)', async () => {
+  it('derives inherited-context wording from a seeded-conversation provider', async () => {
     const ctx = await setup({ provider: 'mock', toolName: 'subagent' }, { inheritsParentContext: true })
     const schema = ctx.tools.schemas().find(s => s.name === 'subagent')!
-    expect(schema.description).toContain('INHERITS this conversation')
+    expect(schema.description).toContain('inherits this conversation')
     expect(schema.description).not.toContain('does not see this conversation')
     const props = (schema.parameters as { properties: Record<string, { description: string }> }).properties
     expect(props['prompt']!.description).toContain('completed turns')
@@ -555,3 +597,268 @@ describe('dsh-tool-subagent', () => {
     await expect(fiber).rejects.toThrow(/names neither `allow` nor `deny`/)
   })
 })
+
+describe('dsh-tool-subagent background mode', () => {
+  /** A live parent with a dedicated scope fiber for structural task cleanup. */
+  function ownerAgent(ctx: Context, sessionId: string, inject: (...args: unknown[]) => void = () => {}): Agent {
+    const scopeFiber = ctx.plugin(() => {})
+    const agent = {
+      id: AgentId(`agent-${sessionId}`),
+      ctx: scopeFiber.ctx,
+      inject,
+      session: { header: { version: 0, id: sessionId, createdAt: 0 } },
+    } as unknown as Agent
+    ctx.agents.register(agent)
+    return agent
+  }
+
+  async function backgroundSetup(toolConfig: tool.Config, mockConfig: Partial<mock.Config> = {}) {
+    const ctx = await setup(toolConfig, mockConfig)
+    await ctx.plugin(AgentRegistry)
+    await ctx.plugin(TaskService)
+    await ctx.plugin(ToolTasks, {})
+    return ctx
+  }
+
+  it('returns a task id immediately and the answer is collected through task_output', async () => {
+    const ctx = await backgroundSetup({ provider: 'mock', agentOptions: { model: 'child-model' } }, { reply: 'background answer' })
+    const parent = ownerAgent(ctx, 'sess-parent')
+
+    const start = await callSubagent(ctx, { description: 'deep research', prompt: 'dig in', run_in_background: true }, { agent: parent })
+    expect(start.isError).toBe(false)
+    expect(text(start)).toBe('started background subagent task subagent-1')
+
+    const collected = await ctx.tools.execute({
+      callId: CallId('collect-1'),
+      name: 'task_output',
+      arguments: { task_id: 'subagent-1', wait: true },
+      agent: parent,
+    })
+    expect(text(collected)).toBe('background answer\n[status: completed]')
+
+    // Final-output reads are idempotent (not consumed).
+    const again = await ctx.tools.execute({
+      callId: CallId('collect-2'),
+      name: 'task_output',
+      arguments: { task_id: 'subagent-1' },
+      agent: parent,
+    })
+    expect(text(again)).toBe('background answer\n[status: completed]')
+  })
+
+  it('fails loud when the tasks runtime is not loaded', async () => {
+    const ctx = await setup({ provider: 'mock' })
+    const result = await callSubagent(ctx, { description: 'd', prompt: 'p', run_in_background: true })
+    expect(result.isError).toBe(true)
+    expect(text(result)).toContain('background tasks unavailable: load @deepseek-ai/dsh-tasks')
+  })
+
+  it('refuses to start when the tool signal is already aborted', async () => {
+    const ctx = await backgroundSetup({ provider: 'mock' })
+    const parent = ownerAgent(ctx, 'sess-parent')
+    const controller = new AbortController()
+    controller.abort()
+    const result = await callSubagent(ctx, { description: 'd', prompt: 'p', run_in_background: true }, { agent: parent, signal: controller.signal })
+    expect(result.isError).toBe(true)
+    expect(text(result)).toContain('subagent delegation aborted')
+  })
+
+  it('settles an asynchronous provider-start failure as a failed task', async () => {
+    const ctx = await backgroundSetup({ provider: 'mock' })
+    const parent = ownerAgent(ctx, 'sess-parent')
+    ctx.subagents.registerProvider({
+      name: 'broken-start',
+      capabilities: { outputSchema: false, depthLimit: false, toolFilter: false, persona: false },
+      inheritsParentContext: false,
+      start: async () => { throw new Error('setup failed') },
+    })
+    tool.apply(ctx, { provider: 'broken-start', toolName: 'subagent_broken' })
+
+    const started = await ctx.tools.execute({
+      callId: CallId('broken-start'),
+      name: 'subagent_broken',
+      arguments: { description: 'broken', prompt: 'p', run_in_background: true },
+      agent: parent,
+    })
+    expect(text(started)).toBe('started background subagent task subagent-1')
+    const output = await ctx.tools.execute({
+      callId: CallId('broken-output'),
+      name: 'task_output',
+      arguments: { task_id: 'subagent-1', wait: true },
+      agent: parent,
+    })
+    expect(text(output)).toContain('[status: failed, Error: setup failed]')
+  })
+
+  it('kills a subagent task while provider readiness is still pending', async () => {
+    const ctx = await backgroundSetup({ provider: 'mock' })
+    const parent = ownerAgent(ctx, 'sess-parent')
+    ctx.subagents.registerProvider({
+      name: 'pending-start',
+      capabilities: { outputSchema: false, depthLimit: false, toolFilter: false, persona: false },
+      inheritsParentContext: false,
+      start: request => new Promise((_resolve, reject) => {
+        request.signal.addEventListener('abort', () => { reject(new Error('startup aborted')) }, { once: true })
+      }),
+    })
+    tool.apply(ctx, { provider: 'pending-start', toolName: 'subagent_pending' })
+
+    await ctx.tools.execute({
+      callId: CallId('pending-start'),
+      name: 'subagent_pending',
+      arguments: { description: 'pending', prompt: 'p', run_in_background: true },
+      agent: parent,
+    })
+    await ctx.tools.execute({
+      callId: CallId('pending-kill'),
+      name: 'task_kill',
+      arguments: { task_id: 'subagent-1', reason: 'no longer needed' },
+      agent: parent,
+    })
+    const output = await ctx.tools.execute({
+      callId: CallId('pending-output'),
+      name: 'task_output',
+      arguments: { task_id: 'subagent-1', wait: true },
+      agent: parent,
+    })
+    expect(text(output)).toBe('(no new output)\n[status: killed]')
+  })
+
+  it('forwards task_kill reasons through the run signal (and defaults one when absent)', async () => {
+    // Use a provider that remains live until its signal is aborted.
+    const ctx = await backgroundSetup({ provider: 'mock', agentOptions: { model: 'child-model' } })
+    const parent = ownerAgent(ctx, 'sess-parent')
+    const cancels: (string | undefined)[] = []
+    let starts = 0
+    ctx.subagents.registerProvider({
+      name: 'hanging',
+      capabilities: { outputSchema: false, depthLimit: false, toolFilter: false, persona: false },
+      inheritsParentContext: false,
+      start: async (request) => {
+        let settle!: (value: { output: { type: 'text'; text: string }[]; stopReason: 'aborted' }) => void
+        const id = AgentId(`hang-${++starts}`)
+        const result = new Promise<{ output: { type: 'text'; text: string }[]; stopReason: 'aborted' }>((res) => { settle = res })
+        request.signal.addEventListener('abort', () => {
+          cancels.push(typeof request.signal.reason === 'string' ? request.signal.reason : undefined)
+          settle({ output: [], stopReason: 'aborted' })
+        }, { once: true })
+        return {
+          id,
+          result,
+          dispose: () => Promise.resolve(),
+        }
+      },
+    })
+    // Direct apply preserves omitted agentOptions instead of applying schema defaults.
+    tool.apply(ctx, { provider: 'hanging', toolName: 'subagent_hang' })
+
+    const startOne = await ctx.tools.execute({ callId: CallId('h1'), name: 'subagent_hang', arguments: { description: 'one', prompt: 'p', run_in_background: true }, agent: parent })
+    const startTwo = await ctx.tools.execute({ callId: CallId('h2'), name: 'subagent_hang', arguments: { description: 'two', prompt: 'p', run_in_background: true }, agent: parent })
+    expect(text(startOne)).toBe('started background subagent task subagent-1')
+    expect(text(startTwo)).toBe('started background subagent task subagent-2')
+
+    const withReason = await ctx.tools.execute({ callId: CallId('k1'), name: 'task_kill', arguments: { task_id: 'subagent-1', reason: 'superseded' }, agent: parent })
+    const withoutReason = await ctx.tools.execute({ callId: CallId('k2'), name: 'task_kill', arguments: { task_id: 'subagent-2' }, agent: parent })
+    expect(text(withReason)).toBe('requested cancellation of task subagent-1')
+    expect(text(withoutReason)).toBe('requested cancellation of task subagent-2')
+    expect(cancels).toEqual(['superseded', 'background subagent task killed'])
+
+    // The aborted children settle as killed tasks.
+    const killed = await ctx.tools.execute({ callId: CallId('w1'), name: 'task_output', arguments: { task_id: 'subagent-1', wait: true }, agent: parent })
+    expect(text(killed)).toBe('(no new output)\n[status: killed]')
+  })
+
+  it('runOutcome maps the stop-reason vocabulary onto task outcomes', () => {
+    const output = [{ type: 'text' as const, text: 'partial' }]
+    expect(runOutcome({ output, stopReason: 'completed' })).toEqual({ status: 'completed', output: 'partial' })
+    expect(runOutcome({ output, stopReason: 'aborted' })).toEqual({ status: 'killed' })
+    expect(runOutcome({ output, stopReason: 'error' })).toEqual({ status: 'failed', detail: 'error' })
+    expect(runOutcome({ output, stopReason: 'max-tokens' })).toEqual({ status: 'failed', detail: 'max-tokens' })
+    expect(runOutcome({ output, stopReason: 'refusal' })).toEqual({ status: 'failed', detail: 'refusal' })
+    // Merge-extensible: an unknown reason is failed-with-detail, never success.
+    expect(runOutcome({ output, stopReason: 'paused' as never })).toEqual({ status: 'failed', detail: 'paused' })
+  })
+
+  it('settleRun disposes the run before reporting, on both result paths', async () => {
+    const order: string[] = []
+    const completed = await settleRun({
+      id: AgentId('child-1'),
+      result: Promise.resolve({ output: [{ type: 'text' as const, text: 'ok' }], stopReason: 'completed' as const }),
+      dispose() { order.push('dispose'); return Promise.resolve() },
+    })
+    order.push('reported')
+    expect(completed).toEqual({ status: 'completed', output: 'ok' })
+    expect(order).toEqual(['dispose', 'reported'])
+
+    // An infrastructure rejection still disposes and reports failed.
+    let disposed = false
+    const failed = await settleRun({
+      id: AgentId('child-2'),
+      result: Promise.reject(new Error('transport gone')),
+      dispose() { disposed = true; return Promise.resolve() },
+    })
+    expect(failed).toEqual({ status: 'failed', detail: 'Error: transport gone' })
+    expect(disposed).toBe(true)
+
+    const disposeFailed = await settleRun({
+      id: AgentId('child-3'),
+      result: Promise.resolve({ output: [], stopReason: 'completed' }),
+      dispose: () => Promise.reject(new Error('reap failed')),
+    })
+    expect(disposeFailed).toEqual({ status: 'failed', detail: 'dispose failed: Error: reap failed' })
+
+    const bothFailed = await settleRun({
+      id: AgentId('child-4'),
+      result: Promise.reject(new Error('result failed')),
+      dispose: () => Promise.reject(new Error('reap failed')),
+    })
+    expect(bothFailed).toEqual({
+      status: 'failed',
+      detail: 'Error: result failed; dispose failed: Error: reap failed',
+    })
+  })
+})
+
+describe('background preflight failure (no orphaned child, by construction)', () => {
+  it('never starts the child when tasks.start preflight throws', async () => {
+    // With no control surface, task preflight fails before the provider can spawn.
+    const ctx = await setup({ provider: 'mock' })
+    await ctx.plugin(AgentRegistry)
+    await ctx.plugin(TaskService)
+    const scopeFiber = ctx.plugin(() => {})
+    const parent = {
+      id: AgentId('agent-sess-p'),
+      ctx: scopeFiber.ctx,
+      inject: () => {},
+      session: { header: { version: 0, id: 'sess-p', createdAt: 0 } },
+    } as unknown as Agent
+    ctx.agents.register(parent)
+
+    let starts = 0
+    ctx.subagents.registerProvider({
+      name: 'probe',
+      capabilities: { outputSchema: false, depthLimit: false, toolFilter: false, persona: false },
+      inheritsParentContext: false,
+      start: async () => {
+        starts += 1
+        return {
+          id: AgentId('probe-child'),
+          result: Promise.resolve({ output: [], stopReason: 'completed' as const }),
+          dispose: () => Promise.resolve(),
+        }
+      },
+    })
+    tool.apply(ctx, { provider: 'probe', toolName: 'subagent_probe' })
+
+    const result = await ctx.tools.execute({
+      callId: CallId('probe-1'),
+      name: 'subagent_probe',
+      arguments: { description: 'd', prompt: 'p', run_in_background: true },
+      agent: parent,
+    })
+    expect(result.isError).toBe(true)
+    expect(text(result)).toContain('no control surface is attached')
+    // Declare-then-execute: the failed preflight means no child ever existed.
+    expect(starts).toBe(0)
+  })
+})
diff --git a/packages/subagent/tool-subagent/tsconfig.json b/packages/subagent/tool-subagent/tsconfig.json
index 0a10bce7c5..2aa9d4f14e 100644
--- a/packages/subagent/tool-subagent/tsconfig.json
+++ b/packages/subagent/tool-subagent/tsconfig.json
@@ -28,6 +28,9 @@
     },
     {
       "path": "../subagent"
+    },
+    {
+      "path": "../../tasks/tasks"
     }
   ]
 }
diff --git a/packages/support/README.md b/packages/support/README.md
index d1bb1883ed..a85fffcdac 100644
--- a/packages/support/README.md
+++ b/packages/support/README.md
@@ -5,9 +5,10 @@ Packages that exist to serve development, testing, and the examples rather than
 | Package | Role | ctx key |
 |---|---|---|
 | `acp-snapshot/` | ACP snapshot suite kit: subprocess scenario harness + golden normalizers + the `defineAcpSnapshotSuite` factory | (library — imported by example `*.snapshot.ts` suites) |
+| `agent-loop-testkit/` | Shared prerequisite mounting for tests that exercise the concrete agent loop | (library — imported by AgentLoop integration tests) |
 | `invariants/` | Runtime event-contract assertions for development diagnostics | (listens on `session/*`, `agent/*`) |
 | `loader-smoke/` | Shared real-Loader subprocess harness for keyless example smokes | (library — imported by example e2e suites) |
 | `llm-replay/` | Record/replay adapter: short-circuits `llm/stream` from a recorded session JSONL (keyless snapshot tests) | (listens on `llm/stream`) |
 | `subagent-mock/` | Scripted `SubagentProvider` for deterministic seam/tool tests | (registers on `ctx.subagents`) |
 
-`invariants` is development support but has no environment guard: it runs wherever registered, and the default `dsh-agent-spine-demo` bundle mounts it unconditionally. `llm-replay` backs the demos and the snapshot test tier under the per-file coverage gate. `acp-snapshot` carries the snapshot tier's harness/normalizer/suite machinery, while `loader-smoke` owns the parallel stdio/Loader process boundary used by keyless example e2e suites. `subagent-mock` exercises the real `ctx.subagents` load path without a model or child agent. A package graduates OUT of `support/` into a product group only when it gains documented product consumers.
+`invariants` is development support but has no environment guard: it runs wherever registered, and the default `dsh-agent-spine-demo` bundle mounts it unconditionally. `agent-loop-testkit` centralizes the mandatory service spine for hand-built AgentLoop tests without owning their loop or scenario. `llm-replay` backs the demos and the snapshot test tier under the per-file coverage gate. `acp-snapshot` carries the snapshot tier's harness/normalizer/suite machinery, while `loader-smoke` owns the parallel stdio/Loader process boundary used by keyless example e2e suites. `subagent-mock` exercises the real `ctx.subagents` load path without a model or child agent. A package graduates OUT of `support/` into a product group only when it gains documented product consumers.
diff --git a/packages/support/acp-snapshot/README.md b/packages/support/acp-snapshot/README.md
index 03254a17a0..222d572043 100644
--- a/packages/support/acp-snapshot/README.md
+++ b/packages/support/acp-snapshot/README.md
@@ -4,7 +4,7 @@ The ACP snapshot suite kit: the shared machinery behind the keyless snapshot tie
 
 Three layers, importable separately:
 
-- **`runScenario` (harness)** — boots the real agent bin as a subprocess via tsx (unbuilt, Loader path), drives it over ACP JSON-RPC stdio from a deterministic `input.json` script, tees raw stdout for the golden + purity check, and harvests every persisted session JSONL (parent + subagent children, primary-first) after a graceful stdin-EOF shutdown. Parameterized by `AgentUnderTest` (`binScript`, `configPath`, `tsconfigPath` — absolute paths; the subprocess cwd is a temp dir outside the repo).
+- **`runScenario` (harness)** — boots the real agent bin as a subprocess via tsx (unbuilt, Loader path), drives it over ACP JSON-RPC stdio from a deterministic `input.json` script, tees raw stdout for the golden + purity check, and harvests every persisted session JSONL (parent + subagent children, primary-first) after a graceful stdin-EOF shutdown. Parameterized by `AgentUnderTest` (`binScript`, `configPath`, `tsconfigPath` — absolute paths; the subprocess cwd is a temp dir outside the repo). Startup failures preserve captured agent stderr in the rejected diagnostic.
 - **Normalizers** — pure functions turning the two captured surfaces into stable text: `normalizeStdout` (JSON-RPC ids → first-seen sequence; UUIDs/cwd → tokens; doubles as the stdout-purity check), `normalizeSessionLog` (times zeroed, `seq` kept), `scrubSystemPrompts` (prompt text → `{{system}}`), `scrubToolSchemas` (schema bulk → `{{tools}}`), and `scrubRequestHeaders` (all header bulk → `{{system}}`/`{{tools}}`/`{{messagePrefix}}` outside each pin, structure kept — [pinned-header RFC](../../../docs/rfc/implemented/testing/2026-07-06-pin-request-header-content-in-one-scenario.md)).
 - **`defineAcpSnapshotSuite` (factory)** — registers the whole describe/it tree for a scenario table: per-scenario golden + re-persisted-log compares, record/refresh fixture write-back, rejection of structured `UNKNOWN_TOOL` results, the per-header-class pin (`system-prompt.golden.md` plus `tool-schemas.golden.json`) with its live uniformity guard, and the fixture guard block (no orphan scenario dirs, required files present, exactly one pin per class, every JSONL prompt/schema-scrubbed, non-pinning fixtures fully header-scrubbed). Must be called at vitest collection time.
 
@@ -35,7 +35,7 @@ defineAcpSnapshotSuite({
 })
 ```
 
-A scenario booting a differently-composed tree sets its own `configPath` (an overlay whose basename still ends in `cordis.yml`, so the bin's replay swap finds the sibling `*cordis.snapshot.yml`) and, when that composition changes the request header, its own `headerClass` with its own pinning scenario — the acp-agent example's Code Mode and filesystem scenarios are templates. Each pinning directory stores the normalized composed prompt in generated `system-prompt.golden.md` and the initial schemas plus schema deltas in generated `tool-schemas.golden.json`; `session.jsonl` stores `"system":"{{system}}","tools":"{{tools}}"` while retaining config, reason, and any model-visible prefix.
+A scenario booting a differently-composed tree sets its own `configPath` (an overlay whose basename still ends in `cordis.yml`, so the bin's replay swap finds the sibling `*cordis.snapshot.yml`) and, when that composition changes the request header, its own `headerClass` with its own pinning scenario — the acp-agent example's Code Mode and filesystem scenarios are templates. Each pinning directory stores the normalized full prompt sequence in generated `system-prompt.golden.md` and the corresponding full tool-schema sequence in generated `tool-schemas.golden.json`; `session.jsonl` stores `"system":"{{system}}","tools":"{{tools}}"` while retaining config, reason, and any model-visible prefix. A pin with legitimate mid-run header changes declares `expectedHeaderChanges`, which fixes the length of both sidecar sequences.
 
 Examples use a `cordis.snapshot.yml` overlay with [`dsh-llm-replay`](../llm-replay/README.md). Recording calls the live model and updates model fixtures; keyless refresh replays those fixtures and updates derived stdout, session-log, prompt, and tool-schema snapshots. See the [snapshot RFC](../../../docs/rfc/implemented/testing/2026-06-19-acp-snapshot-tests.md).
 
diff --git a/packages/support/acp-snapshot/package.json b/packages/support/acp-snapshot/package.json
index d206b08161..eb6b46deaf 100644
--- a/packages/support/acp-snapshot/package.json
+++ b/packages/support/acp-snapshot/package.json
@@ -23,7 +23,7 @@
   "license": "BSD-3-Clause",
   "dependencies": {
     "@agentclientprotocol/sdk": "0.25.1",
-    "tsx": "^4.22.4",
+    "@deepseek-ai/dsh-loader-smoke": "workspace:*",
     "vitest": "^4.1.8"
   },
   "peerDependencies": {
diff --git a/packages/support/acp-snapshot/src/harness.ts b/packages/support/acp-snapshot/src/harness.ts
index 3bae0ba279..e2072b6b17 100644
--- a/packages/support/acp-snapshot/src/harness.ts
+++ b/packages/support/acp-snapshot/src/harness.ts
@@ -11,7 +11,6 @@ import { cp, mkdtemp, readFile, readdir, rm } from 'node:fs/promises'
 import { existsSync } from 'node:fs'
 import { tmpdir } from 'node:os'
 import { join, delimiter } from 'node:path'
-import { fileURLToPath } from 'node:url'
 import { Readable, Writable } from 'node:stream'
 import {
   ClientSideConnection,
@@ -23,12 +22,7 @@ import {
   type RequestPermissionResponse,
   type SessionNotification,
 } from '@agentclientprotocol/sdk'
-
-// Resolve tsx's ESM loader to an ABSOLUTE path once: the child runs with its
-// cwd in a temp dir OUTSIDE the repo, where a bare `--import tsx` would not
-// resolve from node_modules. import.meta.resolve gives this package's tsx
-// regardless of the child cwd.
-const tsxLoader = fileURLToPath(import.meta.resolve('tsx'))
+import { resolveExampleLaunch } from '@deepseek-ai/dsh-loader-smoke'
 
 /**
  * The agent composition a scenario runs against: which bin to boot and which
@@ -37,8 +31,10 @@ const tsxLoader = fileURLToPath(import.meta.resolve('tsx'))
  * them from its own `import.meta.url`.
  */
 export interface AgentUnderTest {
-  /** The agent bin entry (e.g. `packages/examples/acp-demo/src/bin.ts`), run unbuilt via tsx. */
+  /** The agent bin's SOURCE entry (e.g. `packages/examples/acp-demo/src/bin.ts`); the `lib` bin is derived from it. */
   binScript: string
+  /** Explicit plain-Node entry for `lib` mode; intended for test fixtures outside a package `src/` tree. */
+  libBinScript?: string | undefined
   /**
    * The example's live `cordis.yml`. Under `DSH_SNAPSHOT=replay` the bin swaps
    * it for the sibling `cordis.snapshot.yml` (the keyless replay overlay), so
@@ -47,10 +43,8 @@ export interface AgentUnderTest {
   configPath: string
   /**
    * The repo-root tsconfig whose `paths` map resolves the unbuilt workspace
-   * imports. Passed to the child as `TSX_TSCONFIG_PATH`: tsx finds a tsconfig
-   * by searching UP from the child's cwd — a temp dir outside the repo — so
-   * without the explicit pin the dsh-* imports fail before the bin writes a
-   * byte.
+   * imports in `src` mode (passed to the child as `TSX_TSCONFIG_PATH`). Ignored
+   * in `lib` mode, where the example resolves plugins through real `exports`.
    */
   tsconfigPath: string
 }
@@ -168,6 +162,9 @@ export interface RunOptions {
 export async function runScenario(input: InputScript, opts: RunOptions): Promise<RunResult> {
   const cwd = await mkdtemp(join(tmpdir(), 'acp-snap-cwd-'))
   const sessionsRoot = await mkdtemp(join(tmpdir(), 'acp-snap-sessions-'))
+  // Fixed path length: spill-policy budgets the preview against the REAL path
+  // before stdout normalization, so tmpdir() length differences churn goldens.
+  const spillRoot = '/tmp/dsh-acp-snapshot-spill'
   // Everything past the temp-dir creation runs under a try/finally that always
   // removes both dirs — so a failure in workspace seeding, spawn, or any step
   // never leaks them (the "e2e tests own their resources" rule).
@@ -181,24 +178,32 @@ export async function runScenario(input: InputScript, opts: RunOptions): Promise
     if (opts.workspaceDir !== undefined && existsSync(opts.workspaceDir)) {
       await cp(opts.workspaceDir, cwd, { recursive: true })
     }
-    const env: NodeJS.ProcessEnv = {
-      ...process.env,
-      TSX_TSCONFIG_PATH: opts.agent.tsconfigPath,
-      DSH_SNAPSHOT: opts.mode,
-      DSH_SNAPSHOT_FILE: opts.fixtureFile,
-      DSH_SNAPSHOT_SESSIONS_ROOT: sessionsRoot,
-      DSH_HOME: join(cwd, '.dsh'),
-      DSH_AGENTS_HOME: join(cwd, '.agents'),
-      ...opts.overrideFile !== undefined ? { DSH_SNAPSHOT_OVERRIDE: opts.overrideFile } : {},
-      ...opts.childFiles !== undefined && opts.childFiles.length > 0
-        ? { DSH_SNAPSHOT_CHILD_FILES: opts.childFiles.join(delimiter) }
-        : {},
-    }
+    // Boot the agent in the environment's mode (DSH_EXAMPLE_MODE): `src` runs the
+    // source bin under tsx with the paths map; `lib` runs the built bin under plain
+    // Node, resolving plugins through the example's workspace node_modules → lib.
+    const launch = resolveExampleLaunch({
+      srcBin: opts.agent.binScript,
+      libBin: opts.agent.libBinScript,
+      configArgs: ['--config', opts.configPath ?? opts.agent.configPath],
+      tsconfigPath: opts.agent.tsconfigPath,
+      env: {
+        DSH_SNAPSHOT: opts.mode,
+        DSH_SNAPSHOT_FILE: opts.fixtureFile,
+        DSH_SNAPSHOT_SESSIONS_ROOT: sessionsRoot,
+        DSH_SNAPSHOT_SPILL_ROOT: spillRoot,
+        DSH_HOME: join(cwd, '.dsh'),
+        DSH_AGENTS_HOME: join(cwd, '.agents'),
+        ...opts.overrideFile !== undefined ? { DSH_SNAPSHOT_OVERRIDE: opts.overrideFile } : {},
+        ...opts.childFiles !== undefined && opts.childFiles.length > 0
+          ? { DSH_SNAPSHOT_CHILD_FILES: opts.childFiles.join(delimiter) }
+          : {},
+      },
+    })
 
     child = spawn(
-      process.execPath,
-      ['--import', tsxLoader, opts.agent.binScript, '--config', opts.configPath ?? opts.agent.configPath],
-      { cwd, env, stdio: ['pipe', 'pipe', 'pipe'] },
+      launch.command,
+      launch.args,
+      { cwd, env: { ...process.env, ...launch.env }, stdio: ['pipe', 'pipe', 'pipe'] },
     )
 
     child.stderr.setEncoding('utf8')
@@ -281,6 +286,10 @@ export async function runScenario(input: InputScript, opts: RunOptions): Promise
     // Harvest EVERY persisted log (parent + any subagent children) while the
     // temp dirs still exist, ordered primary-first.
     sessionLogs = await harvestSessionLogs(sessionsRoot)
+  } catch (error: unknown) {
+    const stderr = stderrChunks.join('')
+    if (stderr === '') throw error
+    throw new Error(`snapshot-harness: scenario failed: ${String(error)}\nagent stderr:\n${stderr}`, { cause: error })
   } finally {
     // Failure-safe teardown: kill a still-running child and drop the temp dirs
     // even if seeding/spawn/a step/harvest threw, so a flaky run never leaks a
@@ -291,6 +300,7 @@ export async function runScenario(input: InputScript, opts: RunOptions): Promise
     }
     await rm(cwd, { recursive: true, force: true })
     await rm(sessionsRoot, { recursive: true, force: true })
+    await rm(spillRoot, { recursive: true, force: true })
   }
 
   return {
diff --git a/packages/support/acp-snapshot/src/normalize.ts b/packages/support/acp-snapshot/src/normalize.ts
index e3f32792ab..48761864f0 100644
--- a/packages/support/acp-snapshot/src/normalize.ts
+++ b/packages/support/acp-snapshot/src/normalize.ts
@@ -14,6 +14,16 @@ const MESSAGE_PREFIX = '{{messagePrefix}}'
 
 /** A UUID v4 string, the shape `randomUUID()` produces for session ids. */
 const UUID_RE = /[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}/gi
+const LOCAL_SPILL_PATH_RE = new RegExp(
+  String.raw`\{\{cwd\}\}/\.spill/session-[0-9a-f]{12}/[0-9a-f]{12}-([A-Za-z0-9._~-]+?)`
+  + String.raw`(?=\. Use read with offset/limit|[\s)]|$)`,
+  'g',
+)
+const SNAPSHOT_SPILL_PATH_RE = new RegExp(
+  String.raw`/tmp/dsh-acp-snapshot-spill/session-[0-9a-f]{12}/[0-9a-f]{12}-([A-Za-z0-9._~-]+?)`
+  + String.raw`(?=\. Use read with offset/limit|[\s)]|$)`,
+  'g',
+)
 
 /** Inputs the normalizers need to recognize a run's volatile values. */
 export interface NormalizeContext {
@@ -29,6 +39,9 @@ function scrubString(value: string, ctx: NormalizeContext): string {
   // cwd first (longest, most specific), then explicit session ids, then any
   // residual UUID (covers ids that appear in places we didn't enumerate).
   out = out.split(ctx.cwd).join(CWD)
+  out = out.split(`/private${CWD}`).join(CWD)
+  out = out.replace(LOCAL_SPILL_PATH_RE, (_match, name: string) => `{{spillLocator:${name}}}`)
+  out = out.replace(SNAPSHOT_SPILL_PATH_RE, (_match, name: string) => `{{spillLocator:${name}}}`)
   for (const id of ctx.sessionIds) out = out.split(id).join(SESSION_ID)
   out = out.replace(UUID_RE, SESSION_ID)
   return out
@@ -112,8 +125,8 @@ export function normalizeSessionLog(rawLog: string, ctx: NormalizeContext): stri
 }
 
 /**
- * Replace system-prompt content in request headers and header deltas with
- * `{{system}}` tokens while retaining field presence and delta structure.
+ * Replace system-prompt content in request headers with `{{system}}` tokens
+ * while retaining field presence.
  * Other header content stays verbatim, so a header-pinning fixture can keep
  * its complete tool schemas while every JSONL fixture omits the prompt text.
  * Lines without a system payload pass through byte-for-byte; the transform is
@@ -127,11 +140,11 @@ export function scrubSystemPrompts(rawLog: string): string {
 }
 
 /**
- * Replace tool schemas in request headers and header deltas with `{{tools}}`
- * tokens while retaining field presence, tool names, and delta structure.
- * System prompts and session-prefix messages stay verbatim so pinning fixtures
- * can move only schema bulk into their dedicated JSON sidecar. Lines without a
- * tool payload pass through byte-for-byte; the transform is idempotent.
+ * Replace tool schemas in full request-header snapshots with `{{tools}}`
+ * tokens while retaining field presence. System prompts and session-prefix
+ * messages stay verbatim so pinning fixtures can move only schema bulk into
+ * their dedicated JSON sidecar. Lines without a tool payload pass through
+ * byte-for-byte; the transform is idempotent.
  *
  * @param rawLog The raw session `.jsonl` content.
  * @returns The JSONL with tool-schema content tokenized.
@@ -144,9 +157,9 @@ export function scrubToolSchemas(rawLog: string): string {
  * Replace all bulky request-header content in a session JSONL with stable
  * tokens. This includes the system-prompt fields handled by
  * {@link scrubSystemPrompts}, tool schemas, and session-prefix messages. It
- * keeps system-delta line positions and arity, tool-delta names, prefix
- * message counts, field presence, config, and reason. Lines without content
- * to scrub pass through byte-for-byte, and the transform is idempotent.
+ * keeps prefix message counts, field presence, config, and reason. Lines
+ * without content to scrub pass through byte-for-byte, and the transform is
+ * idempotent.
  *
  * @param rawLog The raw session `.jsonl` content.
  * @returns The JSONL with all header bulk tokenized, other lines byte-identical.
@@ -182,33 +195,7 @@ function scrubHeaderContent(rawLog: string, options: HeaderScrubOptions): string
       }
       return touched ? JSON.stringify(record) : line
     }
-    if (record.type === 'request/header-delta') {
-      let touched = false
-      const system = data.system as Record<string, unknown> | null | undefined
-      if (options.system === true && system !== null && typeof system === 'object' && Array.isArray(system.insert)) {
-        system.insert = system.insert.map(() => SYSTEM)
-        touched = true
-      }
-      const tools = data.tools as Record<string, unknown> | null | undefined
-      if (options.tools === true && tools !== null && typeof tools === 'object') {
-        if (Array.isArray(tools.added)) { tools.added = tools.added.map(scrubToolSchema); touched = true }
-        if (Array.isArray(tools.changed)) { tools.changed = tools.changed.map(scrubToolSchema); touched = true }
-      }
-      if (options.prefix === true && Array.isArray(data.messagePrefix)) {
-        data.messagePrefix = data.messagePrefix.map(() => MESSAGE_PREFIX)
-        touched = true
-      }
-      return touched ? JSON.stringify(record) : line
-    }
     return line
   })
   return out.join('\n')
 }
-
-/** Tokenize one tool schema's bulk (description, parameters, anything else), keeping its identifying `name`. */
-function scrubToolSchema(tool: unknown): unknown {
-  if (tool === null || typeof tool !== 'object' || Array.isArray(tool)) return tool
-  const out: Record<string, unknown> = {}
-  for (const [k, v] of Object.entries(tool)) out[k] = k === 'name' ? v : TOOLS
-  return out
-}
diff --git a/packages/support/acp-snapshot/src/suite.ts b/packages/support/acp-snapshot/src/suite.ts
index dc1675eac0..322439bb01 100644
--- a/packages/support/acp-snapshot/src/suite.ts
+++ b/packages/support/acp-snapshot/src/suite.ts
@@ -1,12 +1,17 @@
 /**
- * Keyless-by-default ACP snapshot suite factory. Each scenario drives the real subprocess and
- * compares normalized stdout; comparable session fixtures are both replay input and expected
- * output. Record mode refreshes reproducible model scenarios from the live API, while refresh
- * mode replays committed scripts and rewrites derived artifacts without a key.
+ * Keyless-by-default ACP snapshot suite factory. Each scenario drives the real
+ * subprocess and compares normalized stdout; comparable session fixtures are
+ * both replay input and expected output. Record mode refreshes reproducible
+ * model scenarios from the live API, while refresh mode replays committed
+ * scripts and rewrites derived artifacts without a key.
+ * Replay scenarios run concurrently because each subprocess owns unique temp
+ * cwd and persistence roots and reads only committed fixtures. Record and
+ * refresh stay serial while writing.
  *
- * Exactly one scenario per header-composition class pins the system prompt and tool schemas in
- * dedicated sidecars. Every live header is checked against that pin, so session-dependent
- * composition must declare a separate class instead of escaping coverage.
+ * Exactly one scenario per header-composition class pins the full prompt and
+ * tool-schema sequences in dedicated sidecars. Every live header is checked
+ * against that pin, so session-dependent composition must declare a separate
+ * class instead of escaping coverage.
  * @module @deepseek-ai/dsh-acp-snapshot/suite
  */
 
@@ -80,14 +85,11 @@ export interface Scenario {
    */
   pinsHeader?: boolean
   /**
-   * How many `request/header-delta` events this PINNING scenario's fixture
-   * legitimately carries (default 0). A recorded mid-run header change — a
-   * config-option switch rewriting a prompt section — is part of the pinned
-   * surface, with readable prompt text in Markdown; any OTHER count
-   * still fails, so fixture rot stays caught. Meaningless off the pin (the
-   * live uniformity guard keeps non-pinning scenarios delta-free).
+   * How many changed `request/header` snapshots this PINNING scenario's primary
+   * fixture legitimately carries (default 0). Their full prompt text is kept in
+   * the readable Markdown pin; any other count fails. Meaningless off the pin.
    */
-  expectedHeaderDeltas?: number
+  expectedHeaderChanges?: number
   /**
    * Which header-composition class this scenario belongs to. Scenarios that
    * boot the same config compose the same header; each class has exactly one
@@ -208,114 +210,58 @@ export function normalizedToolSchemas(rawLog: string, ctx: NormalizeContext): un
   })
 }
 
-/**
- * Extract normalized tool-schema edits from request-header deltas in log order.
- * Deltas without an object-valued tools edit are omitted; their remaining
- * structure stays pinned in the session JSONL.
- *
- * @param rawLog The session `.jsonl` content to inspect.
- * @param ctx The volatile values of the run that produced it.
- * @returns The normalized tool-schema edits, in event order.
- */
-export function normalizedToolSchemaDeltas(rawLog: string, ctx: NormalizeContext): unknown[] {
-  return normalizeSessionLog(rawLog, ctx)
-    .split('\n')
-    .filter(line => line.trim().length > 0)
-    .map(line => JSON.parse(line) as { type?: unknown; data?: { tools?: unknown } })
-    .filter(record => record.type === 'request/header-delta')
-    .flatMap((record) => {
-      const tools = record.data?.tools
-      return tools !== null && typeof tools === 'object' && !Array.isArray(tools) ? [tools] : []
-    })
-}
-
 /** The structured contents of a tool-schema sidecar. */
 export interface ToolSchemasSnapshot {
   /** The complete tool schemas from the pinned request header. */
   initial: unknown[]
-  /** Complete tool-schema edits from subsequent request-header deltas. */
-  deltas: unknown[]
+  /** Complete tool schemas from subsequent changed-header snapshots. */
+  changes: unknown[][]
 }
 
 /**
- * Render tool schemas and later schema edits as canonical, readable JSON.
+ * Render the full tool-schema sequence as canonical, readable JSON.
  *
  * @param initial The pinned request header's complete tool schemas.
- * @param deltas Complete tool-schema edits from request-header deltas.
+ * @param changes Complete tool schemas from later changed headers.
  * @returns A pretty-printed JSON snapshot ending in one newline.
  */
-export function formatToolSchemasSnapshot(initial: readonly unknown[], deltas: readonly unknown[] = []): string {
-  return `${JSON.stringify({ initial, deltas }, null, 2)}\n`
+export function formatToolSchemasSnapshot(initial: readonly unknown[], changes: readonly unknown[][] = []): string {
+  return `${JSON.stringify({ initial, changes }, null, 2)}\n`
 }
 
 /**
  * Parse and validate the stable top-level shape of a tool-schema sidecar.
  *
  * @param snapshot The JSON sidecar text.
- * @returns Its initial schemas and schema deltas.
+ * @returns Its initial and changed-header schema sets.
  */
 export function parseToolSchemasSnapshot(snapshot: string): ToolSchemasSnapshot {
   const parsed = JSON.parse(snapshot) as unknown
   if (parsed === null || typeof parsed !== 'object' || Array.isArray(parsed)) {
     throw new Error('acp-snapshot: tool-schema snapshot must be an object')
   }
-  const { initial, deltas } = parsed as { initial?: unknown; deltas?: unknown }
-  if (!Array.isArray(initial) || !Array.isArray(deltas)) {
-    throw new Error('acp-snapshot: tool-schema snapshot must carry array-valued initial and deltas fields')
+  const { initial, changes } = parsed as { initial?: unknown; changes?: unknown }
+  if (!Array.isArray(initial) || !Array.isArray(changes) || !changes.every(Array.isArray)) {
+    throw new Error('acp-snapshot: tool-schema snapshot must carry array-valued initial and changes fields')
   }
-  return { initial, deltas }
+  return { initial, changes }
 }
 
 /**
- * Restore a sidecar's initial schemas into a tokenized pinned header.
+ * Restore one sidecar schema set into a tokenized pinned header.
  *
  * @param header The parsed request header carrying `tools: "{{tools}}"`.
- * @param snapshot The parsed tool-schema sidecar.
- * @returns A copy of the header with its complete initial schemas restored.
+ * @param schemas The complete schemas for this full header snapshot.
+ * @returns A copy of the header with its complete schemas restored.
  */
-export function restorePinnedToolSchemas(header: unknown, snapshot: ToolSchemasSnapshot): unknown {
+export function restorePinnedToolSchemas(header: unknown, schemas: readonly unknown[]): unknown {
   if (header === null || typeof header !== 'object' || Array.isArray(header)) {
     throw new Error('acp-snapshot: pinned request header must be an object')
   }
   if ((header as { tools?: unknown }).tools !== TOOLS_TOKEN) {
     throw new Error(`acp-snapshot: pinned request header tools must equal ${TOOLS_TOKEN}`)
   }
-  return { ...header, tools: snapshot.initial }
-}
-
-/** One normalized system-prompt edit carried by a `request/header-delta`. */
-export interface SystemPromptDeltaSnapshot {
-  /** How many leading lines remain from the prior prompt. */
-  keepStart: number
-  /** How many trailing lines remain from the prior prompt. */
-  keepEnd: number
-  /** The normalized replacement lines inserted between the retained ranges. */
-  insert: string[]
-}
-
-/**
- * Extract normalized system-prompt edits from request-header deltas in log
- * order. Deltas without a well-formed system edit are omitted; their non-prompt
- * structure remains pinned in JSONL.
- *
- * @param rawLog The session `.jsonl` content to inspect.
- * @param ctx The volatile values of the run that produced it.
- * @returns The normalized system-prompt edits, in event order.
- */
-export function normalizedSystemPromptDeltas(rawLog: string, ctx: NormalizeContext): SystemPromptDeltaSnapshot[] {
-  return normalizeSessionLog(rawLog, ctx)
-    .split('\n')
-    .filter(line => line.trim().length > 0)
-    .map(line => JSON.parse(line) as { type?: unknown; data?: { system?: unknown } })
-    .filter(record => record.type === 'request/header-delta')
-    .flatMap((record) => {
-      const system = record.data?.system
-      if (system === null || typeof system !== 'object') return []
-      const { keepStart, keepEnd, insert } = system as { keepStart?: unknown; keepEnd?: unknown; insert?: unknown }
-      if (typeof keepStart !== 'number' || typeof keepEnd !== 'number' || !Array.isArray(insert)) return []
-      if (!insert.every(line => typeof line === 'string')) return []
-      return [{ keepStart, keepEnd, insert: insert }]
-    })
+  return { ...header, tools: schemas }
 }
 
 /**
@@ -324,38 +270,40 @@ export function normalizedSystemPromptDeltas(rawLog: string, ctx: NormalizeConte
  * the committed file follows the repository newline contract.
  *
  * @param prompt The normalized system prompt.
- * @param deltas Normalized prompt edits to append as readable sections.
+ * @param changes Full normalized prompts from later changed-header snapshots.
  * @returns Markdown snapshot text ending in a newline.
  */
 export function formatSystemPromptSnapshot(
   prompt: string,
-  deltas: readonly SystemPromptDeltaSnapshot[] = [],
+  changes: readonly string[] = [],
 ): string {
   let snapshot = prompt.endsWith('\n') ? prompt : `${prompt}\n`
-  for (const [index, delta] of deltas.entries()) {
-    snapshot += `\n<!-- request/header-delta ${index + 1}: keepStart=${delta.keepStart}, keepEnd=${delta.keepEnd} -->\n\n`
-    const insert = delta.insert.join('\n')
-    snapshot += insert.endsWith('\n') ? insert : `${insert}\n`
+  for (const [index, change] of changes.entries()) {
+    snapshot += `\n<!-- request/header change ${index + 1} -->\n\n`
+    snapshot += change.endsWith('\n') ? change : `${change}\n`
   }
   return snapshot
 }
 
-/** Return the initial-prompt portion of a possibly delta-bearing snapshot. */
+/** Return the initial-prompt portion of a possibly multi-header snapshot. */
 function initialSystemPromptSnapshot(snapshot: string): string {
-  const marker = snapshot.indexOf('\n<!-- request/header-delta ')
+  const marker = snapshot.indexOf('\n<!-- request/header change ')
   return marker < 0 ? snapshot : snapshot.slice(0, marker)
 }
 
 /**
- * Count the `request/header-delta` events in a session JSONL.
+ * Count changed `request/header` snapshots in a session JSONL.
  *
  * @param rawLog The session `.jsonl` content.
- * @returns How many `request/header-delta` events the log carries.
+ * @returns How many headers carry reason `change`.
  */
-export function headerDeltaCount(rawLog: string): number {
+export function headerChangeCount(rawLog: string): number {
   return rawLog.split('\n')
     .filter(line => line.trim().length > 0)
-    .filter(line => (JSON.parse(line) as { type?: unknown }).type === 'request/header-delta')
+    .filter((line) => {
+      const record = JSON.parse(line) as { type?: unknown; data?: { reason?: unknown } }
+      return record.type === 'request/header' && record.data?.reason === 'change'
+    })
     .length
 }
 
@@ -461,7 +409,7 @@ export function stabilizeRefreshLog(fresh: string, existing: string, replacement
 }
 
 /**
- * Register the suite: one `describe` per scenario (the golden/log compares and
+ * Register the suite: one test per scenario (the golden/log compares and
  * the header-uniformity guard) plus the fixture guard block (no orphan
  * scenario dirs, required files present, exactly one pin per header class,
  * pinning fixtures well-formed, every JSONL prompt-scrubbed, non-pinning
@@ -477,6 +425,7 @@ export function defineAcpSnapshotSuite(options: SnapshotSuiteOptions): void {
   const RECORDING = mode === 'record'
   const REFRESHING = mode === 'refresh'
   const childMode: 'replay' | 'record' = RECORDING ? 'record' : 'replay'
+  const scenarioSuite = mode === 'replay' ? describe.concurrent : describe
 
   /** The class a scenario's header composition belongs to (see {@link Scenario.headerClass}). */
   const classOf = (scenario: Scenario): string => scenario.headerClass ?? 'default'
@@ -496,11 +445,11 @@ export function defineAcpSnapshotSuite(options: SnapshotSuiteOptions): void {
     }
   }
 
-  for (const scenario of scenarios) {
-    describe(`snapshot: ${scenario.name}`, () => {
+  scenarioSuite('snapshot scenarios', () => {
+    for (const scenario of scenarios) {
       // In RECORD mode, only re-run the `recorded` (live-API) scenarios; the `authored` ones
       // (sidecar-driven errors/cancel) are never re-recorded.
-      it.skipIf(RECORDING && !scenario.recorded)('matches the goldens', async () => {
+      it.skipIf(RECORDING && !scenario.recorded)(`snapshot: ${scenario.name} matches the goldens`, async ({ expect }) => {
         const dir = join(snapshotsDir, scenario.name)
         const input = JSON.parse(await readFile(join(dir, 'input.json'), 'utf8')) as InputScript
         const overrideFile = join(dir, 'replay.override.json')
@@ -564,30 +513,19 @@ export function defineAcpSnapshotSuite(options: SnapshotSuiteOptions): void {
             ))
           }
           if (scenario.pinsHeader === true) {
-            const prompts = result.sessionLogs.flatMap(log => normalizedSystemPrompts(log.content, ctx))
-            expect(prompts.length, `${mode} produced no system prompt to snapshot`).toBeGreaterThan(0)
-            const initialSnapshot = formatSystemPromptSnapshot(prompts[0] as string)
-            for (const prompt of prompts) {
-              expect(formatSystemPromptSnapshot(prompt), 'the pinning run produced divergent system prompts')
-                .toEqual(initialSnapshot)
-            }
             const primary = result.sessionLogs[0] as HarvestedLog
-            const snapshot = formatSystemPromptSnapshot(
-              prompts[0] as string,
-              normalizedSystemPromptDeltas(primary.content, ctx),
-            )
+            const prompts = normalizedSystemPrompts(primary.content, ctx)
+            expect(prompts.length, `${mode} produced no system prompt to snapshot`).toBeGreaterThan(0)
+            const snapshot = formatSystemPromptSnapshot(prompts[0] as string, prompts.slice(1))
             await writeFile(join(dir, SYSTEM_PROMPT_SNAPSHOT), snapshot)
 
-            const schemaSets = result.sessionLogs.flatMap(log => normalizedToolSchemas(log.content, ctx))
+            const schemaSets = normalizedToolSchemas(primary.content, ctx)
             expect(schemaSets.length, `${mode} produced no tool schemas to snapshot`).toBeGreaterThan(0)
-            const initialSchemaSnapshot = formatToolSchemasSnapshot(schemaSets[0] as unknown[])
-            for (const schemas of schemaSets) {
-              expect(formatToolSchemasSnapshot(schemas), 'the pinning run produced divergent tool schemas')
-                .toEqual(initialSchemaSnapshot)
-            }
+            expect(schemaSets.length, `${mode} produced a tool-schema sequence that differs from its prompt sequence`)
+              .toBe(prompts.length)
             await writeFile(join(dir, TOOL_SCHEMAS_SNAPSHOT), formatToolSchemasSnapshot(
               schemaSets[0] as unknown[],
-              normalizedToolSchemaDeltas(primary.content, ctx),
+              schemaSets.slice(1),
             ))
           }
         }
@@ -611,8 +549,8 @@ export function defineAcpSnapshotSuite(options: SnapshotSuiteOptions): void {
           }
         }
 
-        // Header-uniformity guard: every live header in a class must equal the class pin split
-        // across tokenized JSONL plus readable prompt and structured schema sidecars.
+        // Every live full header must equal its class pin reconstructed from
+        // tokenized JSONL plus readable prompt and structured schema sidecars.
         /* v8 ignore next -- construction guarantees the pin exists; a miss would fail the one-header assertion loudly. */
         const pinningScenario = pinningByClass.get(classOf(scenario)) ?? scenario
         const pinningDir = join(snapshotsDir, pinningScenario.name)
@@ -620,17 +558,23 @@ export function defineAcpSnapshotSuite(options: SnapshotSuiteOptions): void {
         const pinned = normalizedHeaders(pinnedFixture, fixtureContext(pinnedFixture))
         const promptSnapshot = await readFile(join(pinningDir, SYSTEM_PROMPT_SNAPSHOT), 'utf8')
         const initialPromptSnapshot = initialSystemPromptSnapshot(promptSnapshot)
+        expect(pinned.length, `the pinning fixture (${pinningScenario.name}) has an unexpected request/header count`)
+          .toBe(1 + (pinningScenario.expectedHeaderChanges ?? 0))
         const toolSchemasSnapshot = await readFile(join(pinningDir, TOOL_SCHEMAS_SNAPSHOT), 'utf8')
         const toolSchemas = parseToolSchemasSnapshot(toolSchemasSnapshot)
-        expect(pinned.length, `the pinning fixture (${pinningScenario.name}) must carry exactly one request/header`)
-          .toBe(1)
-        const pinnedHeader = restorePinnedToolSchemas(pinned[0], toolSchemas)
+        const pinnedSchemaSets = [toolSchemas.initial, ...toolSchemas.changes]
+        expect(pinnedSchemaSets.length, `the pinning fixture (${pinningScenario.name}) has an unexpected tool-schema count`)
+          .toBe(pinned.length)
+        const pinnedHeaders = pinned.map((header, index) => restorePinnedToolSchemas(
+          header,
+          pinnedSchemaSets[index] as unknown[],
+        ))
         for (const [logIndex, log] of result.sessionLogs.entries()) {
-          const expectedDeltas = scenario.pinsHeader === true && logIndex === 0
-            ? scenario.expectedHeaderDeltas ?? 0
+          const expectedChanges = scenario.pinsHeader === true && logIndex === 0
+            ? scenario.expectedHeaderChanges ?? 0
             : 0
-          expect(headerDeltaCount(log.content), `session ${log.id}: request/header-delta count`)
-            .toBe(expectedDeltas)
+          expect(headerChangeCount(log.content), `session ${log.id}: changed request/header count`)
+            .toBe(expectedChanges)
           const headers = normalizedHeaders(scrubSystemPrompts(log.content), ctx)
           const prompts = normalizedSystemPrompts(log.content, ctx)
           const schemaSets = normalizedToolSchemas(log.content, ctx)
@@ -639,27 +583,30 @@ export function defineAcpSnapshotSuite(options: SnapshotSuiteOptions): void {
           expect(schemaSets.length, `session ${log.id}: every request/header must carry an array-valued tools field`)
             .toBe(headers.length)
           for (const [k, header] of headers.entries()) {
+            const expected = expectedChanges > 0 ? pinnedHeaders[k] : pinnedHeaders[0]
             expect(header, `session ${log.id}: request/header #${k + 1} diverged from the pinned (${pinningScenario.name}) header`)
-              .toEqual(pinnedHeader)
-            expect(formatSystemPromptSnapshot(prompts[k] as string), `session ${log.id}: initial system prompt #${k + 1} diverged from ${pinningScenario.name}/${SYSTEM_PROMPT_SNAPSHOT}`)
-              .toEqual(initialPromptSnapshot)
+              .toEqual(expected)
+            if (expectedChanges === 0) {
+              expect(formatSystemPromptSnapshot(prompts[k] as string), `session ${log.id}: initial system prompt #${k + 1} diverged from ${pinningScenario.name}/${SYSTEM_PROMPT_SNAPSHOT}`)
+                .toEqual(initialPromptSnapshot)
+            }
           }
           if (scenario.pinsHeader === true && logIndex === 0) {
             expect(formatSystemPromptSnapshot(
               prompts[0] as string,
-              normalizedSystemPromptDeltas(log.content, ctx),
-            ), `session ${log.id}: system-prompt deltas diverged from ${pinningScenario.name}/${SYSTEM_PROMPT_SNAPSHOT}`)
+              prompts.slice(1),
+            ), `session ${log.id}: changed system prompts diverged from ${pinningScenario.name}/${SYSTEM_PROMPT_SNAPSHOT}`)
               .toEqual(promptSnapshot)
             expect(formatToolSchemasSnapshot(
               schemaSets[0] as unknown[],
-              normalizedToolSchemaDeltas(log.content, ctx),
-            ), `session ${log.id}: tool-schema deltas diverged from ${pinningScenario.name}/${TOOL_SCHEMAS_SNAPSHOT}`)
+              schemaSets.slice(1),
+            ), `session ${log.id}: changed tool schemas diverged from ${pinningScenario.name}/${TOOL_SCHEMAS_SNAPSHOT}`)
               .toEqual(toolSchemasSnapshot)
           }
         }
       })
-    })
-  }
+    }
+  })
 
   describe('snapshot fixtures', () => {
     it('every scenario directory is registered (no orphans)', async () => {
@@ -708,25 +655,30 @@ export function defineAcpSnapshotSuite(options: SnapshotSuiteOptions): void {
       }
     })
 
-    it('every pinning fixture carries one tokenized request/header, two sidecars, and its declared deltas', async () => {
-      // The live uniformity guard runs only in NON-pinning scenarios, so a class made of just
-      // its pinning scenario would otherwise accept a re-recorded pin with several headers or
-      // an undeclared mid-run header-delta — shapes the pin design cannot represent.
+    it('every pinning fixture carries one tokenized header sequence and two sidecars', async () => {
+      // Assert the committed pin directly because a class containing only its
+      // pinning scenario has no non-pinning live run to catch undeclared changes.
       for (const scenario of pinningByClass.values()) {
         const fixture = await readFile(join(snapshotsDir, scenario.name, 'session.jsonl'), 'utf8')
         const headers = normalizedHeaders(fixture, fixtureContext(fixture))
         const promptSnapshot = await readFile(join(snapshotsDir, scenario.name, SYSTEM_PROMPT_SNAPSHOT), 'utf8')
+        expect(headers.length, `${scenario.name}: unexpected request/header count`)
+          .toBe(1 + (scenario.expectedHeaderChanges ?? 0))
         const toolSchemasSnapshot = await readFile(join(snapshotsDir, scenario.name, TOOL_SCHEMAS_SNAPSHOT), 'utf8')
         const toolSchemas = parseToolSchemasSnapshot(toolSchemasSnapshot)
-        expect(headers.length, `${scenario.name}: a pinning fixture must carry exactly one request/header`).toBe(1)
-        expect(() => restorePinnedToolSchemas(headers[0], toolSchemas), `${scenario.name}: tools must use the sidecar token`)
-          .not.toThrow()
+        const schemaSets = [toolSchemas.initial, ...toolSchemas.changes]
+        expect(schemaSets.length, `${scenario.name}: tool-schema sequence must match the header sequence`)
+          .toBe(headers.length)
+        for (const [index, header] of headers.entries()) {
+          expect(() => restorePinnedToolSchemas(header, schemaSets[index] as unknown[]), `${scenario.name}: tools must use the sidecar token`)
+            .not.toThrow()
+        }
         expect(promptSnapshot.length, `${scenario.name}/${SYSTEM_PROMPT_SNAPSHOT} must not be empty`).toBeGreaterThan(0)
         expect(promptSnapshot.endsWith('\n'), `${scenario.name}/${SYSTEM_PROMPT_SNAPSHOT} must end in a newline`).toBe(true)
         expect(toolSchemasSnapshot, `${scenario.name}/${TOOL_SCHEMAS_SNAPSHOT} must use canonical JSON formatting`)
-          .toBe(formatToolSchemasSnapshot(toolSchemas.initial, toolSchemas.deltas))
-        expect(headerDeltaCount(fixture), `${scenario.name}: a pinning fixture must carry exactly its declared request/header-deltas`)
-          .toBe(scenario.expectedHeaderDeltas ?? 0)
+          .toBe(formatToolSchemasSnapshot(toolSchemas.initial, toolSchemas.changes))
+        expect(headerChangeCount(fixture), `${scenario.name}: a pinning fixture must carry exactly its declared changed headers`)
+          .toBe(scenario.expectedHeaderChanges ?? 0)
       }
     })
 
diff --git a/packages/support/acp-snapshot/tests/fixtures/fake-acp-agent.ts b/packages/support/acp-snapshot/tests/fixtures/fake-acp-agent.ts
index d5fcd75a93..fccd1d6fcc 100644
--- a/packages/support/acp-snapshot/tests/fixtures/fake-acp-agent.ts
+++ b/packages/support/acp-snapshot/tests/fixtures/fake-acp-agent.ts
@@ -24,6 +24,8 @@ interface ScriptedLog {
 
 /** The whole scripted behavior for one run. Every field defaults to the least surprising choice. */
 interface Behavior {
+  /** Exit during startup after writing any configured stderr note. */
+  failOnBoot?: boolean
   /** Reject every `session/new` (exercises the expect-error step without extra dirs). */
   rejectNewSession?: boolean
   /** Reject `session/new` only when `additionalDirectories` is non-empty (the real bridge's rule). */
@@ -62,6 +64,7 @@ const behavior: Behavior = fixtureFile === ''
   : JSON.parse(readFileSync(join(dirname(fixtureFile), 'behavior.json'), 'utf8')) as Behavior
 
 if (behavior.stderrNote !== undefined) process.stderr.write(`${behavior.stderrNote}\n`)
+if (behavior.failOnBoot === true) process.exit(7)
 
 let nextOutboundId = 1000
 let sessionId = ''
diff --git a/packages/support/acp-snapshot/tests/fixtures/record-suite/rec-pin/tool-schemas.golden.json b/packages/support/acp-snapshot/tests/fixtures/record-suite/rec-pin/tool-schemas.golden.json
index f96325d18b..952c057186 100644
--- a/packages/support/acp-snapshot/tests/fixtures/record-suite/rec-pin/tool-schemas.golden.json
+++ b/packages/support/acp-snapshot/tests/fixtures/record-suite/rec-pin/tool-schemas.golden.json
@@ -8,5 +8,5 @@
       }
     }
   ],
-  "deltas": []
+  "changes": []
 }
diff --git a/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/behavior.json b/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/behavior.json
index ad47d20380..4b63893a62 100644
--- a/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/behavior.json
+++ b/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/behavior.json
@@ -5,7 +5,7 @@
     "lines": [
       { "type": "session", "id": "{{SID}}", "createdAt": 100, "cwd": "{{CWD}}" },
       { "type": "request/header", "seq": 0, "time": 100, "data": { "header": { "config": { "model": "fake" }, "system": "SYS PROMPT", "tools": [{ "name": "t1", "description": "D1", "parameters": { "type": "object" } }] }, "reason": "initial" } },
-      { "type": "request/header-delta", "seq": 1, "time": 100, "data": { "system": { "keepStart": 1, "keepEnd": 0, "insert": ["NEW PROMPT LINE"] } } },
+      { "type": "request/header", "seq": 1, "time": 100, "data": { "header": { "config": { "model": "fake" }, "system": "SYS PROMPT\n\nNEW PROMPT LINE", "tools": [{ "name": "t1", "description": "D1", "parameters": { "type": "object" } }] }, "reason": "change" } },
       { "type": "turn/start", "seq": 2, "time": 100, "data": { "turn": 1 } }
     ]
   }]
diff --git a/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/session.jsonl b/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/session.jsonl
index 92947ccb9b..54b64b59a3 100644
--- a/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/session.jsonl
+++ b/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/session.jsonl
@@ -1,4 +1,4 @@
 {"type":"session","id":"12121212-3434-4545-8686-787878787878","createdAt":7,"cwd":"/rec/pin-cwd"}
 {"type":"request/header","seq":0,"time":7,"data":{"header":{"config":{"model":"fake"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}
-{"type":"request/header-delta","seq":1,"time":7,"data":{"system":{"keepStart":1,"keepEnd":0,"insert":["{{system}}"]}}}
+{"type":"request/header","seq":1,"time":7,"data":{"header":{"config":{"model":"fake"},"system":"{{system}}","tools":"{{tools}}"},"reason":"change"}}
 {"type":"turn/start","seq":2,"time":7,"data":{"turn":1}}
diff --git a/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/system-prompt.golden.md b/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/system-prompt.golden.md
index 77a5bc4cd7..22925f0dbb 100644
--- a/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/system-prompt.golden.md
+++ b/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/system-prompt.golden.md
@@ -1,5 +1,7 @@
 SYS PROMPT
 
-<!-- request/header-delta 1: keepStart=1, keepEnd=0 -->
+<!-- request/header change 1 -->
+
+SYS PROMPT
 
 NEW PROMPT LINE
diff --git a/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/tool-schemas.golden.json b/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/tool-schemas.golden.json
index f96325d18b..38bc716d49 100644
--- a/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/tool-schemas.golden.json
+++ b/packages/support/acp-snapshot/tests/fixtures/suite/pin-turn/tool-schemas.golden.json
@@ -8,5 +8,15 @@
       }
     }
   ],
-  "deltas": []
+  "changes": [
+    [
+      {
+        "name": "t1",
+        "description": "D1",
+        "parameters": {
+          "type": "object"
+        }
+      }
+    ]
+  ]
 }
diff --git a/packages/support/acp-snapshot/tests/harness.spec.ts b/packages/support/acp-snapshot/tests/harness.spec.ts
index 01b93dc81e..1d953f5e95 100644
--- a/packages/support/acp-snapshot/tests/harness.spec.ts
+++ b/packages/support/acp-snapshot/tests/harness.spec.ts
@@ -14,10 +14,12 @@ import { runScenario, type AgentUnderTest, type InputStep } from '../src/harness
  * assertions read plain `rawStdout`.
  */
 
+const fakeAgent = fileURLToPath(new URL('./fixtures/fake-acp-agent.ts', import.meta.url))
 const AGENT: AgentUnderTest = {
-  binScript: fileURLToPath(new URL('./fixtures/fake-acp-agent.ts', import.meta.url)),
+  binScript: fakeAgent,
+  libBinScript: fakeAgent,
   // The fake bin ignores its config argv; any real path documents the shape.
-  configPath: fileURLToPath(new URL('./fixtures/fake-acp-agent.ts', import.meta.url)),
+  configPath: fakeAgent,
   tsconfigPath: fileURLToPath(new URL('../../../../tsconfig.json', import.meta.url)),
 }
 
@@ -38,6 +40,14 @@ async function scenario(behavior: object): Promise<{ dir: string; fixtureFile: s
 const boot: InputStep[] = [{ op: 'initialize' }, { op: 'newSession' }]
 
 describe('runScenario', () => {
+  it('includes agent stderr when the ACP connection closes during startup', { timeout: 20_000 }, async () => {
+    const { fixtureFile } = await scenario({ failOnBoot: true, stderrNote: 'fake agent requested startup failure' })
+    await expect(runScenario(
+      { steps: [{ op: 'initialize' }] },
+      { agent: AGENT, mode: 'replay', fixtureFile },
+    )).rejects.toThrow(/agent stderr:\nfake agent requested startup failure/)
+  })
+
   it('drives a full turn: initialize (terminal caps), session, prompt, permission stub, harvest', { timeout: 20_000 }, async () => {
     const { fixtureFile } = await scenario({
       permissionProbe: true,
diff --git a/packages/support/acp-snapshot/tests/normalize.spec.ts b/packages/support/acp-snapshot/tests/normalize.spec.ts
index d6c95045f6..2beaba5114 100644
--- a/packages/support/acp-snapshot/tests/normalize.spec.ts
+++ b/packages/support/acp-snapshot/tests/normalize.spec.ts
@@ -93,6 +93,52 @@ describe('normalizeSessionLog', () => {
     expect(out).not.toContain(ctx.cwd)
   })
 
+  it('scrubs random local spill paths under the snapshot cwd', () => {
+    const ev = JSON.stringify({
+      type: 'tool/result', seq: 2, time: 5,
+      data: {
+        content: [{
+          type: 'text',
+          text: `Full formatted result stored at: ${ctx.cwd}/.spill/session-c22bc3f1d2af/8a7b6c5d4e3f-bash.txt. Use read with offset/limit, or grep this path to search within it.`,
+        }],
+      },
+    })
+    const out = normalizeSessionLog(`${header({ cwd: ctx.cwd })}\n${ev}\n`, ctx)
+    expect(out).toContain('{{spillLocator:bash.txt}}')
+    expect(out).not.toContain('session-c22bc3f1d2af')
+    expect(out).not.toContain('8a7b6c5d4e3f')
+  })
+
+  it('scrubs macOS /private aliases for local spill paths', () => {
+    const ev = JSON.stringify({
+      type: 'tool/result', seq: 2, time: 5,
+      data: {
+        content: [{
+          type: 'text',
+          text: `Full formatted result stored at: /private${ctx.cwd}/.spill/session-c22bc3f1d2af/8a7b6c5d4e3f-bash.txt. Use read with offset/limit, or grep this path to search within it.`,
+        }],
+      },
+    })
+    const out = normalizeSessionLog(`${header({ cwd: ctx.cwd })}\n${ev}\n`, ctx)
+    expect(out).toContain('{{spillLocator:bash.txt}}')
+    expect(out).not.toContain('/private{{spillLocator')
+  })
+
+  it('scrubs fixed snapshot spill paths', () => {
+    const ev = JSON.stringify({
+      type: 'tool/result', seq: 2, time: 5,
+      data: {
+        content: [{
+          type: 'text',
+          text: 'Full formatted result stored at: /tmp/dsh-acp-snapshot-spill/session-c22bc3f1d2af/8a7b6c5d4e3f-bash.txt. Use read with offset/limit, or grep this path to search within it.',
+        }],
+      },
+    })
+    const out = normalizeSessionLog(`${header({ cwd: ctx.cwd })}\n${ev}\n`, ctx)
+    expect(out).toContain('{{spillLocator:bash.txt}}')
+    expect(out).not.toContain('/tmp/dsh-acp-snapshot-spill')
+  })
+
   it('scrubs the session id in the header', () => {
     const out = normalizeSessionLog(`${header({ id: ctx.sessionIds[0] })}\n`, ctx)
     expect(out).toContain('{{sessionId}}')
@@ -181,89 +227,19 @@ describe('scrubRequestHeaders', () => {
     expect(scrubRequestHeaders(`${headerLine}\n${odd}\n`)).toContain('"messagePrefix":"weird"')
   })
 
-  it('scrubs a header-delta prefix replacement to one token per message', () => {
-    const delta = JSON.stringify({
-      type: 'request/header-delta', seq: 8, time: 9,
-      data: { messagePrefix: [{ role: 'user', content: [{ type: 'text', text: 'leaked opener' }] }] },
-    })
-    const out = scrubRequestHeaders(`${headerLine}\n${delta}\n`)
-    expect(out).toContain('"messagePrefix":["{{messagePrefix}}"]')
-    expect(out).not.toContain('leaked opener')
-    // The empty-array transition-to-absence stays a structural fact.
-    const toNone = JSON.stringify({ type: 'request/header-delta', seq: 9, time: 9, data: { messagePrefix: [] } })
-    expect(scrubRequestHeaders(`${headerLine}\n${toNone}\n`)).toContain('"messagePrefix":[]')
-  })
-
-  it('leaves a delta with no scrubbable payload byte-identical (config-only, or non-array shapes)', () => {
-    const configOnly = JSON.stringify({ type: 'request/header-delta', seq: 8, time: 9, data: { config: { model: 'm2' } } })
-    const oddShapes = JSON.stringify({ type: 'request/header-delta', seq: 9, time: 9, data: { system: { insert: 'not-an-array' }, tools: null } })
+  it('leaves malformed headers with no scrubbable payload byte-identical', () => {
     const headerless = JSON.stringify({ type: 'request/header', seq: 10, time: 9, data: { reason: 'initial' } })
     const nullData = JSON.stringify({ type: 'request/header', seq: 11, time: 9, data: null })
-    const raw = `${headerLine}\n${configOnly}\n${oddShapes}\n${headerless}\n${nullData}\n`
+    const raw = `${headerLine}\n${headerless}\n${nullData}\n`
     expect(scrubRequestHeaders(raw)).toBe(raw)
   })
 
-  it('scrubs a one-sided tools delta and passes non-object schema entries through', () => {
-    const addedOnly = JSON.stringify({
-      type: 'request/header-delta', seq: 8, time: 9,
-      data: { tools: { added: [null, 'weird', { name: 'x', description: 'D' }] } },
-    })
-    const out = scrubRequestHeaders(`${headerLine}\n${addedOnly}\n`)
-    // Non-object entries survive untouched; the object entry keeps only name.
-    expect(out).toContain('"added":[null,"weird",{"name":"x","description":"{{tools}}"}]')
-    const changedOnly = JSON.stringify({
-      type: 'request/header-delta', seq: 8, time: 9,
-      data: { tools: { changed: [{ name: 'y', parameters: {} }] } },
-    })
-    expect(scrubRequestHeaders(`${headerLine}\n${changedOnly}\n`))
-      .toContain('"changed":[{"name":"y","parameters":"{{tools}}"}]')
-  })
-
-  it('scrubs a header-delta system payload but keeps its line positions and arity', () => {
-    const delta = JSON.stringify({
-      type: 'request/header-delta', seq: 8, time: 9,
-      data: { system: { keepStart: 1, keepEnd: 4, insert: ['leaked prompt line', 'second line'] }, config: { model: 'm2' } },
-    })
-    const out = scrubRequestHeaders(`${headerLine}\n${delta}\n`)
-    // One token PER inserted line: the edit's position AND extent survive.
-    expect(out).toContain('"insert":["{{system}}","{{system}}"]')
-    expect(out).toContain('"keepStart":1')
-    expect(out).toContain('"keepEnd":4')
-    expect(out).toContain('"config":{"model":"m2"}')
-    expect(out).not.toContain('leaked prompt line')
-    expect(out).not.toContain('{{tools}}') // no tools delta → none invented
-  })
-
-  it('scrubs a header-delta tools payload but keeps the added/removed/changed names', () => {
-    const delta = JSON.stringify({
-      type: 'request/header-delta', seq: 8, time: 9,
-      data: {
-        tools: {
-          added: [{ name: 'grep', description: 'Search files.', parameters: { type: 'object' } }],
-          removed: ['bash_kill'],
-          changed: [{ name: 'read', description: 'Read v2.', parameters: { type: 'object' } }],
-        },
-      },
-    })
-    const out = scrubRequestHeaders(`${headerLine}\n${delta}\n`)
-    // WHICH tools changed is behavior and survives; their bulk does not.
-    expect(out).toContain('"added":[{"name":"grep","description":"{{tools}}","parameters":"{{tools}}"}]')
-    expect(out).toContain('"removed":["bash_kill"]')
-    expect(out).toContain('"changed":[{"name":"read","description":"{{tools}}","parameters":"{{tools}}"}]')
-    expect(out).not.toContain('Search files')
-    expect(out).not.toContain('Read v2')
-  })
-
   it('passes every other line through byte-for-byte and is idempotent', () => {
     const other = JSON.stringify({ type: 'assistant/chunk', seq: 4, time: 9, data: { turn: 1, step: 1, chunk: { type: 'text-delta', index: 0, text: 'hi' } } })
-    const delta = JSON.stringify({
-      type: 'request/header-delta', seq: 8, time: 9,
-      data: { system: { keepStart: 0, keepEnd: 0, insert: ['x'] }, tools: { added: [{ name: 't', description: 'd', parameters: {} }], removed: [], changed: [] } },
-    })
-    const raw = `${headerLine}\n${headerEvent({ config: { model: 'm' }, system: 's', tools: [] })}\n${delta}\n${other}\n`
+    const raw = `${headerLine}\n${headerEvent({ config: { model: 'm' }, system: 's', tools: [] })}\n${other}\n`
     const once = scrubRequestHeaders(raw)
     expect(once.split('\n')[0]).toBe(headerLine)
-    expect(once.split('\n')[3]).toBe(other)
+    expect(once.split('\n')[2]).toBe(other)
     expect(scrubRequestHeaders(once)).toBe(once)
   })
 })
@@ -281,12 +257,15 @@ describe('scrubSystemPrompts', () => {
         reason: 'initial',
       },
     })
-    const delta = JSON.stringify({
-      type: 'request/header-delta', seq: 2, time: 3,
+    const changed = JSON.stringify({
+      type: 'request/header', seq: 2, time: 3,
       data: {
-        system: { keepStart: 1, keepEnd: 2, insert: ['new prompt line'] },
-        tools: { changed: [{ name: 'read', description: 'changed schema' }] },
-        messagePrefix: [{ role: 'user', content: [{ type: 'text', text: 'changed prefix' }] }],
+        header: {
+          system: 'new prompt',
+          tools: [{ name: 'read', description: 'changed schema' }],
+          messagePrefix: [{ role: 'user', content: [{ type: 'text', text: 'changed prefix' }] }],
+        },
+        reason: 'change',
       },
     })
     const toolsOnly = JSON.stringify({
@@ -294,11 +273,10 @@ describe('scrubSystemPrompts', () => {
       data: { header: { tools: [{ name: 'read', description: 'schema only' }] }, reason: 'resume' },
     })
 
-    const out = scrubSystemPrompts(`${header}\n${delta}\n${toolsOnly}\n`)
+    const out = scrubSystemPrompts(`${header}\n${changed}\n${toolsOnly}\n`)
     expect(out).toContain('"system":"{{system}}"')
-    expect(out).toContain('"insert":["{{system}}"]')
     expect(out).not.toContain('full prompt')
-    expect(out).not.toContain('new prompt line')
+    expect(out).not.toContain('new prompt')
     expect(out).toContain('full schema')
     expect(out).toContain('full prefix')
     expect(out).toContain('changed schema')
@@ -321,12 +299,15 @@ describe('scrubToolSchemas', () => {
         reason: 'initial',
       },
     })
-    const delta = JSON.stringify({
-      type: 'request/header-delta', seq: 2, time: 3,
+    const changed = JSON.stringify({
+      type: 'request/header', seq: 2, time: 3,
       data: {
-        system: { keepStart: 1, keepEnd: 2, insert: ['new prompt line'] },
-        tools: { added: [{ name: 'grep', description: 'new schema' }], changed: [{ name: 'read', description: 'changed schema' }] },
-        messagePrefix: [{ role: 'user', content: [{ type: 'text', text: 'changed prefix' }] }],
+        header: {
+          system: 'new prompt',
+          tools: [{ name: 'grep', description: 'new schema' }],
+          messagePrefix: [{ role: 'user', content: [{ type: 'text', text: 'changed prefix' }] }],
+        },
+        reason: 'change',
       },
     })
     const systemOnly = JSON.stringify({
@@ -334,15 +315,12 @@ describe('scrubToolSchemas', () => {
       data: { header: { system: 'prompt only' }, reason: 'resume' },
     })
 
-    const out = scrubToolSchemas(`${header}\n${delta}\n${systemOnly}\n`)
-    expect(out).toContain('"tools":"{{tools}}"')
-    expect(out).toContain('"added":[{"name":"grep","description":"{{tools}}"}]')
-    expect(out).toContain('"changed":[{"name":"read","description":"{{tools}}"}]')
+    const out = scrubToolSchemas(`${header}\n${changed}\n${systemOnly}\n`)
+    expect(out.match(/"tools":"{{tools}}"/g)).toHaveLength(2)
     expect(out).not.toContain('full schema')
     expect(out).not.toContain('new schema')
-    expect(out).not.toContain('changed schema')
     expect(out).toContain('full prompt')
-    expect(out).toContain('new prompt line')
+    expect(out).toContain('new prompt')
     expect(out).toContain('full prefix')
     expect(out).toContain('changed prefix')
     expect(out.split('\n')[2]).toBe(systemOnly)
diff --git a/packages/support/acp-snapshot/tests/suite.spec.ts b/packages/support/acp-snapshot/tests/suite.spec.ts
index 204647d8c9..4cedc1cdfb 100644
--- a/packages/support/acp-snapshot/tests/suite.spec.ts
+++ b/packages/support/acp-snapshot/tests/suite.spec.ts
@@ -9,12 +9,10 @@ import {
   childFixturePaths,
   fixtureContext,
   formatSystemPromptSnapshot,
+  headerChangeCount,
   formatToolSchemasSnapshot,
-  headerDeltaCount,
   normalizedHeaders,
-  normalizedSystemPromptDeltas,
   normalizedSystemPrompts,
-  normalizedToolSchemaDeltas,
   normalizedToolSchemas,
   parseToolSchemasSnapshot,
   refreshFixtureReplacements,
@@ -34,9 +32,11 @@ import {
  * spec once with `ACP_SNAPSHOT_SPEC_BOOTSTRAP=1`, then review and commit the resulting tree.
  */
 
+const fakeAgent = fileURLToPath(new URL('./fixtures/fake-acp-agent.ts', import.meta.url))
 const AGENT = {
-  binScript: fileURLToPath(new URL('./fixtures/fake-acp-agent.ts', import.meta.url)),
-  configPath: fileURLToPath(new URL('./fixtures/fake-acp-agent.ts', import.meta.url)),
+  binScript: fakeAgent,
+  libBinScript: fakeAgent,
+  configPath: fakeAgent,
   tsconfigPath: fileURLToPath(new URL('../../../../tsconfig.json', import.meta.url)),
 }
 
@@ -45,7 +45,7 @@ const RECORD_SRC = fileURLToPath(new URL('./fixtures/record-suite', import.meta.
 
 // Replay pins explicit header classes; recording covers the default fallback.
 const REPLAY_SCENARIOS: Scenario[] = [
-  { name: 'pin-turn', hasModelTurn: true, recorded: true, pinsHeader: true, expectedHeaderDeltas: 1, headerClass: 'main' },
+  { name: 'pin-turn', hasModelTurn: true, recorded: true, pinsHeader: true, expectedHeaderChanges: 1, headerClass: 'main' },
   { name: 'plain-turn', hasModelTurn: true, recorded: true, childSessions: 1, headerClass: 'main', configPath: AGENT.configPath },
   { name: 'no-model', hasModelTurn: false, recorded: false, headerClass: 'main' },
   { name: 'blocked-log', hasModelTurn: false, comparesLog: true, recorded: false, headerClass: 'main' },
@@ -76,7 +76,7 @@ afterAll(async () => {
 function staleRefreshFixtures(dir: string): void {
   writeFileSync(join(dir, 'plain-turn', 'stdout.golden.jsonl'), 'stale stdout\n')
   writeFileSync(join(dir, 'pin-turn', 'system-prompt.golden.md'), 'STALE PROMPT\n')
-  writeFileSync(join(dir, 'pin-turn', 'tool-schemas.golden.json'), '{"initial":[{"name":"stale"}],"deltas":[]}\n')
+  writeFileSync(join(dir, 'pin-turn', 'tool-schemas.golden.json'), '{"initial":[{"name":"stale"}],"changes":[]}\n')
 
   const plainBehaviorFile = join(dir, 'plain-turn', 'behavior.json')
   const plainBehavior = JSON.parse(readFileSync(plainBehaviorFile, 'utf8')) as Record<string, unknown>
@@ -127,7 +127,9 @@ describe('defineAcpSnapshotSuite: refresh write-back', () => {
     expect(readFileSync(join(refreshDir, 'pin-turn', 'system-prompt.golden.md'), 'utf8')).toBe([
       'SYS PROMPT',
       '',
-      '<!-- request/header-delta 1: keepStart=1, keepEnd=0 -->',
+      '<!-- request/header change 1 -->',
+      '',
+      'SYS PROMPT',
       '',
       'NEW PROMPT LINE',
       '',
@@ -262,65 +264,41 @@ describe('normalizedToolSchemas', () => {
   })
 })
 
-describe('normalizedToolSchemaDeltas', () => {
-  it('extracts and normalizes object-valued schema edits', () => {
-    const log = [
-      '{"type":"request/header-delta","data":{"tools":{"added":[{"name":"read","description":"work in /w"}]}}}',
-      '{"type":"request/header-delta","data":{"tools":null}}',
-      '{"type":"request/header-delta","data":{"tools":"invalid"}}',
-      '{"type":"request/header-delta","data":{"tools":[]}}',
-      '{"type":"request/header-delta","data":{"system":{"insert":[]}}}',
-      '{"type":"request/header","data":{"tools":{"added":[]}}}',
-      '',
-    ].join('\n')
-    expect(normalizedToolSchemaDeltas(log, { sessionIds: [], cwd: '/w' })).toEqual([
-      { added: [{ name: 'read', description: 'work in {{cwd}}' }] },
-    ])
-  })
-})
-
-describe('normalizedSystemPromptDeltas', () => {
-  it('extracts and normalizes well-formed system edits', () => {
-    const log = [
-      '{"type":"request/header-delta","data":{"system":{"keepStart":1,"keepEnd":0,"insert":["work in /w"]}}}',
-      '{"type":"request/header-delta","data":{"tools":{"replace":[]}}}',
-      '{"type":"request/header-delta","data":{"system":{"keepStart":"1","keepEnd":0,"insert":[]}}}',
-      '{"type":"request/header-delta","data":{"system":{"keepStart":1,"keepEnd":0,"insert":[null]}}}',
-      '',
-    ].join('\n')
-    expect(normalizedSystemPromptDeltas(log, { sessionIds: [], cwd: '/w' })).toEqual([
-      { keepStart: 1, keepEnd: 0, insert: ['work in {{cwd}}'] },
-    ])
-  })
-})
-
 describe('formatSystemPromptSnapshot', () => {
   it('adds a missing terminal newline without changing an existing one', () => {
     expect(formatSystemPromptSnapshot('prompt')).toBe('prompt\n')
     expect(formatSystemPromptSnapshot('prompt\n')).toBe('prompt\n')
   })
 
-  it('renders readable system-prompt delta sections', () => {
-    expect(formatSystemPromptSnapshot('prompt', [
-      { keepStart: 1, keepEnd: 0, insert: ['new', 'lines'] },
-    ])).toBe('prompt\n\n<!-- request/header-delta 1: keepStart=1, keepEnd=0 -->\n\nnew\nlines\n')
+  it('renders readable changed-prompt sections', () => {
+    expect(formatSystemPromptSnapshot('prompt', ['new\nlines']))
+      .toBe('prompt\n\n<!-- request/header change 1 -->\n\nnew\nlines\n')
   })
 
-  it('does not double the newline of a delta insert with a trailing blank line', () => {
-    expect(formatSystemPromptSnapshot('prompt\n', [
-      { keepStart: 2, keepEnd: 1, insert: ['tail', ''] },
-    ])).toBe('prompt\n\n<!-- request/header-delta 1: keepStart=2, keepEnd=1 -->\n\ntail\n')
+  it('does not double the newline of a changed prompt', () => {
+    expect(formatSystemPromptSnapshot('prompt\n', ['changed\n']))
+      .toBe('prompt\n\n<!-- request/header change 1 -->\n\nchanged\n')
+  })
+})
+
+describe('headerChangeCount', () => {
+  it('counts changed request headers, ignoring anchors, blanks, and other lines', () => {
+    const change = JSON.stringify({ type: 'request/header', seq: 2, time: 9, data: { reason: 'change' } })
+    const anchor = JSON.stringify({ type: 'request/header', seq: 0, time: 9, data: { reason: 'initial' } })
+    const other = JSON.stringify({ type: 'turn/start', seq: 1, time: 9, data: {} })
+    expect(headerChangeCount(`${anchor}\n${other}\n\n${change}\n${change}\n`)).toBe(2)
+    expect(headerChangeCount(`${anchor}\n`)).toBe(0)
   })
 })
 
 describe('tool-schema snapshots', () => {
   const snapshot = {
     initial: [{ name: 'read', description: 'Read a file.' }],
-    deltas: [{ added: [{ name: 'grep', description: 'Search files.' }] }],
+    changes: [[{ name: 'grep', description: 'Search files.' }]],
   }
 
   it('formats and parses canonical structured JSON', () => {
-    const formatted = formatToolSchemasSnapshot(snapshot.initial, snapshot.deltas)
+    const formatted = formatToolSchemasSnapshot(snapshot.initial, snapshot.changes)
     expect(formatted).toBe(`${JSON.stringify(snapshot, null, 2)}\n`)
     expect(parseToolSchemasSnapshot(formatted)).toEqual(snapshot)
   })
@@ -329,29 +307,21 @@ describe('tool-schema snapshots', () => {
     expect(() => parseToolSchemasSnapshot('null')).toThrow(/must be an object/)
     expect(() => parseToolSchemasSnapshot('"invalid"')).toThrow(/must be an object/)
     expect(() => parseToolSchemasSnapshot('[]')).toThrow(/must be an object/)
-    expect(() => parseToolSchemasSnapshot('{"initial":{},"deltas":[]}')).toThrow(/array-valued/)
-    expect(() => parseToolSchemasSnapshot('{"initial":[],"deltas":{}}')).toThrow(/array-valued/)
+    expect(() => parseToolSchemasSnapshot('{"initial":{},"changes":[]}')).toThrow(/array-valued/)
+    expect(() => parseToolSchemasSnapshot('{"initial":[],"changes":{}}')).toThrow(/array-valued/)
+    expect(() => parseToolSchemasSnapshot('{"initial":[],"changes":[{}]}')).toThrow(/array-valued/)
   })
 
   it('restores initial schemas into the pinned header token', () => {
-    expect(restorePinnedToolSchemas({ system: '{{system}}', tools: '{{tools}}' }, snapshot))
+    expect(restorePinnedToolSchemas({ system: '{{system}}', tools: '{{tools}}' }, snapshot.initial))
       .toEqual({ system: '{{system}}', tools: snapshot.initial })
   })
 
   it('rejects invalid headers and a missing tool token', () => {
-    expect(() => restorePinnedToolSchemas(null, snapshot)).toThrow(/must be an object/)
-    expect(() => restorePinnedToolSchemas('invalid', snapshot)).toThrow(/must be an object/)
-    expect(() => restorePinnedToolSchemas([], snapshot)).toThrow(/must be an object/)
-    expect(() => restorePinnedToolSchemas({ tools: [] }, snapshot)).toThrow(/must equal/)
-  })
-})
-
-describe('headerDeltaCount', () => {
-  it('counts request/header-delta events, ignoring blanks and other lines', () => {
-    const delta = JSON.stringify({ type: 'request/header-delta', seq: 2, time: 9, data: {} })
-    const other = JSON.stringify({ type: 'request/header', seq: 0, time: 9, data: {} })
-    expect(headerDeltaCount(`${other}\n\n${delta}\n${delta}\n`)).toBe(2)
-    expect(headerDeltaCount(`${other}\n`)).toBe(0)
+    expect(() => restorePinnedToolSchemas(null, snapshot.initial)).toThrow(/must be an object/)
+    expect(() => restorePinnedToolSchemas('invalid', snapshot.initial)).toThrow(/must be an object/)
+    expect(() => restorePinnedToolSchemas([], snapshot.initial)).toThrow(/must be an object/)
+    expect(() => restorePinnedToolSchemas({ tools: [] }, snapshot.initial)).toThrow(/must equal/)
   })
 })
 
diff --git a/packages/support/acp-snapshot/tsconfig.json b/packages/support/acp-snapshot/tsconfig.json
index 749cb0208e..9120df0ad1 100644
--- a/packages/support/acp-snapshot/tsconfig.json
+++ b/packages/support/acp-snapshot/tsconfig.json
@@ -7,5 +7,7 @@
   "include": [
     "src"
   ],
-  "references": []
+  "references": [
+    { "path": "../loader-smoke" }
+  ]
 }
diff --git a/packages/support/agent-loop-testkit/README.md b/packages/support/agent-loop-testkit/README.md
new file mode 100644
index 0000000000..350a8643e1
--- /dev/null
+++ b/packages/support/agent-loop-testkit/README.md
@@ -0,0 +1,27 @@
+# `@deepseek-ai/dsh-agent-loop-testkit`
+
+Shared prerequisite mounting for tests that exercise the concrete `AgentLoop`. `mountAgentLoopTestDependencies(ctx, options?)` installs the LLM, session, system-prompt, tool, and agent services in dependency order, then returns before the loop is mounted.
+
+The caller registers adapters and optional plugins, mounts `AgentLoop` with the configuration under test, and disposes its own Context. System-prompt and tool-registry configuration can be forwarded through `options`; the helper does not provide test defaults beyond those owned by the services. A plugin-load failure rejects the helper call, while services activated earlier in the sequence remain owned by the caller's Context.
+
+```ts
+import { Context } from 'cordis'
+import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
+
+const ctx = new Context()
+
+await mountAgentLoopTestDependencies(ctx)
+// Register the test adapter and any optional plugins here.
+await ctx.plugin(AgentLoop, { agents: [] })
+```
+
+Tests of injection failures, partial topology, service load order, or service teardown mount their dependencies directly instead of using this helper.
+
+## Model Experience
+
+None, as this test-only composition helper neither drives nor modifies model requests.
+
+## Known Limitations and Deferred Work
+
+- **Only the mandatory prerequisite spine is shared** — adapters, optional plugins, `AgentLoop`, agents, and Context teardown remain caller-owned so scenario-specific ordering stays visible.
diff --git a/packages/support/agent-loop-testkit/package.json b/packages/support/agent-loop-testkit/package.json
new file mode 100644
index 0000000000..423bd3e80d
--- /dev/null
+++ b/packages/support/agent-loop-testkit/package.json
@@ -0,0 +1,41 @@
+{
+  "name": "@deepseek-ai/dsh-agent-loop-testkit",
+  "description": "Shared prerequisite mounting for tests that exercise the concrete agent loop",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "@deepseek-ai/dsh-agent": "^0.0.1",
+    "@deepseek-ai/dsh-llm": "^0.0.1",
+    "@deepseek-ai/dsh-session": "^0.0.1",
+    "@deepseek-ai/dsh-system-prompt": "^0.0.1",
+    "@deepseek-ai/dsh-tools": "^0.0.1",
+    "cordis": "^4.0.0-rc.7"
+  },
+  "devDependencies": {
+    "@deepseek-ai/dsh-agent": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-session": "workspace:^",
+    "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-tools": "workspace:^",
+    "cordis": "^4.0.0-rc.7"
+  }
+}
diff --git a/packages/support/agent-loop-testkit/src/index.ts b/packages/support/agent-loop-testkit/src/index.ts
new file mode 100644
index 0000000000..c7b0cb7304
--- /dev/null
+++ b/packages/support/agent-loop-testkit/src/index.ts
@@ -0,0 +1,46 @@
+/**
+ * Shared mounting for the services required before tests load the concrete
+ * agent loop. The caller retains ownership of the context, loop, adapters,
+ * optional plugins, and teardown.
+ * @module @deepseek-ai/dsh-agent-loop-testkit
+ */
+
+import type { Context } from 'cordis'
+import AgentRegistry from '@deepseek-ai/dsh-agent'
+import LlmService from '@deepseek-ai/dsh-llm'
+import SessionStore from '@deepseek-ai/dsh-session'
+import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import type { Config as SystemPromptConfig } from '@deepseek-ai/dsh-system-prompt'
+import ToolRegistry from '@deepseek-ai/dsh-tools'
+import type { Config as ToolRegistryConfig } from '@deepseek-ai/dsh-tools'
+
+/** Configuration forwarded to the prerequisite service plugins. */
+export interface AgentLoopTestDependenciesOptions {
+  /** Configuration for the system-prompt registry. */
+  readonly systemPrompt?: SystemPromptConfig
+  /** Configuration for the tool registry. */
+  readonly tools?: ToolRegistryConfig
+}
+
+/**
+ * Mount the standard prerequisite services for an AgentLoop test.
+ *
+ * The function deliberately does not mount AgentLoop or register an adapter,
+ * so tests retain control of load order and the topology under test. The
+ * context owns every mounted service and remains responsible for disposal. A
+ * plugin-load failure rejects the promise; services activated earlier in the
+ * sequence remain context-owned and unwind with that context.
+ * @param ctx - test context that owns the mounted services.
+ * @param options - optional service configuration forwarded without mutation.
+ * @returns after every prerequisite service has activated.
+ */
+export async function mountAgentLoopTestDependencies(
+  ctx: Context,
+  options: AgentLoopTestDependenciesOptions = {},
+): Promise<void> {
+  await ctx.plugin(LlmService)
+  await ctx.plugin(SessionStore)
+  await ctx.plugin(SystemPrompt, options.systemPrompt ?? {})
+  await ctx.plugin(ToolRegistry, options.tools ?? {})
+  await ctx.plugin(AgentRegistry)
+}
diff --git a/packages/support/agent-loop-testkit/tests/agent-loop-testkit.spec.ts b/packages/support/agent-loop-testkit/tests/agent-loop-testkit.spec.ts
new file mode 100644
index 0000000000..aa125b561f
--- /dev/null
+++ b/packages/support/agent-loop-testkit/tests/agent-loop-testkit.spec.ts
@@ -0,0 +1,20 @@
+import { describe, expect, it } from 'vitest'
+import { Context } from 'cordis'
+import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { renderPrompt } from '@deepseek-ai/dsh-system-prompt'
+import { mountAgentLoopTestDependencies } from '../src/index.ts'
+
+describe('dsh-agent-loop-testkit', () => {
+  it('mounts a configurable prerequisite spine that can activate AgentLoop', async () => {
+    const ctx = new Context()
+    await mountAgentLoopTestDependencies(ctx, {
+      systemPrompt: { persona: 'Test persona.' },
+      tools: { mode: 'native' },
+    })
+
+    expect(renderPrompt(await ctx.systemPrompt.assemble())).toContain('Test persona.')
+    await expect(ctx.plugin(AgentLoop, { agents: [] })).resolves.toBeDefined()
+
+    await ctx.fiber.dispose()
+  })
+})
diff --git a/packages/support/agent-loop-testkit/tsconfig.json b/packages/support/agent-loop-testkit/tsconfig.json
new file mode 100644
index 0000000000..5e5b3c47f2
--- /dev/null
+++ b/packages/support/agent-loop-testkit/tsconfig.json
@@ -0,0 +1,33 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": [
+    "src"
+  ],
+  "references": [
+    {
+      "path": "../../../vendor/cosmokit"
+    },
+    {
+      "path": "../../../vendor/cordis"
+    },
+    {
+      "path": "../../core/agent"
+    },
+    {
+      "path": "../../llm/llm"
+    },
+    {
+      "path": "../../core/session"
+    },
+    {
+      "path": "../../core/system-prompt"
+    },
+    {
+      "path": "../../core/tools"
+    }
+  ]
+}
diff --git a/packages/support/invariants/README.md b/packages/support/invariants/README.md
index db8cf9e231..661e073d13 100644
--- a/packages/support/invariants/README.md
+++ b/packages/support/invariants/README.md
@@ -4,7 +4,7 @@ Runtime event-contract assertions intended for development diagnostics. This pur
 
 The plugin has no environment guard: it is active wherever it is registered. The default [`dsh-agent-spine-demo`](../../examples/agent-spine-demo/README.md) bundle mounts it unconditionally; a custom composition can omit it when the runtime cost is undesirable. It doubles as executable documentation of the event taxonomy — the assertions *are* the contract.
 
-Session itself owns immutable log storage in every composition: it takes one lossless JSON snapshot of each accepted event, deep-freezes that record, and exposes the log through immutable array snapshots. The invariants plugin checks the cross-record and cross-seam rules that storage immutability cannot express.
+Session itself owns immutable, surface-valid log storage in every composition: it takes one lossless JSON snapshot of each candidate, validates the complete surface transition, deep-freezes the accepted record, and exposes the log through immutable array snapshots. The invariants plugin checks the remaining cross-record and cross-seam rules that Session does not own.
 
 Session-log assertions run during Cordis `internal/dispatch`, while `Session.append()` is resolving the `session/event` callback snapshot but before it pushes the candidate into the log. A valid transition is staged by exact event identity and applied to the live trace only when that same committed event reaches the plugin's contained post-commit listener. A later internal dispatch check can therefore veto without advancing either the log or the invariant trace, while ordinary `session/event` observer failures remain observe-only.
 
@@ -32,6 +32,7 @@ Session log (per session):
 - **steps nest in turns** — `step/start` opens a step in the open turn; `step/end` closes the matching step.
 - **chunks belong to an open step** — `step/start` precedes its `assistant/chunk`s.
 - **a `tool/result` needs a prior `tool/call`** — but NOT the converse: a `tool/call` may have no result (a thrown tool-execution pipeline step ends the turn with no `tool/result`, which is legal).
+- **provenance sources are valid and unambiguous** — `sourceEventSeqs` contains unique earlier known seqs; only `assistant/message` may carry an explicit empty list, which denotes a known empty provider stream rather than absent legacy provenance.
 
 Agent status (per agent):
 
@@ -39,7 +40,7 @@ Agent status (per agent):
 
 Model requests (on `llm/stream`):
 
-- **a loop-built request is exactly what the log reconstructs** — a frozen request with a live `sessionId` is rebuilt through a fresh `Session` from the prefix before its in-flight `step/start`; later content belongs to the next request, and hand-built unfrozen one-shots are excluded. Frozen messages must match that derivation, while every other field matches folded `request/header*` events. The prepended check runs before ordinary short-circuiting stream listeners, but correctness comes from the sequence boundary rather than listener timing. See the [reconstructability RFC](../../../docs/rfc/implemented/architecture/2026-07-05-reconstructable-requests.md).
+- **a loop-built request is exactly what the log reconstructs** — a frozen request with a live `sessionId` (the loop-built marker; hand-built one-shots like compaction's summarize are unfrozen and skipped) must carry frozen `messages` deep-equal to the derivation over the log prefix strictly before the in-flight step's `step/start` (rebuilt through a FRESH `Session`, so the live cache cannot vouch for itself — and boundary-correct: content logged after `step/start` legitimately belongs to the next request), and every non-content field must equal the latest logged `request/header` (see [the reconstructability RFC](../../../docs/rfc/implemented/architecture/2026-07-05-reconstructable-requests.md)). Registered with `prepend: true` so a short-circuiting `llm/stream` listener (the replay adapter) cannot silence it; prepend orders it against append-registered listeners only — correctness rests on the seq-bounded rebuild, never listener timing.
 
 On any violation it throws `InvariantError` (`code: 'INVARIANT'`).
 
diff --git a/packages/support/invariants/src/index.ts b/packages/support/invariants/src/index.ts
index d6e2b37946..cdb06edbac 100644
--- a/packages/support/invariants/src/index.ts
+++ b/packages/support/invariants/src/index.ts
@@ -3,7 +3,8 @@
  * turn and step nesting, scoped dispatch, status transitions, and request
  * reconstruction. The plugin has no environment guard and is active wherever
  * mounted, including the default `dsh-agent-spine-demo` bundle; custom compositions
- * may omit it. Sessions still own event snapshots and freezing.
+ * may omit it. Sessions own immutable, surface-valid event storage; this plugin
+ * checks only relationships that event acceptance cannot express.
  * @module @deepseek-ai/dsh-invariants
  */
 
@@ -13,7 +14,7 @@ import { assertNever, HarnessError } from '@deepseek-ai/dsh-llm'
 import type { CallId, GenerateOptions } from '@deepseek-ai/dsh-llm'
 import type { Agent, AgentStatus } from '@deepseek-ai/dsh-agent'
 import { Session, SessionId, foldRequestHeader } from '@deepseek-ai/dsh-session'
-import type { SessionEvent, SurfaceEventType } from '@deepseek-ai/dsh-session'
+import type { SessionEvent } from '@deepseek-ai/dsh-session'
 import { scopedSubjectResolverFor } from './scoped-events.generated.ts'
 
 export const name = 'invariants'
@@ -48,15 +49,6 @@ interface SessionTrace {
    * `step/end` — a result must arrive in the same step as its call.
    */
   pendingCalls: Set<CallId>
-  /** Every seq seen so far — validates `sourceEventSeqs` references. */
-  knownSeqs: Set<number>
-  /**
-   * The seqs currently on the surface linked list, in linked-list order
-   * (head to tail). A replace reorders this relative to seq order (the new
-   * node takes the replaced range's position), so range validation is
-   * positional, not by seq comparison.
-   */
-  surface: number[]
 }
 
 /** One accepted event's deferred mutation of a live session trace. */
@@ -68,12 +60,6 @@ interface SessionTraceTransition {
     | { kind: 'none' }
     | { kind: 'add' | 'delete'; callId: CallId }
     | { kind: 'clear' }
-  /** The event's mutation of the derived surface order. */
-  surface:
-    | { kind: 'none' | 'append' }
-    | { kind: 'replace'; start: number; count: number }
-  /** The committed event sequence to add to the known-sequence set. */
-  seq: number
 }
 
 /** Assert that a step-scoped event names the currently open turn and step. */
@@ -97,73 +83,6 @@ function validateEvent(trace: SessionTrace, event: SessionEvent): SessionTraceTr
   let nextTurn = trace.nextTurn
   let nextStep = trace.nextStep
   let pendingCalls: SessionTraceTransition['pendingCalls'] = { kind: 'none' }
-  let surface: SessionTraceTransition['surface'] = { kind: 'none' }
-
-  // --- Surface invariants ---
-  // Surface metadata (sourceEventSeqs, surfaceOp) is only valid on
-  // surface-eligible event types. The compiler enforces this at append()
-  // call sites; this runtime check catches casts and persisted data.
-  const SURFACE_TYPES = new Set<string>(['user/message', 'assistant/message', 'tool/result', 'context/message', 'steering/message'])
-  // Cast to surface-eligible event type so we can access surfaceOp and
-  // sourceEventSeqs (optional on SessionEvent, mandatory on SurfaceEvent).
-  // SurfaceEvent's mandatory surfaceOp is too strict here — we need to
-  // CHECK whether surface metadata is present, not assume it.
-  const se = event as SessionEvent<SurfaceEventType>
-  if (!SURFACE_TYPES.has(event.type)) {
-    if (se.sourceEventSeqs !== undefined) {
-      throw new InvariantError(`${event.type} cannot carry sourceEventSeqs (non-surface event)`)
-    }
-    if (se.surfaceOp !== undefined) {
-      throw new InvariantError(`${event.type} cannot carry surfaceOp (non-surface event)`)
-    }
-  }
-  if (se.sourceEventSeqs !== undefined) {
-    if (se.sourceEventSeqs.length === 0) {
-      throw new InvariantError('sourceEventSeqs must not be empty when present')
-    }
-    const unique = new Set(se.sourceEventSeqs)
-    if (unique.size !== se.sourceEventSeqs.length) {
-      throw new InvariantError('sourceEventSeqs must not contain duplicates')
-    }
-    for (const ref of se.sourceEventSeqs) {
-      if (ref >= event.seq) {
-        throw new InvariantError(`sourceEventSeqs must reference earlier events: ${ref} >= current seq ${event.seq}`)
-      }
-      if (!trace.knownSeqs.has(ref)) {
-        throw new InvariantError(`sourceEventSeqs references unknown seq ${ref}`)
-      }
-    }
-  }
-  // Fold this event into the tracked surface linked list, validating the
-  // replace contract as we go. `append` adds a tail node; `replace` shadows a
-  // positional range — every shadowed node must appear in sourceEventSeqs.
-  if (se.surfaceOp !== undefined) {
-    if (se.surfaceOp === 'append') {
-      surface = { kind: 'append' }
-    } else {
-      const { start, end } = se.surfaceOp
-      const startIdx = trace.surface.indexOf(start)
-      if (startIdx === -1) {
-        throw new InvariantError(`surface replace: start seq ${start} is not on the surface`)
-      }
-      const endIdx = trace.surface.indexOf(end)
-      if (endIdx === -1) {
-        throw new InvariantError(`surface replace: end seq ${end} is not on the surface`)
-      }
-      if (startIdx > endIdx) {
-        throw new InvariantError(`surface replace: start seq ${start} (pos ${startIdx}) is after end seq ${end} (pos ${endIdx}) on the surface`)
-      }
-      // Every node the replace shadows (surface positions [startIdx, endIdx]
-      // inclusive) must appear in sourceEventSeqs — the provenance contract.
-      const shadowed = trace.surface.slice(startIdx, endIdx + 1)
-      const recorded = new Set(se.sourceEventSeqs ?? [])
-      const missing = shadowed.filter(seq => !recorded.has(seq))
-      if (missing.length > 0) {
-        throw new InvariantError(`surface replace: sourceEventSeqs must include every shadowed surface node; missing ${missing.join(', ')}`)
-      }
-      surface = { kind: 'replace', start: startIdx, count: shadowed.length }
-    }
-  }
 
   // Boundary/step-scoped events have explicit cases; every OTHER event type —
   // including plugin-added (merge-extensible) SessionEventMap keys — is caught
@@ -263,8 +182,6 @@ function validateEvent(trace: SessionTrace, event: SessionEvent): SessionTraceTr
   return {
     scalars: { lastSeq: event.seq, openTurn, openStep, nextTurn, nextStep },
     pendingCalls,
-    surface,
-    seq: event.seq,
   }
 }
 
@@ -287,20 +204,6 @@ function applyTransition(trace: SessionTrace, transition: SessionTraceTransition
     default:
       assertNever(transition.pendingCalls, 'session trace pending-call transition')
   }
-  switch (transition.surface.kind) {
-    case 'none':
-      break
-    case 'append':
-      trace.surface.push(transition.seq)
-      break
-    case 'replace':
-      trace.surface.splice(transition.surface.start, transition.surface.count, transition.seq)
-      break
-    /* v8 ignore next -- validateEvent produces this closed transition union */
-    default:
-      assertNever(transition.surface, 'session trace surface transition')
-  }
-  trace.knownSeqs.add(transition.seq)
 }
 
 /** Validate and apply one event while rebuilding an already-committed log. */
@@ -345,8 +248,6 @@ export function apply(ctx: Context): void {
     nextTurn: 1,
     nextStep: 1,
     pendingCalls: new Set(),
-    knownSeqs: new Set(),
-    surface: [],
   })
 
   /** Build (or rebuild) a session's trace by replaying its whole log. */
@@ -445,7 +346,7 @@ export function apply(ctx: Context): void {
   //   the boundary (an `agent/request`-window inject) is legitimately absent
   //   from this request, and a current-surface comparison would false-fire.
   // - header: every non-content field must equal the fold of the log's
-  //   `request/header*` events — the loop logs the header event BEFORE
+  //   `request/header` events — the loop logs the header event BEFORE
   //   dispatch, so the fold already covers this request.
   //
   // Registered with `prepend: true` so a short-circuiting llm/stream listener
diff --git a/packages/support/invariants/tests/invariants.spec.ts b/packages/support/invariants/tests/invariants.spec.ts
index c451440e8e..d5bd2707ab 100644
--- a/packages/support/invariants/tests/invariants.spec.ts
+++ b/packages/support/invariants/tests/invariants.spec.ts
@@ -48,7 +48,7 @@ describe('session-log invariants', () => {
       session.append('user/message', { content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
       session.append('step/start', { turn: 1, step: 1 })
       session.append('assistant/chunk', { turn: 1, step: 1, chunk: { type: 'text-delta', index: 0, text: 'h' } })
-      session.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'tool-call', id: CallId('c1'), name: 'echo', arguments: '{}' }] }, { surfaceOp: 'append' })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'tool-call', id: CallId('c1'), name: 'echo', arguments: '{}' }] }, { surfaceOp: 'append' })
       session.append('tool/call', { turn: 1, step: 1, callId: CallId('c1'), name: 'echo', arguments: '{}' })
       session.append('tool/result', { turn: 1, step: 1, callId: CallId('c1'), content: [{ type: 'text', text: 'ok' }], isError: false }, { surfaceOp: 'append' })
       session.append('step/end', { turn: 1, step: 1 })
@@ -195,7 +195,7 @@ describe('session-log invariants', () => {
     expect(() => {
       session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
       session.append('step/start', { turn: 1, step: 1 })
-      session.append('assistant/message', { turn: 1, step: 1, content: [
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [
         { type: 'tool-call', id: CallId('crashed'), name: 'bash', arguments: '{}' },
       ] }, { surfaceOp: 'append' })
       session.append('tool/result', {
@@ -251,10 +251,10 @@ describe('session-log invariants', () => {
     expect(() => {
       session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
       session.append('step/start', { turn: 1, step: 1 })
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append' })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: 'append' })
       session.append('step/end', { turn: 1, step: 1 })
       session.append('step/start', { turn: 1, step: 2 })
-      session.append('assistant/message', { turn: 1, step: 2, content: [] }, { surfaceOp: 'append' })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 2, content: [] }, { surfaceOp: 'append' })
       session.append('step/end', { turn: 1, step: 2 })
       session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
       session.append('turn/start', { turn: 2, trigger: { kind: 'message', source: { kind: 'user' } } })
@@ -316,7 +316,7 @@ describe('session-log invariants', () => {
     const session = ctx.sessions.create()
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
     session.append('step/start', { turn: 1, step: 1 })
-    expect(() => session.append('assistant/message', { turn: 1, step: 2, content: [] }, { surfaceOp: 'append' }))
+    expect(() => session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 2, content: [] }, { surfaceOp: 'append' }))
       .toThrow(/open is turn 1\/step 1/)
   })
 })
@@ -464,7 +464,7 @@ describe('HMR safety', () => {
   })
 })
 
-describe('surface invariants', () => {
+describe('surface contract under the invariants composition', () => {
   it('accepts well-formed surface metadata', async () => {
     const { ctx } = await setup()
     const session = ctx.sessions.create()
@@ -473,7 +473,7 @@ describe('surface invariants', () => {
     session.append('step/start', { turn: 1, step: 1 })
     expect(() => {
       session.append('user/message', { content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [1] })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [1] })
     }).not.toThrow()
   })
 
@@ -483,17 +483,21 @@ describe('surface invariants', () => {
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
     session.append('step/start', { turn: 1, step: 1 })
     session.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
-    session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 2, end: 2 }, sourceEventSeqs: [2] })
+    session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 2, end: 2 }, sourceEventSeqs: [2] })
     // no throw — well-formed replace op
   })
 
-  it('rejects empty sourceEventSeqs', async () => {
+  it('accepts known-empty assistant provenance and rejects empty provenance elsewhere', async () => {
     const { ctx } = await setup()
     const session = ctx.sessions.create()
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
+    session.append('step/start', { turn: 1, step: 1 })
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [] })
-    }).toThrow(InvariantError)
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [] })
+    }).not.toThrow()
+    expect(() => {
+      session.append('user/message', { content: [], source: { kind: 'user' } }, { surfaceOp: 'append', sourceEventSeqs: [] })
+    }).toThrow(/must not be empty except on assistant\/message/)
   })
 
   it('rejects duplicate sourceEventSeqs', async () => {
@@ -502,7 +506,7 @@ describe('surface invariants', () => {
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
     session.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [1, 1] })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [1, 1] })
     }).toThrow(/must not contain duplicates/)
   })
 
@@ -513,19 +517,20 @@ describe('surface invariants', () => {
     // The next event is seq 1. Referencing its own seq fails on "must reference
     // earlier events" (the check order is: earlier first, then unknown).
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [1] })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [1] })
     }).toThrow(/must reference earlier/)
   })
 
   it('accepts sourceEventSeqs referencing a valid earlier event', async () => {
-    // Positive test: ref < current seq and ref is in knownSeqs → passes.
+    // Session seqs are contiguous, so every non-negative ref below the current
+    // seq necessarily names an existing earlier event.
     const { ctx } = await setup()
     const session = ctx.sessions.create()
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
     session.append('step/start', { turn: 1, step: 1 })
     // seqs so far: 0, 1. The next event at seq 2 references seq 1 → valid.
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [1] })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [1] })
     }).not.toThrow()
   })
 
@@ -534,27 +539,10 @@ describe('surface invariants', () => {
     const session = ctx.sessions.create()
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [99] })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [99] })
     }).toThrow(/must reference earlier/)
   })
 
-  it('rejects sourceEventSeqs referencing unknown seq (gap in event log)', async () => {
-    // Create an impossible-through-public-API gap so seq 2 is earlier but unknown.
-    const { ctx } = await setup()
-    const session = ctx.sessions.create()
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    session.append('step/start', { turn: 1, step: 1 })
-    ;(session as unknown as { log: unknown[] }).log.push({
-      type: 'assistant/chunk',
-      seq: 3,
-      time: Date.now(),
-      data: { turn: 1, step: 1, chunk: { type: 'text-delta', index: 0, text: 'x' } },
-    })
-    expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: 'append', sourceEventSeqs: [2] })
-    }).toThrow(/unknown seq 2/)
-  })
-
   it('rejects a replace whose start is positioned after its end on the surface', async () => {
     const { ctx } = await setup()
     const session = ctx.sessions.create()
@@ -564,8 +552,8 @@ describe('surface invariants', () => {
     session.append('user/message', { content: [{ type: 'text', text: 'b' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 3
     // Reversed range: start seq 3 is at a later surface position than end seq 2.
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 3, end: 2 }, sourceEventSeqs: [2, 3] })
-    }).toThrow(/is after end seq 2 .* on the surface/)
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 3, end: 2 }, sourceEventSeqs: [2, 3] })
+    }).toThrow(/is after end seq 2/)
   })
 
   it('rejects a replace whose sourceEventSeqs omits a shadowed surface node', async () => {
@@ -577,7 +565,7 @@ describe('surface invariants', () => {
     session.append('user/message', { content: [{ type: 'text', text: 'b' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 3
     // Replace shadows surface nodes [2, 3] but records provenance for only [2].
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'sum' }] }, { surfaceOp: { op: 'replace', start: 2, end: 3 }, sourceEventSeqs: [2] })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'sum' }] }, { surfaceOp: { op: 'replace', start: 2, end: 3 }, sourceEventSeqs: [2] })
     }).toThrow(/must include every shadowed surface node; missing 3/)
   })
 
@@ -589,7 +577,7 @@ describe('surface invariants', () => {
     session.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 2
     session.append('user/message', { content: [{ type: 'text', text: 'b' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 3
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 'sum' }] }, { surfaceOp: { op: 'replace', start: 2, end: 3 }, sourceEventSeqs: [2, 3] })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 'sum' }] }, { surfaceOp: { op: 'replace', start: 2, end: 3 }, sourceEventSeqs: [2, 3] })
     }).not.toThrow()
   })
 
@@ -601,8 +589,8 @@ describe('surface invariants', () => {
     session.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 2
     // seq 1 (step/start) is a real earlier event but never entered the surface.
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 1, end: 2 }, sourceEventSeqs: [1, 2] })
-    }).toThrow(/start seq 1 is not on the surface/)
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 1, end: 2 }, sourceEventSeqs: [1, 2] })
+    }).toThrow(/start seq 1 not found in surface/)
   })
 
   it('rejects a replace naming an end seq that is not on the surface', async () => {
@@ -613,8 +601,8 @@ describe('surface invariants', () => {
     session.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 2
     // start (2) is on the surface but end (99) never entered it.
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 2, end: 99 }, sourceEventSeqs: [2] })
-    }).toThrow(/end seq 99 is not on the surface/)
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 2, end: 99 }, sourceEventSeqs: [2] })
+    }).toThrow(/end seq 99 not found in surface/)
   })
 
   it('rejects a replace whose range is reversed in surface position after a prior replace reordered it', async () => {
@@ -625,13 +613,13 @@ describe('surface invariants', () => {
     session.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 2
     session.append('user/message', { content: [{ type: 'text', text: 'b' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 3
     // Replace node 2 (position 0) with seq 4 — surface is now [4, 3], so seq 4
-    // precedes seq 3 in linked-list order even though 4 > 3 numerically.
-    session.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 's' }] }, { surfaceOp: { op: 'replace', start: 2, end: 2 }, sourceEventSeqs: [2] }) // seq 4
+    // precedes seq 3 in surface order even though 4 > 3 numerically.
+    session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 's' }] }, { surfaceOp: { op: 'replace', start: 2, end: 2 }, sourceEventSeqs: [2] }) // seq 4
     // A replace with start=3, end=4 passes the seq check (3 <= 4) but is
     // reversed positionally (3 is at pos 1, 4 is at pos 0).
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 3, end: 4 }, sourceEventSeqs: [3, 4] }) // seq 5
-    }).toThrow(/is after end seq 4 .* on the surface/)
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 3, end: 4 }, sourceEventSeqs: [3, 4] }) // seq 5
+    }).toThrow(/is after end seq 4/)
   })
 
   it('accepts a replace whose start seq exceeds its end seq when the surface position order is valid', async () => {
@@ -645,9 +633,9 @@ describe('surface invariants', () => {
     // head seq (4) is numerically GREATER than the tail seq (3): the surface is
     // not seq-ordered. A replace spanning start=4 (pos 0) … end=3 (pos 1) is
     // valid positionally and must be accepted even though start seq > end seq.
-    session.append('assistant/message', { turn: 1, step: 1, content: [{ type: 'text', text: 's' }] }, { surfaceOp: { op: 'replace', start: 2, end: 2 }, sourceEventSeqs: [2] }) // seq 4
+    session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [{ type: 'text', text: 's' }] }, { surfaceOp: { op: 'replace', start: 2, end: 2 }, sourceEventSeqs: [2] }) // seq 4
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 4, end: 3 }, sourceEventSeqs: [4, 3] }) // seq 5
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 4, end: 3 }, sourceEventSeqs: [4, 3] }) // seq 5
     }).not.toThrow()
   })
 
@@ -659,7 +647,7 @@ describe('surface invariants', () => {
     session.append('user/message', { content: [{ type: 'text', text: 'a' }], source: { kind: 'user' } }, { surfaceOp: 'append' }) // seq 2
     // A replace with no sourceEventSeqs records no provenance for the node it shadows.
     expect(() => {
-      session.append('assistant/message', { turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 2, end: 2 } })
+      session.append('assistant/message', { provenance: { provider: 'mock', model: 'mock' }, turn: 1, step: 1, content: [] }, { surfaceOp: { op: 'replace', start: 2, end: 2 } })
     }).toThrow(/must include every shadowed surface node; missing 2/)
   })
 
@@ -670,30 +658,11 @@ describe('surface invariants', () => {
       { type: 'step/start' as const, seq: 1, time: 0, data: { turn: 1, step: 1 } },
       { type: 'user/message' as const, seq: 2, time: 0, data: { content: [{ type: 'text' as const, text: 'a' }], source: { kind: 'user' as const } }, surfaceOp: 'append' as const },
       { type: 'user/message' as const, seq: 3, time: 0, data: { content: [{ type: 'text' as const, text: 'b' }], source: { kind: 'user' as const } }, surfaceOp: 'append' as const },
-      { type: 'assistant/message' as const, seq: 4, time: 0, data: { turn: 1, step: 1, content: [{ type: 'text' as const, text: 'sum' }] }, surfaceOp: { op: 'replace' as const, start: 2, end: 3 }, sourceEventSeqs: [2] },
+      { type: 'assistant/message' as const, seq: 4, time: 0, data: { turn: 1, step: 1, content: [{ type: 'text' as const, text: 'sum' }], provenance: { provider: 'mock', model: 'mock' } }, surfaceOp: { op: 'replace' as const, start: 2, end: 3 }, sourceEventSeqs: [2] },
     ]
     expect(() => ctx.sessions.create(undefined, { seed: badSeed })).toThrow(/must include every shadowed surface node; missing 3/)
   })
 
-  it('rejects sourceEventSeqs on a non-surface event', async () => {
-    const { ctx } = await setup()
-    const session = ctx.sessions.create()
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    // Session rejects this at its own acceptance boundary. Emit a hand-built
-    // record to cover the listener's defensive check for alternate producers.
-    const event = { type: 'turn/end', seq: 1, time: 1, data: { turn: 1, reason: { kind: 'completed' } }, sourceEventSeqs: [0] }
-    expect(() => { ctx.emit(scopeTarget(session, undefined), 'session/event', session, event as never) })
-      .toThrow(/cannot carry sourceEventSeqs/)
-  })
-
-  it('rejects surfaceOp on a non-surface event', async () => {
-    const { ctx } = await setup()
-    const session = ctx.sessions.create()
-    session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
-    const event = { type: 'turn/end', seq: 1, time: 1, data: { turn: 1, reason: { kind: 'completed' } }, surfaceOp: 'append' }
-    expect(() => { ctx.emit(scopeTarget(session, undefined), 'session/event', session, event as never) })
-      .toThrow(/cannot carry surfaceOp/)
-  })
 })
 
 describe('request-reconstruction cross-check (llm/stream)', () => {
@@ -705,7 +674,7 @@ describe('request-reconstruction cross-check (llm/stream)', () => {
     session.append('user/message', { content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
     const boundary = session.deriveMessages()
     session.append('step/start', { turn: 1, step: 1 })
-    session.append('request/header', { header: { config: { model: 'm' } }, reason: 'initial' })
+    session.append('request/header', { header: { config: { provider: 'mock', model: 'm' } }, reason: 'initial' })
     return { ctx, session, boundary }
   }
 
@@ -735,7 +704,7 @@ describe('request-reconstruction cross-check (llm/stream)', () => {
   it('expects the folded header\'s session prefix ahead of the derivation (prefix + derived)', async () => {
     const { ctx, session, boundary } = await requestSetup()
     const prefix = { role: 'user' as const, content: [{ type: 'text' as const, text: '<system-reminder>catalog</system-reminder>' }] }
-    session.append('request/header-delta', { messagePrefix: [prefix] })
+    session.append('request/header', { header: { config: { provider: 'mock', model: 'm' }, messagePrefix: [prefix] }, reason: 'change' })
     // The prefixed request matches the fold…
     const prefixed = Object.freeze({ model: 'm', messages: Object.freeze([prefix, ...boundary]), sessionId: session.id })
     expect(() => { dispatch(ctx, prefixed) }).not.toThrow()
@@ -804,7 +773,7 @@ describe('request cross-check ordering (prepend)', () => {
     session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
     session.append('user/message', { content: [{ type: 'text', text: 'hi' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
     session.append('step/start', { turn: 1, step: 1 })
-    session.append('request/header', { header: { config: { model: 'm' } }, reason: 'initial' })
+    session.append('request/header', { header: { config: { provider: 'mock', model: 'm' } }, reason: 'initial' })
 
     const divergent = Object.freeze({
       model: 'm',
diff --git a/packages/support/llm-replay/README.md b/packages/support/llm-replay/README.md
index 847064b160..3272f29b9a 100644
--- a/packages/support/llm-replay/README.md
+++ b/packages/support/llm-replay/README.md
@@ -1,6 +1,6 @@
 # @deepseek-ai/dsh-llm-replay
 
-A replay LLM plugin for keyless snapshot tests. It installs a single `llm/stream` waterfall listener that short-circuits the waterfall (never calls `next()`) and yields model streams reconstructed from a recorded **session JSONL** fixture — so a test can boot the real agent against a fixed model transcript with no API key.
+A replay LLM plugin for keyless snapshot tests. It yields model streams reconstructed from a recorded **session JSONL** fixture, so a test can boot the real agent against a fixed model transcript with no API key. With `providers` configured it registers a replay-only adapter whose catalog is visible to clients such as ACP editors; without `providers` it installs the catch-all `llm/stream` waterfall used by tests that do not need discovery.
 
 Its consumer is the ACP snapshot harness in `examples/acp-agent`, which loads this plugin (via `cordis.snapshot.yml`) in place of a real LLM adapter. The package exists so its derive/parse/replay logic falls under the per-file 100% coverage gate on `packages/*/src` (the same logic, while it lived under `examples/`, was outside the gate).
 
@@ -23,10 +23,18 @@ Replay keys every call by its calling session id (`GenerateOptions.sessionId`, s
 | `file` | string | `$DSH_SNAPSHOT_FILE` | Path to the primary (parent) `session.jsonl` fixture. Required (config or env). |
 | `overrideFile` | string | `$DSH_SNAPSHOT_OVERRIDE` | Optional path to a `ReplayEntry[]` sidecar that replaces the PRIMARY session's derived script. |
 | `childFiles` | string[] | `$DSH_SNAPSHOT_CHILD_FILES` (path-delimited) | Recorded subagent child-session logs for a nested scenario; empty for a single-session scenario. |
+| `providers` | `ReplayProviderConfig[]` | — | Optional replay-only provider and model catalog. Configured routes dispatch through the replay adapter and never perform provider I/O. |
 
 ```yaml
 - id: llm-replay
   name: '@deepseek-ai/dsh-llm-replay'
+  config:
+    providers:
+      - id: deepseek
+        name: DeepSeek
+        models:
+          - id: deepseek-v4-flash
+          - id: deepseek-v4-pro
   # file/overrideFile/childFiles default to $DSH_SNAPSHOT_FILE /
   # $DSH_SNAPSHOT_OVERRIDE / $DSH_SNAPSHOT_CHILD_FILES, set by the snapshot
   # harness per scenario.
@@ -34,11 +42,11 @@ Replay keys every call by its calling session id (`GenerateOptions.sessionId`, s
 
 ## Exports
 
-- `installLlmReplay(ctx, config)` — install the `llm/stream` listener; returns the disposer (HMR safety). Use this in tests to drive replay without the Loader or env vars.
+- `installLlmReplay(ctx, config)` — install the configured replay adapter or catch-all `llm/stream` listener; returns the disposer (HMR safety). Use this in tests to drive replay without the Loader or env vars.
 - `loadSessionScripts(config)` — resolve the ordered `SessionScript[]` (primary + children) for a scenario, ready to bind to live sessions in first-call order.
 - `loadReplayScript(config)` — resolve the `ReplayEntry[]` for the PRIMARY session only (sidecar override if present, else derived from the JSONL; fail-loud if the fixture is missing).
 - `deriveReplayScript(events)` / `parseSessionLog(text)` / `parseSessionHeader(text)` — the pure helpers that turn a recorded session log into a script and read its header `id`/`createdAt`. A derived group must end in a `finish` chunk; a group without one is the fingerprint of a thrown `stream()` and must instead be expressed via an override sidecar.
-- Types `ReplayEntry` / `SessionScript` / `ReplayConfig` / `Config`.
+- Types `ReplayEntry` / `SessionScript` / `ReplayConfig` / `ReplayProviderConfig` / `ReplayModelConfig` / `Config`.
 
 ## Plugin export shape
 
diff --git a/packages/support/llm-replay/src/index.ts b/packages/support/llm-replay/src/index.ts
index 2e509973e5..51e799114f 100644
--- a/packages/support/llm-replay/src/index.ts
+++ b/packages/support/llm-replay/src/index.ts
@@ -10,8 +10,8 @@ import { existsSync, readFileSync } from 'node:fs'
 import { delimiter as pathDelimiter } from 'node:path'
 import type { Context } from 'cordis'
 import type { SessionEvent } from '@deepseek-ai/dsh-session'
-import type { GenerateOptions, StreamChunk } from '@deepseek-ai/dsh-llm'
-import { LlmError, assertNever } from '@deepseek-ai/dsh-llm'
+import type { GenerateOptions, LlmModelInfo, LlmProviderInfo, StreamChunk } from '@deepseek-ai/dsh-llm'
+import { LlmAdapter, LlmError, assertNever } from '@deepseek-ai/dsh-llm'
 
 /**
  * One recorded model call. `throw` may replay prefix chunks before failing;
@@ -23,6 +23,26 @@ export type ReplayEntry =
   | { kind: 'throw'; chunks: StreamChunk[]; message: string; code: string; status?: number }
   | { kind: 'hang' }
 
+/** One model exposed by a replay-only provider catalog. */
+export interface ReplayModelConfig {
+  /** Model id used for replay requests. */
+  id: string
+  /** Selector label; defaults to {@link id}. */
+  name?: string
+  /** Optional selector description. */
+  description?: string
+}
+
+/** One provider route exposed by the replay adapter. */
+export interface ReplayProviderConfig {
+  /** Provider route used for replay requests. */
+  id: string
+  /** Selector label; defaults to {@link id}. */
+  name?: string
+  /** Advisory models exposed to clients such as ACP editors. */
+  models?: ReplayModelConfig[]
+}
+
 /** Resolved plugin configuration. */
 export interface ReplayConfig {
   /**
@@ -45,6 +65,12 @@ export interface ReplayConfig {
    * for a single-session scenario.
    */
   childFiles?: string[]
+  /**
+   * Optional provider catalog. When non-empty, replay registers an adapter for
+   * these routes; when absent or empty, it retains the catch-all waterfall used
+   * by tests that do not need discovery.
+   */
+  providers?: ReplayProviderConfig[]
 }
 
 /**
@@ -203,6 +229,42 @@ export function loadSessionScripts(config: ReplayConfig): SessionScript[] {
   return [primary, ...children]
 }
 
+/** Replay adapter that makes a configured provider catalog discoverable without provider I/O. */
+class ReplayAdapter extends LlmAdapter {
+  private readonly providers: ReadonlyMap<string, ReplayProviderConfig>
+
+  constructor(
+    providers: readonly ReplayProviderConfig[],
+    private readonly replay: (options: GenerateOptions) => AsyncIterable<StreamChunk>,
+  ) {
+    super()
+    this.providers = new Map(providers.map(provider => [provider.id, provider]))
+  }
+
+  override providerInfo(provider: string): LlmProviderInfo {
+    const configured = this.providers.get(provider)
+    /* v8 ignore next -- LlmService only asks about routes registered from this same map. */
+    if (configured === undefined) return super.providerInfo(provider)
+    return { id: provider, name: configured.name ?? provider }
+  }
+
+  override listModels(provider: string): Promise<readonly LlmModelInfo[]> {
+    const configured = this.providers.get(provider)
+    /* v8 ignore next -- LlmService only asks about routes registered from this same map. */
+    if (configured === undefined) return Promise.resolve([])
+    return Promise.resolve((configured.models ?? []).map(model => ({
+      provider,
+      id: model.id,
+      name: model.name ?? model.id,
+      ...model.description === undefined ? {} : { description: model.description },
+    })))
+  }
+
+  override stream(options: GenerateOptions): AsyncIterable<StreamChunk> {
+    return this.replay(options)
+  }
+}
+
 /** Yield a recorded stream back, honoring abort like a real adapter. */
 async function* replayEntry(entry: ReplayEntry, signal: AbortSignal | undefined): AsyncIterable<StreamChunk> {
   switch (entry.kind) {
@@ -243,12 +305,14 @@ async function* replayEntry(entry: ReplayEntry, signal: AbortSignal | undefined)
 /**
  * Install per-session positional replay. A newly seen live session takes the
  * next ordered recorded script, then advances its own cursor synchronously at
- * invocation time; calls without `sessionId` share one anonymous session.
- * Returns the effect disposer for HMR-safe removal.
+ * invocation time; calls without `sessionId` share one anonymous session. A
+ * non-empty provider catalog registers a routed replay adapter; otherwise a
+ * catch-all waterfall intercepts requests. Returns the effect disposer for
+ * HMR-safe removal.
  *
- * @param ctx - the context whose `llm/stream` waterfall the listener short-circuits.
+ * @param ctx - the context whose LLM service receives the replay route or waterfall.
  * @param config - the resolved fixture paths (env-var defaulting is `apply`'s job).
- * @returns the `ctx.on` disposer that removes the listener.
+ * @returns the disposer that removes the registered adapter or listener.
  */
 export function installLlmReplay(ctx: Context, config: ReplayConfig): () => void {
   const scripts = loadSessionScripts(config)
@@ -258,7 +322,7 @@ export function installLlmReplay(ctx: Context, config: ReplayConfig): () => void
   const bound = new Map<string, { entries: ReplayEntry[]; cursor: number }>()
   let nextScript = 0
   const ANON = '\0anon\0' // the key for a call that carries no sessionId
-  return ctx.on('llm/stream', (options: GenerateOptions, _next) => {
+  const replay = (options: GenerateOptions): AsyncIterable<StreamChunk> => {
     const key = options.sessionId ?? ANON
     let state = bound.get(key)
     let unrecorded = false
@@ -296,7 +360,12 @@ export function installLlmReplay(ctx: Context, config: ReplayConfig): () => void
       }
       yield* replayEntry(entry, options.signal)
     })()
-  })
+  }
+  const providers = config.providers ?? []
+  if (providers.length > 0) {
+    return ctx.llm.registerAdapter(providers.map(provider => provider.id), new ReplayAdapter(providers, replay))
+  }
+  return ctx.on('llm/stream', (options: GenerateOptions, _next) => replay(options))
 }
 
 export const name = 'llm-replay'
@@ -314,6 +383,8 @@ export interface Config {
    * a nested-agent scenario; absent/empty for a single-session scenario.
    */
   childFiles?: string[]
+  /** Optional replay-only provider catalog; absent or empty selects catch-all waterfall replay. */
+  providers?: ReplayProviderConfig[]
 }
 
 export function apply(ctx: Context, config: Config = {}): void {
@@ -329,5 +400,6 @@ export function apply(ctx: Context, config: Config = {}): void {
     file,
     ...overrideFile !== undefined && overrideFile.length > 0 ? { overrideFile } : {},
     ...childFiles.length > 0 ? { childFiles } : {},
+    ...config.providers !== undefined ? { providers: config.providers } : {},
   })
 }
diff --git a/packages/support/llm-replay/tests/llm-replay.spec.ts b/packages/support/llm-replay/tests/llm-replay.spec.ts
index ac52ec11c9..c9bd545b36 100644
--- a/packages/support/llm-replay/tests/llm-replay.spec.ts
+++ b/packages/support/llm-replay/tests/llm-replay.spec.ts
@@ -198,7 +198,7 @@ describe('loadReplayScript', () => {
   })
 })
 
-describe('installLlmReplay (through the real waterfall)', () => {
+describe('installLlmReplay (through the real LlmService)', () => {
   function writeLog(...calls: StreamChunk[][]): void {
     let seq = 1
     const events: SessionEvent[] = []
@@ -214,7 +214,41 @@ describe('installLlmReplay (through the real waterfall)', () => {
     await ctx.plugin(LlmService)
     // No adapter registered for 'm' — replay must not reach it.
     installLlmReplay(ctx, { file })
-    expect(await drain(ctx.llm.stream({ model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
+    expect(await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
+  })
+
+  it('registers a replay-only provider catalog when configured', async () => {
+    writeLog(TEXT_CHUNKS)
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    const dispose = installLlmReplay(ctx, {
+      file,
+      providers: [
+        {
+          id: 'deepseek',
+          name: 'DeepSeek',
+          models: [
+            { id: 'flash' },
+            { id: 'pro', name: 'Pro', description: 'Larger model' },
+          ],
+        },
+        { id: 'empty' },
+      ],
+    })
+
+    expect(ctx.llm.listProviders()).toEqual([
+      { id: 'deepseek', name: 'DeepSeek' },
+      { id: 'empty', name: 'empty' },
+    ])
+    await expect(ctx.llm.listModels('deepseek')).resolves.toEqual([
+      { provider: 'deepseek', id: 'flash', name: 'flash' },
+      { provider: 'deepseek', id: 'pro', name: 'Pro', description: 'Larger model' },
+    ])
+    await expect(ctx.llm.listModels('empty')).resolves.toEqual([])
+    expect(await drain(ctx.llm.stream({ provider: 'deepseek', model: 'pro', messages: [] }))).toEqual(TEXT_CHUNKS)
+
+    dispose()
+    expect(ctx.llm.listProviders()).toEqual([])
   })
 
   it('serves the Nth call the Nth derived entry (positional)', async () => {
@@ -227,8 +261,8 @@ describe('installLlmReplay (through the real waterfall)', () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
     installLlmReplay(ctx, { file })
-    expect(await drain(ctx.llm.stream({ model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
-    expect(await drain(ctx.llm.stream({ model: 'm', messages: [] }))).toEqual(second)
+    expect(await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
+    expect(await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))).toEqual(second)
   })
 
   it('replays a sidecar throw-entry as an LlmError with code/status, after its prefix chunks', async () => {
@@ -244,7 +278,7 @@ describe('installLlmReplay (through the real waterfall)', () => {
 
     const seen: StreamChunk[] = []
     await expect((async () => {
-      for await (const c of ctx.llm.stream({ model: 'm', messages: [] })) seen.push(c)
+      for await (const c of ctx.llm.stream({ provider: 'm', model: 'm', messages: [] })) seen.push(c)
     })()).rejects.toMatchObject({ message: 'unauthorized', code: 'AUTH', status: 401 })
     expect(seen).toEqual(partial)
   })
@@ -258,7 +292,7 @@ describe('installLlmReplay (through the real waterfall)', () => {
     installLlmReplay(ctx, { file, overrideFile })
 
     const controller = new AbortController()
-    const iterator = ctx.llm.stream({ model: 'm', messages: [], signal: controller.signal })[Symbol.asyncIterator]()
+    const iterator = ctx.llm.stream({ provider: 'm', model: 'm', messages: [], signal: controller.signal })[Symbol.asyncIterator]()
     // Deterministically consume the two pre-hang chunks (no sleep), then abort
     // and assert the next pull rejects — event-driven, per the no-sleeps rule.
     expect((await iterator.next()).value).toMatchObject({ type: 'block-start' })
@@ -272,8 +306,8 @@ describe('installLlmReplay (through the real waterfall)', () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
     installLlmReplay(ctx, { file })
-    await drain(ctx.llm.stream({ model: 'm', messages: [] }))
-    await expect(drain(ctx.llm.stream({ model: 'm', messages: [] }))).rejects.toThrow(/exhausted/)
+    await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))
+    await expect(drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))).rejects.toThrow(/exhausted/)
   })
 
   it('aborts mid-replay when the signal is already set', async () => {
@@ -283,7 +317,7 @@ describe('installLlmReplay (through the real waterfall)', () => {
     installLlmReplay(ctx, { file })
     const controller = new AbortController()
     controller.abort()
-    await expect(drain(ctx.llm.stream({ model: 'm', messages: [], signal: controller.signal })))
+    await expect(drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [], signal: controller.signal })))
       .rejects.toThrow('aborted')
   })
 
@@ -305,11 +339,11 @@ describe('installLlmReplay (through the real waterfall)', () => {
     }, { inject: ['llm'] }))
 
     // While installed, replay short-circuits to the derived fixture ('hi').
-    expect(await drain(ctx.llm.stream({ model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
+    expect(await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
 
     await fiber.dispose()
     // After dispose the listener is gone; the call reaches the real adapter.
-    expect(await drain(ctx.llm.stream({ model: 'm', messages: [] })))
+    expect(await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] })))
       .toEqual([{ type: 'finish', reason: { kind: 'stop' } }])
   })
 
@@ -321,7 +355,7 @@ describe('installLlmReplay (through the real waterfall)', () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
     installLlmReplay(ctx, { file, overrideFile })
-    await expect(drain(ctx.llm.stream({ model: 'm', messages: [] })))
+    await expect(drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] })))
       .rejects.toThrow(/llm-replay replay entry/)
   })
 
@@ -333,7 +367,7 @@ describe('installLlmReplay (through the real waterfall)', () => {
     await ctx.plugin(LlmService)
     installLlmReplay(ctx, { file, overrideFile })
     const controller = new AbortController()
-    const iterator = ctx.llm.stream({ model: 'm', messages: [], signal: controller.signal })[Symbol.asyncIterator]()
+    const iterator = ctx.llm.stream({ provider: 'm', model: 'm', messages: [], signal: controller.signal })[Symbol.asyncIterator]()
     // Consume the two pre-hang chunks, then start the third pull so the generator
     // is parked inside the await (signal NOT yet aborted — exercises the
     // addEventListener('abort') registration), and only THEN abort.
@@ -359,7 +393,7 @@ describe('installLlmReplay (through the real waterfall)', () => {
     controller.abort()
     // Already aborted: the throw-entry's prefix loop surfaces 'aborted' before
     // it can reach the recorded LlmError.
-    await expect(drain(ctx.llm.stream({ model: 'm', messages: [], signal: controller.signal })))
+    await expect(drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [], signal: controller.signal })))
       .rejects.toThrow('aborted')
   })
 
@@ -373,7 +407,7 @@ describe('installLlmReplay (through the real waterfall)', () => {
     const controller = new AbortController()
     controller.abort()
     // The two pre-hang chunks still flow; the abort surfaces at the await.
-    const iterator = ctx.llm.stream({ model: 'm', messages: [], signal: controller.signal })[Symbol.asyncIterator]()
+    const iterator = ctx.llm.stream({ provider: 'm', model: 'm', messages: [], signal: controller.signal })[Symbol.asyncIterator]()
     await iterator.next()
     await iterator.next()
     await expect(iterator.next()).rejects.toThrow('aborted')
@@ -503,7 +537,7 @@ describe('installLlmReplay (per-session keying)', () => {
   ]
 
   const live = (id: string): GenerateOptions =>
-    ({ model: 'm', messages: [], sessionId: id as NonNullable<GenerateOptions['sessionId']> })
+    ({ provider: 'm', model: 'm', messages: [], sessionId: id as NonNullable<GenerateOptions['sessionId']> })
 
   it('routes each live session to its own script by FIRST-CALL order', async () => {
     const parentFile = writeSession('session.jsonl', { id: 'rec-parent', createdAt: 100 }, [TEXT_CHUNKS])
@@ -540,7 +574,7 @@ describe('installLlmReplay (per-session keying)', () => {
     await ctx.plugin(LlmService)
     installLlmReplay(ctx, { file: parentFile })
     // No sessionId at all — the legacy single-session path.
-    expect(await drain(ctx.llm.stream({ model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
+    expect(await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
   })
 
   it('fails loud when more distinct live sessions call than were recorded', async () => {
@@ -574,12 +608,13 @@ describe('apply (the plugin entry)', () => {
     expect(inject).toEqual(['llm'])
   })
 
-  it('installs replay from an explicit config.file', async () => {
+  it('installs replay and its catalog from explicit config', async () => {
     writeFileSync(file, sessionJsonl(TEXT_CHUNKS.map((c, i) => chunkEvent(i + 1, 1, 1, c))), 'utf8')
     const ctx = new Context()
     await ctx.plugin(LlmService)
-    apply(ctx, { file })
-    expect(await drain(ctx.llm.stream({ model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
+    apply(ctx, { file, providers: [{ id: 'm', models: [{ id: 'm' }] }] })
+    expect(ctx.llm.listProviders()).toEqual([{ id: 'm', name: 'm' }])
+    expect(await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
   })
 
   it('falls back to $DSH_SNAPSHOT_FILE / $DSH_SNAPSHOT_OVERRIDE when config is empty', async () => {
@@ -591,7 +626,7 @@ describe('apply (the plugin entry)', () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
     apply(ctx)
-    expect(await drain(ctx.llm.stream({ model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
+    expect(await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
   })
 
   it('uses only the file when no override path is configured or in the env', async () => {
@@ -601,7 +636,7 @@ describe('apply (the plugin entry)', () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
     apply(ctx)
-    expect(await drain(ctx.llm.stream({ model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
+    expect(await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
   })
 
   it('throws when no fixture path is given by config or env', async () => {
@@ -631,7 +666,7 @@ describe('apply (the plugin entry)', () => {
     await ctx.plugin(LlmService)
     apply(ctx, { file, childFiles: [childFile] })
     const live = (id: string): GenerateOptions =>
-      ({ model: 'm', messages: [], sessionId: id as NonNullable<GenerateOptions['sessionId']> })
+      ({ provider: 'm', model: 'm', messages: [], sessionId: id as NonNullable<GenerateOptions['sessionId']> })
     expect(await drain(ctx.llm.stream(live('A')))).toEqual(TEXT_CHUNKS)
     expect(await drain(ctx.llm.stream(live('B')))).toEqual(childSecond)
   })
@@ -651,7 +686,7 @@ describe('apply (the plugin entry)', () => {
     await ctx.plugin(LlmService)
     apply(ctx)
     const live = (id: string): GenerateOptions =>
-      ({ model: 'm', messages: [], sessionId: id as NonNullable<GenerateOptions['sessionId']> })
+      ({ provider: 'm', model: 'm', messages: [], sessionId: id as NonNullable<GenerateOptions['sessionId']> })
     expect(await drain(ctx.llm.stream(live('A')))).toEqual(TEXT_CHUNKS)
     expect(await drain(ctx.llm.stream(live('B')))).toEqual(childChunks)
   })
@@ -663,6 +698,6 @@ describe('apply (the plugin entry)', () => {
     const ctx = new Context()
     await ctx.plugin(LlmService)
     apply(ctx)
-    expect(await drain(ctx.llm.stream({ model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
+    expect(await drain(ctx.llm.stream({ provider: 'm', model: 'm', messages: [] }))).toEqual(TEXT_CHUNKS)
   })
 })
diff --git a/packages/support/loader-smoke/README.md b/packages/support/loader-smoke/README.md
index ea197b25d0..04f519b4c1 100644
--- a/packages/support/loader-smoke/README.md
+++ b/packages/support/loader-smoke/README.md
@@ -1,10 +1,10 @@
 # `@deepseek-ai/dsh-loader-smoke`
 
-Shared subprocess harness for keyless example smokes that boot the real stdio-agent bin and a real `cordis.yml` through the Cordis Loader. A test supplies absolute bin/config/tsconfig paths, optional environment overrides, and stdin lines; `runLoaderSmoke` owns the isolated cwd, DSH homes, tsx path resolution, 30-second process deadline, captured diagnostics, forced kill, EOF, and cleanup.
+Shared subprocess harness for tests that boot an app and `cordis.yml` through the Cordis Loader. `resolveExampleLaunch` selects local `src` mode (tsx and root tsconfig paths) or CI `lib` mode (plain Node and package exports) from an explicit mode or `DSH_EXAMPLE_MODE`.
 
-Successful runs return stdout and stderr only after a zero exit. Non-zero exits and deadlines reject with both captured streams. `LOADER_SMOKE_TEST_TIMEOUT_MS` leaves Vitest enough room for the process-owned diagnostic timeout to fire first.
+`runLoaderSmoke` owns the isolated cwd, DSH homes, stdin, diagnostics, deadline, termination, and cleanup. It returns both streams after a zero exit and rejects with both streams on failure.
 
-This is support-tier test infrastructure, not product API. The consumers are the Loader-path smokes under `examples/{echo-agent,coding-agent,cordis-agent}`.
+This is support-tier test infrastructure, not product API.
 
 ## Model Experience
 
@@ -12,6 +12,6 @@ None, as this test-only harness boots example processes and inspects their strea
 
 ## Known Limitations and Deferred Work
 
-- **Only the unbuilt tsx/Loader path is exercised** — built-bin artifacts remain the responsibility of their separate e2e smokes.
+- **Built mode requires a prior build** — the config must also resolve every named package upward through `examples/node_modules`.
 - **Captured stdout and stderr are unbounded** — a runaway child can consume memory until the deadline kills it.
 - **Timeout kills only the direct child** — a process tree spawned by a faulty fixture can outlive the smoke and needs external cleanup.
diff --git a/packages/support/loader-smoke/src/index.ts b/packages/support/loader-smoke/src/index.ts
index 72839c6a05..edbe39d8ca 100644
--- a/packages/support/loader-smoke/src/index.ts
+++ b/packages/support/loader-smoke/src/index.ts
@@ -2,6 +2,14 @@
  * Shared subprocess harness for keyless example smokes that boot a real
  * `cordis.yml` through the stdio-agent bin and Cordis Loader.
  *
+ * It also owns the mode-aware launch resolver every example subprocess harness shares
+ * ({@link resolveExampleLaunch}): booting an example bin from TypeScript source under `tsx` (the
+ * zero-build dev path, resolving `@deepseek-ai/dsh-*` / `@cordisjs/*` through the tsconfig `paths`
+ * map) or from built `lib/` under plain Node (resolving bare packages through real `exports`, as an
+ * installed consumer does, while Node type-strips relative example-local TypeScript plugins).
+ * Consolidating that spawn glue here retires the copies in the ACP snapshot harness and the example
+ * e2e drivers (the `TODO(acp-test-harness)`).
+ *
  * @module @deepseek-ai/dsh-loader-smoke
  */
 
@@ -9,26 +17,125 @@ import { spawn } from 'node:child_process'
 import { mkdtemp, rm } from 'node:fs/promises'
 import { tmpdir } from 'node:os'
 import { join } from 'node:path'
-import { fileURLToPath } from 'node:url'
 
 const DEFAULT_PROCESS_TIMEOUT_MS = 30_000
-const TSX_LOADER = fileURLToPath(import.meta.resolve('tsx'))
 
 /** Vitest deadline that leaves room for the subprocess-owned 30-second diagnostic timeout. */
 export const LOADER_SMOKE_TEST_TIMEOUT_MS = DEFAULT_PROCESS_TIMEOUT_MS + 15_000
 
+/** Which artifact an example bin is booted from: unbuilt `src` via tsx, or built `lib` via plain Node. */
+export type ExampleMode = 'src' | 'lib'
+
+/** Environment variable selecting the mode; CI and pre-push set it to `lib`, dev leaves it unset (`src`). */
+export const EXAMPLE_MODE_ENV = 'DSH_EXAMPLE_MODE'
+
+/**
+ * Parse an {@link ExampleMode} from a raw string, defaulting to `src` when absent so an unset
+ * environment reproduces the dev/tsx behavior. Throws on any other value rather than silently
+ * falling back, so a typo in a gate's env fails loud.
+ * @param raw - the raw value; defaults to `process.env.DSH_EXAMPLE_MODE`.
+ * @returns the validated mode.
+ */
+export function resolveExampleMode(raw: string | undefined = process.env[EXAMPLE_MODE_ENV]): ExampleMode {
+  switch (raw) {
+    case undefined:
+    case '':
+    case 'src':
+      return 'src'
+    case 'lib':
+      return 'lib'
+    default:
+      throw new Error(`${EXAMPLE_MODE_ENV} must be 'src' or 'lib', got ${JSON.stringify(raw)}.`)
+  }
+}
+
+/** Inputs to {@link resolveExampleLaunch}. */
+export interface ExampleLaunchOptions {
+  /** Absolute path to the example bin's TypeScript source entry (`<pkg>/src/bin.ts`); the `lib` bin is derived from it. */
+  readonly srcBin: string
+  /** Explicit plain-Node entry for `lib` mode; test fixtures may point this at Node-type-strippable TypeScript. */
+  readonly libBin?: string | undefined
+  /** Arguments passed after the bin — the config, positional (`[configPath]`) or flagged (`['--config', configPath]`). */
+  readonly configArgs?: readonly string[]
+  /** The mode to launch in; defaults to {@link resolveExampleMode} of the environment. */
+  readonly mode?: ExampleMode
+  /** Absolute repo tsconfig whose `paths` map resolves unbuilt workspace imports. Required in `src` mode, ignored in `lib`. */
+  readonly tsconfigPath?: string
+  /** Prepend `--expose-internals` (the Cordis Loader's bare-plugin resolver needs it for some bins); defaults to `false`. */
+  readonly exposeInternals?: boolean
+  /** Extra environment entries the mode-specific ones layer over; the caller then merges the result over `process.env`. */
+  readonly env?: NodeJS.ProcessEnv
+}
+
+/** The resolved spawn: `spawn(command, args, { env: { ...process.env, ...env } })`. */
+export interface ExampleLaunch {
+  /** The executable to spawn — always the current Node binary. */
+  readonly command: string
+  /** Node flags, the resolved bin, then the caller's `configArgs`. */
+  readonly args: string[]
+  /** Mode-specific environment (`TSX_TSCONFIG_PATH` in `src`, nothing added in `lib`) layered over the caller's `env`. */
+  readonly env: NodeJS.ProcessEnv
+}
+
+/** Derive the built-lib bin (`<pkg>/lib/<name>.js`) from a source bin (`<pkg>/src/<name>.ts`). */
+function toLibBin(srcBin: string): string {
+  const markerLength = '/src/'.length
+  const cut = Math.max(srcBin.lastIndexOf('/src/'), srcBin.lastIndexOf('\\src\\'))
+  if (cut === -1) {
+    throw new Error(`resolveExampleLaunch: expected a "/src/" segment or Windows equivalent in bin path ${JSON.stringify(srcBin)}.`)
+  }
+  const separator = srcBin.slice(cut, cut + 1)
+  const tail = srcBin.slice(cut + markerLength).replace(/\.ts$/, '.js')
+  return `${srcBin.slice(0, cut)}${separator}lib${separator}${tail}`
+}
+
+/**
+ * Resolve how to spawn an example bin in the selected mode.
+ *
+ * `src` yields `node [--expose-internals] --import <tsx> <srcBin> <configArgs>` with `TSX_TSCONFIG_PATH`
+ * set so the tsconfig `paths` map resolves workspace imports to source. `lib` yields
+ * `node [--expose-internals] <libBin> <configArgs>` under plain Node with no tsx and no paths map, so
+ * bare package plugins resolve through real package `exports` into built `lib/`; relative example-local
+ * TypeScript plugins remain source files loaded through Node's built-in type stripping. Bare resolution
+ * requires the config to live below a workspace that declares its `cordis.yml` package dependencies.
+ *
+ * @param options - the source bin, config arguments, mode, and environment.
+ * @returns the command, argument vector, and mode-specific environment to spawn with.
+ */
+export function resolveExampleLaunch(options: ExampleLaunchOptions): ExampleLaunch {
+  const mode = options.mode ?? resolveExampleMode()
+  const configArgs = options.configArgs ?? []
+  const flags = options.exposeInternals === true ? ['--expose-internals'] : []
+  const env: NodeJS.ProcessEnv = { ...options.env }
+
+  if (mode === 'src') {
+    if (options.tsconfigPath === undefined) {
+      throw new Error("resolveExampleLaunch: 'src' mode needs tsconfigPath for the workspace paths map.")
+    }
+    const tsxLoader = import.meta.resolve('tsx')
+    env.TSX_TSCONFIG_PATH = options.tsconfigPath
+    return { command: process.execPath, args: [...flags, '--import', tsxLoader, options.srcBin, ...configArgs], env }
+  }
+
+  return { command: process.execPath, args: [...flags, options.libBin ?? toLibBin(options.srcBin), ...configArgs], env }
+}
+
 /** Inputs that vary between real-Loader example smokes. */
 export interface LoaderSmokeOptions {
   /** Human-readable example name used in failure diagnostics. */
   readonly label: string
   /** Prefix for the isolated temporary process cwd. */
   readonly tempDirPrefix: string
-  /** Absolute stdio-agent bin path. */
+  /** Absolute stdio-agent bin SOURCE path (`<pkg>/src/bin.ts`); the `lib` bin is derived from it. */
   readonly binScript: string
+  /** Explicit plain-Node entry for `lib` mode; intended for test fixtures outside a package `src/` tree. */
+  readonly libBinScript?: string | undefined
   /** Absolute real Loader config path. */
   readonly configPath: string
-  /** Absolute repo tsconfig path used for unbuilt workspace-package resolution. */
+  /** Absolute repo tsconfig path used for unbuilt workspace-package resolution (required in `src` mode). */
   readonly tsconfigPath: string
+  /** Boot from source via tsx (`src`) or built lib via plain Node (`lib`); defaults to the environment's mode. */
+  readonly mode?: ExampleMode
   /** Environment overrides layered over the parent and isolated DSH homes. */
   readonly env?: Readonly<NodeJS.ProcessEnv>
   /** Lines written to stdin before EOF; omitted means immediate EOF. */
@@ -48,30 +155,29 @@ export interface LoaderSmokeResult {
 /**
  * Boot one real Loader tree from an isolated cwd, write the requested stdin
  * script, close stdin, and await a clean exit. The helper owns process kill and
- * temp-directory cleanup on every outcome.
- * @param options - example paths, environment, stdin, and diagnostic identity.
+ * temp-directory cleanup on every outcome, and picks src/lib via {@link resolveExampleLaunch}.
+ * @param options - example paths, mode, environment, stdin, and diagnostic identity.
  * @returns captured stdout and stderr after a zero exit.
  */
 export async function runLoaderSmoke(options: LoaderSmokeOptions): Promise<LoaderSmokeResult> {
   const cwd = await mkdtemp(join(tmpdir(), options.tempDirPrefix))
   const processTimeoutMs = options.processTimeoutMs ?? DEFAULT_PROCESS_TIMEOUT_MS
+  const launch = resolveExampleLaunch({
+    srcBin: options.binScript,
+    libBin: options.libBinScript,
+    configArgs: [options.configPath],
+    ...options.mode !== undefined ? { mode: options.mode } : {},
+    tsconfigPath: options.tsconfigPath,
+    exposeInternals: true,
+    env: { DSH_HOME: join(cwd, '.dsh'), DSH_AGENTS_HOME: join(cwd, '.agents'), ...options.env },
+  })
   try {
     return await new Promise((resolve, reject) => {
-      const child = spawn(
-        process.execPath,
-        ['--expose-internals', '--import', TSX_LOADER, options.binScript, options.configPath],
-        {
-          cwd,
-          env: {
-            ...process.env,
-            DSH_HOME: join(cwd, '.dsh'),
-            DSH_AGENTS_HOME: join(cwd, '.agents'),
-            ...options.env,
-            TSX_TSCONFIG_PATH: options.tsconfigPath,
-          },
-          stdio: ['pipe', 'pipe', 'pipe'],
-        },
-      )
+      const child = spawn(launch.command, launch.args, {
+        cwd,
+        env: { ...process.env, ...launch.env },
+        stdio: ['pipe', 'pipe', 'pipe'],
+      })
       let stdout = ''
       let stderr = ''
       let deferredFailure: Error | undefined
diff --git a/packages/support/loader-smoke/tests/example-launch.spec.ts b/packages/support/loader-smoke/tests/example-launch.spec.ts
new file mode 100644
index 0000000000..20520e6fb6
--- /dev/null
+++ b/packages/support/loader-smoke/tests/example-launch.spec.ts
@@ -0,0 +1,111 @@
+import { afterEach, describe, expect, it } from 'vitest'
+import {
+  EXAMPLE_MODE_ENV,
+  resolveExampleLaunch,
+  resolveExampleMode,
+} from '@deepseek-ai/dsh-loader-smoke'
+
+const SRC_BIN = '/repo/packages/examples/stdio-demo/src/bin.ts'
+const TSCONFIG = '/repo/tsconfig.json'
+
+const originalMode = process.env[EXAMPLE_MODE_ENV]
+afterEach(() => {
+  if (originalMode === undefined) Reflect.deleteProperty(process.env, EXAMPLE_MODE_ENV)
+  else process.env[EXAMPLE_MODE_ENV] = originalMode
+})
+
+describe('resolveExampleMode', () => {
+  it('defaults absent/empty/src to src', () => {
+    Reflect.deleteProperty(process.env, EXAMPLE_MODE_ENV)
+    expect(resolveExampleMode()).toBe('src')
+    expect(resolveExampleMode('')).toBe('src')
+    expect(resolveExampleMode('src')).toBe('src')
+  })
+
+  it('accepts lib', () => {
+    expect(resolveExampleMode('lib')).toBe('lib')
+  })
+
+  it('throws on any other value', () => {
+    expect(() => resolveExampleMode('prod')).toThrow(/must be 'src' or 'lib'/)
+  })
+
+  it('reads the environment when no argument is given', () => {
+    process.env[EXAMPLE_MODE_ENV] = 'lib'
+    expect(resolveExampleMode()).toBe('lib')
+    Reflect.deleteProperty(process.env, EXAMPLE_MODE_ENV)
+    expect(resolveExampleMode()).toBe('src')
+  })
+})
+
+describe('resolveExampleLaunch', () => {
+  it('src mode: --import tsx on the source bin with the tsconfig paths env', () => {
+    const { command, args, env } = resolveExampleLaunch({
+      srcBin: SRC_BIN,
+      configArgs: ['./cordis.yml'],
+      mode: 'src',
+      tsconfigPath: TSCONFIG,
+    })
+    expect(command).toBe(process.execPath)
+    expect(args).toContain('--import')
+    expect(args).toContain(SRC_BIN)
+    expect(args[args.length - 1]).toBe('./cordis.yml')
+    expect(args).not.toContain('--expose-internals')
+    expect(env.TSX_TSCONFIG_PATH).toBe(TSCONFIG)
+  })
+
+  it('src mode: throws without a tsconfig path', () => {
+    expect(() => resolveExampleLaunch({ srcBin: SRC_BIN, mode: 'src' })).toThrow(/needs tsconfigPath/)
+  })
+
+  it('lib mode: plain node on the derived lib bin, no tsx and no paths env', () => {
+    const { args, env } = resolveExampleLaunch({
+      srcBin: SRC_BIN,
+      configArgs: ['--config', './cordis.yml'],
+      mode: 'lib',
+      env: { DSH_HOME: '/tmp/home' },
+    })
+    expect(args).not.toContain('--import')
+    expect(args).toContain('/repo/packages/examples/stdio-demo/lib/bin.js')
+    expect(args.slice(-2)).toEqual(['--config', './cordis.yml'])
+    expect(env.TSX_TSCONFIG_PATH).toBeUndefined()
+    expect(env.DSH_HOME).toBe('/tmp/home')
+  })
+
+  it('lib mode: uses an explicit plain-Node bin when provided', () => {
+    const fixture = '/repo/fixture.ts'
+    const { args } = resolveExampleLaunch({ srcBin: fixture, libBin: fixture, mode: 'lib' })
+    expect(args).toContain(fixture)
+  })
+
+  it('prepends --expose-internals when requested', () => {
+    const { args } = resolveExampleLaunch({ srcBin: SRC_BIN, mode: 'lib', exposeInternals: true })
+    expect(args[0]).toBe('--expose-internals')
+  })
+
+  it('lib mode: rewrites only the last /src/ segment', () => {
+    const { args } = resolveExampleLaunch({
+      srcBin: '/repo/src/packages/examples/acp-demo/src/bin.ts',
+      mode: 'lib',
+    })
+    expect(args).toContain('/repo/src/packages/examples/acp-demo/lib/bin.js')
+  })
+
+  it('lib mode: derives the built bin from a Windows source path', () => {
+    const { args } = resolveExampleLaunch({
+      srcBin: String.raw`D:\repo\src\packages\examples\acp-demo\src\bin.ts`,
+      mode: 'lib',
+    })
+    expect(args).toContain(String.raw`D:\repo\src\packages\examples\acp-demo\lib\bin.js`)
+  })
+
+  it('lib mode: throws when the bin has no /src/ segment', () => {
+    expect(() => resolveExampleLaunch({ srcBin: '/repo/lib/bin.js', mode: 'lib' })).toThrow(/"\/src\/" segment/)
+  })
+
+  it('defaults the mode from the environment', () => {
+    process.env[EXAMPLE_MODE_ENV] = 'lib'
+    const { args } = resolveExampleLaunch({ srcBin: SRC_BIN })
+    expect(args).toContain('/repo/packages/examples/stdio-demo/lib/bin.js')
+  })
+})
diff --git a/packages/support/loader-smoke/tests/loader-smoke.spec.ts b/packages/support/loader-smoke/tests/loader-smoke.spec.ts
index 4cc9f878f9..b99d810188 100644
--- a/packages/support/loader-smoke/tests/loader-smoke.spec.ts
+++ b/packages/support/loader-smoke/tests/loader-smoke.spec.ts
@@ -16,6 +16,7 @@ describe('runLoaderSmoke', () => {
       binScript: fixture('success'),
       configPath,
       tsconfigPath,
+      mode: 'src',
       env: { LOADER_SMOKE_MARKER: 'present' },
       stdinLines: ['one', 'two'],
     })
@@ -43,6 +44,7 @@ describe('runLoaderSmoke', () => {
       label: 'failure fixture',
       tempDirPrefix: 'loader-smoke-fail-',
       binScript: fixture('fail'),
+      libBinScript: fixture('fail'),
       configPath,
       tsconfigPath,
     })).rejects.toThrow('failure fixture exited 7. stdout:\n\nstderr:\nfixture failed')
@@ -53,6 +55,7 @@ describe('runLoaderSmoke', () => {
       label: 'hanging fixture',
       tempDirPrefix: 'loader-smoke-hang-',
       binScript: fixture('hang'),
+      libBinScript: fixture('hang'),
       configPath,
       tsconfigPath,
       processTimeoutMs: 100,
diff --git a/packages/tasks/README.md b/packages/tasks/README.md
new file mode 100644
index 0000000000..12e8f41a59
--- /dev/null
+++ b/packages/tasks/README.md
@@ -0,0 +1,10 @@
+# tasks/ — background task capability family
+
+The shared home for background-task ids, owner isolation, reads, cancellation, waiting, and completion notices. Bash, subagents, and future long-running tools use one model-facing protocol. See the [background-task runtime RFC](../../docs/rfc/implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md).
+
+| Package | ctx key | Role |
+|---|---|---|
+| [`tasks`](tasks/README.md) (`@deepseek-ai/dsh-tasks`) | `ctx.tasks` | The registry service: branded `<kind>-N` ids, owner-fenced read/kill/wait/list, settlement bookkeeping, the awaited owner-cleanup path, and the `attachSurface` misconfiguration fence |
+| [`tool-tasks`](tool-tasks/README.md) (`@deepseek-ai/dsh-tool-tasks`) | — | The model-facing control surface: `task_output`, `task_list`, `task_kill`, the completion-notice injection, and the background-habit prompt section |
+
+The registry owns state across producer or surface reloads; the tool package owns presentation. Producers register execution hooks through `ctx.tasks.start` and own whether their config exposes `run_in_background`.
diff --git a/packages/tasks/tasks/README.md b/packages/tasks/tasks/README.md
new file mode 100644
index 0000000000..b061038714
--- /dev/null
+++ b/packages/tasks/tasks/README.md
@@ -0,0 +1,35 @@
+# @deepseek-ai/dsh-tasks
+
+The process-local background task registry (`ctx.tasks`). It gives long-running producers shared ids, owner isolation, reads, cancellation, waiting, notices, and cleanup. Producer plugins extend `TaskKindMap` with their opaque id namespace.
+
+## Service API
+
+- `start(spec): TaskId` validates the control surface, spec, and exact live owner before calling the producer's `run()` once. A starter throw leaves nothing registered; successful return commits without another failable step.
+- `get(id, caller?)` and `list(caller?)` return non-consuming snapshots. Listing includes only caller-owned and unowned tasks.
+- `read(id, caller?)` consumes the single cursor for stream tasks and reads terminal output idempotently for final-output tasks.
+- `kill(id, caller?, reason?)` invokes producer cancellation before changing status. A cancellation throw leaves the task running; success changes it to `stopping` and marks terminal delivery reported.
+- `wait(id, timeoutMs, caller?, signal?)` returns a terminal snapshot or the live snapshot at timeout. Aborting stops only the wait; settlement wins once it has committed terminal delivery to that waiter.
+- `onTaskDone(listener)` observes each terminal record with the exact owner. Listener throws and rejections are contained; listener work is not awaited.
+- `attachSurface(name)` declares a control surface for its effect lifetime. `start()` fails before producer execution when none is attached.
+
+Owned access compares the task's `SessionId` with the caller's. Ids such as `bash-1` are predictable, so this fence is the boundary. Unowned tasks are open to callers and last until service disposal.
+
+## Lifecycle
+
+Tasks belong to their owner and backend, not the producer tool fiber, so producer and surface reloads do not stop them. The first task for an owner attaches one awaited effect to the exact `Agent` scope. Owner disposal cancels that object's tasks, awaits producer quiescence, and removes their snapshots; reused agent or session ids cannot redirect an old cleanup.
+
+Service disposal closes listeners, cancels all live tasks, awaits their records, and detaches effects from surviving owner scopes. If teardown cancellation throws, the service force-fails the record and warns that work may be orphaned instead of deadlocking. A cancellation that returns but never settles `done` remains indistinguishable from a slow stop and can stall teardown.
+
+See the [task type catalog](../../../docs/core-data-structures/tasks.md) and [runtime RFC](../../../docs/rfc/implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md).
+
+## Model Experience
+
+Indirectly, through producer plugins and [`dsh-tool-tasks`](../tool-tasks/README.md), which render task ids, output, status, cancellation, and completion notices.
+
+## Known Limitations and Deferred Work
+
+- **Tasks are process-local** — durable or cross-restart execution needs a separate lifecycle.
+- **The service and implementation are not split** — a second backend must define the lifecycle that shapes that boundary.
+- **Stream output has one consuming cursor** — independent observers need a cursor or snapshot API.
+- **Foreground work cannot be promoted** — producers choose foreground or background before starting.
+- **A silently ineffective cancel can stall teardown** — only an explicit throw can be force-failed safely.
diff --git a/packages/tasks/tasks/package.json b/packages/tasks/tasks/package.json
new file mode 100644
index 0000000000..9e5f824418
--- /dev/null
+++ b/packages/tasks/tasks/package.json
@@ -0,0 +1,38 @@
+{
+  "name": "@deepseek-ai/dsh-tasks",
+  "description": "Background task registry (ctx.tasks) for the DeepSeek Harness \u2014 shared ids, owner isolation, polling, cancellation, and completion listeners for long-running tool work",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "@deepseek-ai/dsh-agent": "^0.0.1",
+    "@deepseek-ai/dsh-brand": "^0.0.1",
+    "@deepseek-ai/dsh-session": "^0.0.1",
+    "@deepseek-ai/dsh-timeout": "^0.0.1",
+    "cordis": "^4.0.0-rc.6"
+  },
+  "devDependencies": {
+    "@deepseek-ai/dsh-agent": "workspace:^",
+    "@deepseek-ai/dsh-brand": "workspace:^",
+    "@deepseek-ai/dsh-session": "workspace:^",
+    "@deepseek-ai/dsh-timeout": "workspace:^",
+    "cordis": "^4.0.0-rc.6"
+  }
+}
diff --git a/packages/tasks/tasks/src/index.ts b/packages/tasks/tasks/src/index.ts
new file mode 100644
index 0000000000..e9c896c3a1
--- /dev/null
+++ b/packages/tasks/tasks/src/index.ts
@@ -0,0 +1,441 @@
+/**
+ * The in-process background task registry (`ctx.tasks`). It owns task ids,
+ * session-scoped access, lifecycle state, completion listeners, and owner
+ * cleanup while producers retain their execution resources.
+ *
+ * Registrations outlive producer and control-surface fibers. Agent or service
+ * disposal cancels live work and awaits compliant producers; a throwing
+ * teardown cancel force-fails only the record and reports a possible orphan.
+ * @module @deepseek-ai/dsh-tasks
+ */
+
+import { Context, Service } from 'cordis'
+import type { Agent } from '@deepseek-ai/dsh-agent'
+import { deadline, timeoutOf } from '@deepseek-ai/dsh-timeout'
+import { TaskId } from './types.ts'
+import type { TaskDoneListener, TaskKind, TaskOutcome, TaskRead, TaskSnapshot, TaskStart, TaskStatus } from './types.ts'
+
+export { TaskId } from './types.ts'
+export type {
+  TaskDoneListener,
+  TaskHooks,
+  TaskKind,
+  TaskKindMap,
+  TaskOutcome,
+  TaskRead,
+  TaskSnapshot,
+  TaskStart,
+  TaskStatus,
+} from './types.ts'
+
+declare module 'cordis' {
+  interface Context {
+    tasks: TaskService
+  }
+}
+
+/** Timeout code that distinguishes a bounded wait from caller cancellation. */
+export const TASK_WAIT_TIMEOUT = 'TASK_WAIT_TIMEOUT'
+
+/** The registry's mutable per-task record (never handed out — see {@link TaskService.snapshot}). */
+interface TrackedTask {
+  id: TaskId
+  kind: TaskKind
+  label: string
+  /** Exact lifecycle owner; session-id authorization is derived from it. */
+  owner: Agent | undefined
+  cancel: (reason?: string) => void
+  readOutput: (() => string) | undefined
+  status: TaskStatus
+  detail: string | undefined
+  output: string | undefined
+  startedAt: number
+  finishedAt: number | undefined
+  reported: boolean
+  /** Resolves once the terminal snapshot is recorded and listeners notified. */
+  settled: Promise<void>
+  /** Resolver for {@link settled}, called by the first effective settlement. */
+  markSettled: () => void
+  /** Live waits; settlement with a waiter marks the task reported. */
+  waiters: number
+  /** Removable resolvers for live waits; timeout/abort unregister before the task settles. */
+  waitResolvers: Set<() => void>
+}
+
+/** True for the three terminal {@link TaskStatus} values. */
+function isTerminal(status: TaskStatus): boolean {
+  return status === 'completed' || status === 'killed' || status === 'failed'
+}
+
+/**
+ * The `tasks` service: the runtime-global background task registry. See the
+ * module doc for the ownership, isolation, and lifecycle contracts.
+ */
+// TODO(task-service-backend): Separate the service contract from this
+// process-local implementation when a second backend defines its lifecycle.
+export class TaskService extends Service {
+  private store = new Map<TaskId, TrackedTask>()
+  private counters = new Map<string, number>()
+  private surfaces = new Set<symbol>()
+  private listeners = new Set<TaskDoneListener>()
+  private listenersClosed = false
+  /** Owner agents with attached scope cleanup, mapped to the exact disposer. */
+  private ownerCleanups = new Map<Agent, () => Promise<void> | void>()
+  /** Service context used by detached settlement continuations and teardown. */
+  private readonly selfCtx: Context
+
+  constructor(ctx: Context) {
+    super(ctx, 'tasks')
+    this.selfCtx = ctx
+    ctx.effect(() => () => this.disposeAll(), 'tasks teardown')
+  }
+
+  /**
+   * Preflight access, validation, and owner cleanup before starting and
+   * atomically registering work. A throwing starter leaves nothing registered;
+   * after it returns, registration cannot fail. Settlement records the outcome,
+   * notifies listeners, and releases waiters.
+   * @param spec - task identity, owner, and synchronous starter.
+   * @returns the registry-issued `<kind>-N` id.
+   */
+  start(spec: TaskStart): TaskId {
+    if (this.surfaces.size === 0) {
+      throw new Error('background tasks unavailable: no control surface is attached (load @deepseek-ai/dsh-tool-tasks)')
+    }
+    if (spec.kind.length === 0) throw new Error('invalid task kind: expected a non-empty string')
+    if (spec.label.length === 0) throw new Error('invalid task label: expected a non-empty string')
+    if (spec.owner !== undefined) this.ensureOwnerCleanup(spec.owner)
+
+    const hooks = spec.run()
+    const count = (this.counters.get(spec.kind) ?? 0) + 1
+    this.counters.set(spec.kind, count)
+    const id = TaskId(`${spec.kind}-${count}`)
+
+    let markSettled!: () => void
+    const settled = new Promise<void>((resolve) => { markSettled = resolve })
+    const task: TrackedTask = {
+      id,
+      kind: spec.kind,
+      label: spec.label,
+      owner: spec.owner,
+      cancel: hooks.cancel.bind(hooks),
+      readOutput: hooks.readOutput?.bind(hooks),
+      status: 'running',
+      detail: undefined,
+      output: undefined,
+      startedAt: Date.now(),
+      finishedAt: undefined,
+      reported: false,
+      settled,
+      markSettled,
+      waiters: 0,
+      waitResolvers: new Set(),
+    }
+    this.store.set(id, task)
+
+    void hooks.done.then(
+      (outcome) => { this.settle(task, outcome) },
+      (error: unknown) => {
+        // Contain a producer contract violation so cleanup and waiters cannot hang.
+        this.selfCtx.logger.warn(`tasks: task ${task.id} 'done' rejected (producer contract violation): ${String(error)}`)
+        this.settle(task, { status: 'failed', detail: String(error) })
+      },
+    )
+    return id
+  }
+
+  /**
+   * List caller-owned and unowned tasks in registration order without exposing
+   * another session's labels.
+   * @param caller - reading agent; a non-agent caller sees only unowned tasks.
+   * @returns fresh snapshots.
+   */
+  list(caller?: Agent): TaskSnapshot[] {
+    const session = caller?.session.header.id
+    return [...this.store.values()]
+      .filter(task => task.owner === undefined || task.owner.session.header.id === session)
+      .map(task => this.snapshot(task))
+  }
+
+  /**
+   * Return a non-consuming snapshot without changing its read cursor or notice
+   * state. Throws for an unknown or foreign task.
+   * @param id - task to look up.
+   * @param caller - reading agent checked against the owner.
+   * @returns a fresh snapshot.
+   */
+  get(id: TaskId, caller?: Agent): TaskSnapshot {
+    const task = this.expect(id)
+    this.assertAccess(task, caller)
+    return this.snapshot(task)
+  }
+
+  /**
+   * Read the next stream delta, or the idempotent final output after settlement.
+   * A terminal read marks the task reported. Throws for an unknown or foreign
+   * task.
+   * @param id - task to read.
+   * @param caller - reading agent checked against the owner.
+   * @returns output text and the post-read snapshot.
+   */
+  read(id: TaskId, caller?: Agent): TaskRead {
+    const task = this.expect(id)
+    this.assertAccess(task, caller)
+    const text = task.readOutput !== undefined
+      ? task.readOutput()
+      : isTerminal(task.status) ? task.output ?? '' : ''
+    if (isTerminal(task.status)) task.reported = true
+    return { text, snapshot: this.snapshot(task) }
+  }
+
+  /**
+   * Request cancellation, then mark the task stopping and reported. A producer
+   * throw propagates without changing task state. Throws for an unknown or
+   * foreign task.
+   * @param id - task to cancel.
+   * @param caller - killing agent checked against the owner.
+   * @param reason - logged reason forwarded to the producer.
+   * @returns `requested` for live work, otherwise `already-finished`.
+   */
+  kill(id: TaskId, caller?: Agent, reason?: string): 'requested' | 'already-finished' {
+    const task = this.expect(id)
+    this.assertAccess(task, caller)
+    if (isTerminal(task.status)) {
+      task.reported = true
+      return 'already-finished'
+    }
+    // Cancel first so a throw leaves both lifecycle and notice state unchanged.
+    task.cancel(reason)
+    task.status = 'stopping'
+    task.reported = true
+    return 'requested'
+  }
+
+  /**
+   * Wait for settlement or timeout without cancelling the task. Caller abort
+   * rejects only while the task is live; after settlement it returns the
+   * terminal snapshot so a notice suppressed for this waiter is still delivered.
+   * Timed-out and aborted waits detach their resolvers. Throws for invalid,
+   * unknown, or foreign input.
+   * @param id - task to wait for.
+   * @param timeoutMs - positive finite wait bound in milliseconds.
+   * @param caller - waiting agent checked against the owner.
+   * @param signal - optional cancellation of the wait itself.
+   * @returns snapshot at settlement or timeout.
+   */
+  async wait(id: TaskId, timeoutMs: number, caller?: Agent, signal?: AbortSignal): Promise<TaskSnapshot> {
+    const task = this.expect(id)
+    this.assertAccess(task, caller)
+    if (!Number.isFinite(timeoutMs) || timeoutMs <= 0) {
+      throw new Error(`invalid wait timeout: expected a positive number of milliseconds, got ${JSON.stringify(timeoutMs)}`)
+    }
+    if (!isTerminal(task.status)) {
+      if (signal?.aborted) throw new Error('wait aborted')
+      // Abort removes the waiter synchronously so same-tick settlement cannot
+      // suppress a notice for a wait that will reject.
+      task.waiters += 1
+      let counted = true
+      const uncount = (): void => {
+        if (!counted) return
+        counted = false
+        task.waiters -= 1
+      }
+      try {
+        // The scoped deadline distinguishes a successful wait timeout from
+        // caller cancellation and clears its timer on every exit.
+        using d = deadline(signal, timeoutMs, TASK_WAIT_TIMEOUT)
+        await new Promise<void>((resolve, reject) => {
+          const onSettled = (): void => {
+            task.waitResolvers.delete(onSettled)
+            d.signal.removeEventListener('abort', onAbort)
+            resolve()
+          }
+          const onAbort = (): void => {
+            task.waitResolvers.delete(onSettled)
+            if (timeoutOf(d.signal, TASK_WAIT_TIMEOUT) !== undefined) {
+              resolve()
+            } else if (isTerminal(task.status)) {
+              // Settlement suppressed the notice for this waiter; deliver it.
+              resolve()
+            } else {
+              uncount()
+              reject(new Error('wait aborted'))
+            }
+          }
+          task.waitResolvers.add(onSettled)
+          d.signal.addEventListener('abort', onAbort, { once: true })
+        })
+      } finally {
+        uncount()
+      }
+    }
+    if (isTerminal(task.status)) task.reported = true
+    return this.snapshot(task)
+  }
+
+  /**
+   * Register an effect-scoped completion listener. Each listener is contained;
+   * returned promises are observed but not awaited. No listener runs after
+   * service disposal.
+   * @param listener - receives each terminal snapshot and its exact owner.
+   * @returns disposer that unregisters the listener.
+   */
+  onTaskDone(listener: TaskDoneListener): () => void {
+    const dispose = this.ctx.effect(() => {
+      this.listeners.add(listener)
+      return () => this.listeners.delete(listener)
+    }, 'tasks.onTaskDone()')
+    return () => void dispose()
+  }
+
+  /**
+   * Attach an effect-scoped surface that can read and stop tasks. {@link start}
+   * refuses work while none is attached.
+   * @param name - diagnostic label; duplicate names remain independent.
+   * @returns disposer that detaches this surface.
+   */
+  attachSurface(name: string): () => void {
+    // One token per call keeps duplicate labels independently disposable.
+    const token = Symbol(name)
+    const dispose = this.ctx.effect(() => {
+      this.surfaces.add(token)
+      return () => this.surfaces.delete(token)
+    }, 'tasks.attachSurface()')
+    return () => void dispose()
+  }
+
+  /** Look up a task or fail loud. */
+  private expect(id: TaskId): TrackedTask {
+    const task = this.store.get(id)
+    if (task === undefined) throw new Error(`unknown task ${id}`)
+    return task
+  }
+
+  /**
+   * The isolation fence: a task with an owner is reachable only by callers
+   * whose session id matches (`!== undefined` semantics — an unowned task is
+   * open, and a no-agent caller can never match an owned one).
+   */
+  private assertAccess(task: TrackedTask, caller?: Agent): void {
+    if (task.owner !== undefined && task.owner.session.header.id !== caller?.session.header.id) {
+      throw new Error(`task ${task.id} belongs to another session`)
+    }
+  }
+
+  /** Project a fresh read-only snapshot from the mutable record. */
+  private snapshot(task: TrackedTask): TaskSnapshot {
+    const ownerSession = task.owner?.session.header.id
+    return {
+      id: task.id,
+      kind: task.kind,
+      label: task.label,
+      ...ownerSession !== undefined ? { ownerSession } : {},
+      status: task.status,
+      ...task.detail !== undefined ? { detail: task.detail } : {},
+      startedAt: task.startedAt,
+      ...task.finishedAt !== undefined ? { finishedAt: task.finishedAt } : {},
+      reported: task.reported,
+    }
+  }
+
+  /**
+   * Record the first terminal outcome, notify contained listeners, and release
+   * waiters. First-wins preserves a teardown force-failure against late producer
+   * settlement. Pending waits mark the task reported before listeners run.
+   */
+  private settle(task: TrackedTask, outcome: TaskOutcome): void {
+    if (isTerminal(task.status)) return
+    task.status = outcome.status
+    task.detail = outcome.detail
+    task.output = outcome.output
+    task.finishedAt = Date.now()
+    if (task.waiters > 0) task.reported = true
+    if (!this.listenersClosed) {
+      const snapshot = this.snapshot(task)
+      for (const listener of this.listeners) {
+        try {
+          const returned = listener(snapshot, task.owner)
+          void Promise.resolve(returned).catch((error: unknown) => {
+            this.selfCtx.logger.warn(`tasks: onTaskDone listener rejected for ${task.id}: ${String(error)}`)
+          })
+        } catch (error: unknown) {
+          this.selfCtx.logger.warn(`tasks: onTaskDone listener threw for ${task.id}: ${String(error)}`)
+        }
+      }
+    }
+    const waitResolvers = [...task.waitResolvers]
+    task.waitResolvers.clear()
+    for (const resolveWait of waitResolvers) resolveWait()
+    task.markSettled()
+  }
+
+  /**
+   * Attach one awaited cleanup through the exact owner's scope. This survives
+   * producer reloads and joins agent quiescence; the retained disposer lets
+   * service teardown detach the cross-fiber effect. Fails when the registry is
+   * absent or the owner is not its currently registered instance.
+   */
+  private ensureOwnerCleanup(owner: Agent): void {
+    const ownerId = owner.id
+    const agents = this.selfCtx.get('agents')
+    if (agents === undefined) {
+      throw new Error('background task ownership requires the agent registry (load @deepseek-ai/dsh-agent)')
+    }
+    if (agents.get(ownerId) !== owner) {
+      throw new Error(`agent "${ownerId}" is not the registered agent instance (background task owner must be live)`)
+    }
+    if (this.ownerCleanups.has(owner)) return
+    // Record only after attach succeeds; a disposing scope rejects new effects.
+    const detach = owner.ctx.effect(() => async () => {
+      this.ownerCleanups.delete(owner)
+      await this.disposeOwned(owner)
+    }, 'tasks.ownerCleanup()')
+    this.ownerCleanups.set(owner, detach)
+  }
+
+  /** Cancel, await terminal records, and drop every task owned by one exact agent lifecycle. */
+  private async disposeOwned(owner: Agent): Promise<void> {
+    const owned = [...this.store.values()].filter(task => task.owner === owner)
+    this.cancelForTeardown(owned, 'owner disposed')
+    await Promise.all(owned.map(task => task.settled))
+    for (const task of owned) this.store.delete(task.id)
+  }
+
+  /**
+   * Close listeners, cancel live tasks, await settlement, and detach owner
+   * effects. Throwing cancels are force-failed to avoid teardown deadlock.
+   */
+  private async disposeAll(): Promise<void> {
+    this.listenersClosed = true
+    this.listeners.clear()
+    const all = [...this.store.values()]
+    this.cancelForTeardown(all, 'tasks service disposed')
+    await Promise.all(all.map(task => task.settled))
+    this.store.clear()
+    // Detach cross-fiber owner effects after the shared store is quiescent.
+    const ownerCleanups = [...this.ownerCleanups.values()]
+    this.ownerCleanups.clear()
+    await Promise.all(ownerCleanups.map(cleanup => Promise.resolve(cleanup())))
+  }
+
+  /**
+   * Cancel tasks during teardown with per-task containment. A throwing cancel
+   * force-fails the record and reports a possible orphan; a cancel that returns
+   * without settling remains indistinguishable from a slow stop and may stall.
+   */
+  private cancelForTeardown(tasks: TrackedTask[], reason: string): void {
+    for (const task of tasks) {
+      if (isTerminal(task.status)) continue
+      try {
+        task.cancel(reason)
+        task.status = 'stopping'
+      } catch (error: unknown) {
+        const detail = `cancel threw during teardown; work may be orphaned: ${String(error)}`
+        this.selfCtx.logger.warn(`tasks: cancel of ${task.id} threw during teardown; task record forced failed and work may be orphaned: ${String(error)}`)
+        this.settle(task, { status: 'failed', detail })
+      }
+    }
+  }
+}
+
+export default TaskService
diff --git a/packages/tasks/tasks/src/types.ts b/packages/tasks/tasks/src/types.ts
new file mode 100644
index 0000000000..96316260ec
--- /dev/null
+++ b/packages/tasks/tasks/src/types.ts
@@ -0,0 +1,152 @@
+/**
+ * Types shared by task producers, the registry, and control surfaces. The
+ * service implementation lives in `./index.ts`.
+ * @module @deepseek-ai/dsh-tasks/types
+ */
+
+import type { Branded } from '@deepseek-ai/dsh-brand'
+import type { Agent } from '@deepseek-ai/dsh-agent'
+import type { SessionId } from '@deepseek-ai/dsh-session'
+
+/**
+ * Identifies a background task. The registry generates `<kind>-N`; predictable
+ * ids rely on owner authorization rather than secrecy.
+ */
+export type TaskId = Branded<'TaskId'>
+
+/**
+ * Brand a string as a {@link TaskId}.
+ * @param id - the raw task-id string (the registry generates `<kind>-N`).
+ * @returns the same string, branded; no validation is performed.
+ */
+export function TaskId(id: string): TaskId {
+  return id as TaskId
+}
+
+/**
+ * Task lifecycle: `running`, optionally `stopping`, then exactly one terminal
+ * status. Producer-specific facts belong in {@link TaskSnapshot.detail}.
+ */
+export type TaskStatus = 'running' | 'stopping' | 'completed' | 'killed' | 'failed'
+
+/**
+ * Producer-defined task kinds. Plugins extend this map by declaration merging;
+ * the registry treats every value as an opaque id namespace.
+ */
+export interface TaskKindMap {
+  bash: 'bash'
+  subagent: 'subagent'
+}
+
+/** The merge-extensible union of registered producer kind names. */
+export type TaskKind = TaskKindMap[keyof TaskKindMap]
+
+/** Terminal result supplied by a producer through {@link TaskHooks.done}. */
+export interface TaskOutcome {
+  /** How the task ended: finished (`completed`), cancelled (`killed`), or broke (`failed`). */
+  status: 'completed' | 'killed' | 'failed'
+  /** Kind-specific detail rendered into status lines ('exit code: 3', 'max-tokens'). */
+  detail?: string
+  /** Final output for tasks without `readOutput`; stream tasks leave it unset. */
+  output?: string
+}
+
+/**
+ * Producer declaration passed to {@link TaskService.start}. The runtime
+ * preflights access and cleanup before invoking {@link run}; the producer owns
+ * execution resources while the runtime owns identity and lifecycle state.
+ */
+export interface TaskStart {
+  /** Producer kind — also the id prefix (`bash`, `subagent`, …). */
+  kind: TaskKind
+  /** One-line model-facing label (the command; the delegation description). */
+  label: string
+  /**
+   * Owning live agent. Access is fenced by its session id, and agent disposal
+   * cancels and awaits the task. The instance must be the one currently
+   * registered under its agent id. Omitting the owner creates an unowned task,
+   * open to any caller until service disposal.
+   */
+  owner?: Agent
+  /**
+   * Start the work after preflight and synchronously return its hooks. Called
+   * once; a throw leaves nothing registered, and the producer must clean up any
+   * partially started resources.
+   */
+  run(): TaskHooks
+}
+
+/** Hooks through which the runtime controls and observes producer work. */
+export interface TaskHooks {
+  /**
+   * Request termination. Must be synchronous, idempotent, and eventually settle
+   * {@link done}; throws propagate. The optional reason is forwarded verbatim.
+   */
+  cancel(reason?: string): void
+  /**
+   * Resolves after the producer releases its resources, not merely when work
+   * finishes. Must not reject; the runtime converts a rejection to `failed`.
+   * If teardown cancellation throws, the runtime may force-fail only the
+   * registry record without claiming that the work stopped.
+   */
+  done: Promise<TaskOutcome>
+  /**
+   * Consume output produced since the previous call. The producer formats
+   * truncation and spill notices. Absence marks a final-output-only task; each
+   * task has one consuming cursor.
+   */
+  readOutput?(): string
+}
+
+/**
+ * A read-only projection of one task, safe to hand to listeners and tools —
+ * a fresh object per call, never live registry state.
+ */
+export interface TaskSnapshot {
+  /** The registry-issued id (`<kind>-N`). */
+  id: TaskId
+  /** The producer kind the task was registered with. */
+  kind: TaskKind
+  /** The producer-supplied one-line label. */
+  label: string
+  /**
+   * Owner session id used for authorization and correlation; absent for
+   * unowned tasks. Completion listeners receive the exact {@link Agent}
+   * separately through {@link TaskDoneListener}.
+   */
+  ownerSession?: SessionId
+  /** Current lifecycle state. */
+  status: TaskStatus
+  /** Kind-specific status detail, present once the producer supplied one (usually terminal). */
+  detail?: string
+  /** Epoch ms when the task was registered. */
+  startedAt: number
+  /** Epoch ms when the task settled; absent while `running`/`stopping`. */
+  finishedAt?: number
+  /**
+   * True when a kill, read, or wait has reported or committed to report the
+   * terminal state. Completion surfaces suppress redundant notices when set.
+   */
+  reported: boolean
+}
+
+/** Output and post-read state returned by {@link TaskService.read}. */
+export interface TaskRead {
+  /**
+   * Stream kinds: the consuming delta since the previous read. Final-output
+   * kinds: empty while live, the terminal {@link TaskOutcome.output} (or
+   * empty) once settled — idempotent, never consumed.
+   */
+  text: string
+  /** The task's state at read time. */
+  snapshot: TaskSnapshot
+}
+
+/**
+ * Completion callback with the exact owner supplied at start, or `undefined`
+ * for an unowned task. Returned promises are observed but not awaited.
+ */
+export type TaskDoneListener = (
+  snapshot: TaskSnapshot,
+  owner: Agent | undefined,
+) => void | PromiseLike<void>
diff --git a/packages/tasks/tasks/tests/tasks.spec.ts b/packages/tasks/tasks/tests/tasks.spec.ts
new file mode 100644
index 0000000000..f21eb9c433
--- /dev/null
+++ b/packages/tasks/tasks/tests/tasks.spec.ts
@@ -0,0 +1,737 @@
+import { describe, expect, expectTypeOf, it, vi } from 'vitest'
+import { Context } from 'cordis'
+import { Session, SessionId } from '@deepseek-ai/dsh-session'
+import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import type { Agent } from '@deepseek-ai/dsh-agent'
+import TaskService, { TaskId } from '@deepseek-ai/dsh-tasks'
+import type { TaskHooks, TaskKind, TaskOutcome, TaskSnapshot, TaskStart } from '@deepseek-ai/dsh-tasks'
+
+declare module '@deepseek-ai/dsh-tasks' {
+  interface TaskKindMap {
+    workflow: 'workflow'
+  }
+}
+
+const agentScopeDisposers = new WeakMap<Agent, () => Promise<void>>()
+
+function stubAgent(ctx: Context, rawId: string, rawSessionId = `${rawId}-session`): Agent {
+  const id = AgentId(rawId)
+  const scopeFiber = ctx.plugin(() => {})
+  const agent = {
+    id,
+    options: {},
+    session: new Session(SessionId(rawSessionId)),
+    status: 'idle' as const,
+    ctx: scopeFiber.ctx,
+    send() {},
+    steer() {},
+    inject() {},
+    cancel() {},
+    whenIdle() { return Promise.resolve() },
+  }
+  agentScopeDisposers.set(agent, async () => { await scopeFiber.dispose() })
+  return agent
+}
+
+async function disposeAgentScope(agent: Agent): Promise<void> {
+  const dispose = agentScopeDisposers.get(agent)
+  if (dispose === undefined) throw new Error(`missing test scope for agent "${agent.id}"`)
+  await dispose()
+}
+
+/** A controllable producer start-spec: settle its `done` on demand, record cancels. */
+function producer(overrides: Partial<Omit<TaskStart, 'run'> & TaskHooks> = {}) {
+  let settle!: (outcome: TaskOutcome) => void
+  let reject!: (error: unknown) => void
+  const cancels: (string | undefined)[] = []
+  const { kind = 'bash', label = 'sleep 60', owner, ...hookOverrides } = overrides
+  const hooks: TaskHooks = {
+    cancel(reason) { cancels.push(reason) },
+    done: new Promise<TaskOutcome>((res, rej) => { settle = res; reject = rej }),
+    ...hookOverrides,
+  }
+  const spec: TaskStart = { kind, label, ...owner !== undefined ? { owner } : {}, run: () => hooks }
+  return { spec, settle, reject, cancels }
+}
+
+async function harness() {
+  const ctx = new Context()
+  await ctx.plugin(AgentRegistry)
+  await ctx.plugin(TaskService)
+  ctx.tasks.attachSurface('test-surface')
+  return ctx
+}
+
+/** Let the settlement continuation (a `done.then`) run. */
+const tick = () => new Promise<void>(r => setTimeout(r, 0))
+
+/** Inspect the internal resolver registry to pin bounded retention while a task stays live. */
+function waitResolverCount(ctx: Context, id: TaskId): number {
+  const service = ctx.tasks as unknown as { store: Map<TaskId, { waitResolvers: Set<() => void> }> }
+  const task = service.store.get(id)
+  if (task === undefined) throw new Error(`missing test task ${id}`)
+  return task.waitResolvers.size
+}
+
+describe('TaskService.start', () => {
+  it('preserves the SessionId brand on public owner snapshots', () => {
+    expectTypeOf<TaskSnapshot['ownerSession']>().toEqualTypeOf<SessionId | undefined>()
+  })
+
+  it('refuses to register while no control surface is attached', async () => {
+    const ctx = new Context()
+    await ctx.plugin(TaskService)
+    expect(() => ctx.tasks.start(producer().spec))
+      .toThrow('background tasks unavailable: no control surface is attached (load @deepseek-ai/dsh-tool-tasks)')
+  })
+
+  it('rejects an empty kind and an empty label', async () => {
+    const ctx = await harness()
+    expect(() => ctx.tasks.start(producer({ kind: '' as TaskKind }).spec)).toThrow('invalid task kind')
+    expect(() => ctx.tasks.start(producer({ label: '' }).spec)).toThrow('invalid task label')
+  })
+
+  it('issues kind-prefixed ids from per-kind counters', async () => {
+    const ctx = await harness()
+    expect(ctx.tasks.start(producer().spec)).toBe('bash-1')
+    expect(ctx.tasks.start(producer().spec)).toBe('bash-2')
+    expect(ctx.tasks.start(producer({ kind: 'subagent' }).spec)).toBe('subagent-1')
+    expect(ctx.tasks.start(producer({ kind: 'workflow' }).spec)).toBe('workflow-1')
+  })
+})
+
+describe('TaskService reads and settlement', () => {
+  it('stream kinds read a consuming delta; terminal reads mark reported', async () => {
+    const ctx = await harness()
+    const chunks = ['first', '', 'rest']
+    const p = producer({ readOutput: () => chunks.shift() ?? '' })
+    const id = ctx.tasks.start(p.spec)
+
+    expect(ctx.tasks.read(id)).toMatchObject({ text: 'first', snapshot: { status: 'running', reported: false } })
+    expect(ctx.tasks.read(id).text).toBe('')
+
+    p.settle({ status: 'completed', detail: 'exit code: 0' })
+    await tick()
+    const read = ctx.tasks.read(id)
+    expect(read.text).toBe('rest')
+    expect(read.snapshot).toMatchObject({ status: 'completed', detail: 'exit code: 0', reported: true })
+    expect(read.snapshot.finishedAt).toBeTypeOf('number')
+  })
+
+  it('final-output kinds read empty while live, the outcome output idempotently once settled', async () => {
+    const ctx = await harness()
+    const p = producer({ kind: 'subagent', label: 'research task' })
+    const id = ctx.tasks.start(p.spec)
+
+    expect(ctx.tasks.read(id)).toMatchObject({ text: '', snapshot: { status: 'running' } })
+
+    p.settle({ status: 'completed', output: 'final answer' })
+    await tick()
+    expect(ctx.tasks.read(id).text).toBe('final answer')
+    expect(ctx.tasks.read(id).text).toBe('final answer') // idempotent, not consumed
+  })
+
+  it('a settled task without output reads as empty text', async () => {
+    const ctx = await harness()
+    const p = producer({ kind: 'subagent' })
+    const id = ctx.tasks.start(p.spec)
+    p.settle({ status: 'failed', detail: 'max-tokens' })
+    await tick()
+    expect(ctx.tasks.read(id)).toMatchObject({ text: '', snapshot: { status: 'failed', detail: 'max-tokens' } })
+  })
+
+  it('throws for unknown task ids', async () => {
+    const ctx = await harness()
+    expect(() => ctx.tasks.read(TaskId('bash-99'))).toThrow('unknown task bash-99')
+  })
+
+  it('notifies onTaskDone once per task with containment across listeners', async () => {
+    const ctx = await harness()
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    const seen: TaskSnapshot[] = []
+    ctx.tasks.onTaskDone(() => { throw new Error('listener boom') })
+    ctx.tasks.onTaskDone(snapshot => void seen.push(snapshot))
+
+    const p = producer()
+    const id = ctx.tasks.start(p.spec)
+    p.settle({ status: 'completed', detail: 'exit code: 0' })
+    await tick()
+
+    expect(seen).toHaveLength(1)
+    expect(seen[0]).toMatchObject({ id, status: 'completed', reported: false })
+    expect(warn).toHaveBeenCalledWith(expect.stringContaining('listener boom'))
+  })
+
+  it('contains a rejecting onTaskDone listener without starving later listeners', async () => {
+    const ctx = await harness()
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    const seen: TaskId[] = []
+    ctx.tasks.onTaskDone(async () => { throw new Error('async listener boom') })
+    ctx.tasks.onTaskDone(snapshot => void seen.push(snapshot.id))
+
+    const p = producer()
+    const id = ctx.tasks.start(p.spec)
+    p.settle({ status: 'completed' })
+    await tick()
+
+    expect(seen).toEqual([id])
+    expect(warn).toHaveBeenCalledWith(expect.stringContaining('onTaskDone listener rejected'))
+    expect(warn).toHaveBeenCalledWith(expect.stringContaining('async listener boom'))
+  })
+
+  it('contains a rejecting done as a failed outcome (producer contract violation)', async () => {
+    const ctx = await harness()
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    const p = producer()
+    const id = ctx.tasks.start(p.spec)
+    p.reject(new Error('transport exploded'))
+    await tick()
+
+    expect(ctx.tasks.read(id).snapshot).toMatchObject({ status: 'failed', detail: 'Error: transport exploded' })
+    expect(warn).toHaveBeenCalledWith(expect.stringContaining('producer contract violation'))
+  })
+
+  it('unregisters onTaskDone listeners with the contributing fiber (HMR safety)', async () => {
+    const ctx = await harness()
+    const seen: string[] = []
+    const fiber = await ctx.plugin(Object.assign((inner: Context) => {
+      inner.tasks.onTaskDone(snapshot => void seen.push(snapshot.id))
+    }, { inject: ['tasks'] }))
+    await fiber.dispose()
+    // The returned disposer detaches too (the non-fiber path).
+    const detach = ctx.tasks.onTaskDone(snapshot => void seen.push(snapshot.id))
+    detach()
+
+    const p = producer()
+    ctx.tasks.start(p.spec)
+    p.settle({ status: 'completed' })
+    await tick()
+    expect(seen).toEqual([])
+  })
+})
+
+describe('TaskService.kill', () => {
+  it('cancels a live task with the forwarded reason and suppresses the notice', async () => {
+    const ctx = await harness()
+    const seen: TaskSnapshot[] = []
+    ctx.tasks.onTaskDone(snapshot => void seen.push(snapshot))
+    const p = producer()
+    const id = ctx.tasks.start(p.spec)
+
+    expect(ctx.tasks.kill(id, undefined, 'no longer needed')).toBe('requested')
+    expect(p.cancels).toEqual(['no longer needed'])
+    expect(ctx.tasks.list()[0]).toMatchObject({ status: 'stopping', reported: true })
+
+    p.settle({ status: 'killed' })
+    await tick()
+    // The listener still fires (telemetry may care), but carries reported: true
+    // so the notice surface suppresses its redundant "finished".
+    expect(seen[0]).toMatchObject({ id, status: 'killed', reported: true })
+  })
+
+  it('reports an already-finished task instead of failing', async () => {
+    const ctx = await harness()
+    const p = producer()
+    const id = ctx.tasks.start(p.spec)
+    p.settle({ status: 'completed' })
+    await tick()
+    expect(ctx.tasks.kill(id)).toBe('already-finished')
+  })
+
+  it('propagates a throwing producer cancel and leaves the task untouched', async () => {
+    const ctx = await harness()
+    const seen: TaskSnapshot[] = []
+    ctx.tasks.onTaskDone(snapshot => void seen.push(snapshot))
+    let broken = true
+    let settle!: (outcome: TaskOutcome) => void
+    const id = ctx.tasks.start({
+      kind: 'bash',
+      label: 'flaky cancel',
+      run: () => ({
+        cancel() { if (broken) throw new Error('cancel boom') },
+        done: new Promise<TaskOutcome>((res) => { settle = res }),
+      }),
+    })
+    expect(() => ctx.tasks.kill(id)).toThrow('cancel boom')
+    // The failed kill mutated NOTHING: still running, notice not suppressed,
+    // and a later (successful) kill still works.
+    expect(ctx.tasks.get(id)).toMatchObject({ status: 'running', reported: false })
+    settle({ status: 'completed' })
+    await tick()
+    expect(seen[0]).toMatchObject({ id, reported: false }) // notice would still fire
+
+    broken = false
+    expect(ctx.tasks.kill(id)).toBe('already-finished')
+  })
+})
+
+describe('TaskService.wait', () => {
+  it('resolves with the terminal snapshot when the task settles, marked reported', async () => {
+    const ctx = await harness()
+    const seen: TaskSnapshot[] = []
+    ctx.tasks.onTaskDone(snapshot => void seen.push(snapshot))
+    const p = producer()
+    const id = ctx.tasks.start(p.spec)
+
+    const wait = ctx.tasks.wait(id, 5_000)
+    p.settle({ status: 'completed', detail: 'exit code: 0' })
+    expect(await wait).toMatchObject({ status: 'completed', reported: true })
+    // A waiting reader claims delivery before completion listeners inspect the snapshot.
+    expect(seen[0]).toMatchObject({ id, reported: true })
+  })
+
+  it('returns the live snapshot on timeout without marking reported', async () => {
+    const ctx = await harness()
+    const id = ctx.tasks.start(producer().spec)
+    expect(await ctx.tasks.wait(id, 5)).toMatchObject({ status: 'running', reported: false })
+  })
+
+  it('unregisters timed-out and aborted wait resolvers while the task remains live', async () => {
+    const ctx = await harness()
+    const id = ctx.tasks.start(producer().spec)
+
+    for (let index = 0; index < 3; index += 1) {
+      const wait = ctx.tasks.wait(id, 5)
+      expect(waitResolverCount(ctx, id)).toBe(1)
+      await expect(wait).resolves.toMatchObject({ status: 'running' })
+      expect(waitResolverCount(ctx, id)).toBe(0)
+    }
+
+    const controller = new AbortController()
+    const wait = ctx.tasks.wait(id, 5_000, undefined, controller.signal)
+    expect(waitResolverCount(ctx, id)).toBe(1)
+    controller.abort()
+    await expect(wait).rejects.toThrow('wait aborted')
+    expect(waitResolverCount(ctx, id)).toBe(0)
+    expect(ctx.tasks.get(id).status).toBe('running')
+  })
+
+  it('returns immediately for an already-finished task', async () => {
+    const ctx = await harness()
+    const p = producer()
+    const id = ctx.tasks.start(p.spec)
+    p.settle({ status: 'completed' })
+    await tick()
+    expect(await ctx.tasks.wait(id, 5_000)).toMatchObject({ status: 'completed', reported: true })
+  })
+
+  it('rejects a non-positive or non-finite timeout', async () => {
+    const ctx = await harness()
+    const id = ctx.tasks.start(producer().spec)
+    await expect(ctx.tasks.wait(id, 0)).rejects.toThrow('invalid wait timeout')
+    await expect(ctx.tasks.wait(id, Number.NaN)).rejects.toThrow('invalid wait timeout')
+  })
+
+  it('an aborted signal rejects the wait only — the task stays alive', async () => {
+    const ctx = await harness()
+    const id = ctx.tasks.start(producer().spec)
+
+    const controller = new AbortController()
+    const wait = ctx.tasks.wait(id, 5_000, undefined, controller.signal)
+    controller.abort()
+    await expect(wait).rejects.toThrow('wait aborted')
+    expect(ctx.tasks.list()[0]).toMatchObject({ status: 'running' })
+
+    const preAborted = new AbortController()
+    preAborted.abort()
+    await expect(ctx.tasks.wait(id, 5_000, undefined, preAborted.signal)).rejects.toThrow('wait aborted')
+  })
+
+  it('an abort racing settlement in the same tick does not swallow the notice', async () => {
+    const ctx = await harness()
+    const seen: TaskSnapshot[] = []
+    ctx.tasks.onTaskDone(snapshot => void seen.push(snapshot))
+    const p = producer()
+    const id = ctx.tasks.start(p.spec)
+
+    const controller = new AbortController()
+    const wait = ctx.tasks.wait(id, 5_000, undefined, controller.signal)
+    // Settlement is queued first, so abort must remove the waiter synchronously;
+    // otherwise settlement suppresses the notice for a reader that receives nothing.
+    p.settle({ status: 'completed', detail: 'exit code: 0' })
+    controller.abort()
+    await expect(wait).rejects.toThrow('wait aborted')
+    expect(seen).toHaveLength(1)
+    expect(seen[0]).toMatchObject({ id, status: 'completed', reported: false })
+  })
+
+  it('an abort landing after settlement still delivers the terminal snapshot it owes', async () => {
+    const ctx = await harness()
+    const controller = new AbortController()
+    const seen: TaskSnapshot[] = []
+    // The listener aborts after settlement has assigned delivery to this waiter
+    // but before its resolve microtask; the waiter must still receive the result.
+    ctx.tasks.onTaskDone((snapshot) => {
+      seen.push(snapshot)
+      controller.abort()
+    })
+    const p = producer()
+    const id = ctx.tasks.start(p.spec)
+
+    const wait = ctx.tasks.wait(id, 5_000, undefined, controller.signal)
+    p.settle({ status: 'completed', detail: 'exit code: 0' })
+    await expect(wait).resolves.toMatchObject({ status: 'completed', reported: true })
+    expect(seen[0]).toMatchObject({ id, reported: true }) // suppression stays honest: the wait delivered
+  })
+})
+
+describe('TaskService owner isolation', () => {
+  it('fences read/kill/wait to the owning session and keeps unowned tasks open', async () => {
+    const ctx = await harness()
+    const owner = stubAgent(ctx, 'owner')
+    ctx.agents.register(owner)
+    const other = stubAgent(ctx, 'other')
+
+    const owned = ctx.tasks.start(producer({ owner }).spec)
+    const open = ctx.tasks.start(producer().spec)
+
+    // The owner and the unowned task are reachable.
+    expect(ctx.tasks.read(owned, owner).snapshot.id).toBe(owned)
+    expect(ctx.tasks.read(open, other).snapshot.id).toBe(open)
+
+    // A different session and a no-agent caller are rejected.
+    expect(() => ctx.tasks.read(owned, other)).toThrow(`task ${owned} belongs to another session`)
+    expect(() => ctx.tasks.kill(owned, other)).toThrow('belongs to another session')
+    await expect(ctx.tasks.wait(owned, 10, other)).rejects.toThrow('belongs to another session')
+    expect(() => ctx.tasks.read(owned)).toThrow('belongs to another session')
+  })
+
+  it('list() shows only caller-owned plus unowned tasks', async () => {
+    const ctx = await harness()
+    const alice = stubAgent(ctx, 'alice')
+    const bob = stubAgent(ctx, 'bob')
+    ctx.agents.register(alice)
+    ctx.agents.register(bob)
+
+    const aliceTask = ctx.tasks.start(producer({ owner: alice }).spec)
+    const bobTask = ctx.tasks.start(producer({ owner: bob }).spec)
+    const openTask = ctx.tasks.start(producer({ kind: 'subagent' }).spec)
+
+    expect(ctx.tasks.list(alice).map(t => t.id)).toEqual([aliceTask, openTask])
+    expect(ctx.tasks.list(bob).map(t => t.id)).toEqual([bobTask, openTask])
+    expect(ctx.tasks.list().map(t => t.id)).toEqual([openTask])
+  })
+
+  it('rejects an owned registration when no agent registry is mounted', async () => {
+    const ctx = new Context()
+    await ctx.plugin(TaskService)
+    ctx.tasks.attachSurface('test-surface')
+    expect(() => ctx.tasks.start(producer({ owner: stubAgent(ctx, 'a') }).spec))
+      .toThrow('background task ownership requires the agent registry')
+    // The failed registration mutated nothing: no stored task, counter untouched.
+    expect(ctx.tasks.list()).toEqual([])
+    expect(ctx.tasks.start(producer().spec)).toBe('bash-1')
+  })
+
+  it('a failed owner-cleanup attach leaves the registry unchanged and does not poison the owner', async () => {
+    const ctx = await harness()
+    const ghost = stubAgent(ctx, 'ghost') // never registered in ctx.agents
+
+    // Exact-instance validation precedes registry mutation and cleanup attachment.
+    expect(() => ctx.tasks.start(producer({ owner: ghost }).spec))
+      .toThrow('is not the registered agent instance')
+    expect(ctx.tasks.list(ghost)).toEqual([])
+
+    // A later valid registration must still attach cleanup for the same object.
+    ctx.agents.register(ghost)
+    const cancels: (string | undefined)[] = []
+    let settle!: (outcome: TaskOutcome) => void
+    const id = ctx.tasks.start({
+      kind: 'bash',
+      label: 'after retry',
+      owner: ghost,
+      run: () => ({
+        cancel(reason) { cancels.push(reason); settle({ status: 'killed' }) },
+        done: new Promise<TaskOutcome>((res) => { settle = res }),
+      }),
+    })
+    expect(id).toBe('bash-1') // the failed attempt burned no counter
+    await disposeAgentScope(ghost)
+    expect(cancels).toEqual(['owner disposed'])
+    expect(ctx.tasks.list(ghost)).toEqual([])
+  })
+
+  it('rejects a stale owner instance after another agent reuses its id', async () => {
+    const ctx = await harness()
+    const staleOwner = stubAgent(ctx, 'owner', 'stale-session')
+    const unregisterStale = ctx.agents.register(staleOwner)
+    unregisterStale()
+
+    const currentOwner = stubAgent(ctx, 'owner', 'current-session')
+    ctx.agents.register(currentOwner)
+    const current = producer({ owner: currentOwner })
+    ctx.tasks.start(current.spec) // Attach the current owner's cleanup first.
+
+    const stale = producer({ owner: staleOwner })
+    const staleRun = vi.fn(() => stale.spec.run())
+    expect(() => ctx.tasks.start({ ...stale.spec, run: staleRun }))
+      .toThrow('is not the registered agent instance')
+    expect(staleRun).not.toHaveBeenCalled()
+    expect(ctx.tasks.list(staleOwner)).toEqual([])
+    expect(ctx.tasks.list(currentOwner)).toHaveLength(1)
+
+    current.settle({ status: 'completed' })
+    await tick()
+    await disposeAgentScope(currentOwner)
+  })
+})
+
+describe('TaskService owner cleanup', () => {
+  it('drains the owner: cancels live tasks, awaits settlement, drops snapshots', async () => {
+    const ctx = await harness()
+    const owner = stubAgent(ctx, 'owner')
+    ctx.agents.register(owner)
+
+    // The producer settles only when cancelled — models a child that stops on request.
+    let settle!: (outcome: TaskOutcome) => void
+    const cancels: (string | undefined)[] = []
+    ctx.tasks.start({
+      kind: 'subagent',
+      label: 'long research',
+      owner,
+      run: () => ({
+        cancel(reason) { cancels.push(reason); settle({ status: 'killed' }) },
+        done: new Promise<TaskOutcome>((res) => { settle = res }),
+      }),
+    })
+    const terminal = producer({ owner })
+    ctx.tasks.start(terminal.spec)
+    terminal.settle({ status: 'completed' })
+    await tick()
+
+    await disposeAgentScope(owner)
+    expect(cancels).toEqual(['owner disposed'])
+    // Snapshots dropped: nothing of the owner's remains, listing is empty.
+    expect(ctx.tasks.list(owner)).toEqual([])
+  })
+
+  it('attaches one cleanup per owner and drains all owned tasks with the scope', async () => {
+    const ctx = await harness()
+    const owner = stubAgent(ctx, 'owner')
+    ctx.agents.register(owner)
+
+    const first = producer({ owner })
+    const second = producer({ owner })
+    ctx.tasks.start(first.spec)
+    ctx.tasks.start(second.spec)
+    first.settle({ status: 'completed' })
+    second.settle({ status: 'completed' })
+    await tick()
+    expect(owner.ctx.fiber.getEffects().filter(effect => effect.label === 'tasks.ownerCleanup()')).toHaveLength(1)
+    await disposeAgentScope(owner)
+    expect(ctx.tasks.list(owner)).toEqual([])
+  })
+
+  it('does not let an old scope cleanup cancel a same-id/session replacement task', async () => {
+    const ctx = await harness()
+    const oldOwner = stubAgent(ctx, 'owner', 'shared-session')
+    const detachOld = ctx.agents.register(oldOwner)
+    const cancels: string[] = []
+
+    function start(owner: Agent, label: string): TaskId {
+      let settle!: (outcome: TaskOutcome) => void
+      return ctx.tasks.start({
+        kind: 'bash',
+        label,
+        owner,
+        run: () => ({
+          cancel() { cancels.push(label); settle({ status: 'killed' }) },
+          done: new Promise<TaskOutcome>((resolve) => { settle = resolve }),
+        }),
+      })
+    }
+
+    start(oldOwner, 'old task')
+    detachOld()
+    const replacement = stubAgent(ctx, 'owner', 'shared-session')
+    ctx.agents.register(replacement)
+    const replacementId = start(replacement, 'replacement task')
+
+    await disposeAgentScope(oldOwner)
+    expect(cancels).toEqual(['old task'])
+    expect(ctx.tasks.list(replacement).map(task => task.id)).toEqual([replacementId])
+
+    await disposeAgentScope(replacement)
+    expect(cancels).toEqual(['old task', 'replacement task'])
+  })
+
+  it('registers owner cleanup on the agent scope rather than the tasks fiber', async () => {
+    const ctx = new Context()
+    await ctx.plugin(AgentRegistry)
+    const tasksFiber = await ctx.plugin(TaskService)
+    ctx.tasks.attachSurface('test-surface')
+    const owner = stubAgent(ctx, 'owner')
+    ctx.agents.register(owner)
+    const ownerCleanupEffects = () => owner.ctx.fiber.getEffects()
+      .filter(effect => effect.label === 'tasks.ownerCleanup()')
+
+    const first = producer({ owner })
+    ctx.tasks.start(first.spec)
+    expect(ownerCleanupEffects()).toHaveLength(1)
+    first.settle({ status: 'completed' })
+    await tick()
+    expect(tasksFiber.getEffects().some(effect => effect.label === 'tasks.ownerCleanup()')).toBe(false)
+    await disposeAgentScope(owner)
+
+    // Only the owner registration is released; the long-lived tasks service
+    // and its own teardown effect remain active.
+    expect(ownerCleanupEffects()).toHaveLength(0)
+    expect(ctx.get('tasks')).toBeDefined()
+    expect(tasksFiber.getEffects().some(effect => effect.label === 'tasks teardown')).toBe(true)
+
+  })
+
+  it('force-fails a throwing teardown cancel without awaiting producer done, first outcome wins', async () => {
+    const ctx = await harness()
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    const owner = stubAgent(ctx, 'owner')
+    ctx.agents.register(owner)
+    const seen: TaskSnapshot[] = []
+    ctx.tasks.onTaskDone(snapshot => void seen.push(snapshot))
+
+    let settle!: (outcome: TaskOutcome) => void
+    ctx.tasks.start({
+      kind: 'bash',
+      label: 'broken producer',
+      owner,
+      run: () => ({
+        cancel() { throw new Error('cancel boom') },
+        done: new Promise<TaskOutcome>((res) => { settle = res }),
+      }),
+    })
+
+    const drain = disposeAgentScope(owner)
+    let drained = false
+    void drain.then(() => { drained = true })
+    await tick()
+    const drainedWithoutProducerDone = drained
+    if (!drainedWithoutProducerDone) {
+      // Release the producer if the assertion fails so the test can finish.
+      settle({ status: 'completed' })
+      await drain
+    } else {
+      // A late producer completion must not replace the failure or notify twice.
+      settle({ status: 'completed' })
+      await tick()
+    }
+
+    expect(drainedWithoutProducerDone).toBe(true)
+    expect(warn).toHaveBeenCalledWith(expect.stringContaining('work may be orphaned'))
+    expect(seen).toHaveLength(1)
+    expect(seen[0]?.status).toBe('failed')
+    expect(seen[0]?.detail).toContain('cancel threw during teardown')
+    expect(ctx.tasks.list(owner)).toEqual([])
+  })
+})
+
+describe('TaskService disposal', () => {
+  it('cancels live tasks, awaits settlement, and silences listeners', async () => {
+    const ctx = new Context()
+    await ctx.plugin(AgentRegistry)
+    const fiber = await ctx.plugin(TaskService)
+    const surface = await ctx.plugin(Object.assign((inner: Context) => {
+      inner.tasks.attachSurface('test-surface')
+    }, { inject: ['tasks'] }))
+    void surface
+
+    const seen: string[] = []
+    ctx.tasks.onTaskDone(snapshot => void seen.push(snapshot.id))
+    let settle!: (outcome: TaskOutcome) => void
+    const cancels: (string | undefined)[] = []
+    ctx.tasks.start({
+      kind: 'bash',
+      label: 'sleep 600',
+      run: () => ({
+        cancel(reason) { cancels.push(reason); settle({ status: 'killed' }) },
+        done: new Promise<TaskOutcome>((res) => { settle = res }),
+      }),
+    })
+
+    await fiber.dispose()
+    expect(cancels).toEqual(['tasks service disposed'])
+    // The teardown kill settles AFTER the listener registry closed: silent.
+    expect(seen).toEqual([])
+  })
+
+  it('force-fails a throwing cancel so service disposal does not await producer done', async () => {
+    const ctx = new Context()
+    await ctx.plugin(AgentRegistry)
+    const fiber = await ctx.plugin(TaskService)
+    ctx.tasks.attachSurface('test-surface')
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    const seen: TaskSnapshot[] = []
+    ctx.tasks.onTaskDone(snapshot => void seen.push(snapshot))
+
+    let settle!: (outcome: TaskOutcome) => void
+    ctx.tasks.start({
+      kind: 'bash',
+      label: 'broken service task',
+      run: () => ({
+        cancel() { throw new Error('service cancel boom') },
+        done: new Promise<TaskOutcome>((resolve) => { settle = resolve }),
+      }),
+    })
+
+    const disposal = fiber.dispose()
+    let disposed = false
+    void disposal.then(() => { disposed = true })
+    await tick()
+    const disposedWithoutProducerDone = disposed
+    if (!disposedWithoutProducerDone) {
+      // Release the producer if the assertion fails so the test can finish.
+      settle({ status: 'completed' })
+      await disposal
+    } else {
+      settle({ status: 'completed' })
+      await tick()
+    }
+
+    expect(disposedWithoutProducerDone).toBe(true)
+    expect(warn).toHaveBeenCalledWith(expect.stringContaining('work may be orphaned'))
+    expect(seen).toEqual([])
+  })
+
+  it('detaches owner effects from still-live agent scopes when the service unloads', async () => {
+    const ctx = new Context()
+    await ctx.plugin(AgentRegistry)
+    const tasksFiber = await ctx.plugin(TaskService)
+    ctx.tasks.attachSurface('test-surface')
+    const owner = stubAgent(ctx, 'owner')
+    ctx.agents.register(owner)
+    let settle!: (outcome: TaskOutcome) => void
+    ctx.tasks.start({
+      kind: 'bash',
+      label: 'owned work',
+      owner,
+      run: () => ({
+        cancel() { settle({ status: 'killed' }) },
+        done: new Promise<TaskOutcome>((resolve) => { settle = resolve }),
+      }),
+    })
+    const ownerEffects = () => owner.ctx.fiber.getEffects()
+      .filter(effect => effect.label === 'tasks.ownerCleanup()')
+    expect(ownerEffects()).toHaveLength(1)
+
+    await tasksFiber.dispose()
+
+    expect(ownerEffects()).toHaveLength(0)
+  })
+
+  it('detaching the last surface re-arms the register fence', async () => {
+    const ctx = new Context()
+    await ctx.plugin(TaskService)
+    const detachA1 = ctx.tasks.attachSurface('a')
+    const detachA2 = ctx.tasks.attachSurface('a') // duplicate name counts independently
+    const fiber = await ctx.plugin(Object.assign((inner: Context) => {
+      inner.tasks.attachSurface('b')
+    }, { inject: ['tasks'] }))
+
+    detachA1()
+    detachA1() // second call of the same disposer is a no-op
+    expect(() => ctx.tasks.start(producer().spec)).not.toThrow() // a ×1 + b remain
+    detachA2()
+    expect(() => ctx.tasks.start(producer().spec)).not.toThrow() // b remains
+    await fiber.dispose() // detaches b with its fiber (HMR safety)
+    expect(() => ctx.tasks.start(producer().spec)).toThrow('no control surface is attached')
+  })
+})
diff --git a/packages/tasks/tasks/tsconfig.json b/packages/tasks/tasks/tsconfig.json
new file mode 100644
index 0000000000..392d3f8d8a
--- /dev/null
+++ b/packages/tasks/tasks/tsconfig.json
@@ -0,0 +1,30 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": [
+    "src"
+  ],
+  "references": [
+    {
+      "path": "../../../vendor/cosmokit"
+    },
+    {
+      "path": "../../../vendor/cordis"
+    },
+    {
+      "path": "../../util/brand"
+    },
+    {
+      "path": "../../core/agent"
+    },
+    {
+      "path": "../../core/session"
+    },
+    {
+      "path": "../../util/timeout"
+    }
+  ]
+}
diff --git a/packages/tasks/tool-tasks/README.md b/packages/tasks/tool-tasks/README.md
new file mode 100644
index 0000000000..520a874df1
--- /dev/null
+++ b/packages/tasks/tool-tasks/README.md
@@ -0,0 +1,56 @@
+# @deepseek-ai/dsh-tool-tasks
+
+The model-facing control surface for `ctx.tasks`: three kind-independent tools, completion notices, and one background-work prompt section. Loading the plugin attaches the surface required by `ctx.tasks.start()`.
+
+## Tools
+
+- `task_output(task_id, wait?, timeout_ms?)` reads without blocking by default. Stream tasks return only the next delta; final-output tasks return their result after settlement. Every response ends with `[status: ...]`. `wait: true` waits up to the configured cap and leaves a still-running task alive on timeout.
+- `task_list()` returns caller-visible tasks as `<id> [<kind>] <status> — <label>`.
+- `task_kill(task_id, reason?)` requests cancellation immediately and forwards the logged reason. Terminal tasks return a non-consuming snapshot.
+
+All three use generic ACP cards: `read` for output and list, `execute` for kill.
+
+## Completion notices
+
+An unreported completion injects `background task <id> (<kind>: <label>) finished [status: ...]. Read its output with task_output.` into the exact owner's session. Injection is durable context for the next request, not a wake-up. A kill or terminal read/wait marks delivery reported and suppresses the redundant notice; owner-disposal races are contained.
+
+## Config
+
+| key | default | meaning |
+|---|---|---|
+| `waitTimeoutMs` | `30000` | wait used when `wait: true` omits `timeout_ms` |
+| `maxWaitTimeoutMs` | `600000` | cap for model-supplied waits |
+
+A default above the cap fails at load.
+
+## Model Experience
+
+### System prompt
+
+**What the model sees**: Every request in this plugin's registration scope contains this guidance. Agent-scoped tool filtering may hide the tools without removing the independently registered prompt section.
+
+**Token effect**: Small fixed input cost per request while active.
+
+#### Background-task guidance
+
+```markdown
+Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task's work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.
+```
+
+### Tool schemas
+
+**What the model sees**: The generated [`task_output`, `task_list`, and `task_kill` schemas](../../../docs/tool-catalog.md#deepseek-aidsh-tool-tasks) while this surface is visible.
+
+**Token effect**: Fixed schema cost on each request where the tools are visible.
+
+### Results and notices
+
+**What the model sees**: Reads return output or `(no new output)` followed by `[status: <status>]` and optional detail. An empty list returns `(no background tasks)`. Kill returns `requested cancellation of task <id>` or the existing terminal status. Unreported owned completion uses the notice above.
+
+**Token effect**: Results and notices remain in parent history until compaction. Stream reads do not repeat consumed output.
+
+## Known Limitations and Deferred Work
+
+- **Completion notices do not wake idle agents** — callers needing an immediate result must use `task_output`.
+- **Stream reads are single-consumer** — independent observers need another runtime API.
+- **Unowned tasks have no session fence** — external surfaces must supply caller policy or avoid them.
diff --git a/packages/tasks/tool-tasks/package.json b/packages/tasks/tool-tasks/package.json
new file mode 100644
index 0000000000..9e0b19d307
--- /dev/null
+++ b/packages/tasks/tool-tasks/package.json
@@ -0,0 +1,43 @@
+{
+  "name": "@deepseek-ai/dsh-tool-tasks",
+  "description": "Model-facing background task control tools (task_output, task_list, task_kill) over the ctx.tasks registry",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "@deepseek-ai/dsh-agent": "^0.0.1",
+    "@deepseek-ai/dsh-system-prompt": "^0.0.1",
+    "@deepseek-ai/dsh-tasks": "^0.0.1",
+    "@deepseek-ai/dsh-tools": "^0.0.1",
+    "cordis": "^4.0.0-rc.6"
+  },
+  "dependencies": {
+    "schemastery": "^3.18.0"
+  },
+  "devDependencies": {
+    "@deepseek-ai/dsh-agent": "workspace:^",
+    "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-session": "workspace:^",
+    "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-tasks": "workspace:^",
+    "@deepseek-ai/dsh-tools": "workspace:^",
+    "cordis": "^4.0.0-rc.6"
+  }
+}
diff --git a/packages/tasks/tool-tasks/src/index.ts b/packages/tasks/tool-tasks/src/index.ts
new file mode 100644
index 0000000000..2d556cada9
--- /dev/null
+++ b/packages/tasks/tool-tasks/src/index.ts
@@ -0,0 +1,148 @@
+/**
+ * Model-facing `task_output`, `task_list`, and `task_kill` tools over
+ * `ctx.tasks`. Loading the plugin attaches the control surface required by
+ * producers. It also injects unreported completions as durable context for the
+ * owner's next request; notices do not wake idle agents.
+ * @module @deepseek-ai/dsh-tool-tasks
+ */
+
+import type { Context } from 'cordis'
+import z from 'schemastery'
+import { defineTool } from '@deepseek-ai/dsh-tools'
+import type { GenericCallView } from '@deepseek-ai/dsh-tools'
+import { TaskId } from '@deepseek-ai/dsh-tasks'
+import type { TaskSnapshot } from '@deepseek-ai/dsh-tasks'
+import type {} from '@deepseek-ai/dsh-system-prompt'
+
+export const name = 'tool-tasks'
+export const inject = ['tools', 'tasks', 'systemPrompt']
+
+/** Configures bounded `task_output` waits. */
+export interface Config {
+  /** Wait duration applied when `task_output` sets `wait` without `timeout_ms` (default 30s). */
+  waitTimeoutMs?: number
+  /** Hard cap on any single wait; a larger model-supplied `timeout_ms` is clamped down to it (default 10min). */
+  maxWaitTimeoutMs?: number
+}
+
+export const Config: z<Config> = z.object({
+  waitTimeoutMs: z.number().min(1).default(30_000),
+  maxWaitTimeoutMs: z.number().min(1).default(600_000),
+})
+
+/**
+ * Render generic status with optional producer detail.
+ * @param snapshot - task state to render.
+ * @returns a bracketed status line.
+ */
+export function statusLine(snapshot: TaskSnapshot): string {
+  return snapshot.detail !== undefined
+    ? `[status: ${snapshot.status}, ${snapshot.detail}]`
+    : `[status: ${snapshot.status}]`
+}
+
+/** Validate the non-empty constraint that SchemaSpec cannot express. */
+function validateTaskId(value: string): TaskId {
+  if (value.length === 0) {
+    throw new Error(`invalid task_id: expected a non-empty string, got ${JSON.stringify(value)}`)
+  }
+  return TaskId(value)
+}
+
+/** Pending presentation shared by the three generic task controls. */
+function presentTaskCall(title: string, kind: 'read' | 'execute', rawInput?: string): GenericCallView {
+  return { card: 'generic', title, kind, ...rawInput !== undefined ? { rawInput } : {} }
+}
+
+export function apply(ctx: Context, config: Config): void {
+  const waitDefault = config.waitTimeoutMs ?? 30_000
+  const waitCap = config.maxWaitTimeoutMs ?? 600_000
+  if (waitDefault > waitCap) {
+    throw new Error(`tool-tasks: waitTimeoutMs (${waitDefault}) exceeds maxWaitTimeoutMs (${waitCap})`)
+  }
+
+  // Producers may start work only while a control surface is attached.
+  ctx.tasks.attachSurface('tool-tasks')
+
+  // Cross-call guidance follows the bash section and precedes product sections.
+  ctx.systemPrompt.section({
+    name: 'tool:tasks',
+    order: 106,
+    text: 'Track every background task id you start. You are notified in-session when a task finishes — do not busy-poll or sleep on one; keep working on independent steps and do not duplicate a running task\'s work. Before giving a final answer, collect every still-relevant task with task_output (set wait: true only when you are genuinely blocked on it), and task_kill tasks that stopped mattering.',
+  })
+
+  // Use the exact lifecycle owner; reusable ids could resolve to a replacement.
+  ctx.tasks.onTaskDone((snapshot, owner) => {
+    if (snapshot.reported || owner === undefined) return
+    try {
+      owner.inject(
+        [{ type: 'text', text: `background task ${snapshot.id} (${snapshot.kind}: ${snapshot.label}) finished ${statusLine(snapshot)}. Read its output with task_output.` }],
+        { source: { kind: 'plugin', plugin: 'tool-tasks' } },
+      )
+    } catch (error: unknown) {
+      // Disposal may win the race after settlement; other injection failures surface.
+      if (error instanceof Error && error.message.includes('is disposed')) return
+      throw error
+    }
+  })
+
+  ctx.tools.register(defineTool({
+    name: 'task_output',
+    description: 'Read a background task. Stream tasks return only output since the previous read; '
+      + 'final-output tasks return their result after settlement. Every response ends with '
+      + '`[status: ...]`. Reads are non-blocking unless `wait: true`, which waits up to the configured cap.',
+    // A timed-out wait returns task state rather than a TOOL_TIMEOUT error, so
+    // this tool owns its deadline instead of using ToolDefinition.timeoutMs.
+    parameters: {
+      task_id: { type: 'string', required: true, description: 'Task id returned by the tool that started the background work.' },
+      wait: { type: 'boolean', description: 'Block until the task reaches a terminal status or the timeout expires. A timed-out wait returns [status: running] and leaves the task alive.' },
+      timeout_ms: { type: 'number', description: 'Max wait in milliseconds (only meaningful with wait: true). Defaults to the configured wait timeout; capped by the configured maximum.' },
+    },
+    async execute(args, exec) {
+      const id = validateTaskId(args.task_id)
+      if (args.wait === true) {
+        const timeout = Math.min(args.timeout_ms ?? waitDefault, waitCap)
+        await ctx.tasks.wait(id, timeout, exec.agent, exec.signal)
+      }
+      const read = ctx.tasks.read(id, exec.agent)
+      const body = read.text.length > 0 ? read.text : '(no new output)'
+      const separator = body.endsWith('\n') ? '' : '\n'
+      return [{ type: 'text', text: `${body}${separator}${statusLine(read.snapshot)}` }]
+    },
+    presentCall: args => presentTaskCall(`Read output from background task ${args.task_id}`, 'read', args.task_id),
+  }))
+
+  ctx.tools.register(defineTool({
+    name: 'task_list',
+    description: 'List your background tasks (running and finished) with their ids, kinds, and statuses.',
+    parameters: {},
+    execute(_args, exec) {
+      const tasks = ctx.tasks.list(exec.agent)
+      const text = tasks.length === 0
+        ? '(no background tasks)'
+        : tasks.map(t => `${t.id} [${t.kind}] ${t.status} — ${t.label}`).join('\n')
+      return Promise.resolve([{ type: 'text', text }])
+    },
+    presentCall: () => presentTaskCall('List background tasks', 'read'),
+  }))
+
+  ctx.tools.register(defineTool({
+    name: 'task_kill',
+    description: 'Request cancellation of a running background task by task id. Returns immediately; the task settles as killed once its work actually stops.',
+    parameters: {
+      task_id: { type: 'string', required: true, description: 'Task id returned by the tool that started the background work.' },
+      reason: { type: 'string', description: 'Optional short reason, recorded in the log and forwarded to the task.' },
+    },
+    execute(args, exec) {
+      const id = validateTaskId(args.task_id)
+      const result = ctx.tasks.kill(id, exec.agent, args.reason)
+      if (result === 'already-finished') {
+        // A snapshot describes terminal state without consuming pending output.
+        const snapshot = ctx.tasks.get(id, exec.agent)
+        return Promise.resolve([{ type: 'text', text: `task ${id} had already finished ${statusLine(snapshot)}` }])
+      }
+      return Promise.resolve([{ type: 'text', text: `requested cancellation of task ${id}` }])
+    },
+    presentCall: args => presentTaskCall(`Kill background task ${args.task_id}`, 'execute', args.task_id),
+  }))
+}
diff --git a/packages/tasks/tool-tasks/tests/tool-tasks.spec.ts b/packages/tasks/tool-tasks/tests/tool-tasks.spec.ts
new file mode 100644
index 0000000000..1b5e221a3c
--- /dev/null
+++ b/packages/tasks/tool-tasks/tests/tool-tasks.spec.ts
@@ -0,0 +1,336 @@
+import { describe, expect, it, vi } from 'vitest'
+import { Context } from 'cordis'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import ToolRegistry from '@deepseek-ai/dsh-tools'
+import AgentRegistry from '@deepseek-ai/dsh-agent'
+import type { Agent } from '@deepseek-ai/dsh-agent'
+import TaskService from '@deepseek-ai/dsh-tasks'
+import type { TaskHooks, TaskOutcome, TaskSnapshot, TaskStart } from '@deepseek-ai/dsh-tasks'
+import * as ToolTasks from '@deepseek-ai/dsh-tool-tasks'
+import { statusLine } from '@deepseek-ai/dsh-tool-tasks'
+
+const agentRegistryDisposers = new WeakMap<Agent, () => void>()
+
+async function setup(config: ToolTasks.Config = {}) {
+  const ctx = new Context()
+  await ctx.plugin(SystemPrompt)
+  await ctx.plugin(ToolRegistry)
+  const agentsFiber = await ctx.plugin(AgentRegistry)
+  await ctx.plugin(TaskService)
+  const toolsFiber = await ctx.plugin(ToolTasks, config)
+  return { ctx, agentsFiber, toolsFiber }
+}
+
+/**
+ * A fake agent whose session token is `sessionId`, registered in `ctx.agents`.
+ * The agent id is deliberately different so session authorization and exact
+ * lifecycle ownership cannot be confused in tests.
+ */
+function fakeAgent(ctx: Context, sessionId: string, inject: (...args: unknown[]) => void = () => {}): Agent {
+  const scopeFiber = ctx.plugin(() => {})
+  const agent = {
+    id: `agent-${sessionId}`,
+    ctx: scopeFiber.ctx,
+    inject,
+    session: { header: { version: 0, id: sessionId, createdAt: 0 } },
+  } as unknown as Agent
+  agentRegistryDisposers.set(agent, ctx.agents.register(agent))
+  return agent
+}
+
+function detachAgent(agent: Agent): void {
+  const dispose = agentRegistryDisposers.get(agent)
+  if (dispose === undefined) throw new Error(`missing registry disposer for agent "${agent.id}"`)
+  dispose()
+}
+
+/** A controllable producer start-spec (settle `done` on demand, record cancels). */
+function producer(overrides: Partial<Omit<TaskStart, 'run'> & TaskHooks> = {}) {
+  let settle!: (outcome: TaskOutcome) => void
+  const cancels: (string | undefined)[] = []
+  const { kind = 'bash', label = 'sleep 60', owner, ...hookOverrides } = overrides
+  const hooks: TaskHooks = {
+    cancel(reason) { cancels.push(reason) },
+    done: new Promise<TaskOutcome>((res) => { settle = res }),
+    ...hookOverrides,
+  }
+  const spec: TaskStart = { kind, label, ...owner !== undefined ? { owner } : {}, run: () => hooks }
+  return { spec, settle, cancels }
+}
+
+let callCounter = 0
+function call(ctx: Context, name: string, args: unknown, agent?: Agent) {
+  return ctx.tools.execute({ callId: CallId(`call-${++callCounter}`), name, arguments: args, ...agent ? { agent } : {} })
+}
+
+function text(result: { content: { type: string; text?: string }[] }): string {
+  return result.content.filter(block => block.type === 'text').map(block => block.text).join('')
+}
+
+const tick = () => new Promise<void>(r => setTimeout(r, 0))
+
+describe('tool-tasks setup', () => {
+  it('attaches the control surface on load and detaches it with the fiber', async () => {
+    const { ctx, toolsFiber } = await setup()
+    expect(() => ctx.tasks.start(producer().spec)).not.toThrow()
+    await toolsFiber.dispose()
+    expect(() => ctx.tasks.start(producer().spec)).toThrow('no control surface is attached')
+  })
+
+  it('rejects a config whose default wait exceeds the cap', async () => {
+    const ctx = new Context()
+    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(TaskService)
+    await expect(ctx.plugin(ToolTasks, { waitTimeoutMs: 100, maxWaitTimeoutMs: 50 }))
+      .rejects.toThrow('waitTimeoutMs (100) exceeds maxWaitTimeoutMs (50)')
+  })
+
+  it('renders status lines with and without producer detail', () => {
+    const base = { id: 'bash-1', kind: 'bash', label: 'x', startedAt: 0, reported: false } as unknown as TaskSnapshot
+    expect(statusLine({ ...base, status: 'running' })).toBe('[status: running]')
+    expect(statusLine({ ...base, status: 'completed', detail: 'exit code: 0' })).toBe('[status: completed, exit code: 0]')
+  })
+
+  it('applies the built-in wait bounds when apply() receives a bare config', async () => {
+    // Bypasses the schemastery defaults on purpose: apply() must stand on its
+    // own `??` fallbacks when embedded programmatically without the schema.
+    const ctx = new Context()
+    await ctx.plugin(SystemPrompt)
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(TaskService)
+    ToolTasks.apply(ctx, {})
+    expect(ctx.tools.get('task_output')).toBeDefined()
+    expect(() => ctx.tasks.start(producer().spec)).not.toThrow()
+  })
+})
+
+describe('task_output', () => {
+  it('reads a consuming delta with a trailing status line', async () => {
+    const { ctx } = await setup()
+    const chunks = ['line one\n', '']
+    ctx.tasks.start(producer({ readOutput: () => chunks.shift() ?? '' }).spec)
+
+    // A body already ending in a newline gets no doubled separator.
+    expect(text(await call(ctx, 'task_output', { task_id: 'bash-1' }))).toBe('line one\n[status: running]')
+    expect(text(await call(ctx, 'task_output', { task_id: 'bash-1' }))).toBe('(no new output)\n[status: running]')
+  })
+
+  it('returns the final output of a settled final-output task', async () => {
+    const { ctx } = await setup()
+    const p = producer({ kind: 'subagent', label: 'research' })
+    ctx.tasks.start(p.spec)
+    expect(text(await call(ctx, 'task_output', { task_id: 'subagent-1' }))).toBe('(no new output)\n[status: running]')
+
+    p.settle({ status: 'completed', detail: 'completed', output: 'the answer' })
+    await tick()
+    expect(text(await call(ctx, 'task_output', { task_id: 'subagent-1' }))).toBe('the answer\n[status: completed, completed]')
+  })
+
+  it('wait: true blocks until settlement and reports the terminal state', async () => {
+    const { ctx } = await setup()
+    const p = producer({ kind: 'subagent', label: 'research' })
+    ctx.tasks.start(p.spec)
+
+    const pending = call(ctx, 'task_output', { task_id: 'subagent-1', wait: true })
+    p.settle({ status: 'completed', output: 'done deal' })
+    expect(text(await pending)).toBe('done deal\n[status: completed]')
+  })
+
+  it('wait: true times out against the configured cap and leaves the task alive', async () => {
+    const { ctx } = await setup({ waitTimeoutMs: 10, maxWaitTimeoutMs: 20 })
+    ctx.tasks.start(producer().spec)
+
+    // A model-supplied timeout far above the cap is clamped: this returns
+    // promptly (≤ the 20ms cap), not after ten minutes.
+    const result = await call(ctx, 'task_output', { task_id: 'bash-1', wait: true, timeout_ms: 600_000 })
+    expect(text(result)).toBe('(no new output)\n[status: running]')
+  })
+
+  it('rejects an empty or unknown task id as an errored result', async () => {
+    const { ctx } = await setup()
+    expect((await call(ctx, 'task_output', { task_id: '' })).isError).toBe(true)
+    const unknown = await call(ctx, 'task_output', { task_id: 'bash-99' })
+    expect(unknown.isError).toBe(true)
+    expect(text(unknown)).toContain('unknown task bash-99')
+  })
+})
+
+describe('task_list', () => {
+  it('lists caller-visible tasks and renders the empty case', async () => {
+    const { ctx } = await setup()
+    expect(text(await call(ctx, 'task_list', {}))).toBe('(no background tasks)')
+
+    const alice = fakeAgent(ctx, 'sess-alice')
+    ctx.tasks.start(producer({ owner: alice, label: 'pnpm test' }).spec)
+    ctx.tasks.start(producer({ kind: 'subagent', label: 'open research' }).spec)
+    const p = producer({ owner: alice, label: 'build' })
+    ctx.tasks.start(p.spec)
+    p.settle({ status: 'completed', detail: 'exit code: 0' })
+    await tick()
+
+    expect(text(await call(ctx, 'task_list', {}, alice))).toBe([
+      'bash-1 [bash] running — pnpm test',
+      'subagent-1 [subagent] running — open research',
+      'bash-2 [bash] completed — build',
+    ].join('\n'))
+    // A different caller sees only the unowned task.
+    const bob = fakeAgent(ctx, 'sess-bob')
+    expect(text(await call(ctx, 'task_list', {}, bob))).toBe('subagent-1 [subagent] running — open research')
+  })
+})
+
+describe('task_kill', () => {
+  it('requests cancellation with the forwarded reason', async () => {
+    const { ctx } = await setup()
+    const p = producer()
+    ctx.tasks.start(p.spec)
+
+    const result = await call(ctx, 'task_kill', { task_id: 'bash-1', reason: 'superseded' })
+    expect(text(result)).toBe('requested cancellation of task bash-1')
+    expect(p.cancels).toEqual(['superseded'])
+  })
+
+  it('reports an already-finished task without consuming its pending delta', async () => {
+    const { ctx } = await setup()
+    let delta = 'unread tail'
+    const p = producer({ readOutput: () => { const d = delta; delta = ''; return d } })
+    ctx.tasks.start(p.spec)
+    p.settle({ status: 'completed', detail: 'exit code: 0' })
+    await tick()
+
+    expect(text(await call(ctx, 'task_kill', { task_id: 'bash-1' })))
+      .toBe('task bash-1 had already finished [status: completed, exit code: 0]')
+    // The kill described the task via a non-consuming snapshot: the delta is intact.
+    expect(text(await call(ctx, 'task_output', { task_id: 'bash-1' }))).toBe('unread tail\n[status: completed, exit code: 0]')
+  })
+
+  it('rejects an empty task id as an errored result', async () => {
+    const { ctx } = await setup()
+    expect((await call(ctx, 'task_kill', { task_id: '' })).isError).toBe(true)
+  })
+})
+
+describe('tool-owned UI presentation (presentCall)', () => {
+  it('renders generic cards for all three control tools', async () => {
+    const { ctx } = await setup()
+    expect(ctx.tools.get('task_output')?.presentCall?.({ task_id: 'bash-1' }))
+      .toEqual({ card: 'generic', title: 'Read output from background task bash-1', kind: 'read', rawInput: 'bash-1' })
+    expect(ctx.tools.get('task_list')?.presentCall?.({}))
+      .toEqual({ card: 'generic', title: 'List background tasks', kind: 'read' })
+    expect(ctx.tools.get('task_kill')?.presentCall?.({ task_id: 'subagent-2' }))
+      .toEqual({ card: 'generic', title: 'Kill background task subagent-2', kind: 'execute', rawInput: 'subagent-2' })
+  })
+})
+
+describe('completion notices', () => {
+  it('injects a notice into the owning agent when an unreported task settles', async () => {
+    const { ctx } = await setup()
+    const inject = vi.fn()
+    const owner = fakeAgent(ctx, 'sess-1', inject)
+    const p = producer({ owner, label: 'pnpm test' })
+    ctx.tasks.start(p.spec)
+
+    p.settle({ status: 'completed', detail: 'exit code: 0' })
+    await tick()
+    expect(inject).toHaveBeenCalledTimes(1)
+    expect(inject).toHaveBeenCalledWith(
+      [{ type: 'text', text: 'background task bash-1 (bash: pnpm test) finished [status: completed, exit code: 0]. Read its output with task_output.' }],
+      { source: { kind: 'plugin', plugin: 'tool-tasks' } },
+    )
+  })
+
+  it('suppresses the notice for a task the model already killed', async () => {
+    const { ctx } = await setup()
+    const inject = vi.fn()
+    const owner = fakeAgent(ctx, 'sess-1', inject)
+    const p = producer({ owner })
+    ctx.tasks.start(p.spec)
+
+    await call(ctx, 'task_kill', { task_id: 'bash-1' }, owner)
+    p.settle({ status: 'killed' })
+    await tick()
+    expect(inject).not.toHaveBeenCalled()
+  })
+
+  it('suppresses the notice when a wait returned the terminal state', async () => {
+    const { ctx } = await setup()
+    const inject = vi.fn()
+    const owner = fakeAgent(ctx, 'sess-1', inject)
+    const p = producer({ owner, kind: 'subagent' })
+    ctx.tasks.start(p.spec)
+
+    const pending = call(ctx, 'task_output', { task_id: 'subagent-1', wait: true }, owner)
+    p.settle({ status: 'completed', output: 'answer' })
+    expect(text(await pending)).toContain('answer')
+    expect(inject).not.toHaveBeenCalled()
+  })
+
+  it('drops the notice for unowned tasks and for a disposed owner (benign race)', async () => {
+    const { ctx } = await setup()
+    // Unowned: settles with nobody to notify — nothing throws.
+    const unowned = producer()
+    ctx.tasks.start(unowned.spec)
+    unowned.settle({ status: 'completed' })
+    await tick()
+
+    // Disposed owner: inject throws the disposed message — contained.
+    const inject = vi.fn(() => { throw new Error('agent "agent-sess-1" is disposed') })
+    const owner = fakeAgent(ctx, 'sess-1', inject)
+    const p = producer({ owner })
+    ctx.tasks.start(p.spec)
+    p.settle({ status: 'completed' })
+    await tick()
+    expect(inject).toHaveBeenCalledTimes(1)
+  })
+
+  it('does not route an old owner completion notice to a same-session replacement', async () => {
+    const { ctx } = await setup()
+    const oldInject = vi.fn(() => { throw new Error('agent "agent-shared" is disposed') })
+    const oldOwner = fakeAgent(ctx, 'shared', oldInject)
+    const p = producer({ owner: oldOwner })
+    ctx.tasks.start(p.spec)
+
+    detachAgent(oldOwner)
+    const replacementInject = vi.fn()
+    fakeAgent(ctx, 'shared', replacementInject)
+    p.settle({ status: 'completed' })
+    await tick()
+
+    expect(oldInject).toHaveBeenCalledTimes(1)
+    expect(replacementInject).not.toHaveBeenCalled()
+  })
+
+  it('propagates a non-disposed inject failure (a real bug must surface)', async () => {
+    const { ctx } = await setup()
+    const warn = vi.spyOn(ctx.logger, 'warn').mockImplementation(() => {})
+    const owner = fakeAgent(ctx, 'sess-1', () => { throw new Error('unexpected inject bug') })
+    const p = producer({ owner })
+    ctx.tasks.start(p.spec)
+    p.settle({ status: 'completed' })
+    await tick()
+    // The throw escapes the notice listener and is contained (logged) by the
+    // registry's per-listener containment — visible, not swallowed.
+    expect(warn).toHaveBeenCalledWith(expect.stringContaining('unexpected inject bug'))
+  })
+
+  it('keeps using the exact owner after the agent registry is gone', async () => {
+    const { ctx, agentsFiber } = await setup()
+    const inject = vi.fn()
+    const owner = fakeAgent(ctx, 'sess-1', inject)
+
+    // Settlement must not depend on a later registry lookup: the exact owner
+    // supplied at start remains the destination while its own scope is live.
+    const p1 = producer({ owner })
+    ctx.tasks.start(p1.spec)
+    const p2 = producer({ owner })
+    ctx.tasks.start(p2.spec)
+
+    await agentsFiber.dispose()
+    p1.settle({ status: 'completed' })
+    p2.settle({ status: 'failed' })
+    await tick()
+    expect(inject).toHaveBeenCalledTimes(2)
+  })
+})
diff --git a/packages/tasks/tool-tasks/tsconfig.json b/packages/tasks/tool-tasks/tsconfig.json
new file mode 100644
index 0000000000..9e5411df25
--- /dev/null
+++ b/packages/tasks/tool-tasks/tsconfig.json
@@ -0,0 +1,33 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": [
+    "src"
+  ],
+  "references": [
+    {
+      "path": "../../../vendor/cosmokit"
+    },
+    {
+      "path": "../../../vendor/cordis"
+    },
+    {
+      "path": "../../../vendor/schemastery"
+    },
+    {
+      "path": "../../core/agent"
+    },
+    {
+      "path": "../../core/system-prompt"
+    },
+    {
+      "path": "../../core/tools"
+    },
+    {
+      "path": "../tasks"
+    }
+  ]
+}
diff --git a/packages/todo/tool-todo/package.json b/packages/todo/tool-todo/package.json
index 9ff69d7c76..bab0f1230c 100644
--- a/packages/todo/tool-todo/package.json
+++ b/packages/todo/tool-todo/package.json
@@ -30,6 +30,7 @@
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
diff --git a/packages/todo/tool-todo/tests/integration.spec.ts b/packages/todo/tool-todo/tests/integration.spec.ts
index 739367a699..97c24a822e 100644
--- a/packages/todo/tool-todo/tests/integration.spec.ts
+++ b/packages/todo/tool-todo/tests/integration.spec.ts
@@ -1,12 +1,9 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
 import type { SessionEvent } from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import { AgentId } from '@deepseek-ai/dsh-agent'
 import AgentLoop, { ReactLoopAgent } from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as ToolTodo from '@deepseek-ai/dsh-tool-todo'
 import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent-loop/tests/mock-adapter.ts'
 
@@ -18,11 +15,7 @@ import { MockAdapter, textResponse, toolCallResponse } from '../../../core/agent
  */
 async function harness(adapter: MockAdapter): Promise<Context> {
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(ToolTodo)
   ctx.llm.registerAdapter(['mock'], adapter)
@@ -64,7 +57,7 @@ describe('todo_write tool through the agent loop', () => {
       textResponse('Plan recorded.'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('it-todo'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('it-todo'), { provider: 'mock', model: 'mock' })
 
     agent.send([{ type: 'text', text: 'plan a two-step task' }])
     await waitForIdle(ctx, agent)
@@ -92,7 +85,7 @@ describe('todo_write tool through the agent loop', () => {
       textResponse('Done planning.'),
     ])
     const ctx = await harness(adapter)
-    const agent = ctx.agentLoop.create(AgentId('it-todo-2'), { model: 'mock' })
+    const agent = ctx.agentLoop.create(AgentId('it-todo-2'), { provider: 'mock', model: 'mock' })
 
     agent.send([{ type: 'text', text: 'plan then update' }])
     await waitForIdle(ctx, agent)
diff --git a/packages/ui/acp/README.md b/packages/ui/acp/README.md
index dd705e8d8d..300cd69023 100644
--- a/packages/ui/acp/README.md
+++ b/packages/ui/acp/README.md
@@ -8,13 +8,14 @@ It is a **client-driver / UI plugin**, the structured analogue of the readline `
 
 `apply(ctx, config)` — wires an `AgentSideConnection` (from `@agentclientprotocol/sdk`) to `process.stdin`/`process.stdout` and implements the ACP `Agent` method surface.
 
-The plugin injects `agents`, `sessions`, `sessionPersistence`, `tools`, and `userInteraction`, never the concrete loop. Persistence backs `session/load`; tool definitions own presentation; user interaction maps agent questions to ACP forms.
+The plugin injects `agents`, `sessions`, `sessionPersistence`, `tools`, `userInteraction`, `llm`, and `systemPrompt`, never the concrete loop. Persistence backs `session/load`; the LLM catalog backs model selection; prompt assembly keeps model variables aligned with routing; tool definitions own presentation; user interaction maps agent questions to ACP forms.
 
 ### Config
 
 | Key | Default | Meaning |
 |---|---|---|
-| `model` | — | Model name for created agents (must have a registered adapter). |
+| `provider` | — | Initial provider route for created agents (must have a registered adapter). |
+| `model` | — | Initial model id for created agents. |
 
 (No persona key: `dsh-system-prompt`'s own `persona` config supplies the global default section, so ACP-created agents render it without the bridge carrying prompt text. An agent-scoped same-name section may still shadow that default.)
 
@@ -32,7 +33,7 @@ The `initialize` handshake reports a fixed server identity (`agentInfo: { name:
 | `session/update` | `session/event` | streams user replay, assistant text/reasoning, and tool render intents |
 | `elicitation/create` | `ctx.userInteraction.ask()` | maps `ask_user_question` questions to ACP form elicitations; option descriptions are shown in enum titles, `multi_select` uses ACP array enums, optionless requests use a required `custom` field, and a non-empty custom answer overrides any selected choice |
 | `session/request_permission` | `approval/request` listener | answers one-shot allow/reject requests for bridge-owned calls; foreign or call-less requests delegate and fail closed if unanswered — see "Permission prompts" |
-| `session/set_config_option` | `ctx.permission.set()` | per-session permission-preset switching over [session config options](https://agentclientprotocol.com/protocol/session-config-options) — see "Session config options" |
+| `session/set_config_option` | agent-scoped request target / `ctx.permission.set()` | per-session provider+model and permission-preset switching over [session config options](https://agentclientprotocol.com/protocol/session-config-options) — see "Session config options" |
 
 ## Multi-session
 
@@ -40,9 +41,11 @@ Forward and reverse indexes route every event, prompt, cancel, and approval to o
 
 ## Session config options
 
-When `ctx.permission` is composed, the bridge advertises one `permission` select in `session/new` and `session/load`. Options come from the deployment's preset table; the current value comes from the session fold, with switch-away-only `custom` for unmatched knobs. `session/set_config_option` accepts advertised presets and writes both sandbox-mode and approval-policy events through `PermissionService.set()`. Open-turn switches append immediately; idle switches overlay responses and anchor at the next `agent/prompt-submit`, before request assembly. A crash before anchoring restores the durable fold. See the [sandbox RFC](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md), [`dsh-permission`](../permission/README.md), and [protocol matrix](acp-feature-support.md#6-session-config-options).
+The bridge advertises a `model`-category select in `session/new` and `session/load` when the session has a complete target whose provider is registered. Values encode the complete provider/model pair, are grouped by provider when more than one group is available, and come from `ctx.llm.listProviders()` / `listModels()`. The configured or last-requested model is added when absent because catalogs are advisory and private adapters may accept unlisted ids. A selection changes only that ACP session. Agent-scoped prompt assembly snapshots the selected pair for one step, supplies matching `{{provider}}` / `{{model}}` variables, and the `agent/request` waterfall applies the same pair; a concurrent selection therefore takes effect on the next step instead of splitting prompt text from routing. The resulting request header is the durable record restored by `session/load`; a selection never used by a request remains in-memory only.
 
-Background bash tasks use the session id as an opaque owner token, so one session cannot inspect or stop another's task. That contract belongs to [`dsh-tool-bash`](../../bash/tool-bash/).
+When `ctx.permission` is composed, the bridge also advertises a `permission` select. Options come from the deployment's preset table; the current value comes from the session fold, with switch-away-only `custom` for unmatched knobs. `session/set_config_option` accepts advertised presets and writes both sandbox-mode and approval-policy events through `PermissionService.set()`. Open-turn switches append immediately; idle switches overlay responses and anchor at the next `agent/prompt-submit`, before request assembly. A crash before anchoring restores the durable fold. See the [model-catalog RFC](../../../docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.md), [sandbox RFC](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md), [`dsh-permission`](../permission/README.md), and [protocol matrix](acp-feature-support.md#6-session-modes--config-options--models).
+
+The shared [`ctx.tasks` runtime](../../tasks/tasks/) fences access to predictable task ids by the owning session; ACP sessions therefore cannot read or stop one another's background work.
 
 ## Per-session cwd
 
@@ -107,6 +110,12 @@ The JSON-RPC frames go on stdout, so this plugin MUST run in an example that loa
 
 **Token effect**: Zero direct tokens from the ACP option or the log-only `permission/preset` event. Downstream cost is limited to the owning plugins' policy prompt, conditional retained change notice, and any changed tool outcome.
 
+### Model switches
+
+**What the model sees**: The ACP selector itself emits no message. The selected provider/model pair supplies the next step's `{{provider}}` / `{{model}}` prompt variables and request routing together; all other call-config fields continue through the `agent/request` waterfall unchanged.
+
+**Token effect**: The selector adds no direct tokens. A changed model may tokenize the same retained prompt/history differently, and any persona text that interpolates provider or model changes accordingly.
+
 ### Loaded sessions
 
 **What the model sees**: `session/load` resumes the persisted log, after which the loop sends its reconstructed history and request header. Replaying that log to the editor is not an extra model message.
@@ -117,6 +126,5 @@ The JSON-RPC frames go on stdout, so this plugin MUST run in an example that loa
 
 - **`additionalDirectories`** — rejected. A session operates in its single `cwd` (see Per-session cwd); widening the tool/filesystem scope to extra roots is a separate sandbox concern, not yet implemented.
 - **Prompt content is `text` + `resource_link` only** — image, audio, and embedded-resource blocks are rejected, as is a non-empty `mcpServers` list at `session/new`.
-- **One configured `model` for every created session** — per-session model selection has no config or protocol surface here yet.
 - **Terminal cards render completed output** — live incremental streaming and command classification are named follow-ups of [the terminal-rendering RFC](../../../docs/rfc/implemented/feature/2026-06-18-acp-terminal-and-tool-rendering.md).
 - **Permission answers are one-shot only** — the bridge offers `allow_once` / `reject_once`; durable `allow_always` grants and their storage/revocation policy remain deferred to the approval seam.
diff --git a/packages/ui/acp/acp-feature-support.md b/packages/ui/acp/acp-feature-support.md
index 121af7cc2f..34e3caae8a 100644
--- a/packages/ui/acp/acp-feature-support.md
+++ b/packages/ui/acp/acp-feature-support.md
@@ -10,7 +10,7 @@ Legend: ✅ supported · ⚠️ partial / fallback · ❌ not yet · — n/a. Th
 
 ## At a glance
 
-The bridge implements the **core prompt-turn loop** for N concurrent sessions: initialize, session new/load, prompt, cancel, streamed assistant/thought chunks, tool-call rendering (including Zed terminal cards), resumable session replay, one-shot permission prompts, and per-session permission presets. The largest **unbuilt** areas are **MCP passthrough**, runtime model selection, **slash commands**, and **agent plans**, plus the client **filesystem** and **terminal** method families (which the adapters mostly do NOT drive either — see rows 43-49). See [Gap summary](#gap-summary).
+The bridge implements the **core prompt-turn loop** for N concurrent sessions: initialize, session new/load, prompt, cancel, streamed assistant/thought chunks, tool-call rendering (including Zed terminal cards), resumable session replay, one-shot permission prompts, per-session model selection, and permission presets. The largest **unbuilt** areas are **MCP passthrough**, **slash commands**, and **agent plans**, plus the client **filesystem** and **terminal** method families (which the adapters mostly do NOT drive either — see rows 43-49). See [Gap summary](#gap-summary).
 
 ## 1. Agent methods (client → agent)
 
@@ -26,8 +26,8 @@ The bridge implements the **core prompt-turn loop** for N concurrent sessions: i
 | `session/prompt` | S | ✅ | ✅ | ✅ | Maps to `agent.send`; one in-flight prompt per session; settles on the owning turn's end. |
 | `session/cancel` | S | ✅ | ✅ | ✅ | Queue-aware `agent.cancel`; settles the in-flight prompt `cancelled`, scoped to the one session. |
 | `session/set_mode` | S | ❌ | ✅ | ✅ | Session modes deliberately skipped: config options are the spec's replacement and modes are slated for removal in ACP v2 (see [§6](#6-session-modes--config-options--models)). |
-| `session/set_config_option` | S | ✅ | ✅ | ✅ | One `permission` select when `ctx.permission` is composed; values come from the deployment preset table, a switch writes its preset event through to both knob events, and the response carries the complete refreshed state ([sandbox RFC § Per-session mode switching](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md)). |
-| model selection | S | ❌ | ✅ | ✅ | No distinct stable `session/set_model` — model is the `model`-category `session/set_config_option`. The bridge fixes the model per-bridge via config; no runtime switch. Codex still uses the legacy `unstable_setSessionModel` ext method. |
+| `session/set_config_option` | S | ✅ | ✅ | ✅ | A provider/model select is present for a complete registered target; one `permission` select is added when `ctx.permission` is composed. Every response carries the complete refreshed state. |
+| model selection | S | ✅ | ✅ | ✅ | No distinct stable `session/set_model` — model is the `model`-category `session/set_config_option`. Values preserve the provider/model pair, catalogs come from `ctx.llm`, selection is per session, and `session/load` restores the last requested pair. Codex also supports the legacy `unstable_setSessionModel` ext method. |
 | `session/list` | S | ❌ | ✅ | ✅ | Gated by `sessionCapabilities.list`. The harness HAS `sessionPersistence.list()` (used internally for load-cwd validation) but does not expose it over ACP. |
 | `session/delete` | S | ❌ | ✅ | ✅ | Gated by `sessionCapabilities.delete`. |
 | `session/fork` | U | ❌ | ✅ | ❌ | Claude ships `unstable_forkSession`; Codex does not. |
@@ -111,7 +111,7 @@ Tool-call presentation is **owned by each tool** (`presentCall` / `presentResult
 
 ## 6. Session modes / config options / models
 
-Config options ✅ (the [sandbox RFC § Per-session mode switching](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md)): when `ctx.permission` is composed, the bridge advertises one `permission` select whose values come from the deployment preset table and whose current value derives from the session log; `session/set_config_option` switches the preset end to end, with idle switches anchoring at the next `agent/prompt-submit` inside its open turn. Session modes stay deliberately unmodeled because config options replace them in ACP v2. Runtime model selection is still not modeled — the harness fixes the model per bridge via `AcpConfig.model` (both reference adapters ship a model selector).
+Config options ✅: the bridge advertises a `model` select from the advisory LLM provider/model catalog, preserving each provider/model pair in an opaque value and grouping multiple providers. A selected pair is isolated to one session, snapshotted with the prompt for each step, applied through `agent/request`, and restored from the logged request header on load. When `ctx.permission` is composed, the bridge also advertises one `permission` select whose values come from the deployment preset table and whose current value derives from the session log; idle permission switches anchor at the next `agent/prompt-submit` inside its open turn. Session modes stay deliberately unmodeled because config options replace them in ACP v2. See the [model-catalog RFC](../../../docs/rfc/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.md) and [sandbox RFC](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md).
 
 ## 7. Content blocks
 
@@ -133,7 +133,7 @@ The bridge rejects unsupported prompt blocks rather than silently dropping them
 | Multi-session (N per connection) | S | ✅ | Strict per-session demux; concurrent streams never interleave. See the [multi-session RFC](../../../docs/rfc/implemented/feature/2026-06-14-acp-multi-session.md). |
 | Disconnect / disposal teardown | S | ✅ | Quiesces every live session on client disconnect or Cordis disposal. |
 | `_meta` extensibility | S | ⚠️ | Consumed (Zed terminal cap) and emitted (terminal `_meta`); no other custom extensions. |
-| Background-task ownership isolation | — | ✅ | `bash_output`/`bash_kill` reject another session's task via an opaque owner token. |
+| Background-task ownership isolation | — | ✅ | Generic `task_output`/`task_kill` reject tasks whose branded owner `SessionId` belongs to another session. |
 | stdout-is-the-protocol guarantee | S | ✅ | The bridge runs in an example with no stdout logger. |
 
 ## Gap summary
@@ -141,13 +141,12 @@ The bridge rejects unsupported prompt blocks rather than silently dropping them
 Ranked by how commonly the reference adapters ship them and how much UX they unlock:
 
 1. **Session lifecycle** — `session/list` + `session/delete` (the persistence layer already lists), then `session/resume` / `session/close`.
-2. **Model selection** — sandbox and approval config options are implemented; selecting the bridge's model at runtime remains open.
-3. **Agent plan** (`sessionUpdate: 'plan'`) — surface the loop's plan as structured entries.
-4. **Slash commands** (`available_commands_update`).
-5. **MCP passthrough** (`mcpServers` on `session/new` + `mcpCapabilities`).
-6. **Richer prompt content** — image / embedded `resource` blocks (needs a multimodal model path).
-7. **Usage reporting** (`usage_update`) — the harness already records token usage internally (on `assistant/message`).
-8. **Editor filesystem delegation** (`fs/read_text_file` / `fs/write_text_file`) — lets the agent see unsaved buffers; lower priority since the harness has direct disk access.
+2. **Agent plan** (`sessionUpdate: 'plan'`) — surface the loop's plan as structured entries.
+3. **Slash commands** (`available_commands_update`).
+4. **MCP passthrough** (`mcpServers` on `session/new` + `mcpCapabilities`).
+5. **Richer prompt content** — image / embedded `resource` blocks (needs a multimodal model path).
+6. **Usage reporting** (`usage_update`) — the harness already records token usage internally (on `assistant/message`).
+7. **Editor filesystem delegation** (`fs/read_text_file` / `fs/write_text_file`) — lets the agent see unsaved buffers; lower priority since the harness has direct disk access.
 
 ## Out of scope
 
diff --git a/packages/ui/acp/package.json b/packages/ui/acp/package.json
index 71efc51a07..84e4dcbda9 100644
--- a/packages/ui/acp/package.json
+++ b/packages/ui/acp/package.json
@@ -34,6 +34,7 @@
     "@deepseek-ai/dsh-sandbox": "^0.0.1",
     "@deepseek-ai/dsh-session": "^0.0.1",
     "@deepseek-ai/dsh-session-persistence": "^0.0.1",
+    "@deepseek-ai/dsh-system-prompt": "^0.0.1",
     "@deepseek-ai/dsh-tools": "^0.0.1",
     "@deepseek-ai/dsh-user-approval": "^0.0.1",
     "@deepseek-ai/dsh-user-interaction": "^0.0.1",
@@ -42,6 +43,7 @@
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-bash": "workspace:^",
     "@deepseek-ai/dsh-bash-local": "workspace:^",
     "@deepseek-ai/dsh-fs-local": "workspace:^",
diff --git a/packages/ui/acp/src/index.ts b/packages/ui/acp/src/index.ts
index c464ffd22b..e8fb326d9b 100644
--- a/packages/ui/acp/src/index.ts
+++ b/packages/ui/acp/src/index.ts
@@ -34,13 +34,15 @@ import {
   type PromptRequest,
   type PromptResponse,
   type SessionConfigOption,
+  type SessionConfigSelectGroup,
+  type SessionConfigSelectOption,
   type SessionNotification,
   type SetSessionConfigOptionRequest,
   type SetSessionConfigOptionResponse,
   type Stream,
   type StopReason,
 } from '@agentclientprotocol/sdk'
-import type { ContentBlock } from '@deepseek-ai/dsh-llm'
+import type { ContentBlock, LlmCallConfig, LlmModelInfo, LlmProviderInfo } from '@deepseek-ai/dsh-llm'
 import { assertNever, CallId } from '@deepseek-ai/dsh-llm'
 import type { Agent } from '@deepseek-ai/dsh-agent'
 import { AgentId } from '@deepseek-ai/dsh-agent'
@@ -52,6 +54,9 @@ import type { ToolCallView, ToolRegistry, ToolResultView, TerminalResultView } f
 // Side-effect type import: declaration-merges `ctx.sessionPersistence` onto
 // Context (the bridge injects it and reads `list()` for load cwd validation).
 import type {} from '@deepseek-ai/dsh-session-persistence'
+// Side-effect type import: declaration-merges prompt assembly onto Context and
+// the scoped waterfall used to keep persona variables aligned with requests.
+import type {} from '@deepseek-ai/dsh-system-prompt'
 // Side-effect type import: declaration-merges the `approval/request` waterfall
 // the bridge answers for its own agents (see the approval answerer below).
 import type {} from '@deepseek-ai/dsh-user-approval'
@@ -73,7 +78,7 @@ import {
 export const name = 'acp'
 // Interface services back advertised loading, tool-owned presentation with a generic fallback, and interaction.
 // TODO(acp-session-inject): remove `sessions`; the bridge never reads it, and ownership is already behind `agents`.
-export const inject = ['agents', 'sessions', 'sessionPersistence', 'tools', 'userInteraction']
+export const inject = ['agents', 'sessions', 'sessionPersistence', 'tools', 'userInteraction', 'llm', 'systemPrompt']
 
 /** Build an ACP invalid-params error with visible human detail. */
 function invalidParams(detail: string): RequestError {
@@ -201,6 +206,8 @@ function stringArrayContent(
 
 /** Plugin config: the agent template ACP sessions are created from. */
 export interface AcpConfig {
+  /** Provider route for created agents. */
+  provider?: string
   /** Model name for created agents (must have a registered adapter). */
   model?: string
   /** Runtime-only transport override for tests; production uses stdio. */
@@ -208,9 +215,35 @@ export interface AcpConfig {
 }
 
 export const Config: Schema<AcpConfig> = Schema.object({
+  provider: Schema.string(),
   model: Schema.string(),
 })
 
+/** Provider/model pair selected for one ACP session. */
+interface LlmTarget {
+  provider: string
+  model: string
+}
+
+/** Mutable target shared by one agent's scoped assembly and request listeners. */
+interface LlmTargetRef {
+  current: LlmTarget | undefined
+  /** Step snapshot captured by prompt assembly so target switches cannot split prompt and request. */
+  assembled: LlmTarget | undefined
+}
+
+/** One resolved ACP model selector plus its opaque value lookup. */
+interface ModelDirectory {
+  option: Extract<SessionConfigOption, { type: 'select' }> | undefined
+  targets: ReadonlyMap<string, LlmTarget>
+}
+
+/** One provider and its adapter-advertised models, detached for one RPC. */
+interface ModelCatalogEntry {
+  provider: LlmProviderInfo
+  models: LlmModelInfo[]
+}
+
 /** Per-session bridge state keyed by ACP session id. */
 interface SessionRecord {
   sessionId: SessionId
@@ -221,6 +254,8 @@ interface SessionRecord {
   presenter: ToolPresenter
   /** Session-creation snapshot of terminal-card support for call/result consistency. */
   terminalEnabled: boolean
+  /** Session-local provider/model selection and the current step snapshot. */
+  target: LlmTargetRef
   /** In-flight prompt and its captured turn number for exact settlement. */
   inflight: {
     resolve: (reason: StopReason) => void
@@ -243,6 +278,7 @@ interface SessionRecord {
 export function apply(ctx: Context, config: AcpConfig): void {
   // Handlers run later outside this injection scope, so capture services now.
   const agents = ctx.agents
+  const llm = ctx.llm
   const sessionPersistence = ctx.sessionPersistence
   const logger = ctx.logger
   const tools = ctx.tools
@@ -250,6 +286,101 @@ export function apply(ctx: Context, config: AcpConfig): void {
   // Presenter failures are logged and contained per session or replay.
   const makePresenter = (agent?: Agent): ToolPresenter => new ToolPresenter(tools, (message) => { logger.warn(message) }, agent)
 
+  /** Resolve a complete target only; partial config remains available to other request listeners. */
+  const configuredTarget = (): LlmTarget | undefined => config.provider !== undefined && config.model !== undefined
+    ? { provider: config.provider, model: config.model }
+    : undefined
+
+  /** Install the ACP target as an agent-scoped prompt/request override. */
+  const installTarget = (agentCtx: Context, target: LlmTargetRef): void => {
+    const agent = agentCtx.agent
+    /* v8 ignore next -- setup is invoked only with the freshly created agent's scoped context. */
+    if (agent === undefined) throw new Error('acp: agent setup has no scoped agent')
+    const logged = agent.session.requestHeader()?.config
+    if (logged !== undefined) target.current = { provider: logged.provider, model: logged.model }
+
+    // Capture once at assembly entry and apply the same pair after downstream
+    // prompt listeners. A selector change during async assembly therefore takes
+    // effect on the following step instead of splitting {{model}} from routing.
+    agentCtx.on('system-prompt/assemble', async (_assembly, _context, next) => {
+      const selected = target.current
+      const assembled = await next()
+      target.assembled = selected
+      if (selected === undefined) return assembled
+      return {
+        ...assembled,
+        variables: {
+          ...assembled.variables,
+          provider: selected.provider,
+          model: selected.model,
+        },
+      }
+    })
+    agentCtx.on('agent/request', async (_agent, _turn, _step, _callConfig, next): Promise<LlmCallConfig> => {
+      const resolved = await next()
+      const selected = target.assembled
+      return selected === undefined ? resolved : {
+        ...resolved,
+        provider: selected.provider,
+        model: selected.model,
+      }
+    })
+  }
+
+  /** Opaque ACP value preserving both routing dimensions. */
+  const targetValue = (target: LlmTarget): string => JSON.stringify([target.provider, target.model])
+
+  /** Read one detached advisory catalog snapshot before mutating session state. */
+  const readModelCatalog = async (): Promise<ModelCatalogEntry[]> => Promise.all(
+    llm.listProviders().map(async provider => ({
+      provider,
+      models: await llm.listModels(provider.id),
+    })),
+  )
+
+  /** Resolve one catalog snapshot into the ACP model selector for a session. */
+  const modelDirectory = (catalog: readonly ModelCatalogEntry[], current: LlmTarget | undefined): ModelDirectory => {
+    if (current === undefined) return { option: undefined, targets: new Map() }
+    const models = catalog.map(entry => ({ provider: entry.provider, models: [...entry.models] }))
+    const currentProvider = models.find(entry => entry.provider.id === current.provider)
+    if (currentProvider === undefined) return { option: undefined, targets: new Map() }
+    if (!currentProvider.models.some(model => model.id === current.model)) {
+      currentProvider.models = [...currentProvider.models, {
+        provider: current.provider,
+        id: current.model,
+        name: current.model,
+      }]
+    }
+
+    const targets = new Map<string, LlmTarget>()
+    const groups = models.flatMap(({ provider, models: entries }) => {
+      if (entries.length === 0) return []
+      const options = entries.map((model): SessionConfigSelectOption => {
+        const target = { provider: model.provider, model: model.id }
+        const value = targetValue(target)
+        targets.set(value, target)
+        return {
+          value,
+          name: model.name,
+          ...model.description === undefined ? {} : { description: model.description },
+        }
+      })
+      return [{ group: provider.id, name: provider.name, options } satisfies SessionConfigSelectGroup]
+    })
+    return {
+      option: {
+        id: 'model',
+        name: 'Model',
+        description: 'Sets this session\'s provider and model.',
+        category: 'model',
+        type: 'select',
+        currentValue: targetValue(current),
+        options: groups.length === 1 ? groups.flatMap(group => group.options) : groups,
+      },
+      targets,
+    }
+  }
+
   // TODO(derive-acp-session-id): derive event ids from `agent.session`, verify ownership, then remove the reverse map.
   // Agent events currently carry only the Agent, so retain `SessionRecord.sessionId` and update both indexes together.
   // Dropping the forward record lets the weak reverse entry expire.
@@ -433,16 +564,17 @@ export function apply(ctx: Context, config: AcpConfig): void {
 
   // --- The ACP Agent method surface -----------------------------------------
 
-  /**
-   * Build the single Permissions option when `ctx.permission` is composed.
-   * Its value comes from the session log, overlaid by an unanchored idle
-   * switch, so `session/load` needs no catch-up state.
-   */
-  const configOptionsFor = (agent: Agent, pending: SessionRecord['pendingSwitches'] = {}): SessionConfigOption[] => {
+  /** Build every ACP session option from the model directory and live services. */
+  const configOptionsFor = (
+    agent: Agent,
+    directory: ModelDirectory,
+    pending: SessionRecord['pendingSwitches'] = {},
+  ): SessionConfigOption[] => {
+    const options = directory.option === undefined ? [] : [directory.option]
     const presets = ctx.get('permission')
-    if (presets === undefined) return []
+    if (presets === undefined) return options
     const currentValue = pending.preset ?? presets.current(agent.session.events)
-    return [{
+    return [...options, {
       id: 'permission',
       name: 'Permissions',
       description: 'Sets this session\'s sandbox and approval behavior.',
@@ -541,11 +673,15 @@ export function apply(ctx: Context, config: AcpConfig): void {
         validateWorkspaceParams(params)
         validateMcpServers(params)
         const sessionId = SessionId(randomUUID())
+        const target: LlmTargetRef = { current: configuredTarget(), assembled: undefined }
+        const directory = modelDirectory(await readModelCatalog(), target.current)
+        assertOpen()
         const handle = await agents.create({
           agentId: AgentId(sessionId),
           sessionId,
           meta: { cwd: params.cwd },
           agentOptions: agentOptions(config),
+          setup: (agentCtx) => { installTarget(agentCtx, target) },
         })
         // Creation awaits the unpublished setup transaction. A client disconnect
         // can therefore close this bridge
@@ -564,10 +700,11 @@ export function apply(ctx: Context, config: AcpConfig): void {
           dispose: () => handle.dispose(),
           presenter: makePresenter(handle.agent),
           terminalEnabled: terminalOutputCap,
+          target,
           inflight: undefined,
           pendingSwitches: {},
         })
-        const configOptions = configOptionsFor(handle.agent)
+        const configOptions = configOptionsFor(handle.agent, directory)
         return { sessionId, ...configOptions.length > 0 ? { configOptions } : {} }
       },
 
@@ -612,10 +749,14 @@ export function apply(ctx: Context, config: AcpConfig): void {
               throw invalidParams(`session ${sessionId} cwd mismatch: persisted ${persistedCwd}, requested ${params.cwd}`)
             }
           }
+          const catalog = await readModelCatalog()
+          assertOpen()
+          const target: LlmTargetRef = { current: configuredTarget(), assembled: undefined }
           const handle = await agents.resume({
             agentId: AgentId(sessionId),
             resumeSessionId: sessionId,
             agentOptions: agentOptions(config),
+            setup: (agentCtx) => { installTarget(agentCtx, target) },
           })
           // The bridge may have torn down (disposal / client disconnect) while
           // resume() was pending. Its listeners are gone, so installing a record
@@ -631,6 +772,7 @@ export function apply(ctx: Context, config: AcpConfig): void {
             await handle.dispose()
             throw invalidParams('connection closed during session/load')
           }
+          const directory = modelDirectory(catalog, target.current)
           const agent = handle.agent
           bySession.set(agent, sessionId)
           // Snapshot the terminal capability ONCE for this session (used by both
@@ -643,6 +785,7 @@ export function apply(ctx: Context, config: AcpConfig): void {
             dispose: () => handle.dispose(),
             presenter: makePresenter(agent),
             terminalEnabled,
+            target,
             inflight: undefined,
             pendingSwitches: {},
           }
@@ -668,7 +811,7 @@ export function apply(ctx: Context, config: AcpConfig): void {
           for (const event of agent.session.events) {
             streamSessionEventUpdate(sessionId, event, notify, replayPresenter, replayTerminal)
           }
-          const configOptions = configOptionsFor(agent)
+          const configOptions = configOptionsFor(agent, directory)
           return configOptions.length > 0 ? { configOptions } : {}
         } finally {
           loadingIds.delete(sessionId)
@@ -722,18 +865,35 @@ export function apply(ctx: Context, config: AcpConfig): void {
         return Promise.resolve()
       },
 
-      setSessionConfigOption(params: SetSessionConfigOptionRequest): Promise<SetSessionConfigOptionResponse> {
+      async setSessionConfigOption(params: SetSessionConfigOptionRequest): Promise<SetSessionConfigOptionResponse> {
         assertOpen()
         const rec = requireSession(SessionId(params.sessionId))
-        // The advertised option is a select, so the boolean-shaped variant of
-        // the request is a protocol misuse regardless of configId.
+        // Every advertised option is a select, so the boolean-shaped variant
+        // is a protocol misuse regardless of configId.
         if (typeof params.value !== 'string') {
           throw invalidParams(`config option ${params.configId} is a select; boolean values are not accepted`)
         }
+        let directory = modelDirectory(await readModelCatalog(), rec.target.current)
         // Open-turn switches append immediately; idle switches wait for the
         // next prompt-submit. Only values advertised by this composition are
         // accepted, and the session log remains the durable store.
         switch (params.configId) {
+          case 'model': {
+            const target = directory.targets.get(params.value)
+            if (target === undefined) {
+              throw invalidParams(`unknown model value ${JSON.stringify(params.value)}`)
+            }
+            rec.target.current = { ...target }
+            const option = directory.option
+            /* v8 ignore next -- `targets` is populated only while constructing
+               this selector; a found target therefore proves it exists. */
+            if (option === undefined) throw internalError('model directory target has no selector')
+            directory = {
+              ...directory,
+              option: { ...option, currentValue: params.value },
+            }
+            break
+          }
           case 'permission': {
             const presets = ctx.get('permission')
             if (presets === undefined) {
@@ -755,7 +915,7 @@ export function apply(ctx: Context, config: AcpConfig): void {
         }
         // The spec requires the COMPLETE refreshed config state in the response
         // (a change may cascade); ours are independent, but the contract holds.
-        return Promise.resolve({ configOptions: configOptionsFor(rec.agent, rec.pendingSwitches) })
+        return { configOptions: configOptionsFor(rec.agent, directory, rec.pendingSwitches) }
       },
     }
   }
@@ -851,11 +1011,12 @@ export function apply(ctx: Context, config: AcpConfig): void {
  * Build per-agent options from the plugin config, omitting absent fields
  * (exactOptionalPropertyTypes: never assign `undefined` to an optional key).
  * Exported for unit coverage of both the present and absent branches.
- * @param config - the plugin config carrying the optional model name.
- * @returns the per-agent options, with `model` present only when configured.
+ * @param config - the plugin config carrying the optional provider/model target.
+ * @returns the per-agent options, with each configured target field present.
  */
-export function agentOptions(config: AcpConfig): { model?: string } {
+export function agentOptions(config: AcpConfig): { provider?: string; model?: string } {
   return {
+    ...config.provider !== undefined ? { provider: config.provider } : {},
     ...config.model !== undefined ? { model: config.model } : {},
   }
 }
diff --git a/packages/ui/acp/tests/config-options.spec.ts b/packages/ui/acp/tests/config-options.spec.ts
index f914004e8b..806aaa3ab6 100644
--- a/packages/ui/acp/tests/config-options.spec.ts
+++ b/packages/ui/acp/tests/config-options.spec.ts
@@ -40,6 +40,26 @@ function permissionOption(currentValue: string): object {
   }
 }
 
+function modelValue(provider = 'mock', model = 'mock'): string {
+  return JSON.stringify([provider, model])
+}
+
+function modelOption(currentValue = modelValue()): object {
+  return {
+    id: 'model',
+    name: 'Model',
+    description: 'Sets this session\'s provider and model.',
+    category: 'model',
+    type: 'select',
+    currentValue,
+    options: [{ value: modelValue(), name: 'Mock' }],
+  }
+}
+
+function optionsWithPermission(currentValue: string): object[] {
+  return [modelOption(), permissionOption(currentValue)]
+}
+
 describe('acp bridge — session config options', () => {
   let storageDir: string
   let h: BridgeHarness | undefined
@@ -64,19 +84,111 @@ describe('acp bridge — session config options', () => {
     return harness
   }
 
-  it('advertises no configOptions without the permission service — even with both knobs composed', async () => {
+  it('advertises the model selector without requiring the permission service', async () => {
     h = await makeBridgeHarness({ storageDir })
     await h.ctx.plugin(SandboxedLocalExecutor, { timeoutMs: 10_000 })
     await h.ctx.plugin(ApprovalService)
     await h.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
     const res = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
-    expect(res.configOptions).toBeUndefined()
+    expect(res.configOptions).toEqual([modelOption()])
+  })
+
+  it('groups models by provider and switches routing plus prompt variables as one session target', async () => {
+    h = await makeBridgeHarness({
+      storageDir,
+      script: [textResponse('ok')],
+      config: { provider: 'alpha', model: 'a1' },
+      persona: 'Route {{provider}} / {{model}}',
+      catalog: {
+        providers: [{ id: 'alpha', name: 'Alpha' }, { id: 'beta', name: 'Beta' }],
+        models: [
+          { provider: 'alpha', id: 'a1', name: 'Alpha One', description: 'Fast' },
+          { provider: 'beta', id: 'b1', name: 'Beta One' },
+        ],
+      },
+    })
+    await h.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
+    const created = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
+    expect(created.configOptions).toEqual([{
+      id: 'model',
+      name: 'Model',
+      description: 'Sets this session\'s provider and model.',
+      category: 'model',
+      type: 'select',
+      currentValue: modelValue('alpha', 'a1'),
+      options: [
+        { group: 'alpha', name: 'Alpha', options: [{ value: modelValue('alpha', 'a1'), name: 'Alpha One', description: 'Fast' }] },
+        { group: 'beta', name: 'Beta', options: [{ value: modelValue('beta', 'b1'), name: 'Beta One' }] },
+      ],
+    }])
+
+    const switched = await h.client.setSessionConfigOption({
+      sessionId: created.sessionId,
+      configId: 'model',
+      value: modelValue('beta', 'b1'),
+    })
+    expect(switched.configOptions?.[0]).toMatchObject({ currentValue: modelValue('beta', 'b1') })
+    await h.client.prompt({ sessionId: created.sessionId, prompt: [{ type: 'text', text: 'use beta' }] })
+    expect(h.adapter.requests[0]).toMatchObject({
+      provider: 'beta',
+      model: 'b1',
+    })
+    expect(h.adapter.requests[0]?.system).toContain('Route beta / b1')
+    expect(h.ctx.agents.list()[0]?.session.requestHeader()?.config).toMatchObject({ provider: 'beta', model: 'b1' })
+  })
+
+  it('adds the configured private model to an advisory catalog and ignores empty non-current groups', async () => {
+    h = await makeBridgeHarness({
+      storageDir,
+      config: { provider: 'alpha', model: 'private-model' },
+      catalog: {
+        providers: [{ id: 'alpha', name: 'Alpha' }, { id: 'empty', name: 'Empty' }],
+        models: [{ provider: 'alpha', id: 'public-model', name: 'Public Model' }],
+      },
+    })
+    await h.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
+    const res = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
+    expect(res.configOptions?.[0]).toMatchObject({
+      currentValue: modelValue('alpha', 'private-model'),
+      options: [
+        { value: modelValue('alpha', 'public-model'), name: 'Public Model' },
+        { value: modelValue('alpha', 'private-model'), name: 'private-model' },
+      ],
+    })
+  })
+
+  it('omits model selection without a complete or registered current target', async () => {
+    h = await makeBridgeHarness({ storageDir, config: { model: undefined } })
+    await h.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
+    const missing = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
+    expect(missing.configOptions).toBeUndefined()
+    await h.dispose()
+
+    h = await makeBridgeHarness({ storageDir, config: { provider: 'unregistered', model: 'm' } })
+    await h.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
+    const unknown = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
+    expect(unknown.configOptions).toBeUndefined()
+  })
+
+  it('leaves model-less agents available to another agent/request supplier', async () => {
+    h = await makeBridgeHarness({ storageDir, config: { model: undefined }, script: [textResponse('ok')] })
+    await h.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
+    const { sessionId } = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
+    const agent = h.ctx.agents.list()[0]
+    if (agent === undefined) throw new Error('expected an agent')
+    agent.ctx.on('agent/request', async (_agent, _turn, _step, callConfig, _next) => ({
+      ...callConfig,
+      provider: 'mock',
+      model: 'mock',
+    }))
+    await h.client.prompt({ sessionId, prompt: [{ type: 'text', text: 'supplied elsewhere' }] })
+    expect(h.adapter.requests[0]).toMatchObject({ provider: 'mock', model: 'mock' })
   })
 
   it('advertises the Permissions select with the default preset current', async () => {
     h = await presetStack()
     const res = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
-    expect(res.configOptions).toEqual([permissionOption('workspace-write')])
+    expect(res.configOptions).toEqual(optionsWithPermission('workspace-write'))
   })
 
   it('an idle switch is pending (overlaid, not yet logged), then anchors inside the next prompt\'s turn', async () => {
@@ -84,7 +196,7 @@ describe('acp bridge — session config options', () => {
     const { sessionId } = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
 
     const after = await h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'danger-full-access' })
-    expect(after.configOptions).toEqual([permissionOption('danger-full-access')])
+    expect(after.configOptions).toEqual(optionsWithPermission('danger-full-access'))
 
     const session = h.ctx.agents.list()[0]?.session
     expect(session?.events.some(e => e.type === 'permission/preset' || e.type === 'bash/sandbox-mode' || e.type === 'approval/policy')).toBe(false)
@@ -105,7 +217,7 @@ describe('acp bridge — session config options', () => {
     const { sessionId } = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
     await h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'danger-full-access' })
     const again = await h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'danger-full-access' })
-    expect(again.configOptions).toEqual([permissionOption('danger-full-access')])
+    expect(again.configOptions).toEqual(optionsWithPermission('danger-full-access'))
     await h.client.prompt({ sessionId, prompt: [{ type: 'text', text: 'anchor' }] })
     const events = h.ctx.agents.list()[0]?.session.events ?? []
     expect(events.filter(e => e.type === 'permission/preset')).toHaveLength(1)
@@ -119,7 +231,7 @@ describe('acp bridge — session config options', () => {
     const { sessionId } = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
     await h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'danger-full-access' })
     const back = await h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'workspace-write' })
-    expect(back.configOptions).toEqual([permissionOption('workspace-write')])
+    expect(back.configOptions).toEqual(optionsWithPermission('workspace-write'))
     await h.client.prompt({ sessionId, prompt: [{ type: 'text', text: 'anchor' }] })
     const events = h.ctx.agents.list()[0]?.session.events ?? []
     expect(events.some(e => e.type === 'permission/preset' || e.type === 'bash/sandbox-mode' || e.type === 'approval/policy')).toBe(false)
@@ -129,10 +241,10 @@ describe('acp bridge — session config options', () => {
     h = await presetStack({ script: [textResponse('ok')] })
     const { sessionId } = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
     const echo = await h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'workspace-write' })
-    expect(echo.configOptions).toEqual([permissionOption('workspace-write')])
+    expect(echo.configOptions).toEqual(optionsWithPermission('workspace-write'))
     await h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'danger-full-access' })
     const repeat = await h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'danger-full-access' })
-    expect(repeat.configOptions).toEqual([permissionOption('danger-full-access')])
+    expect(repeat.configOptions).toEqual(optionsWithPermission('danger-full-access'))
     await h.client.prompt({ sessionId, prompt: [{ type: 'text', text: 'anchor' }] })
     const events = h.ctx.agents.list()[0]?.session.events ?? []
     expect(events.filter(e => e.type === 'permission/preset').map(e => e.data)).toEqual([{ preset: 'danger-full-access' }])
@@ -167,6 +279,8 @@ describe('acp bridge — session config options', () => {
     // This composition never advertised `permission`.
     await expect(h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'danger-full-access' }))
       .rejects.toThrow(/unknown permission value/)
+    await expect(h.client.setSessionConfigOption({ sessionId, configId: 'model', value: modelValue('mock', 'missing') }))
+      .rejects.toThrow(/unknown model value/)
     await expect(h.client.setSessionConfigOption({ sessionId, configId: 'permission', type: 'boolean', value: true }))
       .rejects.toThrow(/select; boolean values are not accepted/)
   })
@@ -184,9 +298,31 @@ describe('acp bridge — session config options', () => {
     const b = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
     await h.client.setSessionConfigOption({ sessionId: a.sessionId, configId: 'permission', value: 'danger-full-access' })
     const bAfter = await h.client.setSessionConfigOption({ sessionId: b.sessionId, configId: 'permission', value: 'workspace-write' })
-    expect(bAfter.configOptions).toEqual([permissionOption('workspace-write')])
+    expect(bAfter.configOptions).toEqual(optionsWithPermission('workspace-write'))
     const aAfter = await h.client.setSessionConfigOption({ sessionId: a.sessionId, configId: 'permission', value: 'danger-full-access' })
-    expect(aAfter.configOptions).toEqual([permissionOption('danger-full-access')])
+    expect(aAfter.configOptions).toEqual(optionsWithPermission('danger-full-access'))
+  })
+
+  it('keeps model targets isolated across concurrent sessions', async () => {
+    h = await makeBridgeHarness({
+      storageDir,
+      script: [textResponse('a'), textResponse('b')],
+      config: { provider: 'mock', model: 'one' },
+      catalog: {
+        providers: [{ id: 'mock', name: 'Mock' }],
+        models: [
+          { provider: 'mock', id: 'one', name: 'One' },
+          { provider: 'mock', id: 'two', name: 'Two' },
+        ],
+      },
+    })
+    await h.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
+    const a = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
+    const b = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
+    await h.client.setSessionConfigOption({ sessionId: a.sessionId, configId: 'model', value: modelValue('mock', 'two') })
+    await h.client.prompt({ sessionId: a.sessionId, prompt: [{ type: 'text', text: 'a' }] })
+    await h.client.prompt({ sessionId: b.sessionId, prompt: [{ type: 'text', text: 'b' }] })
+    expect(h.adapter.requests.map(request => request.model)).toEqual(['two', 'one'])
   })
 
   it('a knob drifted outside the table derives a visible-but-untargetable custom current', async () => {
@@ -199,12 +335,12 @@ describe('acp bridge — session config options', () => {
     agent.session.append('bash/sandbox-mode', { mode: 'read-only' })
     agent.session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
     const echo = await h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'custom' })
-    const option = echo.configOptions?.[0]
+    const option = echo.configOptions?.find(entry => entry.id === 'permission')
     expect(option).toMatchObject({ currentValue: 'custom' })
     if (option === undefined || !('options' in option)) throw new Error('expected a select option')
     expect(option.options.map(o => 'value' in o ? o.value : o)).toEqual(['workspace-write', 'danger-full-access', 'custom'])
     const away = await h.client.setSessionConfigOption({ sessionId, configId: 'permission', value: 'danger-full-access' })
-    const afterOption = away.configOptions?.[0]
+    const afterOption = away.configOptions?.find(entry => entry.id === 'permission')
     expect(afterOption).toMatchObject({ currentValue: 'danger-full-access' })
     if (afterOption === undefined || !('options' in afterOption)) throw new Error('expected a select option')
     expect(afterOption.options.map(o => 'value' in o ? o.value : o)).toEqual(['workspace-write', 'danger-full-access'])
@@ -223,6 +359,52 @@ describe('acp bridge — session config options', () => {
 
     loader = await presetStack()
     const res = await loader.client.loadSession({ sessionId, cwd: process.cwd(), mcpServers: [] })
-    expect(res.configOptions).toEqual([permissionOption('danger-full-access')])
+    expect(res.configOptions).toEqual(optionsWithPermission('danger-full-access'))
+  })
+
+  it('session/load restores the last requested provider/model from the request header', async () => {
+    const catalog = {
+      providers: [{ id: 'mock', name: 'Mock' }],
+      models: [
+        { provider: 'mock', id: 'one', name: 'One' },
+        { provider: 'mock', id: 'two', name: 'Two' },
+      ],
+    }
+    h = await makeBridgeHarness({
+      storageDir,
+      script: [textResponse('ok')],
+      config: { provider: 'mock', model: 'one' },
+      catalog,
+    })
+    await h.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
+    const { sessionId } = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
+    await h.client.setSessionConfigOption({ sessionId, configId: 'model', value: modelValue('mock', 'two') })
+    await h.client.prompt({ sessionId, prompt: [{ type: 'text', text: 'persist target' }] })
+    await h.dispose()
+    h = undefined
+
+    loader = await makeBridgeHarness({ storageDir, config: { provider: 'mock', model: 'one' }, catalog })
+    await loader.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
+    const loaded = await loader.client.loadSession({ sessionId, cwd: process.cwd(), mcpServers: [] })
+    expect(loaded.configOptions?.find(option => option.id === 'model')).toMatchObject({
+      currentValue: modelValue('mock', 'two'),
+    })
+  })
+
+  it('session/load omits config options when the persisted session has no target or permission service', async () => {
+    h = await makeBridgeHarness({ storageDir, config: { model: undefined } })
+    await h.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
+    const { sessionId } = await h.client.newSession({ cwd: process.cwd(), mcpServers: [] })
+    const agent = h.ctx.agents.list()[0]
+    if (agent === undefined) throw new Error('expected an agent')
+    agent.inject([{ type: 'text', text: 'checkpoint' }], { source: { kind: 'plugin', plugin: 'test' } })
+    await agent.whenIdle()
+    await h.dispose()
+    h = undefined
+
+    loader = await makeBridgeHarness({ storageDir, config: { model: undefined } })
+    await loader.client.initialize({ protocolVersion: PROTOCOL_VERSION, clientCapabilities: {} })
+    const loaded = await loader.client.loadSession({ sessionId, cwd: process.cwd(), mcpServers: [] })
+    expect(loaded.configOptions).toBeUndefined()
   })
 })
diff --git a/packages/ui/acp/tests/dispose.spec.ts b/packages/ui/acp/tests/dispose.spec.ts
index e63658f12a..30faf9aafe 100644
--- a/packages/ui/acp/tests/dispose.spec.ts
+++ b/packages/ui/acp/tests/dispose.spec.ts
@@ -175,10 +175,10 @@ describe('acp bridge — disposal & HMR safety', () => {
     // published, which guards against context-wide teardown.
     const harness = await makeBridgeHarness({ storageDir, script: [] })
     const handleA = await harness.ctx.agents.create({
-      agentId: AgentId('sib-a'), sessionId: SessionId('sib-a'), agentOptions: { model: 'mock' },
+      agentId: AgentId('sib-a'), sessionId: SessionId('sib-a'), agentOptions: { provider: 'mock', model: 'mock' },
     })
     const handleB = await harness.ctx.agents.create({
-      agentId: AgentId('sib-b'), sessionId: SessionId('sib-b'), agentOptions: { model: 'mock' },
+      agentId: AgentId('sib-b'), sessionId: SessionId('sib-b'), agentOptions: { provider: 'mock', model: 'mock' },
     })
     expect(harness.ctx.agents.get(AgentId('sib-a'))).toBe(handleA.agent)
     expect(harness.ctx.agents.get(AgentId('sib-b'))).toBe(handleB.agent)
@@ -199,7 +199,7 @@ describe('acp bridge — disposal & HMR safety', () => {
     const harness = await makeBridgeHarness({ storageDir, script: [textResponse('ok')] })
     harness.ctx.on('agent/disposed', () => { throw new Error('boom disposed listener') })
     const handle = await harness.ctx.agents.create({
-      agentId: AgentId('guard-a'), sessionId: SessionId('guard-a'), agentOptions: { model: 'mock' },
+      agentId: AgentId('guard-a'), sessionId: SessionId('guard-a'), agentOptions: { provider: 'mock', model: 'mock' },
     })
     handle.agent.send([{ type: 'text', text: 'go' }])
     await handle.agent.whenIdle()
@@ -216,7 +216,7 @@ describe('acp bridge — disposal & HMR safety', () => {
     // clears. AgentHandle must memoize the whole async teardown so every caller awaits quiescence.
     const harness = await makeBridgeHarness({ storageDir, script: ['hang'] })
     const handle = await harness.ctx.agents.create({
-      agentId: AgentId('conc-a'), sessionId: SessionId('conc-a'), agentOptions: { model: 'mock' },
+      agentId: AgentId('conc-a'), sessionId: SessionId('conc-a'), agentOptions: { provider: 'mock', model: 'mock' },
     })
     // A hanging turn makes disposal produce a final flush; gate it so the second call arrives while
     // teardown is observably in flight.
diff --git a/packages/ui/acp/tests/edges.spec.ts b/packages/ui/acp/tests/edges.spec.ts
index e86fb9fc94..de3a7a6f03 100644
--- a/packages/ui/acp/tests/edges.spec.ts
+++ b/packages/ui/acp/tests/edges.spec.ts
@@ -27,7 +27,7 @@ describe('acp bridge — demux & config edges', () => {
     await harness.client.newSession({ cwd: process.cwd(), mcpServers: [] })
     const before = harness.updates.length
 
-    const { agent: foreign } = await harness.ctx.agents.create({ agentId: AgentId('foreign'), sessionId: SessionId('foreign-session'), agentOptions: { model: 'mock' } })
+    const { agent: foreign } = await harness.ctx.agents.create({ agentId: AgentId('foreign'), sessionId: SessionId('foreign-session'), agentOptions: { provider: 'mock', model: 'mock' } })
     foreign.send([{ type: 'text', text: 'hi' }])
     await foreign.whenIdle()
     await new Promise(r => setTimeout(r, 10))
diff --git a/packages/ui/acp/tests/harness.ts b/packages/ui/acp/tests/harness.ts
index 440120943f..1796cbf868 100644
--- a/packages/ui/acp/tests/harness.ts
+++ b/packages/ui/acp/tests/harness.ts
@@ -5,13 +5,10 @@
  */
 
 import { Context } from 'cordis'
-import LlmService, { CallId, type GenerateOptions, type StreamChunk } from '@deepseek-ai/dsh-llm'
+import { CallId, type GenerateOptions, type LlmModelInfo, type LlmProviderInfo, type StreamChunk } from '@deepseek-ai/dsh-llm'
 import { LlmAdapter } from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry from '@deepseek-ai/dsh-agent'
 import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
 import { LocalBashExecutor } from '@deepseek-ai/dsh-bash-local'
 import LocalFileSystem from '@deepseek-ai/dsh-fs-local'
@@ -39,10 +36,24 @@ import { type AcpConfig } from '../src/index.ts'
 /** A scripted mock adapter (mirrors the agent-loop test adapter). */
 class MockAdapter extends LlmAdapter {
   requests: GenerateOptions[] = []
-  constructor(private script: (StreamChunk[] | 'hang')[]) {
+  constructor(
+    private script: (StreamChunk[] | 'hang')[],
+    private readonly providers: readonly LlmProviderInfo[],
+    private readonly models: readonly LlmModelInfo[],
+  ) {
     super()
   }
 
+  override providerInfo(provider: string): LlmProviderInfo {
+    const info = this.providers.find(entry => entry.id === provider)
+    if (info === undefined) throw new Error(`MockAdapter: unknown provider ${provider}`)
+    return info
+  }
+
+  override listModels(provider: string): Promise<readonly LlmModelInfo[]> {
+    return Promise.resolve(this.models.filter(model => model.provider === provider))
+  }
+
   async * stream(options: GenerateOptions): AsyncIterable<StreamChunk> {
     this.requests.push(options)
     const entry = this.script.shift()
@@ -142,6 +153,9 @@ export interface BridgeHarness {
   storageDir: string
 }
 
+/** Test-only overrides preserve explicit undefined to suppress harness defaults. */
+type AcpConfigOverrides = { [K in keyof AcpConfig]?: AcpConfig[K] | undefined }
+
 /**
  * Build the bridge + a connected client over an in-memory transport pair.
  *
@@ -150,12 +164,13 @@ export interface BridgeHarness {
  * The bridge's `apply` receives the agent-side `Stream` via `config.stream`;
  * the test holds the `ClientSideConnection`.
  *
- * Pass `config: { model: undefined }` to override the default `model: 'mock'`
- * (the model key is dropped entirely when explicitly undefined).
+ * Pass an explicit undefined route field to suppress its mock default.
  */
 export async function makeBridgeHarness(options: {
   script?: (StreamChunk[] | 'hang')[]
-  config?: Partial<AcpConfig>
+  config?: AcpConfigOverrides
+  /** Provider-neutral directory exposed to ACP model-selection tests. */
+  catalog?: { providers: LlmProviderInfo[]; models: LlmModelInfo[] }
   /** Deployment persona for the tree (the system-prompt plugin's config). */
   persona?: string
   storageDir: string
@@ -185,14 +200,16 @@ export async function makeBridgeHarness(options: {
   withFs?: boolean
   fsCwd?: string
 } = { storageDir: '' }): Promise<BridgeHarness> {
-  const adapter = new MockAdapter(options.script ?? [])
+  const catalog = options.catalog ?? {
+    providers: [{ id: 'mock', name: 'Mock' }],
+    models: [{ provider: 'mock', id: 'mock', name: 'Mock' }],
+  }
+  const adapter = new MockAdapter(options.script ?? [], catalog.providers, catalog.models)
 
   const ctx = new Context()
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt, { persona: options.persona ?? '' })
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx, {
+    systemPrompt: { persona: options.persona ?? '' },
+  })
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(SessionPersistenceJsonl, { root: options.storageDir })
   await ctx.plugin(UserInteractionService)
@@ -211,7 +228,7 @@ export async function makeBridgeHarness(options: {
     await ctx.plugin(FsPolicy)
     await ctx.plugin(ToolFs)
   }
-  ctx.llm.registerAdapter(['mock'], adapter)
+  ctx.llm.registerAdapter(catalog.providers.map(provider => provider.id), adapter)
 
   // Two identity byte pipes cross-wired into the two ndJsonStreams: bytes the agent writes flow
   // to the client's reader and vice versa. (ndJsonStream takes (output, input): the agent
@@ -271,9 +288,9 @@ export async function makeBridgeHarness(options: {
     },
   })
 
-  // Default to `mock` only when the caller omitted the key; explicit `model: undefined` means no
-  // model and must survive the object spread.
-  const cfg: AcpConfig = { stream: agentStream, ...options.config }
+  // Default route fields only when the caller omitted them; explicit undefined values must survive.
+  const cfg = { stream: agentStream, ...options.config } as AcpConfig
+  if (!(options.config && 'provider' in options.config)) cfg.provider = 'mock'
   if (!(options.config && 'model' in options.config)) cfg.model = 'mock'
   // Mount the bridge the way production does: as a cordis plugin (via `ctx.plugin` with the
   // real `inject`), not `AcpPlugin.apply(ctx, cfg)` on the ungated root. Later JSON-RPC callbacks run
diff --git a/packages/ui/acp/tests/stream-update.spec.ts b/packages/ui/acp/tests/stream-update.spec.ts
index 2afa49e1d4..415e9afb33 100644
--- a/packages/ui/acp/tests/stream-update.spec.ts
+++ b/packages/ui/acp/tests/stream-update.spec.ts
@@ -794,5 +794,6 @@ describe('agentOptions', () => {
   it('includes only the fields present in config', () => {
     expect(agentOptions({})).toEqual({})
     expect(agentOptions({ model: 'm' })).toEqual({ model: 'm' })
+    expect(agentOptions({ provider: 'p', model: 'm' })).toEqual({ provider: 'p', model: 'm' })
   })
 })
diff --git a/packages/ui/app-boot/README.md b/packages/ui/app-boot/README.md
index e7145c24fc..5fed7b5023 100644
--- a/packages/ui/app-boot/README.md
+++ b/packages/ui/app-boot/README.md
@@ -12,7 +12,9 @@ Shared boot glue for the app bins ([`dsh-stdio-demo`](../../examples/stdio-demo/
 
 Two failure classes the guards handle: `loader.await()` swallows init rejections (`Promise.allSettled`) — Node still exits non-zero on the resulting unhandled rejection, and `installFailLoud` replaces the noisy dump with one labelled line and a guaranteed `exit(1)`; a failed plugin IMPORT is only logged by the Loader (the process would otherwise exit 0 on a usable config typo), leaving a fiber-less entry that `assertEntriesLoaded` turns into a `boot()` rejection.
 
-Bare plugin specifiers in a config (`@deepseek-ai/dsh-*`) resolve through the cordis Loader's internal module loader, using `node --expose-internals` or the optional `node-addon-require-builtin` fallback. the bins' subprocess smokes exercise that path, while this package's unit suite drives `boot()` in-process against configs with relative specifiers.
+Bare plugin specifiers in a config (`@deepseek-ai/dsh-*`, npm packages) resolve through the cordis Loader's internal module loader when Node runs with `--expose-internals` or the optional `node-addon-require-builtin` fallback is installed; without either, consumers must install plugins where plain Node import resolution can find them. Relative specifiers resolve against the config directory with no flag. The bins' subprocess smokes exercise the internal-loader path, while this package's unit suite drives `boot()` in-process against configs with relative specifiers.
+
+This package carries no loader hooks and no dev-mode surface: the `dsh-scripts` launcher ([`sdk/scripts`](../../sdk/scripts/README.md), with the shared project model in [`sdk/helper`](../../sdk/helper/README.md)) owns process startup, tsx registration, and local-plugin source resolution, and consumes these helpers for the boot sequence itself.
 
 ## Model Experience
 
diff --git a/packages/ui/jsonrpc/README.md b/packages/ui/jsonrpc/README.md
index bf686ea800..c814919079 100644
--- a/packages/ui/jsonrpc/README.md
+++ b/packages/ui/jsonrpc/README.md
@@ -4,7 +4,7 @@ Stdio JSON-RPC plugin for out-of-process SDK clients such as Python `deepseek_ha
 
 ## Wiring
 
-`inject: ['agents']`. The server gets or creates one agent per `sessionId` on `session/prompt` and demuxes `subagent/end` through the registry. If `initialize.model` lacks a registered adapter, it mounts `dsh-llm-deepseek` using `$DEEPSEEK_API_KEY` and `$DEEPSEEK_BASE_URL`; a config-registered adapter wins. Persistence, tools, and other adapters come from the surrounding `cordis.yml`.
+`inject: ['agents']`. The server gets or creates one agent per `sessionId` from the `initialize.provider`/`initialize.model` pair and demuxes `subagent/end` through the registry. A registered owner for the provider route wins; an unowned `deepseek` route mounts `dsh-llm-deepseek` using `$DEEPSEEK_API_KEY` and `$DEEPSEEK_BASE_URL`, while any other unowned provider fails initialization. Persistence, tools, and other adapters come from the surrounding `cordis.yml`.
 
 ## Config
 
diff --git a/packages/ui/jsonrpc/src/server.ts b/packages/ui/jsonrpc/src/server.ts
index 91ef655c0c..2b3646c731 100644
--- a/packages/ui/jsonrpc/src/server.ts
+++ b/packages/ui/jsonrpc/src/server.ts
@@ -21,6 +21,8 @@ import type { JsonRpcTransportPeer } from './transport.ts'
 export interface InitializeParams {
   /** Working directory recorded on every SDK-created session's header. */
   cwd: string
+  /** Provider route every SDK-created agent runs on. */
+  provider: string
   /** Model name every SDK-created agent runs on (see {@link HarnessSdkServer.initialize} for adapter fallback). */
   model: string
 }
@@ -66,6 +68,7 @@ interface SubagentRecord {
  */
 export class HarnessSdkServer {
   private cwd = process.cwd()
+  private provider = 'deepseek'
   private model = 'deepseek'
   private llmFiber: { dispose(): Promise<void> } | undefined
   private readonly sessions = new Map<string, SessionRecord>()
@@ -124,16 +127,18 @@ export class HarnessSdkServer {
   }
 
   /**
-   * Record cwd and model, mounting the DeepSeek adapter only when the config
-   * registered no adapter for that model.
+   * Record cwd and provider/model, mounting the DeepSeek adapter only when the
+   * `deepseek` provider route has no configured owner.
    * @param params - the SDK handshake parameters.
    * @returns the server identity for the handshake.
    */
   async initialize(params: InitializeParams): Promise<InitializeResult> {
     this.cwd = resolve(params.cwd)
+    this.provider = params.provider
     this.model = params.model
-    if (!this.llmFiber && !this.hasAdapterFor(this.model)) {
-      this.llmFiber = await this.ctx.plugin(LlmDeepSeek, { models: [this.model] })
+    if (!this.hasAdapterFor(this.provider)) {
+      if (this.provider !== 'deepseek') throw new Error(`no adapter registered for provider "${this.provider}"`)
+      this.llmFiber = await this.ctx.plugin(LlmDeepSeek, {})
     }
     return { serverInfo: { name: 'deepseek-harness-sdk-runtime', version: '0.0.1' } }
   }
@@ -244,7 +249,7 @@ export class HarnessSdkServer {
       agentId: AgentId(sessionId),
       sessionId: SessionId(sessionId),
       meta: { cwd: this.cwd },
-      agentOptions: { model: this.model },
+      agentOptions: { provider: this.provider, model: this.model },
     })
     const rec: SessionRecord = { handle, lastTurnEnd: undefined, activePrompt: false }
     this.sessions.set(sessionId, rec)
@@ -256,7 +261,7 @@ export class HarnessSdkServer {
     return reason.kind === 'completed' ? 'ok' : 'error'
   }
 
-  private hasAdapterFor(model: string): boolean {
-    return this.ctx.get('llm')?.models().includes(model) ?? false
+  private hasAdapterFor(provider: string): boolean {
+    return this.ctx.get('llm')?.listProviders().some(entry => entry.id === provider) ?? false
   }
 }
diff --git a/packages/ui/jsonrpc/tests/plugin-apply.spec.ts b/packages/ui/jsonrpc/tests/plugin-apply.spec.ts
index 3b4e585ff6..0eef295911 100644
--- a/packages/ui/jsonrpc/tests/plugin-apply.spec.ts
+++ b/packages/ui/jsonrpc/tests/plugin-apply.spec.ts
@@ -59,7 +59,7 @@ async function mountPlugin(
   options: { writeDelayMs?: number; failFlush?: boolean } = {},
 ): Promise<ApplyHarness> {
   const ctx = new Context()
-  await ctx.plugin(agentCore)
+  await ctx.plugin(agentCore, { workspaceContext: false })
   await ctx.plugin(SessionPersistenceJsonl, { root: storageDir })
   await new Promise(resolve => setTimeout(resolve, 50))
 
@@ -153,7 +153,7 @@ describe('dsh-jsonrpc plugin apply', () => {
     vi.stubEnv('DEEPSEEK_API_KEY', 'test-key')
     const harness = await mountPlugin(storageDir)
     try {
-      harness.send({ jsonrpc: '2.0', id: 'init-1', method: 'initialize', params: { cwd: storageDir, model: 'apply-model' } })
+      harness.send({ jsonrpc: '2.0', id: 'init-1', method: 'initialize', params: { cwd: storageDir, provider: 'deepseek', model: 'apply-model' } })
 
       const response = await harness.waitForFrame(frame => frame.id === 'init-1', 'initialize response')
       expect(response).toEqual({
@@ -175,7 +175,7 @@ describe('dsh-jsonrpc plugin apply', () => {
     vi.stubEnv('DEEPSEEK_BASE_URL', llmServer.url)
     const harness = await mountPlugin(storageDir)
     try {
-      harness.send({ jsonrpc: '2.0', id: 1, method: 'initialize', params: { cwd: storageDir, model: 'dsagent-model' } })
+      harness.send({ jsonrpc: '2.0', id: 1, method: 'initialize', params: { cwd: storageDir, provider: 'deepseek', model: 'dsagent-model' } })
       await harness.waitForFrame(frame => frame.id === 1, 'initialize response')
 
       harness.send({
@@ -236,7 +236,7 @@ describe('dsh-jsonrpc plugin apply', () => {
       expect(harness.exits()).toEqual([0])
 
       const before = harness.frames().length
-      harness.send({ jsonrpc: '2.0', id: 'after-exit', method: 'initialize', params: { cwd: storageDir, model: 'x' } })
+      harness.send({ jsonrpc: '2.0', id: 'after-exit', method: 'initialize', params: { cwd: storageDir, provider: 'deepseek', model: 'x' } })
       await settle()
       expect(harness.frames().length).toBe(before)
     } finally {
@@ -257,7 +257,7 @@ describe('dsh-jsonrpc plugin apply', () => {
       expect(harness.outputErrors.map(error => error.message)).toEqual(['flush callback failed'])
 
       const before = harness.frames().length
-      harness.send({ jsonrpc: '2.0', id: 'after-flush-failure', method: 'initialize', params: { cwd: storageDir, model: 'x' } })
+      harness.send({ jsonrpc: '2.0', id: 'after-flush-failure', method: 'initialize', params: { cwd: storageDir, provider: 'deepseek', model: 'x' } })
       await settle()
       expect(harness.frames().length).toBe(before)
     } finally {
@@ -281,7 +281,7 @@ describe('dsh-jsonrpc plugin apply', () => {
       await harness.fiber.dispose()
 
       const before = harness.frames().length
-      harness.send({ jsonrpc: '2.0', id: 'probe-2', method: 'initialize', params: { cwd: storageDir, model: 'x' } })
+      harness.send({ jsonrpc: '2.0', id: 'probe-2', method: 'initialize', params: { cwd: storageDir, provider: 'deepseek', model: 'x' } })
       await settle()
       expect(harness.frames().length).toBe(before)
       expect(harness.exits()).toEqual([])
diff --git a/packages/ui/jsonrpc/tests/server.spec.ts b/packages/ui/jsonrpc/tests/server.spec.ts
index 8c024c6e81..22f01178ec 100644
--- a/packages/ui/jsonrpc/tests/server.spec.ts
+++ b/packages/ui/jsonrpc/tests/server.spec.ts
@@ -58,7 +58,7 @@ async function mockCompletionServer(): Promise<{ url: string; requests: unknown[
 
 async function makeHarness(storageDir: string) {
   const ctx = new Context()
-  await ctx.plugin(agentCore)
+  await ctx.plugin(agentCore, { workspaceContext: false })
   await ctx.plugin(SubagentService)
   await ctx.plugin(SessionPersistenceJsonl, { root: storageDir })
   await new Promise(resolve => setTimeout(resolve, 50))
@@ -107,6 +107,7 @@ describe('HarnessSdkServer', () => {
 
       const init = await server.handleRequest('initialize', {
         cwd: storageDir,
+        provider: 'deepseek',
         model: 'dsagent-model',
       }) as { serverInfo: { name: string } }
       expect(init.serverInfo.name).toBe('deepseek-harness-sdk-runtime')
@@ -138,7 +139,7 @@ describe('HarnessSdkServer', () => {
         agentId: AgentId('orphan-agent'),
         sessionId: SessionId('orphan-session'),
         meta: { cwd: storageDir },
-        agentOptions: { model: 'dsagent-model' },
+        agentOptions: { provider: 'deepseek', model: 'dsagent-model' },
       })
       orphanHandle.agent.send([{ type: 'text', text: 'outside the sdk session map' }])
       await orphanHandle.agent.whenIdle()
@@ -241,7 +242,7 @@ describe('HarnessSdkServer', () => {
     try {
       const server = new HarnessSdkServer(ctx, new FakeTransport())
 
-      await server.initialize({ cwd: storageDir, model: 'plain-model' })
+      await server.initialize({ cwd: storageDir, provider: 'deepseek', model: 'plain-model' })
       await server.prompt({
         sessionId: 'plain',
         contentBlocks: [{ type: 'text', text: 'hello' }],
@@ -266,13 +267,13 @@ describe('HarnessSdkServer', () => {
         agentId: AgentId('parent-agent'),
         sessionId: SessionId('main'),
         meta: { cwd: storageDir },
-        agentOptions: { model: 'deepseek' },
+        agentOptions: { provider: 'deepseek', model: 'deepseek' },
       })
       const handle = await ctx.agents.create({
         agentId: AgentId('child-agent'),
         sessionId: SessionId('child-session'),
         meta: { cwd: storageDir, parentSession: SessionId('main') },
-        agentOptions: { model: 'deepseek' },
+        agentOptions: { provider: 'deepseek', model: 'deepseek' },
       })
       await settleSubagent(ctx, parentHandle.agent, {
         provider: 'spawn',
@@ -314,19 +315,19 @@ describe('HarnessSdkServer', () => {
         agentId: AgentId('fallback-parent-agent'),
         sessionId: SessionId('fallback-parent'),
         meta: { cwd: storageDir },
-        agentOptions: { model: 'deepseek' },
+        agentOptions: { provider: 'deepseek', model: 'deepseek' },
       })
       handle = await ctx.agents.create({
         agentId: AgentId('fallback-child-agent'),
         sessionId: SessionId('fallback-child-session'),
         meta: { cwd: storageDir, parentSession: SessionId('fallback-parent') },
-        agentOptions: { model: 'deepseek' },
+        agentOptions: { provider: 'deepseek', model: 'deepseek' },
       })
       failedHandle = await ctx.agents.create({
         agentId: AgentId('failed-child-agent'),
         sessionId: SessionId('failed-child-session'),
         meta: { cwd: storageDir },
-        agentOptions: { model: 'deepseek' },
+        agentOptions: { provider: 'deepseek', model: 'deepseek' },
       })
       const transport = new FakeTransport()
       const server = new HarnessSdkServer(ctx, transport)
@@ -385,20 +386,20 @@ describe('HarnessSdkServer', () => {
     }
   })
 
-  it('does not re-register an LLM adapter that already exists', async () => {
+  it('does not re-register an LLM adapter whose provider already has an owner', async () => {
     const storageDir = await mkdtemp(join(tmpdir(), 'dsh-jsonrpc-existing-llm-'))
     const ctx = await makeHarness(storageDir)
     vi.stubEnv('DEEPSEEK_API_KEY', 'test-key')
-    await ctx.plugin(LlmDeepSeek, { models: ['preinstalled-model'] })
+    await ctx.plugin(LlmDeepSeek)
     try {
       const server = new HarnessSdkServer(ctx, new FakeTransport())
-      const inspect = server as unknown as { hasAdapterFor(model: string): boolean }
+      const inspect = server as unknown as { hasAdapterFor(provider: string): boolean }
 
-      expect(inspect.hasAdapterFor('preinstalled-model')).toBe(true)
-      expect(inspect.hasAdapterFor('missing-model')).toBe(false)
-      await server.initialize({ cwd: storageDir, model: 'preinstalled-model' })
+      expect(inspect.hasAdapterFor('deepseek')).toBe(true)
+      expect(inspect.hasAdapterFor('missing-provider')).toBe(false)
+      await server.initialize({ cwd: storageDir, provider: 'deepseek', model: 'preinstalled-model' })
 
-      expect(ctx.get('llm')?.models().filter(model => model === 'preinstalled-model')).toEqual(['preinstalled-model'])
+      expect(ctx.get('llm')?.listProviders().filter(provider => provider.id === 'deepseek')).toEqual([{ id: 'deepseek', name: 'DeepSeek' }])
       await server.shutdown()
     } finally {
       await ctx.fiber.dispose()
@@ -406,17 +407,18 @@ describe('HarnessSdkServer', () => {
     }
   })
 
-  it('registers a missing model when an LLM service already exists', async () => {
+  it('rejects a missing non-DeepSeek provider when an LLM service already exists', async () => {
     const storageDir = await mkdtemp(join(tmpdir(), 'dsh-jsonrpc-new-llm-'))
     const ctx = await makeHarness(storageDir)
     vi.stubEnv('DEEPSEEK_API_KEY', 'test-key')
-    await ctx.plugin(LlmDeepSeek, { models: ['other-model'] })
+    await ctx.plugin(LlmDeepSeek)
     try {
       const server = new HarnessSdkServer(ctx, new FakeTransport())
 
-      await server.initialize({ cwd: storageDir, model: 'new-model' })
+      await expect(server.initialize({ cwd: storageDir, provider: 'private', model: 'new-model' }))
+        .rejects.toThrow('no adapter registered for provider "private"')
 
-      expect(ctx.get('llm')?.models()).toEqual(expect.arrayContaining(['other-model', 'new-model']))
+      expect(ctx.get('llm')?.listProviders()).toEqual([{ id: 'deepseek', name: 'DeepSeek' }])
       await server.shutdown()
     } finally {
       await ctx.fiber.dispose()
@@ -517,15 +519,15 @@ describe('HarnessSdkServer', () => {
     const ctx = {
       on: vi.fn(() => () => undefined),
       agents: { create, get: () => undefined },
-      get: () => ({ models: () => ['model'] }),
+      get: () => ({ listProviders: () => [{ id: 'mock', name: 'Mock' }] }),
     } as unknown as Context
     const server = new HarnessSdkServer(ctx, new FakeTransport()) as unknown as {
-      initialize(params: { cwd: string; model: string }): Promise<unknown>
+      initialize(params: { cwd: string; provider: string; model: string }): Promise<unknown>
       getOrCreateSession(sessionId: string): Promise<unknown>
       shutdown(): Promise<Record<string, never>>
     }
 
-    await server.initialize({ cwd: '.', model: 'model' })
+    await server.initialize({ cwd: '.', provider: 'mock', model: 'model' })
     await server.getOrCreateSession('relative')
 
     expect(create).toHaveBeenCalledWith(expect.objectContaining({ meta: { cwd: process.cwd() } }))
diff --git a/packages/ui/user-approval/README.md b/packages/ui/user-approval/README.md
index 571dc0e7b8..cebd25275e 100644
--- a/packages/ui/user-approval/README.md
+++ b/packages/ui/user-approval/README.md
@@ -6,7 +6,7 @@ Each request must belong to an open agent turn. The service appends a paired `ap
 
 Answerers are `approval/request` waterfall listeners. Return an outcome to answer for an owned agent or call `next()` to delegate. Agent-scoped listeners receive only that agent's requests; compose one terminal answerer per deployment because sibling listener order is not a policy priority mechanism. The ACP bridge is the shipped human answerer.
 
-`ApprovalPolicy` is `'ask'` or `'never'`. The effective value is the last `approval/policy` event, falling back to config; `setApprovalPolicy()` is the write path. `'never'` rejects before interactive dispatch and is the only policy stated in the prompt. Switches produce at most one coalesced notice whose header marker distinguishes user changes from operator/config changes.
+`ApprovalPolicy` is `'ask'` or `'never'`. The effective value is the last `approval/policy` event, falling back to config; `setApprovalPolicy()` is the write path. `'never'` rejects before interactive dispatch and is the only policy stated in the prompt. Switches produce at most one coalesced notice, attributed to the user when the override follows the last `request/header` and to operator/config otherwise.
 
 The tools pipeline routes `ask` decisions through this seam and fails closed when it is absent; the sandboxed bash tool also uses it for escalated retries. The ACP bridge is the shipped human answerer for calls it owns. Audit events remain log-only, so the model sees only the asking consumer's result. See the [approval-seam RFC](../../../docs/rfc/implemented/feature/2026-07-06-approval-seam.md) and [sandbox RFC](../../../docs/rfc/implemented/feature/2026-07-06-sandbox.md).
 
diff --git a/packages/ui/user-approval/src/index.ts b/packages/ui/user-approval/src/index.ts
index b9d757964d..69a243a854 100644
--- a/packages/ui/user-approval/src/index.ts
+++ b/packages/ui/user-approval/src/index.ts
@@ -63,7 +63,7 @@ declare module '@deepseek-ai/dsh-session' {
      * from the prompt section and the narrator's notices). The LAST such
      * event is the session's override ({@link effectiveApprovalPolicy});
      * who asked for it is derivable from position (an event after the log's
-     * last `request/header*` was a runtime switch by the user).
+     * last `request/header` was a runtime switch by the user).
      */
     'approval/policy': { policy: ApprovalPolicy }
   }
@@ -258,7 +258,7 @@ export class ApprovalService extends Service {
     // narrated no later than the next step. What each session was last told
     // is in-memory with a log-derived fallback (the folded header's system
     // text), so restarts lose nothing. Attribution is positional: an
-    // override event after the log's last `request/header*` was a runtime
+    // override event after the log's last `request/header` was a runtime
     // switch by the user; otherwise the configured default moved under the
     // session (operator/config).
     const narrated = new WeakMap<Agent['session'], ApprovalPolicy>()
@@ -271,7 +271,7 @@ export class ApprovalService extends Service {
         const event = events[index] as (typeof events)[number]
         if (overrideIndex < 0 && event.type === 'approval/policy') {
           overrideIndex = index
-        } else if (headerIndex < 0 && (event.type === 'request/header' || event.type === 'request/header-delta')) {
+        } else if (headerIndex < 0 && event.type === 'request/header') {
           headerIndex = index
         }
       }
diff --git a/packages/ui/user-approval/tests/approval.spec.ts b/packages/ui/user-approval/tests/approval.spec.ts
index f248e5c468..9219734c9f 100644
--- a/packages/ui/user-approval/tests/approval.spec.ts
+++ b/packages/ui/user-approval/tests/approval.spec.ts
@@ -375,7 +375,7 @@ describe('approval policy (the approval/policy fold)', () => {
 
   /** Append a `request/header` snapshot whose system text is exactly `system`. */
   function appendHeader(session: Session, system: string): void {
-    session.append('request/header', { header: { config: { model: 'mock' }, system }, reason: 'initial' })
+    session.append('request/header', { header: { config: { provider: 'mock', model: 'mock' }, system }, reason: 'initial' })
   }
 
   it('folds to the last event, or undefined without one', () => {
diff --git a/packages/util/README.md b/packages/util/README.md
index 45afe7b0a9..4b4f23ddf7 100644
--- a/packages/util/README.md
+++ b/packages/util/README.md
@@ -5,8 +5,15 @@ Zero-dependency primitives shared across the other groups. A package lands here
 | Package | Role |
 |---|---|
 | `brand/` | The type-only `Branded<B>` nominal-typing primitive (no runtime code, no harness deps) |
+| `home/` | Canonical `DSH_HOME` resolution from explicit config, environment, or `~/.dsh` (no harness deps) |
+| `paths/` | Shared filesystem path constants and helpers for harness user data |
 | `timeout/` | The timing/classification half of a timeout — `clampTimeout`/`deadline`/`timeoutOf`/`TimeoutReason` (pure functions, no harness deps); termination stays in each capability |
+| `retention/` | Bounded model-facing output — `ItemRetainer`/`TextRetainer` + neutral notice helpers (pure, no harness deps); business semantics stay in each tool |
 
-`dsh-brand` is the canonical case: it owns ONLY the `Branded<B>` helper, so a capability package can brand the ids it owns (`dsh-bash`'s `BashTaskId`/`OwnerToken`, `dsh-session`'s `SessionId`, …) by depending on `dsh-brand` alone, without pulling in an unrelated package just to reach `Branded`.
+`dsh-brand` is the canonical case: it owns ONLY the `Branded<B>` helper, so a capability package can brand the ids it owns (`dsh-tasks`'s `TaskId`, `dsh-session`'s `SessionId`, …) by depending on `dsh-brand` alone, without pulling in an unrelated package just to reach `Branded`.
+
+`dsh-home` gives every package the same configurable Harness home without assigning that cross-cutting fact to bash, skills, or a composition bundle. It resolves an explicit value before `$DSH_HOME`, falls back to `~/.dsh`, and returns an absolute path without caching, creating, or mutating anything.
 
 `dsh-timeout` follows the same shape for the timeout family: `dsh-bash` and `dsh-web-fetch-local` each fuse a caller's cancellation with a deadline and later classify "timed out" vs "cancelled" by depending on `dsh-timeout` alone. It deliberately owns only the timing/classification half — the *termination* (SIGKILL a process group, tear down a fetch socket) stays in each capability, because no shared layer can own every capability's kill (see [the timeout-library RFC](../../docs/rfc/implemented/architecture/2026-07-06-timeout-deadline-library.md)).
+
+`dsh-retention` is the same split for bounded tool output: a tool (`glob`/`grep`/`bash`/`web_fetch`/`web_search`) feeds items or text into a retainer and gets back what it kept and exactly what it omitted — while grouping, exit codes, provider errors, and recovery prose stay tool-owned. It deliberately owns only the retention mechanic; `truncated` is a budget fact, never an "incomplete inspection" state (see [the retention-library RFC](../../docs/rfc/implemented/architecture/2026-07-06-tool-result-retention-library.md)).
diff --git a/packages/util/brand/README.md b/packages/util/brand/README.md
index 8f7943def7..46e6c1ff98 100644
--- a/packages/util/brand/README.md
+++ b/packages/util/brand/README.md
@@ -17,10 +17,10 @@ export function SessionId(id: string): SessionId {
 }
 ```
 
-Construction goes through the per-id factory in the OWNING package (a plain cast inside — zero runtime cost). Comparison, logging, JSON serialization, and the wire format all behave exactly as for an ordinary string; the brand is erased at compile time.
+Construction goes through the per-id factory in the owning package. Comparison, logging, JSON serialization, and the wire format behave as for an ordinary string; the brand is erased at compile time.
 
 ## Policy: brand ids that cross package boundaries
 
-A package brands the ids it OWNS — `CallId` in `dsh-llm` (tool-call correlation), `SessionId` in `dsh-session`, `AgentId` in `dsh-agent`, `BashTaskId`/`OwnerToken` in `dsh-bash`. Branding is for ids that cross package boundaries and could plausibly be confused; **not every string needs a brand.**
+A package brands the ids it owns — `CallId` in `dsh-llm`, `SessionId` in `dsh-session`, `AgentId` in `dsh-agent`, and `TaskId` in `dsh-tasks`. Brand cross-package ids that could plausibly be confused; not every string needs one.
 
-This package owns ONLY the primitive — no concrete id, no runtime code beyond the (erased) type. Keeping the primitive dependency-free is the point: a capability package can brand its ids without depending on an unrelated package. `dsh-bash`, for example, brands `BashTaskId`/`OwnerToken` by depending on `dsh-brand` alone — it never pulls in `dsh-llm` (or `dsh-session`) just to reach `Branded`.
+This package owns only the primitive. Keeping it dependency-free lets `dsh-tasks`, for example, brand `TaskId` without importing an unrelated capability package merely to reach `Branded`.
diff --git a/packages/util/brand/src/index.ts b/packages/util/brand/src/index.ts
index 0e880c52da..0c669e2416 100644
--- a/packages/util/brand/src/index.ts
+++ b/packages/util/brand/src/index.ts
@@ -1,9 +1,23 @@
 /**
- * Dependency-free nominal typing for cross-boundary identifiers. Structurally identical runtime
- * strings become non-interchangeable statically while retaining ordinary comparison, logging, and
- * serialization. Each owning package defines its concrete id and zero-cost factory; brand ids that
- * can plausibly be confused across packages, not arbitrary strings. This package exports only the
- * erased primitive so an owner need not depend on another capability package.
+ * The `Branded<B>` nominal-typing primitive — a type-only utility (no runtime
+ * code, no harness-package dependency) shared by every package that owns a
+ * cross-boundary id.
+ *
+ * A brand makes structurally-identical strings non-interchangeable at the type
+ * level: an `AgentId` cannot be passed where a `CallId` is expected, even
+ * though both are plain strings at runtime. Construction goes through a per-id
+ * factory in the OWNING package (a plain cast inside — zero runtime cost);
+ * comparison, logging, and serialization all behave as ordinary strings.
+ *
+ * Policy: a package brands the ids it owns — `CallId` in dsh-llm (tool-call
+ * correlation), `SessionId` in dsh-session, `AgentId` in dsh-agent, and
+ * `TaskId` in dsh-tasks. Branding is for ids that cross package boundaries and
+ * could plausibly be confused; not every string needs a brand.
+ * This package owns ONLY the primitive — no concrete id, no runtime code beyond
+ * the (erased) type — so the brand vocabulary stays dependency-free and a
+ * package can brand its ids without depending on an unrelated capability
+ * package.
+ *
  * @module @deepseek-ai/dsh-brand
  */
 
diff --git a/packages/util/home/README.md b/packages/util/home/README.md
new file mode 100644
index 0000000000..f9107a8706
--- /dev/null
+++ b/packages/util/home/README.md
@@ -0,0 +1,17 @@
+# @deepseek-ai/dsh-home
+
+`@deepseek-ai/dsh-home` is the single owner of DeepSeek Harness home-directory resolution. `resolveDshHome(configured?)` returns an absolute path using this precedence:
+
+1. The explicit `configured` path.
+2. The `DSH_HOME` environment variable.
+3. The `.dsh` directory under the current user's home directory.
+
+The resolver reads its inputs at call time. It does not cache a result, create the directory, or mutate `process.env`; consumers keep ownership of their own configuration fields and pass the configured value when resolving the shared home.
+
+## Model Experience
+
+Indirectly, through `dsh-tool-bash`, which exposes the resolved path to model bash as `DSH_HOME` without adding a prompt section.
+
+## Known Limitations and Deferred Work
+
+- **Resolution only** — the resolver makes a path absolute but does not create it, check access, or canonicalize symlinks; each consumer owns those filesystem decisions.
diff --git a/packages/util/home/package.json b/packages/util/home/package.json
new file mode 100644
index 0000000000..efeaf4832c
--- /dev/null
+++ b/packages/util/home/package.json
@@ -0,0 +1,30 @@
+{
+  "name": "@deepseek-ai/dsh-home",
+  "description": "Canonical DeepSeek Harness home-directory resolver",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "cordis": "^4.0.0-rc.6"
+  },
+  "devDependencies": {
+    "cordis": "^4.0.0-rc.6"
+  }
+}
diff --git a/packages/util/home/src/index.ts b/packages/util/home/src/index.ts
new file mode 100644
index 0000000000..4e3d56b54b
--- /dev/null
+++ b/packages/util/home/src/index.ts
@@ -0,0 +1,23 @@
+/**
+ * Canonical DeepSeek Harness home-directory resolution.
+ *
+ * @module @deepseek-ai/dsh-home
+ */
+
+import { homedir } from 'node:os'
+import { join, resolve } from 'node:path'
+
+const DEFAULT_DSH_HOME_DIRNAME = '.dsh'
+
+/** Environment variable that overrides the default Harness home directory. */
+export const DSH_HOME_ENV = 'DSH_HOME' as const
+
+/**
+ * Resolve the DeepSeek Harness home directory without caching or mutating the environment.
+ *
+ * @param configured - Optional configured path, which takes precedence over the environment.
+ * @returns The absolute configured path, `$DSH_HOME`, or `~/.dsh`, in that order.
+ */
+export function resolveDshHome(configured?: string): string {
+  return resolve(configured ?? process.env[DSH_HOME_ENV] ?? join(homedir(), DEFAULT_DSH_HOME_DIRNAME))
+}
diff --git a/packages/util/home/tests/home.spec.ts b/packages/util/home/tests/home.spec.ts
new file mode 100644
index 0000000000..3ebde50bee
--- /dev/null
+++ b/packages/util/home/tests/home.spec.ts
@@ -0,0 +1,26 @@
+import { homedir } from 'node:os'
+import { join, resolve } from 'node:path'
+import { afterEach, describe, expect, it, vi } from 'vitest'
+import { DSH_HOME_ENV, resolveDshHome } from '@deepseek-ai/dsh-home'
+
+afterEach(() => vi.unstubAllEnvs())
+
+describe('resolveDshHome', () => {
+  it('prefers an explicit configured path and resolves it absolutely', () => {
+    vi.stubEnv(DSH_HOME_ENV, './environment-home')
+
+    expect(resolveDshHome('./configured-home')).toBe(resolve('./configured-home'))
+  })
+
+  it('uses DSH_HOME when no configured path is supplied', () => {
+    vi.stubEnv(DSH_HOME_ENV, './environment-home')
+
+    expect(resolveDshHome()).toBe(resolve('./environment-home'))
+  })
+
+  it('defaults to the .dsh directory under the user home', () => {
+    vi.stubEnv(DSH_HOME_ENV, undefined)
+
+    expect(resolveDshHome()).toBe(join(homedir(), '.dsh'))
+  })
+})
diff --git a/packages/util/home/tsconfig.json b/packages/util/home/tsconfig.json
new file mode 100644
index 0000000000..9770ef25d6
--- /dev/null
+++ b/packages/util/home/tsconfig.json
@@ -0,0 +1,9 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": ["src"],
+  "references": []
+}
diff --git a/packages/util/paths/README.md b/packages/util/paths/README.md
new file mode 100644
index 0000000000..2668e679b6
--- /dev/null
+++ b/packages/util/paths/README.md
@@ -0,0 +1,18 @@
+# dsh-paths
+
+Shared filesystem path helpers for DeepSeek Harness user data.
+
+## DSH home
+
+`DSH_HOME_DIR_NAME` owns the default user-data directory name: `.dsh`.
+
+`defaultDshHome()` returns the default DeepSeek Harness home by joining the operating-system home directory with `.dsh`, using Node's platform path rules.
+
+`expandHomePath()` expands `~`, `~/...`, and Windows-style `~\...` prefixes against the operating-system home directory. It leaves non-tilde paths and `~user/...` untouched.
+
+This package is intentionally small and harness-dep-free so product packages can share user-data path conventions without depending on one another.
+
+## Known Limitations and Deferred Work
+
+- **Expansion is deliberately narrow** — only bare `~`, `~/...`, and `~\...` use the current operating-system home; named-user forms such as `~alice/...`, environment variables, and shell expressions remain unchanged.
+- **Helpers do not touch the filesystem** — callers still own directory creation, existence checks, permissions, and trust policy for the resulting path.
diff --git a/packages/util/paths/package.json b/packages/util/paths/package.json
new file mode 100644
index 0000000000..b4f760afe9
--- /dev/null
+++ b/packages/util/paths/package.json
@@ -0,0 +1,30 @@
+{
+  "name": "@deepseek-ai/dsh-paths",
+  "description": "Shared filesystem path helpers for the DeepSeek Harness",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "cordis": "^4.0.0-rc.6"
+  },
+  "devDependencies": {
+    "cordis": "^4.0.0-rc.6"
+  }
+}
diff --git a/packages/util/paths/src/index.ts b/packages/util/paths/src/index.ts
new file mode 100644
index 0000000000..89e188cedd
--- /dev/null
+++ b/packages/util/paths/src/index.ts
@@ -0,0 +1,47 @@
+/**
+ * Shared filesystem path helpers for DeepSeek Harness user data.
+ *
+ * @module @deepseek-ai/dsh-paths
+ */
+
+import { homedir } from 'node:os'
+import { join, resolve } from 'node:path'
+
+/** Directory name for the default DeepSeek Harness home under the OS home. */
+export const DSH_HOME_DIR_NAME = '.dsh'
+
+/** Stable user-facing display form for the default DeepSeek Harness home. */
+export const DEFAULT_DSH_HOME_DISPLAY = `~/${DSH_HOME_DIR_NAME}`
+
+/** Environment variable that overrides the default DeepSeek Harness home. */
+export const DSH_HOME_ENV = 'DSH_HOME'
+
+/**
+ * Resolve the default DeepSeek Harness home using Node's platform path rules.
+ * @returns the absolute default harness home path.
+ */
+export function defaultDshHome(): string {
+  return join(homedir(), DSH_HOME_DIR_NAME)
+}
+
+/**
+ * Expand supported tilde prefixes against the operating-system home.
+ * @param path - configured path that may begin with `~`, `~/`, or `~\`.
+ * @returns the expanded path, or the original value when no supported prefix is present.
+ */
+export function expandHomePath(path: string): string {
+  if (path === '~') return homedir()
+  if (path.startsWith('~/') || path.startsWith('~\\')) return join(homedir(), path.slice(2))
+  return path
+}
+
+/**
+ * Resolve an explicitly configured, environment-selected, or default DSH home.
+ * @param configured - explicit harness-home override, which has highest precedence.
+ * @param env - environment mapping used to read `DSH_HOME`.
+ * @returns the normalized absolute harness home path.
+ */
+export function resolveDshHome(configured?: string, env: Record<string, string | undefined> = process.env): string {
+  const selected = configured ?? env[DSH_HOME_ENV] ?? defaultDshHome()
+  return resolve(expandHomePath(selected))
+}
diff --git a/packages/util/paths/tests/paths.spec.ts b/packages/util/paths/tests/paths.spec.ts
new file mode 100644
index 0000000000..97e91a556e
--- /dev/null
+++ b/packages/util/paths/tests/paths.spec.ts
@@ -0,0 +1,34 @@
+import { homedir } from 'node:os'
+import { join } from 'node:path'
+import { describe, expect, it } from 'vitest'
+import {
+  DEFAULT_DSH_HOME_DISPLAY,
+  DSH_HOME_DIR_NAME,
+  defaultDshHome,
+  expandHomePath,
+  resolveDshHome,
+} from '@deepseek-ai/dsh-paths'
+
+describe('dsh path helpers', () => {
+  it('owns the shared default DSH home directory name', () => {
+    expect(DSH_HOME_DIR_NAME).toBe('.dsh')
+    expect(DEFAULT_DSH_HOME_DISPLAY).toBe('~/.dsh')
+    expect(defaultDshHome()).toBe(join(homedir(), '.dsh'))
+  })
+
+  it('expands tilde paths without changing non-tilde paths', () => {
+    expect(expandHomePath('~')).toBe(homedir())
+    expect(expandHomePath('~/.dsh')).toBe(join(homedir(), '.dsh'))
+    expect(expandHomePath('~\\.dsh')).toBe(join(homedir(), '.dsh'))
+    expect(expandHomePath('/tmp/.dsh')).toBe('/tmp/.dsh')
+    expect(expandHomePath('~other/.dsh')).toBe('~other/.dsh')
+  })
+
+  it('resolves explicit DSH home before environment and default locations', () => {
+    const envHome = join(homedir(), 'env-dsh')
+
+    expect(resolveDshHome(undefined, { DSH_HOME: '~/env-dsh' })).toBe(envHome)
+    expect(resolveDshHome('/tmp/explicit-dsh', { DSH_HOME: '~/env-dsh' })).toBe('/tmp/explicit-dsh')
+    expect(resolveDshHome(undefined, {})).toBe(defaultDshHome())
+  })
+})
diff --git a/packages/util/paths/tsconfig.json b/packages/util/paths/tsconfig.json
new file mode 100644
index 0000000000..749cb0208e
--- /dev/null
+++ b/packages/util/paths/tsconfig.json
@@ -0,0 +1,11 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": [
+    "src"
+  ],
+  "references": []
+}
diff --git a/packages/util/retention/README.md b/packages/util/retention/README.md
new file mode 100644
index 0000000000..6ca095be41
--- /dev/null
+++ b/packages/util/retention/README.md
@@ -0,0 +1,91 @@
+# dsh-retention
+
+A dependency-light **retention** library: bounded model-facing output for tools that must cap how much context they return. A caller feeds items or text chunks into a bounded object, then gets the retained content plus exact omission metadata.
+
+The library owns **only** the mechanical question *"what did we keep, and what did we omit?"*. Tool-specific code keeps its business semantics: file grouping, line numbering, exit codes, provider error states, per-line preview truncation, spill files, and the model-facing prose. This is the boundary the [RFC](../../../docs/rfc/implemented/architecture/2026-07-06-tool-result-retention-library.md) draws.
+
+It is a **library, not a service or plugin**: no `ctx`, registers nothing, emits no events. The only state is per-retainer (one accumulation), never cross-call. Tool packages import it directly.
+
+## Surface
+
+```ts
+import {
+  ItemRetainer, TextRetainer,
+  describeOmitted, formatRetentionNotice,
+} from '@deepseek-ai/dsh-retention'
+import type {
+  Omitted, PushDecision, RetainedItems, RetainedText,
+  ItemRetentionStrategy, TextRetentionStrategy, RetentionNotice,
+} from '@deepseek-ai/dsh-retention'
+```
+
+| Export | Role |
+|---|---|
+| `ItemRetainer<T>` | Bounds ordered logical units (paths, grep matches, sources). `head` only in v1. `push()` → `PushDecision`; `finish()` → `RetainedItems<T>`. |
+| `TextRetainer` | Bounds a byte-oriented text stream. `head` / `tail` / `headTail`, UTF-8 boundaries preserved at `finish()`. `push()` → `PushDecision`; `finish()` → `RetainedText`. |
+| `describeOmitted(omitted, unit)` | Standardized omission clause (`exact` prints a count; `unknown` does not). |
+| `formatRetentionNotice(notice, recovery)` | Joins the standardized omission clause with the tool's own recovery guidance. |
+| `Omitted` | `none` / `exact` / `unknown` — how much was omitted. |
+| `PushDecision` | `{ kept, truncated }` — the per-push retention result. |
+
+## Resource Modes
+
+The two retainers are separate names, not one generic collector, because they differ in **resource model**.
+
+- **`ItemRetainer` bounds ordered logical units.** A search tool can collect a full result set for spill-file recovery while retaining only the first `maxItems` for the model-facing preview. The omission count is exact because the caller keeps feeding every observed item.
+- **`TextRetainer` bounds byte-oriented text.** `head`, `tail`, and `headTail` preserve UTF-8 boundaries at `finish()`; `headTail` is the shape `dsh-spill-policy` uses to build a bounded preview around a spill-file notice.
+
+## `truncated` is a budget fact, never "incomplete"
+
+`truncated` means *the retainer omitted otherwise-available content because of a budget*. It does **not** mean the upstream was incomplete. Permission failures, skipped binary files, provider partial failures, unreadable candidates, and invalid UTF-8 stay in tool-domain fields — never folded into `truncated`. Conflating the two is the bug this library's naming most invites; keep them separate.
+
+## Bytes, not characters
+
+Text caps and `omittedBytes` count **bytes**, for process/body safety (a child's pipe and an HTTP body are byte streams). A chunk that straddles a codepoint is handled: `finish()` trims a partial codepoint at each cut so the returned text never introduces a replacement char at the boundary, and the two sides are decoded separately so a codepoint is never reconstructed across the omitted middle. Character- or line-level preview budgets are a separate, tool-owned concern.
+
+## Tool mappings
+
+Every current retention consumer maps to the library below. A broad migration is out of scope for the library's first landing — these are the intended shapes.
+
+| Tool | Retainer & strategy | Notes |
+|---|---|---|
+| `glob` | `ItemRetainer<FsGlobEntry>`, `head` | Collect the full sorted path list for a spill file while retaining the first page inline. Path mapping, skipped candidates, and `incomplete` stay outside. |
+| `grep` | `ItemRetainer<FlatGrepMatch>`, `head` | Collect matches for a spill file while retaining the first page inline. Per-match preview truncation, grouping, sorting, and `incomplete` stay outside. |
+| `bash` | `TextRetainer`, `tail` or `headTail` | Executor still owns spill files, exit status, signal, timeout, and background tasks. |
+| `web_fetch` | `TextRetainer`, `head` or `headTail` | Provider/resource caps stay provider facts; the retainer supplies only retained text and omission metadata. |
+| `web_search` | `ItemRetainer<WebSearchSource>`, `head` | Standardizes the "sources capped" notice when providers return more sources than the model-facing result should include. |
+
+`read` is **intentionally out of scope for v1.** Its `read-render` helper owns a file-specific pagination contract — `offset`/`limit`, line numbers, `totalLines`, offset-out-of-range errors, per-line preview truncation, a byte cap over the selected window — which is a line-window renderer, not generic retention. A single `Omitted` count cannot represent both sides of a line window.
+
+## Usage shape
+
+```ts ignore-check
+// glob: keep the first page inline while still collecting the full list for spill.
+const retainer = new ItemRetainer<FsGlobEntry>({ kind: 'head', maxItems: globMaxResults })
+const allEntries: FsGlobEntry[] = []
+for await (const entry of candidates) {
+  allEntries.push(entry)
+  retainer.push(entry)
+}
+const { items, truncated, omitted } = retainer.finish()
+
+// bash: keep a head + tail, read to process exit.
+const out = new TextRetainer({ kind: 'headTail', headBytes: headCap, tailBytes: tailCap })
+child.stdout.on('data', (chunk: Buffer) => { out.push(chunk) })
+const { text, omittedBytes } = out.finish()
+
+// A footer: the library standardizes the omission clause; the tool owns recovery words.
+const footer = formatRetentionNotice(
+  { scope: 'grep', strategy: 'head', unit: 'items', limit: grepMaxMatches, kept: items.length, omitted },
+  ({ kept }) => `Results capped at ${kept}. Narrow the pattern, path, or include to see more.`,
+)
+```
+
+## Model Experience
+
+Indirectly, through tool consumers that render retained content and omission metadata.
+
+## Known Limitations and Deferred Work
+
+- **Item retention supports `head` only** — tail, head/tail, pagination, grouping, and provider-completeness semantics remain tool-owned.
+- **Text retention is byte-oriented** — line and character windows such as `read` pagination require a separate renderer, and a cut may discard partial UTF-8 boundary bytes to keep returned text valid.
diff --git a/packages/util/retention/package.json b/packages/util/retention/package.json
new file mode 100644
index 0000000000..db8bab3342
--- /dev/null
+++ b/packages/util/retention/package.json
@@ -0,0 +1,30 @@
+{
+  "name": "@deepseek-ai/dsh-retention",
+  "description": "Zero-dependency bounded-retention primitive: ItemRetainer/TextRetainer + neutral notice helpers (what did we keep, what did we omit)",
+  "version": "0.0.1",
+  "private": true,
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "lib/types/**/*.d.ts.map",
+    "src"
+  ],
+  "license": "BSD-3-Clause",
+  "peerDependencies": {
+    "cordis": "^4.0.0-rc.6"
+  },
+  "devDependencies": {
+    "cordis": "^4.0.0-rc.6"
+  }
+}
diff --git a/packages/util/retention/src/index.ts b/packages/util/retention/src/index.ts
new file mode 100644
index 0000000000..07547a7d93
--- /dev/null
+++ b/packages/util/retention/src/index.ts
@@ -0,0 +1,444 @@
+/**
+ * A dependency-light **retention** library: bounded model-facing output for
+ * tools that must cap how much context they return. A caller feeds items or
+ * text chunks into a bounded object, then gets the retained content plus exact
+ * omission metadata ({@link RetainedItems} / {@link RetainedText}).
+ *
+ * The library owns ONLY the mechanical question "what did we keep, what did we
+ * omit?". Tool-specific code still owns
+ * business semantics: file grouping, line numbering, exit codes, provider error
+ * states, per-line preview truncation, spill files, and the model-facing prose.
+ * In particular {@link RetainedText.truncated}/{@link RetainedItems.truncated}
+ * means "the retainer omitted otherwise-available content because of a budget" —
+ * NOT "the upstream was incomplete". Permission failures, skipped binaries,
+ * provider partial failures, and unreadable candidates stay in tool-domain
+ * fields, never folded into `truncated`.
+ *
+ * This is deliberately a library, not a cordis service or plugin: it takes no
+ * `ctx`, registers nothing, and emits no events. The two retainers are the only
+ * stateful pieces and their state is per-instance (one accumulation), never
+ * cross-call. Tool packages import it directly when they need bounded output.
+ *
+ * The two retainers differ in resource model, which is why they are two names
+ * rather than one generic collector:
+ * - {@link ItemRetainer} bounds ordered logical units (paths, grep matches,
+ *   search sources). `head` retention only in v1.
+ * - {@link TextRetainer} bounds byte-oriented text streams (bash stdout/stderr,
+ *   web bodies). `head` / `tail` / `headTail`, preserving UTF-8 boundaries at
+ *   {@link TextRetainer.finish}.
+ *
+ * @module @deepseek-ai/dsh-retention
+ */
+
+/**
+ * How much content the retainer omitted.
+ *
+ * `exact` is the normal retainer shape: every unit/byte was observed, so the
+ * omitted count is precise. `unknown` is reserved for a caller that omits
+ * without a count; the retainers themselves never return it.
+ */
+export type Omitted =
+  | { kind: 'none' }
+  | { kind: 'exact'; count: number }
+  | { kind: 'unknown' }
+
+/**
+ * The caller receives this after each `push()`.
+ */
+export interface PushDecision {
+  /** Was this whole unit / all of this chunk's bytes retained (nothing dropped)? */
+  kept: boolean
+  /** Cumulative: has the retainer omitted anything due to the budget yet? */
+  truncated: boolean
+}
+
+/**
+ * Final result for ordered logical units.
+ *
+ * `seen` means units OBSERVED by the retainer, not necessarily the total in the
+ * upstream source. `kept` is `items.length`, surfaced explicitly so a notice
+ * formatter need not re-count.
+ */
+export interface RetainedItems<T> {
+  items: T[]
+  truncated: boolean
+  seen: number
+  kept: number
+  omitted: Omitted
+}
+
+/**
+ * Final result for text streams.
+ *
+ * The returned `text` is safe to hand to a formatter: the retainer adds no
+ * tool-specific headers, exit markers, XML tags, or recovery instructions, and
+ * `omittedBytes` counts BYTES (not characters or lines) — text retention is
+ * byte-oriented for process/body safety. UTF-8 boundaries at each cut are
+ * preserved, so `text` never carries a replacement char introduced by the cut
+ * itself.
+ */
+export interface RetainedText {
+  text: string
+  truncated: boolean
+  omittedBytes: Omitted
+}
+
+/** Item retention strategy. Only `head` in v1; windows/grouped budgets wait for a second consumer. */
+export type ItemRetentionStrategy = {
+  /** Keep the first `maxItems` units. Use for `glob`, `grep`, and web sources. */
+  kind: 'head'
+  maxItems: number
+}
+
+/** Text retention strategy: keep a prefix, a suffix, or both, counted in bytes. */
+export type TextRetentionStrategy =
+  | {
+    /** Keep the first `maxBytes` bytes. */
+    kind: 'head'
+    maxBytes: number
+  }
+  | {
+    /** Keep the final `maxBytes` bytes. Requires reading to the end. */
+    kind: 'tail'
+    maxBytes: number
+  }
+  | {
+    /** Keep a stable prefix and suffix, omitting the middle. Requires reading to the end. */
+    kind: 'headTail'
+    headBytes: number
+    tailBytes: number
+  }
+
+/**
+ * A neutral, tool-agnostic description of one retention outcome — the input to
+ * {@link formatRetentionNotice}. It carries the mechanical facts (strategy,
+ * unit, limit, kept count, {@link Omitted}); the tool supplies the recovery
+ * words, because only the tool knows the recovery action ("narrow the pattern",
+ * "fetch a more specific URL", "read the spill file").
+ */
+export interface RetentionNotice {
+  /** Tool/scope label, e.g. `grep`, `web_fetch`, `bash stdout`. */
+  scope: string
+  strategy: 'head' | 'tail' | 'headTail'
+  unit: 'items' | 'bytes' | 'chars' | 'lines'
+  limit: number | { head: number; tail: number }
+  kept: number
+  omitted: Omitted
+}
+
+/** Assert a budget field is a non-negative integer (the retainer request contract). */
+function assertBudget(value: number, name: string): void {
+  if (!Number.isInteger(value) || value < 0) {
+    throw new Error(`${name} must be a non-negative integer`)
+  }
+}
+
+/**
+ * Bounds an ordered stream of logical units, keeping the first `maxItems`
+ * ({@link ItemRetentionStrategy} `head`). `push()` reports, per unit, whether it
+ * was kept and whether the retained result is now truncated.
+ *
+ * Grouping, sorting, path mapping, per-unit preview truncation, and any
+ * `incomplete` state stay OUTSIDE the retainer: it counts and keeps, nothing
+ * more. The caller pushes already-shaped units and, after {@link finish},
+ * groups/sorts the retained subset itself.
+ */
+export class ItemRetainer<T> {
+  private readonly maxItems: number
+  private readonly items: T[] = []
+  private seen = 0
+  private omittedCount = 0
+
+  /** @param strategy Head strategy: `maxItems` (non-negative integer). */
+  constructor(strategy: ItemRetentionStrategy) {
+    assertBudget(strategy.maxItems, 'maxItems')
+    this.maxItems = strategy.maxItems
+  }
+
+  /**
+   * Offer one unit. Kept when the retainer is below `maxItems`; otherwise dropped
+   * and counted as omitted. Callers keep pushing all observed units, so the final
+   * {@link Omitted} count is exact.
+   *
+   * @param item The already-shaped logical unit (path, flat match, source).
+   * @returns The per-push {@link PushDecision}.
+   */
+  push(item: T): PushDecision {
+    this.seen++
+    if (this.items.length < this.maxItems) {
+      // Reached only below the cap, before any omission (items only grow, the
+      // cap is fixed), so nothing has been dropped yet: truncated is always false.
+      this.items.push(item)
+      return { kept: true, truncated: false }
+    }
+    this.omittedCount++
+    return {
+      kept: false,
+      truncated: true,
+    }
+  }
+
+  /**
+   * Finalize and report what was kept and omitted.
+   *
+   * @returns The {@link RetainedItems} snapshot (safe to group/sort downstream).
+   */
+  finish(): RetainedItems<T> {
+    const truncated = this.omittedCount > 0
+    return {
+      items: this.items,
+      truncated,
+      seen: this.seen,
+      kept: this.items.length,
+      omitted: truncated
+        ? { kind: 'exact', count: this.omittedCount }
+        : { kind: 'none' },
+    }
+  }
+}
+
+const encoder = new TextEncoder()
+const decoder = new TextDecoder() // utf-8, non-fatal: internal malformed bytes → U+FFFD
+
+/**
+ * Drop a trailing incomplete UTF-8 sequence so a prefix cut never emits a
+ * replacement char at the boundary. Walks back over continuation bytes
+ * (`10xxxxxx`) to the lead byte; if fewer bytes follow it than the lead byte's
+ * length declares, the sequence is incomplete and is trimmed. A complete tail,
+ * or a run too long/short to be a valid lead, is returned untouched (any
+ * genuinely malformed interior is left for the decoder to replace).
+ */
+function trimTrailingPartialUtf8(bytes: Uint8Array): Uint8Array {
+  let i = bytes.length - 1
+  // Continuation bytes are 0b10xxxxxx; scan back at most 3 (max sequence is 4).
+  // Indices are bounds-checked by the loop guard, so the reads are in range (a
+  // cast, not `!`, per the repo's no-non-null-assertion rule).
+  while (i >= 0 && ((bytes[i] as number) & 0xc0) === 0x80 && bytes.length - i <= 3) i--
+  if (i < 0) return bytes
+  const lead = bytes[i] as number
+  const expected = lead < 0x80 ? 1 : lead < 0xe0 ? 2 : lead < 0xf0 ? 3 : lead < 0xf8 ? 4 : 0
+  // expected 0 → not a lead byte (stray continuation / invalid): leave it.
+  if (expected === 0) return bytes
+  return bytes.length - i < expected ? bytes.subarray(0, i) : bytes
+}
+
+/**
+ * Drop leading continuation bytes (`10xxxxxx`) so a suffix cut starts on a
+ * lead/ASCII byte instead of mid-codepoint.
+ */
+function trimLeadingContinuationUtf8(bytes: Uint8Array): Uint8Array {
+  let i = 0
+  // i < length guards the read; cast rather than `!` (no-non-null-assertion).
+  while (i < bytes.length && ((bytes[i] as number) & 0xc0) === 0x80) i++
+  return bytes.subarray(i)
+}
+
+/**
+ * Bounds a byte-oriented text stream, keeping a prefix, a suffix, or both
+ * ({@link TextRetentionStrategy}). All three strategies share one prefix/suffix
+ * accumulator: `head` is prefix-only, `tail` is suffix-only, `headTail` is both.
+ *
+ * Bytes, not characters: caps and `omittedBytes` are byte counts for process/
+ * body safety. Chunks that straddle a codepoint are handled — {@link finish}
+ * trims a partial codepoint at each cut so the returned text never introduces a
+ * replacement char at the boundary. The retainer holds at most
+ * `prefixCap + tailBytes + one chunk` in memory (old suffix chunks are dropped
+ * as they slide out), so a large stream does not accumulate unbounded.
+ */
+export class TextRetainer {
+  private readonly prefixCap: number
+  private readonly suffixCap: number
+  private readonly prefixChunks: Uint8Array[] = []
+  private prefixHeld = 0
+  private readonly suffixChunks: Uint8Array[] = []
+  private suffixHeld = 0
+  private total = 0
+
+  /** @param strategy One of the {@link TextRetentionStrategy} shapes; byte budgets must be non-negative integers. */
+  constructor(strategy: TextRetentionStrategy) {
+    switch (strategy.kind) {
+      case 'head':
+        assertBudget(strategy.maxBytes, 'maxBytes')
+        this.prefixCap = strategy.maxBytes
+        this.suffixCap = 0
+        break
+      case 'tail':
+        assertBudget(strategy.maxBytes, 'maxBytes')
+        this.prefixCap = 0
+        this.suffixCap = strategy.maxBytes
+        break
+      case 'headTail':
+        assertBudget(strategy.headBytes, 'headBytes')
+        assertBudget(strategy.tailBytes, 'tailBytes')
+        this.prefixCap = strategy.headBytes
+        this.suffixCap = strategy.tailBytes
+        break
+    }
+  }
+
+  /**
+   * Offer one chunk (a `Uint8Array`, or a `string` encoded as UTF-8). Prefix
+   * bytes fill up to the prefix cap then stop; suffix bytes roll so only the
+   * last `suffixCap` bytes are retained. `kept` is `true` only when no byte of
+   * this chunk was dropped.
+   *
+   * @param chunk The next bytes of the stream (`Uint8Array` or UTF-8 `string`).
+   * @returns The per-push {@link PushDecision}.
+   */
+  push(chunk: Uint8Array | string): PushDecision {
+    const bytes = typeof chunk === 'string' ? encoder.encode(chunk) : chunk
+    const before = this.total
+    this.total += bytes.length
+
+    // Prefix: take only up to the cap; the rest of this chunk is "not prefixed".
+    const room = this.prefixCap - this.prefixHeld
+    const take = Math.max(0, Math.min(room, bytes.length))
+    if (take > 0) {
+      this.prefixChunks.push(bytes.subarray(0, take))
+      this.prefixHeld += take
+    }
+
+    // Suffix: append the whole chunk, then drop whole leading chunks that have
+    // fully slid out of the last `suffixCap` bytes (bounded memory).
+    if (this.suffixCap > 0) {
+      this.suffixChunks.push(bytes)
+      this.suffixHeld += bytes.length
+      let head = this.suffixChunks[0]
+      while (head !== undefined && this.suffixHeld - head.length >= this.suffixCap) {
+        this.suffixChunks.shift()
+        this.suffixHeld -= head.length
+        head = this.suffixChunks[0]
+      }
+      // The head chunk can still hold leading bytes beyond the last `suffixCap`
+      // — a single chunk LARGER than the window is retained whole by the loop
+      // above (dropping the only chunk would leave < cap). Trim those leading
+      // bytes so the accumulator (and finish()'s concat) stays bounded by
+      // `suffixCap` instead of allocating/copying the full chunk again;
+      // finish() only ever reads the last `suffixLen ≤ suffixCap` bytes, so this
+      // drops nothing it would return. (head.length > excess by the loop
+      // invariant `suffixHeld - head.length < suffixCap`, so the slice is non-empty.)
+      if (head !== undefined && this.suffixHeld > this.suffixCap) {
+        const excess = this.suffixHeld - this.suffixCap
+        this.suffixChunks[0] = head.subarray(excess)
+        this.suffixHeld -= excess
+      }
+    }
+
+    // Dropped = bytes that no side can keep. Compute cumulative omission the
+    // SAME way finish() does (via omittedAt), so push and finish never disagree;
+    // per-push we only need whether THIS chunk pushed the total past what the
+    // two caps hold.
+    const droppedThisChunk = this.omittedAt(this.total) > this.omittedAt(before)
+    return {
+      kept: !droppedThisChunk,
+      truncated: this.omittedAt(this.total) > 0,
+    }
+  }
+
+  /** Bytes omitted once `total` bytes have been seen: `total − keptPrefix − keptSuffix`. */
+  private omittedAt(total: number): number {
+    const prefixLen = Math.min(total, this.prefixCap)
+    const suffixLen = Math.min(total - prefixLen, this.suffixCap)
+    return total - prefixLen - suffixLen
+  }
+
+  /**
+   * Finalize: decode the retained prefix and suffix (each trimmed to a UTF-8
+   * boundary at its cut) and report the exact omitted byte count.
+   *
+   * @returns The {@link RetainedText} snapshot (safe to hand to a formatter).
+   */
+  finish(): RetainedText {
+    const prefixLen = Math.min(this.total, this.prefixCap)
+    const suffixLen = Math.min(this.total - prefixLen, this.suffixCap)
+
+    const prefix = concat(this.prefixChunks) // exactly prefixLen bytes (prefixHeld === prefixLen)
+    const suffix = concat(this.suffixChunks).subarray(this.suffixHeld - suffixLen)
+
+    // With nothing omitted by budget, prefix and suffix are ADJACENT slices of
+    // one stream (prefixLen + suffixLen === total), so the head|tail split is
+    // artificial: a codepoint may span it. Decode the contiguous whole as one
+    // buffer — trimming or decoding the halves separately here would corrupt a
+    // boundary-spanning codepoint though no content was dropped. Only a real
+    // omitted gap makes each side a true cut: trim each to a UTF-8 boundary and
+    // decode separately so a codepoint is never reconstructed across the gap.
+    const budgetOmitted = this.omittedAt(this.total)
+    const [keptPrefix, keptSuffix] = budgetOmitted > 0
+      ? [trimTrailingPartialUtf8(prefix), trimLeadingContinuationUtf8(suffix)]
+      : [prefix, suffix]
+    const text = budgetOmitted > 0
+      ? decoder.decode(keptPrefix) + decoder.decode(keptSuffix)
+      : decoder.decode(concat([prefix, suffix]))
+
+    // Report omission against the bytes ACTUALLY returned, not the pre-trim
+    // budget: a boundary trim drops partial-codepoint bytes too, so an exact
+    // count derived from the budget alone would overstate the retained text (and
+    // any "Omitted N bytes" notice built from it would be a lie).
+    const omitted = this.total - keptPrefix.length - keptSuffix.length
+    const truncated = omitted > 0
+
+    return {
+      text,
+      truncated,
+      omittedBytes: truncated
+        ? { kind: 'exact', count: omitted }
+        : { kind: 'none' },
+    }
+  }
+}
+
+/** Concatenate chunks into one contiguous buffer (their exact total length). */
+function concat(chunks: readonly Uint8Array[]): Uint8Array {
+  let length = 0
+  for (const chunk of chunks) length += chunk.length
+  const out = new Uint8Array(length)
+  let offset = 0
+  for (const chunk of chunks) {
+    out.set(chunk, offset)
+    offset += chunk.length
+  }
+  return out
+}
+
+/**
+ * Standardized, false-precision-safe wording for one {@link Omitted} value —
+ * the "may standardize omission wording" half the library owns. `exact` prints
+ * the count (`Omitted 3 items`); `unknown` prints NO count because the caller
+ * did not provide one. `none` is the empty string.
+ *
+ * @param omitted The omission metadata from a retainer result.
+ * @param unit The noun for the omitted quantity (`items`, `bytes`, `chars`, `lines`).
+ * @returns A neutral clause (no trailing space), or `''` when nothing was omitted.
+ */
+export function describeOmitted(omitted: Omitted, unit: RetentionNotice['unit']): string {
+  switch (omitted.kind) {
+    case 'none':
+      return ''
+    case 'exact':
+      return `Omitted ${omitted.count} ${unit}.`
+    case 'unknown':
+      return `More ${unit} were omitted.`
+  }
+}
+
+/**
+ * Turn a {@link RetentionNotice} into a one-line footer: the library-owned
+ * standardized omission clause ({@link describeOmitted}) followed by the tool's
+ * own recovery guidance. The library never owns recovery words — only the tool
+ * knows the action ("narrow the pattern", "fetch a more specific URL", "read the
+ * spill file") — so `recovery` supplies them and receives the full notice to
+ * phrase from (`kept`, `limit`, `omitted`, …). Either half may be empty; the two
+ * are joined with a single space.
+ *
+ * @param notice The neutral retention outcome.
+ * @param recovery Tool-supplied guidance builder; receives the notice, returns a sentence (or `''`).
+ * @returns The combined footer line.
+ */
+export function formatRetentionNotice(
+  notice: RetentionNotice,
+  recovery: (notice: RetentionNotice) => string,
+): string {
+  return [describeOmitted(notice.omitted, notice.unit), recovery(notice)]
+    .filter(part => part.length > 0)
+    .join(' ')
+}
diff --git a/packages/util/retention/tests/retention.spec.ts b/packages/util/retention/tests/retention.spec.ts
new file mode 100644
index 0000000000..8fac7d8575
--- /dev/null
+++ b/packages/util/retention/tests/retention.spec.ts
@@ -0,0 +1,376 @@
+import { describe, expect, it } from 'vitest'
+import {
+  describeOmitted,
+  formatRetentionNotice,
+  ItemRetainer,
+  type Omitted,
+  type RetentionNotice,
+  TextRetainer,
+} from '@deepseek-ai/dsh-retention'
+
+/** Decode a RetainedText via a round-trip helper for readable UTF-8 assertions. */
+const utf8 = (s: string): Uint8Array => new TextEncoder().encode(s)
+
+describe('ItemRetainer — head retention', () => {
+  it('keeps the first maxItems while callers keep draining for an exact omitted count', () => {
+    const r = new ItemRetainer<string>({ kind: 'head', maxItems: 2 })
+    expect(r.push('a')).toEqual({ kept: true, truncated: false })
+    expect(r.push('b')).toEqual({ kept: true, truncated: false })
+    expect(r.push('c')).toEqual({ kept: false, truncated: true })
+
+    const result = r.finish()
+    expect(result.items).toEqual(['a', 'b'])
+    expect(result.kept).toBe(2)
+    expect(result.seen).toBe(3)
+    expect(result.truncated).toBe(true)
+    expect(result.omitted).toEqual<Omitted>({ kind: 'exact', count: 1 })
+  })
+
+  it('reports none when everything fits', () => {
+    const r = new ItemRetainer<number>({ kind: 'head', maxItems: 3 })
+    r.push(1)
+    r.push(2)
+    const result = r.finish()
+    expect(result.items).toEqual([1, 2])
+    expect(result.truncated).toBe(false)
+    expect(result.omitted).toEqual<Omitted>({ kind: 'none' })
+  })
+  it('keeps draining past the cap and reports an exact omitted count', () => {
+    const r = new ItemRetainer<string>({ kind: 'head', maxItems: 1 })
+    expect(r.push('a')).toEqual({ kept: true, truncated: false })
+    expect(r.push('b')).toEqual({ kept: false, truncated: true })
+    expect(r.push('c')).toEqual({ kept: false, truncated: true })
+
+    const result = r.finish()
+    expect(result.items).toEqual(['a'])
+    expect(result.seen).toBe(3)
+    expect(result.omitted).toEqual<Omitted>({ kind: 'exact', count: 2 })
+  })
+})
+
+describe('ItemRetainer — zero budget', () => {
+  it('keeps nothing and counts every pushed item as omitted', () => {
+    const r = new ItemRetainer<string>({ kind: 'head', maxItems: 0 })
+    expect(r.push('a')).toEqual({ kept: false, truncated: true })
+    const result = r.finish()
+    expect(result.items).toEqual([])
+    expect(result.kept).toBe(0)
+    expect(result.omitted).toEqual<Omitted>({ kind: 'exact', count: 1 })
+  })
+
+  it('rejects a non-integer / negative maxItems', () => {
+    expect(() => new ItemRetainer({ kind: 'head', maxItems: -1 }))
+      .toThrow(/maxItems must be a non-negative integer/)
+    expect(() => new ItemRetainer({ kind: 'head', maxItems: 1.5 }))
+      .toThrow(/maxItems must be a non-negative integer/)
+  })
+})
+
+describe('TextRetainer — head (exact omission, reads to end)', () => {
+  it('keeps the prefix and counts omitted bytes exactly', () => {
+    const r = new TextRetainer({ kind: 'head', maxBytes: 5 })
+    expect(r.push('abc')).toEqual({ kept: true, truncated: false })
+    // 'de' fills the cap exactly (5 bytes) — still fully kept.
+    expect(r.push('de')).toEqual({ kept: true, truncated: false })
+    expect(r.push('fgh')).toEqual({ kept: false, truncated: true })
+
+    const result = r.finish()
+    expect(result.text).toBe('abcde')
+    expect(result.truncated).toBe(true)
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'exact', count: 3 })
+  })
+
+  it('flags a partially-dropped chunk as not fully kept', () => {
+    const r = new TextRetainer({ kind: 'head', maxBytes: 4 })
+    r.push('ab')
+    // 'cde' straddles the cap: 'c','d' fit, 'e' drops → kept:false.
+    expect(r.push('cde')).toEqual({ kept: false, truncated: true })
+    expect(r.finish().text).toBe('abcd')
+  })
+
+  it('keeps draining past the cap', () => {
+    const r = new TextRetainer({ kind: 'head', maxBytes: 3 })
+    r.push('abc')
+    expect(r.push('defg')).toEqual({ kept: false, truncated: true })
+    const result = r.finish()
+    expect(result.text).toBe('abc')
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'exact', count: 4 })
+  })
+})
+
+describe('TextRetainer — tail (exact omission, reads to end)', () => {
+  it('keeps the final maxBytes and reports exact omission', () => {
+    const r = new TextRetainer({ kind: 'tail', maxBytes: 4 })
+    expect(r.push('hello')).toEqual({ kept: false, truncated: true })
+    r.push('world')
+    const result = r.finish()
+    expect(result.text).toBe('orld') // last 4 bytes of 'helloworld'
+    expect(result.truncated).toBe(true)
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'exact', count: 6 })
+  })
+
+  it('keeps everything when the stream is under the cap', () => {
+    const r = new TextRetainer({ kind: 'tail', maxBytes: 100 })
+    r.push('short')
+    const result = r.finish()
+    expect(result.text).toBe('short')
+    expect(result.truncated).toBe(false)
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'none' })
+  })
+
+  it('drops old chunks as they slide out of the tail window', () => {
+    const r = new TextRetainer({ kind: 'tail', maxBytes: 3 })
+    for (const c of ['11', '22', '33', '44']) r.push(c)
+    // Only the final 3 bytes survive; earlier whole chunks are dropped.
+    expect(r.finish().text).toBe('344')
+  })
+})
+
+describe('TextRetainer — headTail (prefix + suffix, omit the middle)', () => {
+  it('keeps a stable head and tail, omitting the middle exactly', () => {
+    const r = new TextRetainer({ kind: 'headTail', headBytes: 3, tailBytes: 3 })
+    r.push('abcdefghij') // 10 bytes: head 'abc', tail 'hij', middle 'defg' omitted
+    const result = r.finish()
+    expect(result.text).toBe('abchij')
+    expect(result.truncated).toBe(true)
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'exact', count: 4 })
+  })
+
+  it('does not double-count when head+tail cover the whole stream', () => {
+    const r = new TextRetainer({ kind: 'headTail', headBytes: 3, tailBytes: 3 })
+    r.push('abcdef') // exactly head(3) + tail(3), nothing omitted
+    const result = r.finish()
+    expect(result.text).toBe('abcdef')
+    expect(result.truncated).toBe(false)
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'none' })
+  })
+
+  it('does not drop a codepoint that spans the head|tail split when nothing is omitted', () => {
+    // Regression: with head+tail covering the whole stream, the split is
+    // artificial — a multibyte codepoint may straddle it. 'éab' is C3 A9 61 62
+    // (4 bytes); headBytes 1 + tailBytes 3 covers all 4 with omitted === 0, but
+    // the split falls INSIDE 'é'. The bytes are contiguous, so the full 'éab'
+    // must survive — not be trimmed to 'ab'.
+    const r = new TextRetainer({ kind: 'headTail', headBytes: 1, tailBytes: 3 })
+    r.push('éab')
+    const result = r.finish()
+    expect(result.text).toBe('éab')
+    expect(result.truncated).toBe(false)
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'none' })
+  })
+
+  it('still trims boundary partials once a real middle is omitted', () => {
+    // With a genuine gap the two sides ARE true cuts: '€' (3 bytes) split across
+    // the omitted middle must not resurface as a replacement char on either side.
+    const r = new TextRetainer({ kind: 'headTail', headBytes: 2, tailBytes: 2 })
+    r.push('a€€b') // 8 bytes; head 'a'+partial, tail partial+'b', middle omitted
+    const result = r.finish()
+    expect(result.truncated).toBe(true)
+    expect(result.text).not.toContain('�')
+    expect(result.text.startsWith('a')).toBe(true)
+    expect(result.text.endsWith('b')).toBe(true)
+  })
+})
+
+describe('TextRetainer — zero budgets', () => {
+  it('head maxBytes 0 keeps nothing and counts every byte exactly', () => {
+    const r = new TextRetainer({ kind: 'head', maxBytes: 0 })
+    expect(r.push('x')).toEqual({ kept: false, truncated: true })
+    const result = r.finish()
+    expect(result.text).toBe('')
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'exact', count: 1 })
+  })
+
+  it('an empty stream omits nothing', () => {
+    const r = new TextRetainer({ kind: 'headTail', headBytes: 2, tailBytes: 2 })
+    const result = r.finish()
+    expect(result.text).toBe('')
+    expect(result.truncated).toBe(false)
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'none' })
+  })
+
+  it('rejects non-integer / negative byte budgets', () => {
+    expect(() => new TextRetainer({ kind: 'head', maxBytes: -1 }))
+      .toThrow(/maxBytes must be a non-negative integer/)
+    expect(() => new TextRetainer({ kind: 'tail', maxBytes: 2.5 }))
+      .toThrow(/maxBytes must be a non-negative integer/)
+    expect(() => new TextRetainer({ kind: 'headTail', headBytes: -1, tailBytes: 2 }))
+      .toThrow(/headBytes must be a non-negative integer/)
+    expect(() => new TextRetainer({ kind: 'headTail', headBytes: 2, tailBytes: 1.1 }))
+      .toThrow(/tailBytes must be a non-negative integer/)
+  })
+})
+
+describe('TextRetainer — UTF-8 boundary handling', () => {
+  it('trims a partial codepoint at the head cut instead of emitting U+FFFD', () => {
+    // '€' is 3 bytes (E2 82 AC). A 2-byte head cap keeps 'a' (61) + the first
+    // byte of '€' (E2); that partial lead byte must be trimmed, not decoded to
+    // a replacement char.
+    const r = new TextRetainer({ kind: 'head', maxBytes: 2 })
+    r.push('a€b') // bytes: 61 E2 82 AC 62
+    const result = r.finish()
+    expect(result.text).toBe('a') // partial '€' dropped, no U+FFFD
+    expect(result.text).not.toContain('�')
+    // Omission counts bytes ACTUALLY absent from the returned text, including
+    // the partial 'E2' the boundary trim dropped: 5 total − 1 retained = 4
+    // (not the pre-trim budget of 3, which would overstate what was kept).
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'exact', count: 4 })
+  })
+
+  it('trims a leading partial codepoint at the tail cut', () => {
+    // Tail cap 2 over 'a€b' (5 bytes) keeps AC 62 — AC is a continuation byte
+    // (the middle of '€'); the leading continuation byte is dropped so the tail
+    // begins on a boundary.
+    const r = new TextRetainer({ kind: 'tail', maxBytes: 2 })
+    r.push('a€b')
+    const result = r.finish()
+    expect(result.text).toBe('b') // partial '€' at the front dropped
+    expect(result.text).not.toContain('�')
+    // Honest count: 5 total − 1 retained ('b') = 4, including the trimmed AC.
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'exact', count: 4 })
+  })
+
+  it('omitted count matches the bytes actually absent, across a headTail boundary trim', () => {
+    // Regression: the exact count must equal total − retained (post-trim), never
+    // the pre-trim budget. 'a€€b' is 8 bytes (61 E2828C… ×2 61? no: 61 E2 82 AC
+    // E2 82 AC 62). headBytes 2 keeps 'a'+partial-E2 → trims to 'a' (1 byte);
+    // tailBytes 2 keeps partial-AC+'b' → trims to 'b' (1 byte). Retained text is
+    // 2 bytes, so omitted must be 8 − 2 = 6 — not the budget's 8 − 2 − 2 = 4.
+    const r = new TextRetainer({ kind: 'headTail', headBytes: 2, tailBytes: 2 })
+    r.push('a€€b')
+    const result = r.finish()
+    const retainedBytes = new TextEncoder().encode(result.text).length
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'exact', count: 8 - retainedBytes })
+  })
+
+  it('preserves a whole multibyte codepoint that fits exactly', () => {
+    const r = new TextRetainer({ kind: 'head', maxBytes: 3 })
+    r.push('€x') // '€' is exactly 3 bytes
+    expect(r.finish().text).toBe('€')
+  })
+
+  it('does not reconstruct a codepoint across the omitted middle', () => {
+    // headBytes ends mid-'€' and tailBytes starts mid-another '€'; neither cut
+    // may glue a valid codepoint across the gap.
+    const r = new TextRetainer({ kind: 'headTail', headBytes: 2, tailBytes: 2 })
+    r.push('€€€') // 9 bytes
+    const result = r.finish()
+    expect(result.text).not.toContain('�')
+    expect(result.truncated).toBe(true)
+  })
+
+  it('accepts a raw Uint8Array chunk', () => {
+    const r = new TextRetainer({ kind: 'head', maxBytes: 2 })
+    r.push(utf8('xy'))
+    r.push(utf8('z'))
+    expect(r.finish().text).toBe('xy')
+  })
+
+  it('trims a partial 2-byte codepoint at the head cut', () => {
+    // 'é' is 2 bytes (C3 A9). A 2-byte head cap over 'aé' keeps 'a' (61) + the
+    // lead byte of 'é' (C3) — an incomplete 2-byte sequence to trim.
+    const r = new TextRetainer({ kind: 'head', maxBytes: 2 })
+    r.push('aé') // bytes: 61 C3 A9
+    const result = r.finish()
+    expect(result.text).toBe('a')
+    expect(result.text).not.toContain('�')
+  })
+
+  it('trims a partial 4-byte codepoint (emoji) at the head cut', () => {
+    // '😀' is 4 bytes (F0 9F 98 80). A 3-byte head cap keeps 'a' + the first two
+    // bytes of the emoji — an incomplete 4-byte sequence that must be trimmed.
+    const r = new TextRetainer({ kind: 'head', maxBytes: 3 })
+    r.push('a😀') // bytes: 61 F0 9F 98 80
+    const result = r.finish()
+    expect(result.text).toBe('a')
+    expect(result.text).not.toContain('�')
+  })
+
+  it('keeps a whole 4-byte codepoint that fits exactly', () => {
+    const r = new TextRetainer({ kind: 'head', maxBytes: 4 })
+    r.push('😀x')
+    expect(r.finish().text).toBe('😀')
+  })
+
+  it('leaves a head cut ending on a stray continuation run untouched', () => {
+    // A cut whose trailing bytes are ALL continuation bytes with no lead in
+    // reach is not a trimmable incomplete sequence — the trimmer bails (no lead
+    // byte found) and leaves them for the non-fatal decoder to replace.
+    const r = new TextRetainer({ kind: 'head', maxBytes: 2 })
+    // 0x80 0x80 are bare continuation bytes; 'z' follows so the head keeps just
+    // the two continuation bytes and the cut lands right after them.
+    r.push(new Uint8Array([0x80, 0x80, 0x7a]))
+    const result = r.finish()
+    // The trimmer did not throw and did not eat the bytes as a partial sequence;
+    // only the trailing 'z' is omitted by the 2-byte cap.
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'exact', count: 1 })
+  })
+
+  it('leaves a head cut ending on an invalid lead byte untouched', () => {
+    // 0xF8 is not a valid UTF-8 lead byte (only 0x00–0xF7 lead). The trimmer
+    // recognizes it as "not a lead" (expected length 0) and leaves the byte in
+    // place rather than trimming a phantom partial sequence.
+    const r = new TextRetainer({ kind: 'head', maxBytes: 1 })
+    r.push(new Uint8Array([0xf8, 0x61])) // 0xF8 kept, 'a' dropped by the 1-byte cap
+    const result = r.finish()
+    expect(result.omittedBytes).toEqual<Omitted>({ kind: 'exact', count: 1 })
+  })
+})
+
+describe('describeOmitted — false precision safety', () => {
+  it('prints an exact count for exact omission', () => {
+    expect(describeOmitted({ kind: 'exact', count: 3 }, 'items')).toBe('Omitted 3 items.')
+    expect(describeOmitted({ kind: 'exact', count: 12 }, 'bytes')).toBe('Omitted 12 bytes.')
+  })
+
+  it('prints NO count for unknown omission', () => {
+    expect(describeOmitted({ kind: 'unknown' }, 'lines')).toBe('More lines were omitted.')
+  })
+
+  it('returns empty string when nothing was omitted', () => {
+    expect(describeOmitted({ kind: 'none' }, 'chars')).toBe('')
+  })
+})
+
+describe('formatRetentionNotice', () => {
+  const notice = (omitted: Omitted): RetentionNotice => ({
+    scope: 'grep',
+    strategy: 'head',
+    unit: 'items',
+    limit: 100,
+    kept: 100,
+    omitted,
+  })
+
+  it('joins the standardized omission clause with the tool recovery guidance', () => {
+    const out = formatRetentionNotice(
+      notice({ kind: 'exact', count: 25 }),
+      ({ kept }) => `Results capped at ${kept}. Narrow the pattern, path, or include to see more.`,
+    )
+    expect(out).toBe('Omitted 25 items. Results capped at 100. Narrow the pattern, path, or include to see more.')
+  })
+
+  it('omits the empty half when nothing was omitted', () => {
+    const out = formatRetentionNotice(notice({ kind: 'none' }), () => 'Recovery text.')
+    expect(out).toBe('Recovery text.')
+  })
+
+  it('omits the empty half when the tool supplies no recovery text', () => {
+    const out = formatRetentionNotice(notice({ kind: 'exact', count: 2 }), () => '')
+    expect(out).toBe('Omitted 2 items.')
+  })
+
+  it('passes the full notice to the recovery builder (limit as a head/tail pair)', () => {
+    const headTail: RetentionNotice = {
+      scope: 'bash stdout',
+      strategy: 'headTail',
+      unit: 'bytes',
+      limit: { head: 2_000, tail: 2_000 },
+      kept: 4_000,
+      omitted: { kind: 'exact', count: 500 },
+    }
+    const out = formatRetentionNotice(headTail, n =>
+      typeof n.limit === 'object' ? `Kept ${n.limit.head}B head + ${n.limit.tail}B tail.` : '')
+    expect(out).toBe('Omitted 500 bytes. Kept 2000B head + 2000B tail.')
+  })
+})
diff --git a/packages/util/retention/tsconfig.json b/packages/util/retention/tsconfig.json
new file mode 100644
index 0000000000..749cb0208e
--- /dev/null
+++ b/packages/util/retention/tsconfig.json
@@ -0,0 +1,11 @@
+{
+  "extends": "../../../tsconfig.base.json",
+  "compilerOptions": {
+    "rootDir": "src",
+    "outDir": "lib/types"
+  },
+  "include": [
+    "src"
+  ],
+  "references": []
+}
diff --git a/packages/util/timeout/README.md b/packages/util/timeout/README.md
index ff56b65535..61a2e95ed8 100644
--- a/packages/util/timeout/README.md
+++ b/packages/util/timeout/README.md
@@ -25,12 +25,19 @@ import { clampTimeout, deadline, timeoutOf, TimeoutReason } from '@deepseek-ai/d
 
 ## Usage shape
 
-```ts ignore-check
+```ts
+import { deadline, timeoutOf } from '@deepseek-ai/dsh-timeout'
+
+declare function runWork(options: { signal: AbortSignal }): Promise<unknown>
+
 // Scope-lifetime consumer (foreground bash, one fetch): `using` disposes the timer.
-using d = deadline(upstream, timeoutMs, 'BASH_TIMEOUT')
-const outcome = await runWork({ signal: d.signal })            // work listens on d.signal and terminates itself
-const timedOut = timeoutOf(d.signal, 'BASH_TIMEOUT') !== undefined  // classify the first abort, scoped to OUR code
-const aborted = d.signal.aborted && !timedOut                  // mutually exclusive: timeout won, or cancel did
+export async function runWithDeadline(upstream: AbortSignal | undefined, timeoutMs: number): Promise<unknown> {
+  using d = deadline(upstream, timeoutMs, 'BASH_TIMEOUT')
+  const outcome = await runWork({ signal: d.signal })               // work listens on d.signal and terminates itself
+  const timedOut = timeoutOf(d.signal, 'BASH_TIMEOUT') !== undefined // classify the first abort, scoped to OUR code
+  const aborted = d.signal.aborted && !timedOut                     // mutually exclusive: timeout won, or cancel did
+  return { outcome, timedOut, aborted }
+}
 ```
 
 The signal only *notifies* — the caller MUST attach its own termination (`d.signal.addEventListener('abort', kill)`, or hand `d.signal` to `fetch`). Racing a promise against a timer would resolve the tool-call while the child process or socket leaks on; handing out a signal forces a real termination path to exist.
diff --git a/packages/web/tool-web/README.md b/packages/web/tool-web/README.md
index 4935751082..ecc3d3e218 100644
--- a/packages/web/tool-web/README.md
+++ b/packages/web/tool-web/README.md
@@ -11,6 +11,8 @@ Each tool is registered independently; a product that wants only one disables th
 | `web_search` | `query` (string) | Discovery. Returns an optional answer plus source URLs. `max_results` is **not** model-facing — the tool sets the bound (the `searchMaxResults` config, default 8) and passes it to the seam. |
 | `web_fetch` | `url` (string) | Retrieves a specific URL. HTML bodies are rendered to markdown-ish text; text bodies pass through. A non-2xx status is reported, not an error. The tool-call timeout is deployment policy (`dsh-timeout-policy`), not a model argument. |
 
+Both tools opt into concurrent scheduling because provider reads return content without mutating parent-agent state.
+
 ## Config
 
 | Key | Default | Meaning |
diff --git a/packages/web/tool-web/package.json b/packages/web/tool-web/package.json
index f6b5791f0f..23ed7157a2 100644
--- a/packages/web/tool-web/package.json
+++ b/packages/web/tool-web/package.json
@@ -35,6 +35,8 @@
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
     "@deepseek-ai/dsh-session": "workspace:^",
+    "@deepseek-ai/dsh-spill-local": "workspace:^",
+    "@deepseek-ai/dsh-spill-policy": "workspace:^",
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
     "@deepseek-ai/dsh-tools": "workspace:^",
     "@deepseek-ai/dsh-timeout-policy": "workspace:^",
diff --git a/packages/web/tool-web/src/fetch.ts b/packages/web/tool-web/src/fetch.ts
index a6b8883f4d..83358251fb 100644
--- a/packages/web/tool-web/src/fetch.ts
+++ b/packages/web/tool-web/src/fetch.ts
@@ -92,6 +92,8 @@ export function applyWebFetchTool(ctx: Context, timeoutMs: number): void {
       url: { type: 'string', required: true, description: 'The HTTP(S) URL to fetch.' },
     },
     timeoutMs,
+    // Provider reads do not mutate parent-agent state.
+    isConcurrencySafe: () => true,
     async execute(args, exec): Promise<ContentBlock[]> {
       const input = parseFetchArgs(args)
       const result = await ctx.web.fetch(
diff --git a/packages/web/tool-web/src/search.ts b/packages/web/tool-web/src/search.ts
index 6db829b7fc..af8753720d 100644
--- a/packages/web/tool-web/src/search.ts
+++ b/packages/web/tool-web/src/search.ts
@@ -109,6 +109,8 @@ export function applyWebSearchTool(ctx: Context, maxResults: number, timeoutMs:
       query: { type: 'string', required: true, description: 'The search query.' },
     },
     timeoutMs,
+    // Provider reads do not mutate parent-agent state.
+    isConcurrencySafe: () => true,
     async execute(args, exec): Promise<ContentBlock[]> {
       const input = parseSearchArgs(args)
       const result = await ctx.web.search(
diff --git a/packages/web/tool-web/tests/spill.spec.ts b/packages/web/tool-web/tests/spill.spec.ts
new file mode 100644
index 0000000000..58599d2c54
--- /dev/null
+++ b/packages/web/tool-web/tests/spill.spec.ts
@@ -0,0 +1,95 @@
+/**
+ * Showcase integration: the real `web_fetch` tool + the real spill stack
+ * (`dsh-spill-local` backend + `dsh-spill-policy`), exercised through
+ * `ctx.tools.execute()`. Proves the RFC's default local-backend path — a large
+ * formatted fetch result is automatically retained and spilled with NO
+ * tool-specific spill code, and the model-facing text changes ONLY by the
+ * deliberate spill notice (the full formatted result lands in the spill file).
+ */
+
+import { afterEach, beforeEach, describe, expect, it } from 'vitest'
+import { createServer, type IncomingMessage, type Server, type ServerResponse } from 'node:http'
+import { AddressInfo } from 'node:net'
+import { mkdtempSync, readFileSync, rmSync } from 'node:fs'
+import { tmpdir } from 'node:os'
+import { join } from 'node:path'
+import { Context } from 'cordis'
+import { CallId } from '@deepseek-ai/dsh-llm'
+import { SessionId } from '@deepseek-ai/dsh-session'
+import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
+import ToolRegistry from '@deepseek-ai/dsh-tools'
+import type { ToolExecution } from '@deepseek-ai/dsh-tools'
+import WebService from '@deepseek-ai/dsh-web'
+import * as WebFetchLocal from '@deepseek-ai/dsh-web-fetch-local'
+import LocalSpillStore from '@deepseek-ai/dsh-spill-local'
+import * as SpillPolicy from '@deepseek-ai/dsh-spill-policy'
+import * as ToolWeb from '@deepseek-ai/dsh-tool-web'
+
+type Handler = (req: IncomingMessage, res: ServerResponse) => void
+
+let server: Server
+let base: string
+let handler: Handler
+let spillRoot: string
+let ctx: Context
+
+const BODY = 'X'.repeat(4000) // formatted result is well over the policy cap
+const MAX_INLINE_BYTES = 1000 // leaves room for a head/tail preview beside the notice
+
+beforeEach(async () => {
+  handler = (_req, res) => { res.writeHead(200, { 'content-type': 'text/plain' }); res.end(BODY) }
+  server = createServer((req, res) => { handler(req, res) })
+  await new Promise<void>(resolve => server.listen(0, '127.0.0.1', resolve))
+  base = `http://127.0.0.1:${(server.address() as AddressInfo).port}`
+  spillRoot = mkdtempSync(join(tmpdir(), 'dsh-spill-web-'))
+
+  ctx = new Context()
+  await ctx.plugin(SystemPrompt)
+  await ctx.plugin(ToolRegistry)
+  await ctx.plugin(WebService, { fetchProvider: WebFetchLocal.LOCAL_FETCH_PROVIDER_ID })
+  // Provider cap generous so the tool returns a large formatted result; the
+  // policy cap is what triggers the spill (the RFC's separation of concerns).
+  await ctx.plugin(WebFetchLocal, { maxBodyChars: 500_000 })
+  await ctx.plugin(LocalSpillStore, { root: spillRoot })
+  await ctx.plugin(SpillPolicy, { maxInlineBytes: MAX_INLINE_BYTES })
+  await ctx.plugin(ToolWeb)
+})
+
+afterEach(async () => {
+  await new Promise<void>(resolve => server.close(() => { resolve() }))
+  rmSync(spillRoot, { recursive: true, force: true })
+})
+
+/** A web_fetch call carrying a session owner (so the policy can scope the spill). */
+function fetchCall(): Promise<{ isError: boolean; content: { type: string; text?: string }[] }> {
+  const agent = { session: { header: { id: SessionId('web-sess') } } }
+  const exec = { callId: CallId('call-1'), name: 'web_fetch', arguments: { url: base }, agent } as unknown as ToolExecution
+  return ctx.tools.execute(exec)
+}
+
+describe('web_fetch spill showcase', () => {
+  it('spills a large formatted result and returns a preview + spill locator', async () => {
+    const out = await fetchCall()
+    expect(out.isError).toBe(false)
+    const text = out.content.map(b => b.text).join('')
+
+    // Model-facing text is a preview + notice within the cap, NOT the full body.
+    expect(text.length).toBeLessThan(BODY.length)
+    expect(Buffer.byteLength(text, 'utf8')).toBeLessThanOrEqual(MAX_INLINE_BYTES)
+    expect(text).toContain(`Fetched ${base}`) // the head of the formatted result survives
+    expect(text).toContain('Full formatted result stored at:')
+    expect(text).toContain('Use read with offset/limit, or grep this path')
+
+    // The spill file holds the FULL formatted result the tool returned.
+    const match = /stored at: (\S+?)\. Use read/.exec(text)
+    expect(match).not.toBeNull()
+    const spillPath = match![1]!
+    const saved = readFileSync(spillPath, 'utf8')
+    // The provider cap was generous, so the tool did not truncate: the spill file
+    // holds the full formatted result (header + the complete body), far larger
+    // than the model-facing preview.
+    expect(saved).toContain('(HTTP 200)')
+    expect(saved).toContain(BODY)
+    expect(saved.length).toBeGreaterThan(text.length)
+  })
+})
diff --git a/packages/web/tool-web/tests/tool-web.spec.ts b/packages/web/tool-web/tests/tool-web.spec.ts
index a9f66d3746..088ace395e 100644
--- a/packages/web/tool-web/tests/tool-web.spec.ts
+++ b/packages/web/tool-web/tests/tool-web.spec.ts
@@ -166,6 +166,10 @@ describe('tool-web registration', () => {
     const names = ctx.tools.schemas().map(s => s.name)
     expect(names).toContain('web_search')
     expect(names).toContain('web_fetch')
+    expect(ctx.tools.executionMode({ callId: CallId('search-safe'), name: 'web_search', arguments: { query: 'q' } }))
+      .toEqual({ kind: 'parallel' })
+    expect(ctx.tools.executionMode({ callId: CallId('fetch-safe'), name: 'web_fetch', arguments: { url: 'https://a.test' } }))
+      .toEqual({ kind: 'parallel' })
     await fiber.dispose()
     expect(ctx.tools.schemas().map(s => s.name)).not.toContain('web_search')
   })
diff --git a/packages/web/web-fetch-local/src/index.ts b/packages/web/web-fetch-local/src/index.ts
index 471c61af2d..6e1e6b2bf6 100644
--- a/packages/web/web-fetch-local/src/index.ts
+++ b/packages/web/web-fetch-local/src/index.ts
@@ -20,8 +20,6 @@ export {
   LocalFetchProvider,
 } from './provider.ts'
 export type { LocalFetchLimits } from './provider.ts'
-export { classifyContentType, decoderForCharset, isSameOrigin, parseCharset, validateFetchUrl } from './policy.ts'
-export type { FetchableKind } from './policy.ts'
 
 /** Default `User-Agent`: an explicit product agent, never a browser disguise. */
 export const DEFAULT_USER_AGENT = 'deepseek-harness/0.0.1 (+https://github.com/deepseek-ai)'
diff --git a/packages/web/web-fetch-local/tests/fetch-local.spec.ts b/packages/web/web-fetch-local/tests/fetch-local.spec.ts
index 3158111134..092e384d94 100644
--- a/packages/web/web-fetch-local/tests/fetch-local.spec.ts
+++ b/packages/web/web-fetch-local/tests/fetch-local.spec.ts
@@ -3,9 +3,10 @@ import { createServer, type IncomingMessage, type Server, type ServerResponse }
 import { AddressInfo } from 'node:net'
 import { Context } from 'cordis'
 import WebService from '@deepseek-ai/dsh-web'
-import { LocalFetchProvider, LOCAL_FETCH_PROVIDER_ID, classifyContentType, decoderForCharset, isSameOrigin, parseCharset, validateFetchUrl } from '@deepseek-ai/dsh-web-fetch-local'
+import { LocalFetchProvider, LOCAL_FETCH_PROVIDER_ID } from '@deepseek-ai/dsh-web-fetch-local'
 import type { LocalFetchLimits } from '@deepseek-ai/dsh-web-fetch-local'
 import * as fetchPlugin from '@deepseek-ai/dsh-web-fetch-local'
+import { classifyContentType, decoderForCharset, isSameOrigin, parseCharset, validateFetchUrl } from '../src/policy.ts'
 
 const limits: LocalFetchLimits = {
   maxUrlLength: 2048,
diff --git a/packages/web/web-search-deepseek/src/index.ts b/packages/web/web-search-deepseek/src/index.ts
index 9666dde720..a24bb3f95f 100644
--- a/packages/web/web-search-deepseek/src/index.ts
+++ b/packages/web/web-search-deepseek/src/index.ts
@@ -25,8 +25,6 @@ export {
   DEEPSEEK_DEFAULT_MAX_USES,
   DEEPSEEK_DEFAULT_MODEL,
   DEEPSEEK_PROVIDER_ID,
-  citationSnippets,
-  mapAnthropicResponse,
 } from './provider.ts'
 export type { DeepSeekSearchProviderOptions } from './provider.ts'
 
diff --git a/packages/web/web-search-deepseek/tests/deepseek.spec.ts b/packages/web/web-search-deepseek/tests/deepseek.spec.ts
index 3d8d4be597..afca4ecce2 100644
--- a/packages/web/web-search-deepseek/tests/deepseek.spec.ts
+++ b/packages/web/web-search-deepseek/tests/deepseek.spec.ts
@@ -4,11 +4,10 @@ import Loader from '@cordisjs/plugin-loader'
 import WebService from '@deepseek-ai/dsh-web'
 import {
   DeepSeekSearchProvider,
-  citationSnippets,
-  mapAnthropicResponse,
   DEEPSEEK_PROVIDER_ID,
 } from '@deepseek-ai/dsh-web-search-deepseek'
 import * as deepseekPlugin from '@deepseek-ai/dsh-web-search-deepseek'
+import { citationSnippets, mapAnthropicResponse } from '../src/provider.ts'
 import type { AnthropicResponse } from '@deepseek-ai/dsh-web-search-deepseek/src/types.ts'
 
 const options = {
diff --git a/packages/web/web-search-exa/src/index.ts b/packages/web/web-search-exa/src/index.ts
index fc284dd79a..67b2eed574 100644
--- a/packages/web/web-search-exa/src/index.ts
+++ b/packages/web/web-search-exa/src/index.ts
@@ -24,8 +24,6 @@ export {
   EXA_DEFAULT_SEARCH_TYPE,
   EXA_PROVIDER_ID,
   ExaSearchProvider,
-  mapExaResponse,
-  mapExaResult,
 } from './provider.ts'
 export type { ExaSearchProviderOptions } from './provider.ts'
 
diff --git a/packages/web/web-search-exa/tests/exa.spec.ts b/packages/web/web-search-exa/tests/exa.spec.ts
index cf66ea67f9..40c7afda6f 100644
--- a/packages/web/web-search-exa/tests/exa.spec.ts
+++ b/packages/web/web-search-exa/tests/exa.spec.ts
@@ -1,8 +1,9 @@
 import { afterEach, describe, expect, it, vi } from 'vitest'
 import { Context } from 'cordis'
 import WebService from '@deepseek-ai/dsh-web'
-import { ExaSearchProvider, mapExaResponse, mapExaResult, EXA_PROVIDER_ID } from '@deepseek-ai/dsh-web-search-exa'
+import { ExaSearchProvider, EXA_PROVIDER_ID } from '@deepseek-ai/dsh-web-search-exa'
 import * as exaPlugin from '@deepseek-ai/dsh-web-search-exa'
+import { mapExaResponse, mapExaResult } from '../src/provider.ts'
 
 const options = { apiKey: 'exa-key', baseURL: 'https://api.exa.test', searchType: 'auto' as const, highlightsPerResult: 1 }
 
diff --git a/packages/web/web-search-perplexity/src/index.ts b/packages/web/web-search-perplexity/src/index.ts
index b71d5052d6..d673f575c8 100644
--- a/packages/web/web-search-perplexity/src/index.ts
+++ b/packages/web/web-search-perplexity/src/index.ts
@@ -18,8 +18,6 @@ export {
   PERPLEXITY_DEFAULT_MODEL,
   PERPLEXITY_PROVIDER_ID,
   PerplexitySearchProvider,
-  mapPerplexityResponse,
-  mapPerplexityResult,
 } from './provider.ts'
 export type { PerplexityRecency, PerplexitySearchProviderOptions } from './provider.ts'
 
diff --git a/packages/web/web-search-perplexity/tests/perplexity.spec.ts b/packages/web/web-search-perplexity/tests/perplexity.spec.ts
index e769989d73..9662226bc6 100644
--- a/packages/web/web-search-perplexity/tests/perplexity.spec.ts
+++ b/packages/web/web-search-perplexity/tests/perplexity.spec.ts
@@ -3,10 +3,10 @@ import { Context } from 'cordis'
 import WebService from '@deepseek-ai/dsh-web'
 import {
   PerplexitySearchProvider,
-  mapPerplexityResponse,
   PERPLEXITY_PROVIDER_ID,
 } from '@deepseek-ai/dsh-web-search-perplexity'
 import * as perplexityPlugin from '@deepseek-ai/dsh-web-search-perplexity'
+import { mapPerplexityResponse } from '../src/provider.ts'
 
 const options = { apiKey: 'pplx-key', baseURL: 'https://api.perplexity.test', model: 'sonar', maxTokens: 1024 }
 
diff --git a/packages/workflow/tool-workflow/src/index.ts b/packages/workflow/tool-workflow/src/index.ts
index 33f2939164..2a8ef96682 100644
--- a/packages/workflow/tool-workflow/src/index.ts
+++ b/packages/workflow/tool-workflow/src/index.ts
@@ -44,10 +44,10 @@ type ResolvedConfig = Required<Config>
  */
 const DESCRIPTION = `Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn.
 
-The workflow's identity rides the \`meta\` parameter as JSON: required \`name\` (short kebab-case) and \`description\` strings, optional \`whenToUse\` string and \`phases\` array (\`{title, detail?, model?}\`). The \`script\` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO \`export const meta\` statement — meta is a parameter, not code), running with top-level await; end with \`return <value>\` — the value must be JSON-serializable and is this tool's result.
+The workflow's identity rides the \`meta\` parameter as JSON: required \`name\` (short kebab-case) and \`description\` strings, optional \`whenToUse\` string and \`phases\` array (\`{title, detail?, provider?, model?}\`). The \`script\` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO \`export const meta\` statement — meta is a parameter, not code), running with top-level await; end with \`return <value>\` — the value must be JSON-serializable and is this tool's result.
 
 Script-body hooks:
-- \`agent(prompt, opts?): Promise<any>\` — run one subagent to completion. Without \`opts.schema\` it resolves to the child's final text; with \`opts.schema\` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves \`null\` when the child fails (filter with \`.filter(Boolean)\`). Other opts: \`label\` (display), \`phase\` (progress group), \`model\` (override). Anything else (\`effort\`/\`isolation\`/\`agentType\`) is rejected loudly.
+- \`agent(prompt, opts?): Promise<any>\` — run one subagent to completion. Without \`opts.schema\` it resolves to the child's final text; with \`opts.schema\` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves \`null\` when the child fails (filter with \`.filter(Boolean)\`). Other opts: \`label\` (display), \`phase\` (progress group), and independent \`provider\`/\`model\` LLM target overrides (either may be provided alone). Anything else (\`effort\`/\`isolation\`/\`agentType\`) is rejected loudly.
 - \`pipeline(items, ...stages): Promise<any[]>\` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives \`(prev, item, index)\`. An ordinary stage throw drops that ITEM to \`null\` and skips its remaining stages.
 - \`parallel(thunks): Promise<any[]>\` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to \`null\`.
 - \`phase(title)\` — start a progress phase; \`log(message)\` — narrate progress; \`args\` — the tool call's \`args\` input, verbatim.
@@ -58,7 +58,12 @@ Constraints: concurrency and total-agent caps apply; no filesystem, network, tim
 
 type WorkflowCallArgs = {
   script: string
-  meta: { name: string; description: string; whenToUse?: string; phases?: { title: string; detail?: string; model?: string }[] }
+  meta: {
+    name: string
+    description: string
+    whenToUse?: string
+    phases?: { title: string; detail?: string; provider?: string; model?: string }[]
+  }
   args?: Record<string, unknown>
 }
 
@@ -140,6 +145,7 @@ export function apply(ctx: Context, config: Config): void {
               properties: {
                 title: { type: 'string', required: true, description: 'The phase title phase() calls match by exact string.' },
                 detail: { type: 'string', description: 'Optional one-line description of the phase.' },
+                provider: { type: 'string', description: 'Optional provider override this phase is expected to use.' },
                 model: { type: 'string', description: 'Optional model override this phase is expected to use.' },
               },
             },
diff --git a/packages/workflow/workflow-workerthread/package.json b/packages/workflow/workflow-workerthread/package.json
index 485e252c97..1cd6e07e17 100644
--- a/packages/workflow/workflow-workerthread/package.json
+++ b/packages/workflow/workflow-workerthread/package.json
@@ -42,6 +42,7 @@
   "devDependencies": {
     "@deepseek-ai/dsh-agent": "workspace:^",
     "@deepseek-ai/dsh-agent-loop": "workspace:^",
+    "@deepseek-ai/dsh-agent-loop-testkit": "workspace:^",
     "@deepseek-ai/dsh-brand": "workspace:^",
     "@deepseek-ai/dsh-invariants": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
diff --git a/packages/workflow/workflow-workerthread/src/host.ts b/packages/workflow/workflow-workerthread/src/host.ts
index 34d0fce2d8..57570a5098 100644
--- a/packages/workflow/workflow-workerthread/src/host.ts
+++ b/packages/workflow/workflow-workerthread/src/host.ts
@@ -326,7 +326,14 @@ export class WorkerRun implements WorkflowRun {
         parent: this.parent,
         signal: this.controller.signal,
         ...request.schema !== undefined ? { outputSchema: request.schema } : {},
-        ...request.model !== undefined ? { agentOptions: { model: request.model } } : {},
+        ...request.provider !== undefined || request.model !== undefined
+          ? {
+            agentOptions: {
+              ...request.provider !== undefined ? { provider: request.provider } : {},
+              ...request.model !== undefined ? { model: request.model } : {},
+            },
+          }
+          : {},
       })
     } catch (error: unknown) {
       const failure = this.childAdmissionFailure()
diff --git a/packages/workflow/workflow-workerthread/src/meta.ts b/packages/workflow/workflow-workerthread/src/meta.ts
index c223be28fa..5412345178 100644
--- a/packages/workflow/workflow-workerthread/src/meta.ts
+++ b/packages/workflow/workflow-workerthread/src/meta.ts
@@ -35,15 +35,17 @@ function validateMetaShape(meta: unknown): { meta?: WorkflowMeta; violations: st
         }
         const entry = phase as Record<string, unknown>
         for (const key of Object.keys(entry)) {
-          if (!['title', 'detail', 'model'].includes(key)) violations.push(`meta.phases[${index}].${key} is not a recognized field`)
+          if (!['title', 'detail', 'provider', 'model'].includes(key)) violations.push(`meta.phases[${index}].${key} is not a recognized field`)
         }
         if (typeof entry.title !== 'string' || entry.title.length === 0) violations.push(`meta.phases[${index}].title must be a non-empty string`)
         if (entry.detail !== undefined && typeof entry.detail !== 'string') violations.push(`meta.phases[${index}].detail must be a string`)
+        if (entry.provider !== undefined && typeof entry.provider !== 'string') violations.push(`meta.phases[${index}].provider must be a string`)
         if (entry.model !== undefined && typeof entry.model !== 'string') violations.push(`meta.phases[${index}].model must be a string`)
         if (violations.length === 0) {
           phases.push({
             title: entry.title as string,
             ...entry.detail !== undefined ? { detail: entry.detail as string } : {},
+            ...entry.provider !== undefined ? { provider: entry.provider as string } : {},
             ...entry.model !== undefined ? { model: entry.model as string } : {},
           })
         }
diff --git a/packages/workflow/workflow-workerthread/src/runtime.ts b/packages/workflow/workflow-workerthread/src/runtime.ts
index a99ce495a4..b5dbed1744 100644
--- a/packages/workflow/workflow-workerthread/src/runtime.ts
+++ b/packages/workflow/workflow-workerthread/src/runtime.ts
@@ -36,7 +36,7 @@ export interface ExecutionObserver {
 }
 
 /** The `agent()` options the script may pass; everything else rejects loud. */
-const SUPPORTED_AGENT_OPTIONS = new Set(['label', 'phase', 'schema', 'model'])
+const SUPPORTED_AGENT_OPTIONS = new Set(['label', 'phase', 'schema', 'provider', 'model'])
 /** Deferred Claude Code options we name explicitly in the rejection message. */
 const DEFERRED_AGENT_OPTIONS = new Set(['effort', 'isolation', 'agentType'])
 
@@ -277,6 +277,7 @@ export class WorkflowExecution {
         run = await this.children.startAgent({
           prompt: rawPrompt,
           ...opts.schema !== undefined ? { schema: opts.schema } : {},
+          ...opts.provider !== undefined ? { provider: opts.provider } : {},
           ...opts.model !== undefined ? { model: opts.model } : {},
         })
       } catch (error: unknown) {
@@ -344,7 +345,13 @@ export class WorkflowExecution {
   }
 
   /** Materialize + validate the `agent()` options bag from the realm. */
-  private readAgentOptions(rawOpts: unknown): { label?: string; phase?: string; model?: string; schema?: StructuredOutputSchema } {
+  private readAgentOptions(rawOpts: unknown): {
+    label?: string
+    phase?: string
+    provider?: string
+    model?: string
+    schema?: StructuredOutputSchema
+  } {
     if (rawOpts === undefined) return {}
     let opts: unknown
     try {
@@ -361,11 +368,11 @@ export class WorkflowExecution {
     for (const key of Object.keys(record)) {
       if (SUPPORTED_AGENT_OPTIONS.has(key)) continue
       if (DEFERRED_AGENT_OPTIONS.has(key)) {
-        throw new WorkflowError(`agent() option "${key}" is deferred and not supported by this engine (supported: label, phase, schema, model)`, 'UNSUPPORTED_OPTION')
+        throw new WorkflowError(`agent() option "${key}" is deferred and not supported by this engine (supported: label, phase, schema, provider, model)`, 'UNSUPPORTED_OPTION')
       }
-      throw new WorkflowError(`agent() option "${key}" is not recognized (supported: label, phase, schema, model)`, 'UNSUPPORTED_OPTION')
+      throw new WorkflowError(`agent() option "${key}" is not recognized (supported: label, phase, schema, provider, model)`, 'UNSUPPORTED_OPTION')
     }
-    for (const key of ['label', 'phase', 'model'] as const) {
+    for (const key of ['label', 'phase', 'provider', 'model'] as const) {
       if (record[key] !== undefined && typeof record[key] !== 'string') {
         throw new WorkflowError(`agent() option "${key}" must be a string`, 'INVALID_ARGUMENT')
       }
@@ -384,6 +391,7 @@ export class WorkflowExecution {
     return {
       ...record.label !== undefined ? { label: record.label as string } : {},
       ...record.phase !== undefined ? { phase: record.phase as string } : {},
+      ...record.provider !== undefined ? { provider: record.provider as string } : {},
       ...record.model !== undefined ? { model: record.model as string } : {},
       ...schema !== undefined ? { schema } : {},
     }
diff --git a/packages/workflow/workflow-workerthread/src/types.ts b/packages/workflow/workflow-workerthread/src/types.ts
index 6adb422b2a..7e2bfff36c 100644
--- a/packages/workflow/workflow-workerthread/src/types.ts
+++ b/packages/workflow/workflow-workerthread/src/types.ts
@@ -42,6 +42,8 @@ export interface ChildStartRequest {
   prompt: string
   /** The structured-output schema, if the call passed one (already subset-checked). */
   schema?: StructuredOutputSchema
+  /** The per-child provider override, if the call passed one. */
+  provider?: string
   /** The per-child model override, if the call passed one. */
   model?: string
 }
diff --git a/packages/workflow/workflow-workerthread/tests/integration.spec.ts b/packages/workflow/workflow-workerthread/tests/integration.spec.ts
index 0e1727f877..25aa76cf48 100644
--- a/packages/workflow/workflow-workerthread/tests/integration.spec.ts
+++ b/packages/workflow/workflow-workerthread/tests/integration.spec.ts
@@ -1,11 +1,8 @@
 import { describe, expect, it } from 'vitest'
 import { Context } from 'cordis'
-import LlmService from '@deepseek-ai/dsh-llm'
-import SessionStore from '@deepseek-ai/dsh-session'
-import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
-import ToolRegistry from '@deepseek-ai/dsh-tools'
-import AgentRegistry, { AgentId } from '@deepseek-ai/dsh-agent'
+import { AgentId } from '@deepseek-ai/dsh-agent'
 import AgentLoop from '@deepseek-ai/dsh-agent-loop'
+import { mountAgentLoopTestDependencies } from '@deepseek-ai/dsh-agent-loop-testkit'
 import * as Invariants from '@deepseek-ai/dsh-invariants'
 import SubagentService from '@deepseek-ai/dsh-subagent'
 import * as spawn from '@deepseek-ai/dsh-subagent-spawn'
@@ -26,18 +23,14 @@ type Script = ConstructorParameters<typeof MockAdapter>[0]
 async function setup(script: Script) {
   const ctx = new Context()
   const adapter = new MockAdapter(script)
-  await ctx.plugin(LlmService)
-  await ctx.plugin(SessionStore)
-  await ctx.plugin(SystemPrompt)
-  await ctx.plugin(ToolRegistry)
-  await ctx.plugin(AgentRegistry)
+  await mountAgentLoopTestDependencies(ctx)
   await ctx.plugin(Invariants)
   await ctx.plugin(AgentLoop, { agents: [] })
   await ctx.plugin(SubagentService)
   await ctx.plugin(spawn, { providerName: 'spawn' })
   await ctx.plugin(WorkerWorkflowEngine, {})
   ctx.llm.registerAdapter(['mock'], adapter)
-  const parent = ctx.agentLoop.create(AgentId('parent'), { model: 'mock' })
+  const parent = ctx.agentLoop.create(AgentId('parent'), { provider: 'mock', model: 'mock' })
   return { ctx, parent, adapter }
 }
 
diff --git a/packages/workflow/workflow-workerthread/tests/meta.spec.ts b/packages/workflow/workflow-workerthread/tests/meta.spec.ts
index 37b86440be..00505a6435 100644
--- a/packages/workflow/workflow-workerthread/tests/meta.spec.ts
+++ b/packages/workflow/workflow-workerthread/tests/meta.spec.ts
@@ -33,7 +33,7 @@ describe('validateMeta', () => {
       description: 'migrate call sites',
       whenToUse: 'large mechanical sweeps',
       phases: [
-        { title: 'Discover' },
+        { title: 'Discover', provider: 'openai' },
         { title: 'Transform', detail: 'one agent per file', model: 'deepseek-v4-pro' },
       ],
     })
@@ -42,7 +42,7 @@ describe('validateMeta', () => {
       description: 'migrate call sites',
       whenToUse: 'large mechanical sweeps',
       phases: [
-        { title: 'Discover' },
+        { title: 'Discover', provider: 'openai' },
         { title: 'Transform', detail: 'one agent per file', model: 'deepseek-v4-pro' },
       ],
     })
@@ -73,6 +73,7 @@ describe('validateMeta', () => {
     expectInvalid({ name: 'x', description: 'd', phases: [{ title: '' }] }, 'meta.phases[0].title must be a non-empty string')
     expectInvalid({ name: 'x', description: 'd', phases: [{ title: 'Scan', order: 1 }] }, 'meta.phases[0].order is not a recognized field')
     expectInvalid({ name: 'x', description: 'd', phases: [{ title: 'Scan', detail: 9 }] }, 'meta.phases[0].detail must be a string')
+    expectInvalid({ name: 'x', description: 'd', phases: [{ title: 'Scan', provider: 9 }] }, 'meta.phases[0].provider must be a string')
     expectInvalid({ name: 'x', description: 'd', phases: [{ title: 'Scan', model: 9 }] }, 'meta.phases[0].model must be a string')
   })
 
diff --git a/packages/workflow/workflow-workerthread/tests/session.spec.ts b/packages/workflow/workflow-workerthread/tests/session.spec.ts
index 07a9bac8d0..102bcbd9a0 100644
--- a/packages/workflow/workflow-workerthread/tests/session.spec.ts
+++ b/packages/workflow/workflow-workerthread/tests/session.spec.ts
@@ -35,7 +35,7 @@ interface FakeHost {
 
 interface FakeHostOptions {
   /** Auto-respond to child-start: reply started + settled per child index. Omit a reply to leave the child pending. */
-  reply?: (request: { prompt: string; schema?: unknown; model?: string }, index: number) => ChildResult | undefined
+  reply?: (request: { prompt: string; schema?: unknown; provider?: string; model?: string }, index: number) => ChildResult | undefined
   /** Reject the start instead (child-start-error) when returning a string. */
   refuse?: (index: number) => string | undefined
   /** Auto-send `go` on `ready` (default true). */
@@ -143,6 +143,17 @@ describe('runWorkerSession over an in-process MessageChannel', () => {
     host.close()
   })
 
+  it('agent({provider}) forwards a provider without inventing a model', async () => {
+    const host = fakeHost({ reply: () => text('ok') })
+    void runWorkerSession(host.port, init("return await agent('route me', { provider: 'openai' })"))
+    const result = await host.result()
+    expect(result.value).toBe('ok')
+    const start = host.ofType(WorkerToHostType.ChildStart)[0]!
+    expect(start.request.provider).toBe('openai')
+    expect(start.request.model).toBeUndefined()
+    host.close()
+  })
+
   it('a schema child completing WITHOUT a structured value resolves null with a failed outcome', async () => {
     const host = fakeHost({ reply: () => text('prose, no structure') })
     void runWorkerSession(host.port, init("return await agent('p', { schema: { type: 'object' } })"))
@@ -330,7 +341,7 @@ describe('runWorkerSession over an in-process MessageChannel', () => {
       ["return await agent('p', { label: 3 })", '"label" must be a string'],
       ["return await agent('p', { get label() { throw new Error('read failed') } })", 'options must be plain JSON data'],
       ["return await agent('p', { bogus: true })", '"bogus" is not recognized'],
-      ["return await agent('p', { effort: 'high' })", '"effort" is deferred'],
+      ["return await agent('p', { effort: 'high' })", '"effort" is deferred and not supported by this engine (supported: label, phase, schema, provider, model)'],
       ["return await agent('p', { schema: { type: 'object', oneOf: [] } })", 'outside the supported subset'],
       ['return await parallel([() => 1, () => 2, () => 3])', 'over the per-call cap (2)'],
       ['return await pipeline([1, 2, 3], (x) => x)', 'maxItemsPerCall'],
diff --git a/packages/workflow/workflow-workerthread/tests/workflow-workerthread.e2e.ts b/packages/workflow/workflow-workerthread/tests/workflow-workerthread.e2e.ts
index 1d6f61432a..d49633636d 100644
--- a/packages/workflow/workflow-workerthread/tests/workflow-workerthread.e2e.ts
+++ b/packages/workflow/workflow-workerthread/tests/workflow-workerthread.e2e.ts
@@ -35,7 +35,7 @@ async function harness(): Promise<Context> {
   await built.plugin(ToolRegistry)
   await built.plugin(AgentRegistry)
   await built.plugin(AgentLoop, { agents: [] })
-  await built.plugin(LlmDeepSeek, { models: ['deepseek-v4-flash'] })
+  await built.plugin(LlmDeepSeek)
   await built.plugin(SubagentService)
   await built.plugin(Spawn, { providerName: 'spawn' })
   await built.plugin(WorkerWorkflowEngine, { provider: 'spawn' })
@@ -64,7 +64,7 @@ describe.skipIf(!process.env.DEEPSEEK_API_KEY)('worker workflow engine with-key
     const parentHandle = await ctx.agents.create({
       agentId: AgentId('wf-worker-e2e-parent'),
       sessionId: 'wf-worker-e2e-session' as never,
-      agentOptions: { model: 'deepseek-v4-flash' },
+      agentOptions: { provider: 'deepseek', model: 'deepseek-v4-flash' },
     })
 
     const events: string[] = []
diff --git a/packages/workflow/workflow-workerthread/tests/workflow-workerthread.spec.ts b/packages/workflow/workflow-workerthread/tests/workflow-workerthread.spec.ts
index 0e12f6e4d9..dffb5c3f9d 100644
--- a/packages/workflow/workflow-workerthread/tests/workflow-workerthread.spec.ts
+++ b/packages/workflow/workflow-workerthread/tests/workflow-workerthread.spec.ts
@@ -223,6 +223,14 @@ describe('dsh-workflow-workerthread', () => {
       expect(provider.runs[0]!.request.parent).toBeDefined()
     })
 
+    it('agent({provider}) forwards provider-only agentOptions across the thread', async () => {
+      const { ctx, parent, provider } = await setup()
+      const result = await run(ctx, parent, scripted("return await agent('route me', { provider: 'openai' })"))
+
+      expect(result.value).toBe('stub reply')
+      expect(provider.runs[0]!.request.agentOptions).toEqual({ provider: 'openai' })
+    })
+
     it('a fatal hook error inside the worker kills the script and reports the error', async () => {
       const { ctx, parent } = await setup()
       const result = await run(ctx, parent, scripted("return await parallel([() => agent('x', { isolation: 'worktree' })])"))
diff --git a/packages/workflow/workflow/src/types.ts b/packages/workflow/workflow/src/types.ts
index 5a00cfbb6d..21d8b4a4fb 100644
--- a/packages/workflow/workflow/src/types.ts
+++ b/packages/workflow/workflow/src/types.ts
@@ -30,6 +30,8 @@ export interface WorkflowPhase {
   title: string
   /** Optional one-line description of what the phase does. */
   detail?: string
+  /** Optional provider override this phase is expected to use (informational). */
+  provider?: string
   /** Optional model override this phase is expected to use (informational). */
   model?: string
 }
diff --git a/pnpm-lock.yaml b/pnpm-lock.yaml
index c212d56ce6..d2a6b91ef6 100644
--- a/pnpm-lock.yaml
+++ b/pnpm-lock.yaml
@@ -68,6 +68,9 @@ importers:
       publint:
         specifier: ^0.3.21
         version: 0.3.21
+      saxes:
+        specifier: ^6.0.0
+        version: 6.0.0
       tsdown:
         specifier: ^0.22.2
         version: 0.22.2(oxc-resolver@11.20.0)(publint@0.3.21)(tsx@4.22.4)(typescript@6.0.3)
@@ -87,11 +90,122 @@ importers:
         specifier: ^4.1.8
         version: 4.1.8(@types/node@22.20.0)(@vitest/coverage-v8@4.1.8)(jsdom@29.1.1)(vite@8.0.16(@types/node@22.20.0)(esbuild@0.28.1)(jiti@2.7.0)(tsx@4.22.4)(yaml@2.9.0))
 
+  examples:
+    dependencies:
+      '@cordisjs/plugin-hmr':
+        specifier: workspace:*
+        version: link:../vendor/hmr
+      '@cordisjs/plugin-include':
+        specifier: workspace:*
+        version: link:../vendor/include
+      '@deepseek-ai/dsh-acp-demo':
+        specifier: workspace:*
+        version: link:../packages/examples/acp-demo
+      '@deepseek-ai/dsh-bash-local':
+        specifier: workspace:*
+        version: link:../packages/bash/bash-local
+      '@deepseek-ai/dsh-bash-sandbox':
+        specifier: workspace:*
+        version: link:../packages/bash/bash-sandbox
+      '@deepseek-ai/dsh-code-runtime-worker':
+        specifier: workspace:*
+        version: link:../packages/code-runtime/code-runtime-worker
+      '@deepseek-ai/dsh-compact-basic':
+        specifier: workspace:*
+        version: link:../packages/compact/compact-basic
+      '@deepseek-ai/dsh-fs-local':
+        specifier: workspace:*
+        version: link:../packages/fs/fs-local
+      '@deepseek-ai/dsh-fs-policy':
+        specifier: workspace:*
+        version: link:../packages/fs/fs-policy
+      '@deepseek-ai/dsh-hooks-claude':
+        specifier: workspace:*
+        version: link:../packages/hooks/hooks-claude
+      '@deepseek-ai/dsh-hooks-codex':
+        specifier: workspace:*
+        version: link:../packages/hooks/hooks-codex
+      '@deepseek-ai/dsh-llm':
+        specifier: workspace:*
+        version: link:../packages/llm/llm
+      '@deepseek-ai/dsh-llm-deepseek':
+        specifier: workspace:*
+        version: link:../packages/llm/llm-deepseek
+      '@deepseek-ai/dsh-llm-replay':
+        specifier: workspace:*
+        version: link:../packages/support/llm-replay
+      '@deepseek-ai/dsh-permission':
+        specifier: workspace:*
+        version: link:../packages/ui/permission
+      '@deepseek-ai/dsh-repeat-tool-guard':
+        specifier: workspace:*
+        version: link:../packages/guard/repeat-tool-guard
+      '@deepseek-ai/dsh-sandbox-local':
+        specifier: workspace:*
+        version: link:../packages/sandbox/sandbox-local
+      '@deepseek-ai/dsh-spill-local':
+        specifier: workspace:*
+        version: link:../packages/spill/spill-local
+      '@deepseek-ai/dsh-spill-policy':
+        specifier: workspace:*
+        version: link:../packages/spill/spill-policy
+      '@deepseek-ai/dsh-stdio-demo':
+        specifier: workspace:*
+        version: link:../packages/examples/stdio-demo
+      '@deepseek-ai/dsh-subagent':
+        specifier: workspace:*
+        version: link:../packages/subagent/subagent
+      '@deepseek-ai/dsh-subagent-fork':
+        specifier: workspace:*
+        version: link:../packages/subagent/subagent-fork
+      '@deepseek-ai/dsh-subagent-spawn':
+        specifier: workspace:*
+        version: link:../packages/subagent/subagent-spawn
+      '@deepseek-ai/dsh-time-context':
+        specifier: workspace:*
+        version: link:../packages/context/time-context
+      '@deepseek-ai/dsh-timeout-policy':
+        specifier: workspace:*
+        version: link:../packages/timeout/timeout-policy
+      '@deepseek-ai/dsh-token-meter':
+        specifier: workspace:*
+        version: link:../packages/llm/token-meter
+      '@deepseek-ai/dsh-tool-cordis':
+        specifier: workspace:*
+        version: link:../packages/cordis/tool-cordis
+      '@deepseek-ai/dsh-tool-fs':
+        specifier: workspace:*
+        version: link:../packages/fs/tool-fs
+      '@deepseek-ai/dsh-tool-fs-search':
+        specifier: workspace:*
+        version: link:../packages/fs/tool-fs-search
+      '@deepseek-ai/dsh-tool-subagent':
+        specifier: workspace:*
+        version: link:../packages/subagent/tool-subagent
+      '@deepseek-ai/dsh-tool-todo':
+        specifier: workspace:*
+        version: link:../packages/todo/tool-todo
+      '@deepseek-ai/dsh-tool-workflow':
+        specifier: workspace:*
+        version: link:../packages/workflow/tool-workflow
+      '@deepseek-ai/dsh-tools':
+        specifier: workspace:*
+        version: link:../packages/core/tools
+      '@deepseek-ai/dsh-user-approval':
+        specifier: workspace:*
+        version: link:../packages/ui/user-approval
+      '@deepseek-ai/dsh-web':
+        specifier: workspace:*
+        version: link:../packages/web/web
+      '@deepseek-ai/dsh-web-fetch-local':
+        specifier: workspace:*
+        version: link:../packages/web/web-fetch-local
+      '@deepseek-ai/dsh-workflow-workerthread':
+        specifier: workspace:*
+        version: link:../packages/workflow/workflow-workerthread
+
   packages/bash/bash:
     devDependencies:
-      '@deepseek-ai/dsh-brand':
-        specifier: workspace:^
-        version: link:../../util/brand
       '@deepseek-ai/dsh-sandbox':
         specifier: workspace:^
         version: link:../../sandbox/sandbox
@@ -144,6 +258,10 @@ importers:
         version: 0.0.0-test.0
 
   packages/bash/tool-bash:
+    dependencies:
+      schemastery:
+        specifier: ^3.18.0
+        version: 3.18.0
     devDependencies:
       '@deepseek-ai/dsh-agent':
         specifier: workspace:^
@@ -151,30 +269,42 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-bash':
         specifier: workspace:^
         version: link:../bash
       '@deepseek-ai/dsh-bash-local':
         specifier: workspace:^
         version: link:../bash-local
-      '@deepseek-ai/dsh-bash-sandbox':
+      '@deepseek-ai/dsh-home':
         specifier: workspace:^
-        version: link:../bash-sandbox
+        version: link:../../util/home
       '@deepseek-ai/dsh-llm':
         specifier: workspace:^
         version: link:../../llm/llm
       '@deepseek-ai/dsh-sandbox':
         specifier: workspace:^
         version: link:../../sandbox/sandbox
-      '@deepseek-ai/dsh-sandbox-local':
-        specifier: workspace:^
-        version: link:../../sandbox/sandbox-local
       '@deepseek-ai/dsh-session':
         specifier: workspace:^
         version: link:../../core/session
+      '@deepseek-ai/dsh-session-persistence':
+        specifier: workspace:^
+        version: link:../../session-persistence/session-persistence
+      '@deepseek-ai/dsh-session-persistence-jsonl':
+        specifier: workspace:^
+        version: link:../../session-persistence/session-persistence-jsonl
       '@deepseek-ai/dsh-system-prompt':
         specifier: workspace:^
         version: link:../../core/system-prompt
+      '@deepseek-ai/dsh-tasks':
+        specifier: workspace:^
+        version: link:../../tasks/tasks
+      '@deepseek-ai/dsh-tool-tasks':
+        specifier: workspace:^
+        version: link:../../tasks/tool-tasks
       '@deepseek-ai/dsh-tools':
         specifier: workspace:^
         version: link:../../core/tools
@@ -217,13 +347,26 @@ importers:
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
 
   packages/compact/compact-basic:
+    dependencies:
+      schemastery:
+        specifier: ^3.18.0
+        version: 3.18.0
     devDependencies:
+      '@cordisjs/plugin-include':
+        specifier: workspace:^
+        version: link:../../../vendor/include
+      '@cordisjs/plugin-loader':
+        specifier: workspace:^
+        version: link:../../../vendor/loader
       '@deepseek-ai/dsh-agent':
         specifier: workspace:^
         version: link:../../core/agent
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-compact':
         specifier: workspace:^
         version: link:../compact
@@ -236,15 +379,15 @@ importers:
       '@deepseek-ai/dsh-session':
         specifier: workspace:^
         version: link:../../core/session
-      '@deepseek-ai/dsh-system-prompt':
+      '@deepseek-ai/dsh-token-meter':
         specifier: workspace:^
-        version: link:../../core/system-prompt
+        version: link:../../llm/token-meter
       '@deepseek-ai/dsh-tools':
         specifier: workspace:^
         version: link:../../core/tools
       cordis:
         specifier: ^4.0.0-rc.7
-        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@vendor+include)(@cordisjs/plugin-loader@vendor+loader)
 
   packages/context/time-context:
     dependencies:
@@ -258,9 +401,15 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-llm':
         specifier: workspace:^
         version: link:../../llm/llm
+      '@deepseek-ai/dsh-loader-smoke':
+        specifier: workspace:^
+        version: link:../../support/loader-smoke
       '@deepseek-ai/dsh-session':
         specifier: workspace:^
         version: link:../../core/session
@@ -274,6 +423,52 @@ importers:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
 
+  packages/context/workspace-context:
+    dependencies:
+      schemastery:
+        specifier: ^3.18.0
+        version: 3.18.0
+    devDependencies:
+      '@cordisjs/plugin-loader':
+        specifier: workspace:^
+        version: link:../../../vendor/loader
+      '@deepseek-ai/dsh-agent':
+        specifier: workspace:^
+        version: link:../../core/agent
+      '@deepseek-ai/dsh-agent-loop':
+        specifier: workspace:^
+        version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-fs':
+        specifier: workspace:^
+        version: link:../../fs/fs
+      '@deepseek-ai/dsh-fs-local':
+        specifier: workspace:^
+        version: link:../../fs/fs-local
+      '@deepseek-ai/dsh-llm':
+        specifier: workspace:^
+        version: link:../../llm/llm
+      '@deepseek-ai/dsh-llm-deepseek':
+        specifier: workspace:^
+        version: link:../../llm/llm-deepseek
+      '@deepseek-ai/dsh-paths':
+        specifier: workspace:^
+        version: link:../../util/paths
+      '@deepseek-ai/dsh-session':
+        specifier: workspace:^
+        version: link:../../core/session
+      '@deepseek-ai/dsh-system-prompt':
+        specifier: workspace:^
+        version: link:../../core/system-prompt
+      '@deepseek-ai/dsh-tool-fs':
+        specifier: workspace:^
+        version: link:../../fs/tool-fs
+      '@deepseek-ai/dsh-tools':
+        specifier: workspace:^
+        version: link:../../core/tools
+      cordis:
+        specifier: ^4.0.0-rc.6
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@vendor+loader)
+
   packages/cordis/tool-cordis:
     dependencies:
       schemastery:
@@ -292,6 +487,9 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-llm':
         specifier: workspace:^
         version: link:../../llm/llm
@@ -469,6 +667,9 @@ importers:
       '@deepseek-ai/dsh-user-interaction':
         specifier: workspace:^
         version: link:../../ui/user-interaction
+      '@deepseek-ai/dsh-workspace-context':
+        specifier: workspace:^
+        version: link:../../context/workspace-context
       cordis:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@vendor+include)(@cordisjs/plugin-loader@vendor+loader)
@@ -491,6 +692,12 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-fs-local':
+        specifier: workspace:^
+        version: link:../../fs/fs-local
+      '@deepseek-ai/dsh-home':
+        specifier: workspace:^
+        version: link:../../util/home
       '@deepseek-ai/dsh-invariants':
         specifier: workspace:^
         version: link:../../support/invariants
@@ -509,15 +716,24 @@ importers:
       '@deepseek-ai/dsh-system-prompt':
         specifier: workspace:^
         version: link:../../core/system-prompt
+      '@deepseek-ai/dsh-tasks':
+        specifier: workspace:^
+        version: link:../../tasks/tasks
       '@deepseek-ai/dsh-tool-bash':
         specifier: workspace:^
         version: link:../../bash/tool-bash
       '@deepseek-ai/dsh-tool-skill':
         specifier: workspace:^
         version: link:../../skill/tool-skill
+      '@deepseek-ai/dsh-tool-tasks':
+        specifier: workspace:^
+        version: link:../../tasks/tool-tasks
       '@deepseek-ai/dsh-tools':
         specifier: workspace:^
         version: link:../../core/tools
+      '@deepseek-ai/dsh-workspace-context':
+        specifier: workspace:^
+        version: link:../../context/workspace-context
       cordis:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
@@ -576,6 +792,9 @@ importers:
       '@deepseek-ai/dsh-user-interaction':
         specifier: workspace:^
         version: link:../../ui/user-interaction
+      '@deepseek-ai/dsh-workspace-context':
+        specifier: workspace:^
+        version: link:../../context/workspace-context
       cordis:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@vendor+include)(@cordisjs/plugin-loader@vendor+loader)
@@ -638,6 +857,9 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-fs':
         specifier: workspace:^
         version: link:../fs
@@ -666,6 +888,43 @@ importers:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
 
+  packages/fs/tool-fs-search:
+    dependencies:
+      schemastery:
+        specifier: ^3.18.0
+        version: 3.18.0
+    devDependencies:
+      '@deepseek-ai/dsh-agent':
+        specifier: workspace:^
+        version: link:../../core/agent
+      '@deepseek-ai/dsh-bash':
+        specifier: workspace:^
+        version: link:../../bash/bash
+      '@deepseek-ai/dsh-bash-local':
+        specifier: workspace:^
+        version: link:../../bash/bash-local
+      '@deepseek-ai/dsh-llm':
+        specifier: workspace:^
+        version: link:../../llm/llm
+      '@deepseek-ai/dsh-retention':
+        specifier: workspace:^
+        version: link:../../util/retention
+      '@deepseek-ai/dsh-session':
+        specifier: workspace:^
+        version: link:../../core/session
+      '@deepseek-ai/dsh-spill':
+        specifier: workspace:^
+        version: link:../../spill/spill
+      '@deepseek-ai/dsh-system-prompt':
+        specifier: workspace:^
+        version: link:../../core/system-prompt
+      '@deepseek-ai/dsh-tools':
+        specifier: workspace:^
+        version: link:../../core/tools
+      cordis:
+        specifier: ^4.0.0-rc.6
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
   packages/guard/repeat-tool-guard:
     dependencies:
       schemastery:
@@ -678,15 +937,15 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-llm':
         specifier: workspace:^
         version: link:../../llm/llm
       '@deepseek-ai/dsh-session':
         specifier: workspace:^
         version: link:../../core/session
-      '@deepseek-ai/dsh-system-prompt':
-        specifier: workspace:^
-        version: link:../../core/system-prompt
       '@deepseek-ai/dsh-tools':
         specifier: workspace:^
         version: link:../../core/tools
@@ -718,6 +977,9 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-bash':
         specifier: workspace:^
         version: link:../../bash/bash
@@ -733,12 +995,15 @@ importers:
       '@deepseek-ai/dsh-session':
         specifier: workspace:^
         version: link:../../core/session
+      '@deepseek-ai/dsh-session-persistence':
+        specifier: workspace:^
+        version: link:../../session-persistence/session-persistence
+      '@deepseek-ai/dsh-session-persistence-jsonl':
+        specifier: workspace:^
+        version: link:../../session-persistence/session-persistence-jsonl
       '@deepseek-ai/dsh-subagent':
         specifier: workspace:^
         version: link:../../subagent/subagent
-      '@deepseek-ai/dsh-system-prompt':
-        specifier: workspace:^
-        version: link:../../core/system-prompt
       '@deepseek-ai/dsh-tools':
         specifier: workspace:^
         version: link:../../core/tools
@@ -758,6 +1023,9 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-bash':
         specifier: workspace:^
         version: link:../../bash/bash
@@ -773,9 +1041,12 @@ importers:
       '@deepseek-ai/dsh-session':
         specifier: workspace:^
         version: link:../../core/session
-      '@deepseek-ai/dsh-system-prompt':
+      '@deepseek-ai/dsh-session-persistence':
         specifier: workspace:^
-        version: link:../../core/system-prompt
+        version: link:../../session-persistence/session-persistence
+      '@deepseek-ai/dsh-session-persistence-jsonl':
+        specifier: workspace:^
+        version: link:../../session-persistence/session-persistence-jsonl
       '@deepseek-ai/dsh-tools':
         specifier: workspace:^
         version: link:../../core/tools
@@ -824,6 +1095,22 @@ importers:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
 
+  packages/llm/token-meter:
+    dependencies:
+      schemastery:
+        specifier: ^3.18.0
+        version: 3.18.0
+    devDependencies:
+      '@deepseek-ai/dsh-llm':
+        specifier: workspace:^
+        version: link:../llm
+      '@deepseek-ai/dsh-session':
+        specifier: workspace:^
+        version: link:../../core/session
+      cordis:
+        specifier: ^4.0.0-rc.7
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
   packages/mcp/mcp-client:
     dependencies:
       '@modelcontextprotocol/sdk':
@@ -880,6 +1167,87 @@ importers:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
 
+  packages/sdk/create-sdk:
+    dependencies:
+      '@deepseek-ai/dsh-helper':
+        specifier: workspace:^
+        version: link:../helper
+      commander:
+        specifier: ^15.0.0
+        version: 15.0.0
+    devDependencies:
+      cordis:
+        specifier: ^4.0.0-rc.7
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
+  packages/sdk/helper:
+    dependencies:
+      '@clack/core':
+        specifier: ^1.4.3
+        version: 1.4.3
+      '@clack/prompts':
+        specifier: ^1.7.0
+        version: 1.7.0
+      handlebars:
+        specifier: ^4.7.9
+        version: 4.7.9
+      jsonc-parser:
+        specifier: ^3.3.1
+        version: 3.3.1
+      yaml:
+        specifier: ^2.9.0
+        version: 2.9.0
+    devDependencies:
+      '@deepseek-ai/dsh-brand':
+        specifier: workspace:^
+        version: link:../../util/brand
+      '@deepseek-ai/dsh-hooks-claude':
+        specifier: workspace:^
+        version: link:../../hooks/hooks-claude
+      '@deepseek-ai/dsh-hooks-codex':
+        specifier: workspace:^
+        version: link:../../hooks/hooks-codex
+      '@deepseek-ai/dsh-session-persistence-jsonl':
+        specifier: workspace:^
+        version: link:../../session-persistence/session-persistence-jsonl
+      '@deepseek-ai/dsh-session-persistence-sqlite':
+        specifier: workspace:^
+        version: link:../../session-persistence/session-persistence-sqlite
+      '@deepseek-ai/dsh-tool-subagent':
+        specifier: workspace:^
+        version: link:../../subagent/tool-subagent
+      '@deepseek-ai/dsh-tool-web':
+        specifier: workspace:^
+        version: link:../../web/tool-web
+      cordis:
+        specifier: ^4.0.0-rc.7
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
+  packages/sdk/scripts:
+    dependencies:
+      '@deepseek-ai/dsh-helper':
+        specifier: workspace:^
+        version: link:../helper
+      commander:
+        specifier: ^15.0.0
+        version: 15.0.0
+      node-addon-require-builtin:
+        specifier: ^0.1.0
+        version: 0.1.0
+    devDependencies:
+      '@deepseek-ai/dsh-app-boot':
+        specifier: workspace:^
+        version: link:../../ui/app-boot
+      cordis:
+        specifier: ^4.0.0-rc.7
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+      tsdown:
+        specifier: ^0.22.2
+        version: 0.22.2(oxc-resolver@11.20.0)(publint@0.3.21)(tsx@4.22.4)(typescript@6.0.3)
+      tsx:
+        specifier: ^4.22.4
+        version: 4.22.4
+
   packages/session-persistence/session-persistence:
     devDependencies:
       '@deepseek-ai/dsh-session':
@@ -962,6 +1330,9 @@ importers:
       '@deepseek-ai/dsh-fs':
         specifier: workspace:^
         version: link:../../fs/fs
+      '@deepseek-ai/dsh-home':
+        specifier: workspace:^
+        version: link:../../util/home
       '@deepseek-ai/dsh-skill':
         specifier: workspace:^
         version: link:../skill
@@ -997,6 +1368,71 @@ importers:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
 
+  packages/spill/spill:
+    devDependencies:
+      '@deepseek-ai/dsh-brand':
+        specifier: workspace:^
+        version: link:../../util/brand
+      '@deepseek-ai/dsh-llm':
+        specifier: workspace:^
+        version: link:../../llm/llm
+      '@deepseek-ai/dsh-session':
+        specifier: workspace:^
+        version: link:../../core/session
+      cordis:
+        specifier: ^4.0.0-rc.6
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
+  packages/spill/spill-local:
+    dependencies:
+      schemastery:
+        specifier: ^3.18.0
+        version: 3.18.0
+    devDependencies:
+      '@deepseek-ai/dsh-brand':
+        specifier: workspace:^
+        version: link:../../util/brand
+      '@deepseek-ai/dsh-llm':
+        specifier: workspace:^
+        version: link:../../llm/llm
+      '@deepseek-ai/dsh-session':
+        specifier: workspace:^
+        version: link:../../core/session
+      '@deepseek-ai/dsh-spill':
+        specifier: workspace:^
+        version: link:../spill
+      cordis:
+        specifier: ^4.0.0-rc.6
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
+  packages/spill/spill-policy:
+    dependencies:
+      schemastery:
+        specifier: ^3.18.0
+        version: 3.18.0
+    devDependencies:
+      '@deepseek-ai/dsh-agent':
+        specifier: workspace:^
+        version: link:../../core/agent
+      '@deepseek-ai/dsh-llm':
+        specifier: workspace:^
+        version: link:../../llm/llm
+      '@deepseek-ai/dsh-retention':
+        specifier: workspace:^
+        version: link:../../util/retention
+      '@deepseek-ai/dsh-session':
+        specifier: workspace:^
+        version: link:../../core/session
+      '@deepseek-ai/dsh-spill':
+        specifier: workspace:^
+        version: link:../spill
+      '@deepseek-ai/dsh-tools':
+        specifier: workspace:^
+        version: link:../../core/tools
+      cordis:
+        specifier: ^4.0.0-rc.6
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
   packages/subagent/subagent:
     devDependencies:
       '@deepseek-ai/dsh-agent':
@@ -1033,6 +1469,9 @@ importers:
       '@deepseek-ai/dsh-llm':
         specifier: workspace:^
         version: link:../../llm/llm
+      '@deepseek-ai/dsh-loader-smoke':
+        specifier: workspace:^
+        version: link:../../support/loader-smoke
       '@deepseek-ai/dsh-subagent':
         specifier: workspace:^
         version: link:../subagent
@@ -1058,6 +1497,9 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-invariants':
         specifier: workspace:^
         version: link:../../support/invariants
@@ -1076,12 +1518,6 @@ importers:
       '@deepseek-ai/dsh-subagent-spawn':
         specifier: workspace:^
         version: link:../subagent-spawn
-      '@deepseek-ai/dsh-system-prompt':
-        specifier: workspace:^
-        version: link:../../core/system-prompt
-      '@deepseek-ai/dsh-tools':
-        specifier: workspace:^
-        version: link:../../core/tools
       cordis:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
@@ -1094,6 +1530,9 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-invariants':
         specifier: workspace:^
         version: link:../../support/invariants
@@ -1131,6 +1570,9 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-bash-local':
         specifier: workspace:^
         version: link:../../bash/bash-local
@@ -1152,18 +1594,12 @@ importers:
       '@deepseek-ai/dsh-subagent-inprocess':
         specifier: workspace:^
         version: link:../subagent-inprocess
-      '@deepseek-ai/dsh-system-prompt':
-        specifier: workspace:^
-        version: link:../../core/system-prompt
       '@deepseek-ai/dsh-tool-bash':
         specifier: workspace:^
         version: link:../../bash/tool-bash
       '@deepseek-ai/dsh-tool-subagent':
         specifier: workspace:^
         version: link:../tool-subagent
-      '@deepseek-ai/dsh-tools':
-        specifier: workspace:^
-        version: link:../../core/tools
       cordis:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
@@ -1198,6 +1634,12 @@ importers:
       '@deepseek-ai/dsh-system-prompt':
         specifier: workspace:^
         version: link:../../core/system-prompt
+      '@deepseek-ai/dsh-tasks':
+        specifier: workspace:^
+        version: link:../../tasks/tasks
+      '@deepseek-ai/dsh-tool-tasks':
+        specifier: workspace:^
+        version: link:../../tasks/tool-tasks
       '@deepseek-ai/dsh-tools':
         specifier: workspace:^
         version: link:../../core/tools
@@ -1210,9 +1652,9 @@ importers:
       '@agentclientprotocol/sdk':
         specifier: 0.25.1
         version: 0.25.1(zod@4.4.3)
-      tsx:
-        specifier: ^4.22.4
-        version: 4.22.4
+      '@deepseek-ai/dsh-loader-smoke':
+        specifier: workspace:*
+        version: link:../loader-smoke
       vitest:
         specifier: ^4.1.8
         version: 4.1.8(@types/node@25.9.3)(@vitest/coverage-v8@4.1.8)(jsdom@29.1.1)(vite@8.0.16(@types/node@25.9.3)(esbuild@0.28.1)(jiti@2.7.0)(tsx@4.22.4)(yaml@2.9.0))
@@ -1221,6 +1663,30 @@ importers:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
 
+  packages/support/agent-loop-testkit:
+    devDependencies:
+      '@deepseek-ai/dsh-agent':
+        specifier: workspace:^
+        version: link:../../core/agent
+      '@deepseek-ai/dsh-agent-loop':
+        specifier: workspace:^
+        version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-llm':
+        specifier: workspace:^
+        version: link:../../llm/llm
+      '@deepseek-ai/dsh-session':
+        specifier: workspace:^
+        version: link:../../core/session
+      '@deepseek-ai/dsh-system-prompt':
+        specifier: workspace:^
+        version: link:../../core/system-prompt
+      '@deepseek-ai/dsh-tools':
+        specifier: workspace:^
+        version: link:../../core/tools
+      cordis:
+        specifier: ^4.0.0-rc.7
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
   packages/support/invariants:
     devDependencies:
       '@deepseek-ai/dsh-agent':
@@ -1295,6 +1761,52 @@ importers:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
 
+  packages/tasks/tasks:
+    devDependencies:
+      '@deepseek-ai/dsh-agent':
+        specifier: workspace:^
+        version: link:../../core/agent
+      '@deepseek-ai/dsh-brand':
+        specifier: workspace:^
+        version: link:../../util/brand
+      '@deepseek-ai/dsh-session':
+        specifier: workspace:^
+        version: link:../../core/session
+      '@deepseek-ai/dsh-timeout':
+        specifier: workspace:^
+        version: link:../../util/timeout
+      cordis:
+        specifier: ^4.0.0-rc.6
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
+  packages/tasks/tool-tasks:
+    dependencies:
+      schemastery:
+        specifier: ^3.18.0
+        version: 3.18.0
+    devDependencies:
+      '@deepseek-ai/dsh-agent':
+        specifier: workspace:^
+        version: link:../../core/agent
+      '@deepseek-ai/dsh-llm':
+        specifier: workspace:^
+        version: link:../../llm/llm
+      '@deepseek-ai/dsh-session':
+        specifier: workspace:^
+        version: link:../../core/session
+      '@deepseek-ai/dsh-system-prompt':
+        specifier: workspace:^
+        version: link:../../core/system-prompt
+      '@deepseek-ai/dsh-tasks':
+        specifier: workspace:^
+        version: link:../tasks
+      '@deepseek-ai/dsh-tools':
+        specifier: workspace:^
+        version: link:../../core/tools
+      cordis:
+        specifier: ^4.0.0-rc.6
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
   packages/timeout/timeout-policy:
     devDependencies:
       '@deepseek-ai/dsh-llm':
@@ -1318,6 +1830,9 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-llm':
         specifier: workspace:^
         version: link:../../llm/llm
@@ -1352,6 +1867,9 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-bash':
         specifier: workspace:^
         version: link:../../bash/bash
@@ -1573,6 +2091,24 @@ importers:
         specifier: ^4.0.0-rc.7
         version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
 
+  packages/util/home:
+    devDependencies:
+      cordis:
+        specifier: ^4.0.0-rc.6
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
+  packages/util/paths:
+    devDependencies:
+      cordis:
+        specifier: ^4.0.0-rc.6
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
+  packages/util/retention:
+    devDependencies:
+      cordis:
+        specifier: ^4.0.0-rc.6
+        version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5)
+
   packages/util/timeout:
     devDependencies:
       cordis:
@@ -1594,6 +2130,12 @@ importers:
       '@deepseek-ai/dsh-session':
         specifier: workspace:^
         version: link:../../core/session
+      '@deepseek-ai/dsh-spill-local':
+        specifier: workspace:^
+        version: link:../../spill/spill-local
+      '@deepseek-ai/dsh-spill-policy':
+        specifier: workspace:^
+        version: link:../../spill/spill-policy
       '@deepseek-ai/dsh-system-prompt':
         specifier: workspace:^
         version: link:../../core/system-prompt
@@ -1748,6 +2290,9 @@ importers:
       '@deepseek-ai/dsh-agent-loop':
         specifier: workspace:^
         version: link:../../core/agent-loop
+      '@deepseek-ai/dsh-agent-loop-testkit':
+        specifier: workspace:^
+        version: link:../../support/agent-loop-testkit
       '@deepseek-ai/dsh-brand':
         specifier: workspace:^
         version: link:../../util/brand
@@ -1838,6 +2383,9 @@ importers:
       '@deepseek-ai/dsh-fs-policy':
         specifier: workspace:^
         version: link:../../packages/fs/fs-policy
+      '@deepseek-ai/dsh-home':
+        specifier: workspace:^
+        version: link:../../packages/util/home
       '@deepseek-ai/dsh-hook-protocol':
         specifier: workspace:^
         version: link:../../packages/hooks/hook-protocol
@@ -1865,6 +2413,9 @@ importers:
       '@deepseek-ai/dsh-llm-pi-ai':
         specifier: workspace:^
         version: link:../../packages/llm/llm-pi-ai
+      '@deepseek-ai/dsh-paths':
+        specifier: workspace:^
+        version: link:../../packages/util/paths
       '@deepseek-ai/dsh-permission':
         specifier: workspace:^
         version: link:../../packages/ui/permission
@@ -1916,12 +2467,18 @@ importers:
       '@deepseek-ai/dsh-system-prompt':
         specifier: workspace:^
         version: link:../../packages/core/system-prompt
+      '@deepseek-ai/dsh-tasks':
+        specifier: workspace:^
+        version: link:../../packages/tasks/tasks
       '@deepseek-ai/dsh-timeout':
         specifier: workspace:^
         version: link:../../packages/util/timeout
       '@deepseek-ai/dsh-timeout-policy':
         specifier: workspace:^
         version: link:../../packages/timeout/timeout-policy
+      '@deepseek-ai/dsh-token-meter':
+        specifier: workspace:^
+        version: link:../../packages/llm/token-meter
       '@deepseek-ai/dsh-tool-ask-user':
         specifier: workspace:^
         version: link:../../packages/ui/tool-ask-user
@@ -1940,6 +2497,9 @@ importers:
       '@deepseek-ai/dsh-tool-subagent':
         specifier: workspace:^
         version: link:../../packages/subagent/tool-subagent
+      '@deepseek-ai/dsh-tool-tasks':
+        specifier: workspace:^
+        version: link:../../packages/tasks/tool-tasks
       '@deepseek-ai/dsh-tool-todo':
         specifier: workspace:^
         version: link:../../packages/todo/tool-todo
@@ -1979,6 +2539,9 @@ importers:
       '@deepseek-ai/dsh-workflow-workerthread':
         specifier: workspace:^
         version: link:../../packages/workflow/workflow-workerthread
+      '@deepseek-ai/dsh-workspace-context':
+        specifier: workspace:^
+        version: link:../../packages/context/workspace-context
       cordis:
         specifier: workspace:^
         version: link:../../vendor/cordis
@@ -2407,6 +2970,14 @@ packages:
   '@chevrotain/types@11.1.2':
     resolution: {integrity: sha512-U+HFai5+zmJCkK86QsaJtoITlboZHBqrVketcO2ROv865xfCMSFpELQoz1GkX5GzME8pTa+3kbKrZHQtI0gdbw==}
 
+  '@clack/core@1.4.3':
+    resolution: {integrity: sha512-/kr3UWNtdJfxZtPgDqUOmG2pvwlmcLGheex5yiZKdwbzZJxhV+HMNR9QNmyY5cGwTNV6LrR7Jtp+KjhUAP1qBQ==}
+    engines: {node: '>= 20.12.0'}
+
+  '@clack/prompts@1.7.0':
+    resolution: {integrity: sha512-y7/yvZ2TPAnR9+jnc00klvNNLkJiXFFrQA/hlLCcxA9a2A4zQIOimyFQ9XfwYKiGD1fb5GY8vbKIIgO8d5Tb2A==}
+    engines: {node: '>= 20.12.0'}
+
   '@cordisjs/plugin-include@1.0.4':
     resolution: {integrity: sha512-b1Hm1wmue0v7d/jayoXoBjCV2J14XWTL5yyDZEYeL2L9HgcyTq6JbCw99ozSbei98uAbkwq/pBhimsp/HsySeg==}
     peerDependencies:
@@ -4148,6 +4719,10 @@ packages:
   comma-separated-tokens@2.0.3:
     resolution: {integrity: sha512-Fu4hJdvzeylCfQPp9SGWidpzrMs7tTrlu6Vb8XGaRGck8QSNZJJp538Wrb60Lax4fPwR64ViY468OIUTbRlGZg==}
 
+  commander@15.0.0:
+    resolution: {integrity: sha512-z67u4ZhzCL/Tydu1lJARtEZYWbWaN7oYLHbsuzocr6y4N6WZAagG3RQ4FW61V1/0+jImpj293XfrcYnd1qxtPg==}
+    engines: {node: '>=22.12.0'}
+
   commander@7.2.0:
     resolution: {integrity: sha512-QrWXB+ZQSVPmIWIhtEO9H+gwHaMGYiF5ChvoJ+K9ZGHG/sVsa6yiesAD1GC/x46sET00Xlwo1u49RVVVzvcSkw==}
     engines: {node: '>= 10'}
@@ -4628,9 +5203,18 @@ packages:
   fast-levenshtein@2.0.6:
     resolution: {integrity: sha512-DCXu6Ifhqcks7TZKY3Hxp3y6qphY5SJZmrWMDrKcERSOXWQdMhU9Ig/PYrzyw/ul9jOIyh0N4M0tbC5hodg8dw==}
 
+  fast-string-truncated-width@3.0.3:
+    resolution: {integrity: sha512-0jjjIEL6+0jag3l2XWWizO64/aZVtpiGE3t0Zgqxv0DPuxiMjvB3M24fCyhZUO4KomJQPj3LTSUnDP3GpdwC0g==}
+
+  fast-string-width@3.0.2:
+    resolution: {integrity: sha512-gX8LrtNEI5hq8DVUfRQMbr5lpaS4nMIWV+7XEbXk2b8kiQIizgnlr12B4dA3ZEx3308ze0O4Q1R+cHts8kyUJg==}
+
   fast-uri@3.1.3:
     resolution: {integrity: sha512-i70LwGWUduXqzicKXWshooq+sWL1K3WUU5rKZNG/0i3a1OSoX3HqhH5WbWwTmqWfor4urUakGPiRQcleRZTwOg==}
 
+  fast-wrap-ansi@0.2.2:
+    resolution: {integrity: sha512-7F2Fl+TjRSenLqlU3UjSH0iyqopqoZIu7eZVpEirP2g1GtWa2G/ecEmBdgz31+Mxr+ELclgg6sokpSFIQiZ02Q==}
+
   fast-xml-builder@1.2.0:
     resolution: {integrity: sha512-00aAWieqff+ZJhsXA4g1g7M8k+7AYoMUUHF+/zFb5U6Uv/P0Vl4QZo84/IcufzYalLuEj9928bXN9PbbFzMF0Q==}
 
@@ -4762,6 +5346,11 @@ packages:
   hachure-fill@0.5.2:
     resolution: {integrity: sha512-3GKBOn+m2LX9iq+JC1064cSFprJY4jL1jCXTcpnfER5HYE2l/4EfWSGzkPa/ZDBmYI0ZOEj5VHV/eKnPGkHuOg==}
 
+  handlebars@4.7.9:
+    resolution: {integrity: sha512-4E71E0rpOaQuJR2A3xDZ+GM1HyWYv1clR58tC8emQNeQe3RH7MAzSbat+V0wG78LQBo6m6bzSG/L4pBuCsgnUQ==}
+    engines: {node: '>=0.4.7'}
+    hasBin: true
+
   has-flag@4.0.0:
     resolution: {integrity: sha512-EykJT/Q1KjTWctppgIAgfSO0tKVuZUjhgMr17kqTumMl6Afv3EISleU7qZUzoXDFTAHTDC4NOoG/ZxU3EvlMPQ==}
     engines: {node: '>=8'}
@@ -4994,6 +5583,9 @@ packages:
   json-stable-stringify-without-jsonify@1.0.1:
     resolution: {integrity: sha512-Bdboy+l7tA3OGW6FjyFHWkP5LuByj1Tk33Ljyq0axyzdk9//JSi2u3fP1QSmd1KNwq6VOKYGlAu87CisVir6Pw==}
 
+  jsonc-parser@3.3.1:
+    resolution: {integrity: sha512-HUgH65KyejrUFPvHFPbqOY0rsFip3Bo5wb4ngvdi1EpCYWUQDC5V+Y7mZws+DLkr4M//zQJoanu1SP+87Dv1oQ==}
+
   jsx-ast-utils-x@0.1.0:
     resolution: {integrity: sha512-eQQBjBnsVtGacsG9uJNB8qOr3yA8rga4wAaGG1qRcBzSIvfhERLrWxMAM1hp5fcS6Abo8M4+bUBTekYR0qTPQw==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
@@ -5361,6 +5953,9 @@ packages:
     resolution: {integrity: sha512-OBwBN9AL4dqmETlpS2zasx+vTeWclWzkblfZk7KTA5j3jeOONz/tRCnZomUyvNg83wL5Zv9Ss6HMJXAgL8R2Yg==}
     engines: {node: '>=16 || 14 >=14.17'}
 
+  minimist@1.2.8:
+    resolution: {integrity: sha512-2yyAR8qBkN3YuheJanUpWC5U3bb5osDywNB8RzDVlDwDHbocAJveqqj1u8+SVD7jkWT4yvsHCpWqqWqAxb0zCA==}
+
   minipass@7.1.3:
     resolution: {integrity: sha512-tEBHqDnIoM/1rXME1zgka9g6Q2lcoCkxHLuc7ODJ5BxbP5d4c2Z5cGgtXAku59200Cx7diuHTOYfSBD8n6mm8A==}
     engines: {node: '>=16 || 14 >=14.17'}
@@ -5390,6 +5985,9 @@ packages:
     resolution: {integrity: sha512-8Ofs/AUQh8MaEcrlq5xOX0CQ9ypTF5dl78mjlMNfOK08fzpgTHQRQPBxcPlEtIw0yRpws+Zo/3r+5WRby7u3Gg==}
     engines: {node: '>= 0.6'}
 
+  neo-async@2.6.2:
+    resolution: {integrity: sha512-Yd3UES5mWCSqR+qNT93S3UoYUkqAZ9lLg8a7g9rimsWmYGK8cVToA4/sF3RrshdyV3sAGMXVUmpMYOw+dLpOuw==}
+
   node-addon-landlock-run-linux-arm64@0.0.0-test.0:
     resolution: {integrity: sha512-oJsXcC33qKl9mWYx0n9YPJ2pUAoY39PoIX0Gx4lDrSCTEvENFrEaODAsQYNY+eEGpn9YMN7E+FOftvea3/1FqQ==}
     engines: {node: '>=20'}
@@ -5813,6 +6411,9 @@ packages:
     resolution: {integrity: sha512-bzyZ1e88w9O1iNJbKnOlvYTrWPDl46O1bG0D3XInv+9tkPrxrN8jUUTiFlDkkmKWgn1M6CfIA13SuGqOa9Korw==}
     engines: {node: '>=14'}
 
+  sisteransi@1.0.5:
+    resolution: {integrity: sha512-bLGGlR1QxBcynn2d5YmDX4MGjlZvy2MRBDRNHLJ8VI6l6+9FUiyTFNJ0IveOSP0bcXgVDPRcfGqA0pjaqUpfVg==}
+
   smol-toml@1.6.1:
     resolution: {integrity: sha512-dWUG8F5sIIARXih1DTaQAX4SsiTXhInKf1buxdY9DIg4ZYPZK5nGM1VRIYmEbDbsHt7USo99xSLFu5Q1IqTmsg==}
     engines: {node: '>= 18'}
@@ -5821,6 +6422,10 @@ packages:
     resolution: {integrity: sha512-UXWMKhLOwVKb728IUtQPXxfYU+usdybtUrK/8uGE8CQMvrhOpwvzDBwj0QhSL7MQc7vIsISBG8VQ8+IDQxpfQA==}
     engines: {node: '>=0.10.0'}
 
+  source-map@0.6.1:
+    resolution: {integrity: sha512-UjgapumWlbMhkBgzT7Ykc5YXUT46F0iKu8SGXq0bcwP5dz/h0Plj6enJqjz1Zbq2l5WaqYnrVbwWOWMyF3F47g==}
+    engines: {node: '>=0.10.0'}
+
   space-separated-tokens@2.0.2:
     resolution: {integrity: sha512-PEGlAwrG8yXGXRjW32fGbg66JAlOAwbObuqVoJpv/mRgoWDQfgH1wDPvtzWyUSNAXBGSk8h755YDbbcEy3SH2Q==}
 
@@ -6018,6 +6623,11 @@ packages:
     engines: {node: '>=14.17'}
     hasBin: true
 
+  uglify-js@3.19.3:
+    resolution: {integrity: sha512-v3Xu+yuwBXisp6QYTcH4UbH+xYJXqnq2m/LtQVWKWzYc1iehYnLixoQDN9FH6/j9/oybfd6W9Ghwkl8+UMKTKQ==}
+    engines: {node: '>=0.8.0'}
+    hasBin: true
+
   unbash@3.0.0:
     resolution: {integrity: sha512-FeFPZ/WFT0mbRCuydiZzpPFlrYN8ZUpphQKoq4EeElVIYjYyGzPMxQR/simUwCOJIyVhpFk4RbtyO7RuMpMnHA==}
     engines: {node: '>=14'}
@@ -6258,6 +6868,9 @@ packages:
     resolution: {integrity: sha512-BN22B5eaMMI9UMtjrGd5g5eCYPpCPDUy0FJXbYsaT5zYxjFOckS53SQDE3pWkVoWpHXVb3BrYcEN4Twa55B5cA==}
     engines: {node: '>=0.10.0'}
 
+  wordwrap@1.0.0:
+    resolution: {integrity: sha512-gvVzJFlPycKc5dZN4yPkP8w7Dc37BtP1yczEneOb4uq34pXZcvrtRTmWV8W+Ume+XCxKgbjM+nevkyFPMybd4Q==}
+
   wrap-ansi@7.0.0:
     resolution: {integrity: sha512-YVGIj2kamLSTxw6NsZjoBxfSwsn0ycdesmc4p+Q21c5zPuZ1pl+NfxVdxPtdHvmNVOQ6XSYG4AUtyt/Fi7D16Q==}
     engines: {node: '>=10'}
@@ -6740,6 +7353,18 @@ snapshots:
 
   '@chevrotain/types@11.1.2': {}
 
+  '@clack/core@1.4.3':
+    dependencies:
+      fast-wrap-ansi: 0.2.2
+      sisteransi: 1.0.5
+
+  '@clack/prompts@1.7.0':
+    dependencies:
+      '@clack/core': 1.4.3
+      fast-string-width: 3.0.2
+      fast-wrap-ansi: 0.2.2
+      sisteransi: 1.0.5
+
   '@cordisjs/plugin-include@1.0.4(@cordisjs/plugin-loader@1.0.0-rc.5)(cordis@4.0.0-rc.7)':
     dependencies:
       '@cordisjs/plugin-loader': 1.0.0-rc.5(cordis@4.0.0-rc.7)(node-addon-require-builtin@0.1.0)
@@ -8255,6 +8880,8 @@ snapshots:
 
   comma-separated-tokens@2.0.3: {}
 
+  commander@15.0.0: {}
+
   commander@7.2.0: {}
 
   commander@8.3.0: {}
@@ -8822,8 +9449,18 @@ snapshots:
 
   fast-levenshtein@2.0.6: {}
 
+  fast-string-truncated-width@3.0.3: {}
+
+  fast-string-width@3.0.2:
+    dependencies:
+      fast-string-truncated-width: 3.0.3
+
   fast-uri@3.1.3: {}
 
+  fast-wrap-ansi@0.2.2:
+    dependencies:
+      fast-string-width: 3.0.2
+
   fast-xml-builder@1.2.0:
     dependencies:
       path-expression-matcher: 1.5.0
@@ -8980,6 +9617,15 @@ snapshots:
 
   hachure-fill@0.5.2: {}
 
+  handlebars@4.7.9:
+    dependencies:
+      minimist: 1.2.8
+      neo-async: 2.6.2
+      source-map: 0.6.1
+      wordwrap: 1.0.0
+    optionalDependencies:
+      uglify-js: 3.19.3
+
   has-flag@4.0.0: {}
 
   has-symbols@1.1.0: {}
@@ -9197,6 +9843,8 @@ snapshots:
 
   json-stable-stringify-without-jsonify@1.0.1: {}
 
+  jsonc-parser@3.3.1: {}
+
   jsx-ast-utils-x@0.1.0: {}
 
   jszip@3.10.1:
@@ -9735,6 +10383,8 @@ snapshots:
     dependencies:
       brace-expansion: 2.1.2
 
+  minimist@1.2.8: {}
+
   minipass@7.1.3: {}
 
   minisearch@7.2.0: {}
@@ -9751,6 +10401,8 @@ snapshots:
 
   negotiator@1.0.0: {}
 
+  neo-async@2.6.2: {}
+
   node-addon-landlock-run-linux-arm64@0.0.0-test.0:
     optional: true
 
@@ -10278,10 +10930,14 @@ snapshots:
 
   signal-exit@4.1.0: {}
 
+  sisteransi@1.0.5: {}
+
   smol-toml@1.6.1: {}
 
   source-map-js@1.2.1: {}
 
+  source-map@0.6.1: {}
+
   space-separated-tokens@2.0.2: {}
 
   speakingurl@14.0.1: {}
@@ -10446,6 +11102,9 @@ snapshots:
 
   typescript@6.0.3: {}
 
+  uglify-js@3.19.3:
+    optional: true
+
   unbash@3.0.0: {}
 
   unconfig-core@7.5.0:
@@ -10710,6 +11369,8 @@ snapshots:
 
   word-wrap@1.2.5: {}
 
+  wordwrap@1.0.0: {}
+
   wrap-ansi@7.0.0:
     dependencies:
       ansi-styles: 4.3.0
diff --git a/pnpm-workspace.yaml b/pnpm-workspace.yaml
index 07182c9959..8f93814899 100644
--- a/pnpm-workspace.yaml
+++ b/pnpm-workspace.yaml
@@ -2,6 +2,13 @@ packages:
   - vendor/*
   - packages/*/*
   - website
+  # The runnable demo leaves join as ONE workspace member: examples/package.json
+  # declares the union of every leaf's cordis.yml plugins as workspace:*, so a
+  # plain-node (`:lib`) boot of any leaf (examples/<leaf>/cordis.yml) resolves its
+  # plugins through real package `exports`→lib by walking up to examples/node_modules.
+  # Members for DEPENDENCY RESOLUTION only — NOT build targets: tsdown's explicit
+  # globs (vendor/*, packages/*/*) exclude them. See the example-execute-over-tsx RFC.
+  - examples
   # Deploy root of the single-exe build: a pure dependency manifest whose
   # closure is what the exe bundles and what the Python runtime distributes.
   - python/sdk-runtime
diff --git a/python/sdk-runtime/README.i18n.yaml b/python/sdk-runtime/README.i18n.yaml
index b4e7981f10..00f0fe0035 100644
--- a/python/sdk-runtime/README.i18n.yaml
+++ b/python/sdk-runtime/README.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-README.md: 5525e8a7b88df3f686bb1fa3a08c3556acc2e655
-README.zh.md: 30723d6d78b84e16261898a88ed22c3c58fc83bf
+README.md: cdf38d4474a0e0148a4804e14c12971c76b38e27
+README.zh.md: 99d57c6f900371a46b94c5ea80b2a1665fd5e8d1
diff --git a/python/sdk-runtime/README.md b/python/sdk-runtime/README.md
index 5525e8a7b8..cdf38d4474 100644
--- a/python/sdk-runtime/README.md
+++ b/python/sdk-runtime/README.md
@@ -26,4 +26,4 @@ Each wheel contains exactly one executable. The fixed tags are `py3-none-manylin
 
 ## Zero-config design
 
-The runtime binary always demands an explicit config (`$DSH_CORDIS_CONFIG`, or a config path as an argv positional argument) and exits loudly without one — that hard semantic is part of the runtime's design and this package does not soften it. The bin (`dsh-jsonrpc-agent`) boots only the plugins the config lists; the serving surface (the stdio JSON-RPC server) is itself one of its entries (`@deepseek-ai/dsh-jsonrpc`), and without it the booted agent has no channel to the outside. This package checks in `runtime/cordis.yml` with the JSON-RPC serving entry, agent core, a preloaded DeepSeek adapter, JSONL persistence, and local bash. The adapter reads `DEEPSEEK_API_KEY` and `DEEPSEEK_BASE_URL`, while persistence and bash use `DSH_SESSION_ROOT` and `DSH_CWD` with manual-run fallbacks. When the caller uses no explicit config channel, the `deepseek_harness` client injects that file's path via `DSH_CORDIS_CONFIG` (injection conditions: [sdk README](../sdk/README.md)). Zero-config is thus an explicit, visible parameter pass in the wrapper, not a hidden fallback in the runtime.
+The runtime binary always demands an explicit config (`$DSH_CORDIS_CONFIG`, or a config path as an argv positional argument) and exits loudly without one — that hard semantic is part of the runtime's design and this package does not soften it. The bin (`dsh-jsonrpc-agent`) boots only the plugins the config lists; the serving surface (the stdio JSON-RPC server) is itself one of its entries (`@deepseek-ai/dsh-jsonrpc`), and without it the booted agent has no channel to the outside. This package checks in `runtime/cordis.yml` with the JSON-RPC serving entry, agent core, a preloaded DeepSeek adapter, JSONL persistence, local bash, and a local filesystem provider for bounded workspace-instruction loading. The adapter reads `DEEPSEEK_API_KEY` and `DEEPSEEK_BASE_URL`, while persistence, bash, and the filesystem provider use `DSH_SESSION_ROOT` and `DSH_CWD` with manual-run fallbacks. When the caller uses no explicit config channel, the `deepseek_harness` client injects that file's path via `DSH_CORDIS_CONFIG` (injection conditions: [sdk README](../sdk/README.md)). Zero-config is thus an explicit, visible parameter pass in the wrapper, not a hidden fallback in the runtime.
diff --git a/python/sdk-runtime/README.zh.md b/python/sdk-runtime/README.zh.md
index 30723d6d78..99d57c6f90 100644
--- a/python/sdk-runtime/README.zh.md
+++ b/python/sdk-runtime/README.zh.md
@@ -26,4 +26,4 @@ exe 缺失时抛出 `FileNotFoundError`，并写明两种获取途径：在 deep
 
 ## 零配置设计
 
-运行时二进制始终要求显式配置（`$DSH_CORDIS_CONFIG`，或作为 argv 位置参数的配置路径），缺了就报错退出——这一硬语义是运行时设计的一部分，本包不软化它。`bin`（`dsh-jsonrpc-agent`）只启动配置里列出的插件；对外服务接口（stdio JSON-RPC 服务器）也是其中一个条目（`@deepseek-ai/dsh-jsonrpc`），缺了它，启动出的 agent 就没有对外通道。本包检入的 `runtime/cordis.yml` 包含 JSON-RPC 服务条目、`agent-core`、预载的 DeepSeek 适配器、JSONL 持久化与本地 bash。DeepSeek 适配器读取 `DEEPSEEK_API_KEY` 与 `DEEPSEEK_BASE_URL`，持久化与 bash 则使用 `DSH_SESSION_ROOT` 和 `DSH_CWD`，并为手动运行提供回退值。调用方未使用任何显式配置通道时，`deepseek_harness` 客户端把该文件路径注入 `DSH_CORDIS_CONFIG`（注入条件见 [sdk README](../sdk/README.md)）。因此，零配置是包装层中一次显式、可见的参数传递，而不是运行时中的隐藏回退。
+运行时二进制始终要求显式配置（`$DSH_CORDIS_CONFIG`，或作为 argv 位置参数的配置路径），缺了就报错退出——这一硬语义是运行时设计的一部分，本包不软化它。`bin`（`dsh-jsonrpc-agent`）只启动配置里列出的插件；对外服务接口（stdio JSON-RPC 服务器）也是其中一个条目（`@deepseek-ai/dsh-jsonrpc`），缺了它，启动出的 agent 就没有对外通道。本包检入的 `runtime/cordis.yml` 包含 JSON-RPC 服务条目、`agent-core`、预载的 DeepSeek 适配器、JSONL 持久化、本地 bash，以及用于有界加载工作区指令的本地文件系统 provider。DeepSeek 适配器读取 `DEEPSEEK_API_KEY` 与 `DEEPSEEK_BASE_URL`，持久化、bash 和文件系统 provider 则使用 `DSH_SESSION_ROOT` 和 `DSH_CWD`，并为手动运行提供回退值。调用方未使用任何显式配置通道时，`deepseek_harness` 客户端把该文件路径注入 `DSH_CORDIS_CONFIG`（注入条件见 [sdk README](../sdk/README.md)）。因此，零配置是包装层中一次显式、可见的参数传递，而不是运行时中的隐藏回退。
diff --git a/python/sdk-runtime/package.json b/python/sdk-runtime/package.json
index 491d975e04..2d3560ec5c 100644
--- a/python/sdk-runtime/package.json
+++ b/python/sdk-runtime/package.json
@@ -23,6 +23,7 @@
     "@deepseek-ai/dsh-fs": "workspace:^",
     "@deepseek-ai/dsh-fs-local": "workspace:^",
     "@deepseek-ai/dsh-fs-policy": "workspace:^",
+    "@deepseek-ai/dsh-home": "workspace:^",
     "@deepseek-ai/dsh-hook-protocol": "workspace:^",
     "@deepseek-ai/dsh-hooks-claude": "workspace:^",
     "@deepseek-ai/dsh-hooks-codex": "workspace:^",
@@ -30,9 +31,11 @@
     "@deepseek-ai/dsh-jsonrpc": "workspace:^",
     "@deepseek-ai/dsh-jsonrpc-demo": "workspace:^",
     "@deepseek-ai/dsh-llm": "workspace:^",
+    "@deepseek-ai/dsh-token-meter": "workspace:^",
     "@deepseek-ai/dsh-llm-deepseek": "workspace:^",
     "@deepseek-ai/dsh-llm-pi-ai": "workspace:^",
     "@deepseek-ai/dsh-permission": "workspace:^",
+    "@deepseek-ai/dsh-paths": "workspace:^",
     "@deepseek-ai/dsh-repeat-tool-guard": "workspace:^",
     "@deepseek-ai/dsh-sandbox": "workspace:^",
     "@deepseek-ai/dsh-scope": "workspace:^",
@@ -49,6 +52,7 @@
     "@deepseek-ai/dsh-subagent-spawn": "workspace:^",
     "@deepseek-ai/dsh-subagent-subprocess": "workspace:^",
     "@deepseek-ai/dsh-system-prompt": "workspace:^",
+    "@deepseek-ai/dsh-tasks": "workspace:^",
     "@deepseek-ai/dsh-timeout": "workspace:^",
     "@deepseek-ai/dsh-timeout-policy": "workspace:^",
     "@deepseek-ai/dsh-tool-ask-user": "workspace:^",
@@ -57,6 +61,7 @@
     "@deepseek-ai/dsh-tool-fs": "workspace:^",
     "@deepseek-ai/dsh-tool-skill": "workspace:^",
     "@deepseek-ai/dsh-tool-subagent": "workspace:^",
+    "@deepseek-ai/dsh-tool-tasks": "workspace:^",
     "@deepseek-ai/dsh-tool-todo": "workspace:^",
     "@deepseek-ai/dsh-tool-web": "workspace:^",
     "@deepseek-ai/dsh-tool-workflow": "workspace:^",
@@ -68,6 +73,7 @@
     "@deepseek-ai/dsh-web-search-deepseek": "workspace:^",
     "@deepseek-ai/dsh-web-search-exa": "workspace:^",
     "@deepseek-ai/dsh-web-search-perplexity": "workspace:^",
+    "@deepseek-ai/dsh-workspace-context": "workspace:^",
     "@deepseek-ai/dsh-workflow": "workspace:^",
     "@deepseek-ai/dsh-workflow-workerthread": "workspace:^",
     "cordis": "workspace:^"
diff --git a/python/sdk-runtime/src/deepseek_harness_runtime/runtime/cordis.yml b/python/sdk-runtime/src/deepseek_harness_runtime/runtime/cordis.yml
index a47dcc26a7..a0eccdf483 100644
--- a/python/sdk-runtime/src/deepseek_harness_runtime/runtime/cordis.yml
+++ b/python/sdk-runtime/src/deepseek_harness_runtime/runtime/cordis.yml
@@ -9,6 +9,9 @@
 # Agent spine; the SDK server creates agents per sessionId.
 - id: agent-core
   name: '@deepseek-ai/dsh-agent-spine-demo'
+  config:
+    workspaceContext:
+      maxBytes: 65536
 
 # Stock DeepSeek adapters. Loading requires an API key; initialize and shutdown
 # may use a dummy key because they do not call the model.
@@ -17,9 +20,6 @@
   config:
     apiKey: !!js process.env.DEEPSEEK_API_KEY
     baseURL: !!js process.env.DEEPSEEK_BASE_URL
-    models:
-      - deepseek-v4-flash
-      - deepseek-v4-pro
 
 # JSONL persistence; $DSH_SESSION_ROOT wins over ./.sessions in the process cwd.
 - id: sessions
@@ -32,3 +32,10 @@
   name: '@deepseek-ai/dsh-bash-local'
   config:
     cwd: !!js process.env.DSH_CWD ?? process.cwd()
+
+# Local filesystem provider for workspace instruction loading. This does not
+# expose model-facing file tools by itself.
+- id: fs-local
+  name: '@deepseek-ai/dsh-fs-local'
+  config:
+    cwd: !!js process.env.DSH_CWD ?? process.cwd()
diff --git a/python/sdk/README.i18n.yaml b/python/sdk/README.i18n.yaml
index c6f8e4de1d..bb012a90db 100644
--- a/python/sdk/README.i18n.yaml
+++ b/python/sdk/README.i18n.yaml
@@ -2,5 +2,5 @@
 # side as of the last confirmed-consistent state. Both languages carry equal authority;
 # after editing either side, bring the other along and re-record with:
 #   pnpm run verify-translation-pairing --write
-README.md: 441b335b9e850c221fbd6a657c7de539ceca070d
-README.zh.md: 65134c0e856b933c793510c2ed528e97f940f47d
+README.md: 80c9d1f50d26fc4f4670800fd7d7f5ea442ad891
+README.zh.md: ffedb5eb30f17388fe589863dbc654b22716b40c
diff --git a/python/sdk/README.md b/python/sdk/README.md
index 441b335b9e..80c9d1f50d 100644
--- a/python/sdk/README.md
+++ b/python/sdk/README.md
@@ -25,12 +25,15 @@ By default, the SDK launches the bundled single-file `dsh-jsonrpc-agent` executa
 from deepseek_harness import DeepSeekHarness
 
 with DeepSeekHarness(
+    provider="deepseek",
     model="deepseek-v4-flash",
     cordis="examples/dsbench-coding-agent/cordis.yml",
 ) as harness:
     result = harness.run("Make the requested code change.")
 ```
 
+`provider` selects a provider route registered by the chosen Cordis composition; `model` is the model id resolved by that adapter. The bundled default composition registers `deepseek`. A custom composition can mount `llm-pi-ai`, configure provider-specific credentials/endpoints there, and select any provider/model present in pi-ai's installed catalog.
+
 `TurnResult.final_response` is the text content from the last
 `assistant/message` event in the turn. Use `TurnResult.events` for the complete
 event stream, including intermediate assistant messages and tool activity.
diff --git a/python/sdk/README.zh.md b/python/sdk/README.zh.md
index 65134c0e85..ffedb5eb30 100644
--- a/python/sdk/README.zh.md
+++ b/python/sdk/README.zh.md
@@ -21,12 +21,15 @@ with DeepSeekHarness() as harness:
 from deepseek_harness import DeepSeekHarness
 
 with DeepSeekHarness(
+    provider="deepseek",
     model="deepseek-v4-flash",
     cordis="examples/dsbench-coding-agent/cordis.yml",
 ) as harness:
     result = harness.run("Make the requested code change.")
 ```
 
+`provider` 用于选择当前 Cordis 组合已注册的提供方路由；`model` 是该适配器解析的模型 ID。内置默认组合注册 `deepseek`。自定义组合可以挂载 `llm-pi-ai`，在其中配置各提供方的凭据与端点，再选择 pi-ai 已安装目录中的任意提供方/模型组合。
+
 `TurnResult.final_response` 是本轮次最后一个 `assistant/message` 事件的文本内容。完整的事件流（包括中间的助手消息与工具活动）用 `TurnResult.events` 获取。
 
 同样的行为也可以通过 `DSH_CORDIS_CONFIG` 为运行时子进程选定。注入逻辑位于 `HarnessClient.start()`，因此底层客户端的默认启动也具有此行为：当启动解析到内置运行时，且 `cordis` 与非空的 `DSH_CORDIS_CONFIG` 均未设置时（运行时把空值视为缺省，注入检查与之一致），使用内置的默认配置；显式给出 `runtime_bin`、`bridge_bin` 或 `launch_args_override` 则完全禁用注入。运行时载体（生产用 exe 与仅限开发的 `node` 闭包）及其获取方式见 [sdk-runtime README](../sdk-runtime/README.md)。
diff --git a/python/sdk/src/deepseek_harness/api.py b/python/sdk/src/deepseek_harness/api.py
index 3743371cea..2b44a50c64 100644
--- a/python/sdk/src/deepseek_harness/api.py
+++ b/python/sdk/src/deepseek_harness/api.py
@@ -18,6 +18,7 @@ class DeepSeekHarnessConfig:
     intentionally override or inject variables for a subprocess.
     """
 
+    provider: str = "deepseek"
     model: str = "deepseek-v4-flash"
     cwd: str | None = None
     runtime_cwd: str | None = None
@@ -97,6 +98,7 @@ class DeepSeekHarness:
         self._client.start()
         self._client.initialize(
             cwd=self._cwd,
+            provider=self.config.provider,
             model=self.config.model,
         )
         self._initialized = True
diff --git a/python/sdk/src/deepseek_harness/client.py b/python/sdk/src/deepseek_harness/client.py
index f5126c21c6..e552c8b685 100644
--- a/python/sdk/src/deepseek_harness/client.py
+++ b/python/sdk/src/deepseek_harness/client.py
@@ -115,10 +115,12 @@ class HarnessClient:
         self,
         *,
         cwd: str,
+        provider: str,
         model: str,
     ) -> InitializeResponse:
         payload: JsonObject = {
             "cwd": str(Path(cwd).resolve()),
+            "provider": provider,
             "model": model,
         }
         try:
diff --git a/python/sdk/tests/test_bundled_runtime.py b/python/sdk/tests/test_bundled_runtime.py
index 0ee373102e..bc1da849b2 100644
--- a/python/sdk/tests/test_bundled_runtime.py
+++ b/python/sdk/tests/test_bundled_runtime.py
@@ -22,6 +22,8 @@ _CORDIS_YML = """\
   name: '@deepseek-ai/dsh-jsonrpc'
 - id: agent-core
   name: '@deepseek-ai/dsh-agent-spine-demo'
+  config:
+    workspaceContext: false
 - id: sessions
   name: '@deepseek-ai/dsh-session-persistence-jsonl'
   config:
@@ -66,7 +68,7 @@ def test_bundled_runtime_boots_a_cordis_config(tmp_path: Path, mode: str) -> Non
     (tmp_path / "cordis.yml").write_text(_CORDIS_YML)
 
     with _client(tmp_path, launch_args) as client:
-        init = client.initialize(cwd=str(tmp_path), model="deepseek-v4-pro")
+        init = client.initialize(provider="deepseek", cwd=str(tmp_path), model="deepseek-v4-pro")
 
     assert init.serverInfo is not None
     assert init.serverInfo.name == "deepseek-harness-sdk-runtime"
@@ -83,7 +85,7 @@ def test_bundled_runtime_surfaces_unbundled_plugin_failure(tmp_path: Path, mode:
     client.start()
     try:
         with pytest.raises((TransportClosedError, TimeoutError)) as excinfo:
-            client.initialize(cwd=str(tmp_path), model="deepseek-v4-pro")
+            client.initialize(provider="deepseek", cwd=str(tmp_path), model="deepseek-v4-pro")
     finally:
         client.close()
 
diff --git a/python/sdk/tests/test_client.py b/python/sdk/tests/test_client.py
index f7d0cfa8a3..f66abf4b36 100644
--- a/python/sdk/tests/test_client.py
+++ b/python/sdk/tests/test_client.py
@@ -330,7 +330,7 @@ for line in sys.stdin:
     with HarnessClient(
         HarnessConfig(launch_args_override=(sys.executable, str(script)))
     ) as client:
-        init = client.initialize(cwd="/workspace", model="dsagent")
+        init = client.initialize(provider="deepseek", cwd="/workspace", model="dsagent")
         assert init.serverInfo.name == "fake-dsh"
 
         client.session_prompt("main", [{"type": "text", "text": "fix it"}])
@@ -382,7 +382,7 @@ for line in sys.stdin:
         raise RuntimeError("bad notification filter")
 
     with HarnessClient(HarnessConfig(launch_args_override=(sys.executable, str(script)))) as client:
-        client.initialize(cwd="/workspace", model="dsagent")
+        client.initialize(provider="deepseek", cwd="/workspace", model="dsagent")
         with (
             client.subscribe_notifications(broken_filter) as broken,
             client.subscribe_notifications(lambda notification: notification.method == "tick") as healthy,
@@ -419,7 +419,7 @@ for line in sys.stdin:
     )
 
     with HarnessClient(HarnessConfig(launch_args_override=(sys.executable, str(script)))) as client:
-        client.initialize(cwd="/workspace", model="dsagent")
+        client.initialize(provider="deepseek", cwd="/workspace", model="dsagent")
         with pytest.raises(ValueError):
             client.session_prompt("main", [{"type": "text", "text": "fix it"}])
 
@@ -448,7 +448,7 @@ for line in sys.stdin:
     with HarnessClient(
         HarnessConfig(launch_args_override=(sys.executable, str(script)))
     ) as client:
-        client.initialize(cwd="/workspace", model="dsagent")
+        client.initialize(provider="deepseek", cwd="/workspace", model="dsagent")
 
         request = client.next_request()
         assert request.id == "bridge-req-1"
@@ -482,7 +482,7 @@ for line in sys.stdin:
     with HarnessClient(
         HarnessConfig(launch_args_override=(sys.executable, str(script)))
     ) as client:
-        init = client.initialize(cwd="/workspace", model="dsagent")
+        init = client.initialize(provider="deepseek", cwd="/workspace", model="dsagent")
         assert init.serverInfo.name == "fake-dsh"
 
 
@@ -504,7 +504,7 @@ time.sleep(60)
     ) as client:
         start = time.monotonic()
         try:
-            client.initialize(cwd="/workspace", model="dsagent")
+            client.initialize(provider="deepseek", cwd="/workspace", model="dsagent")
         except TimeoutError:
             assert time.monotonic() - start < 2
         else:
@@ -540,7 +540,7 @@ for line in sys.stdin:
     client.start()
     proc = client._proc
     assert proc is not None
-    client.initialize(cwd="/workspace", model="dsagent")
+    client.initialize(provider="deepseek", cwd="/workspace", model="dsagent")
     start = time.monotonic()
     client.close()
     assert time.monotonic() - start < 2
@@ -571,7 +571,7 @@ for line in sys.stdin:
     assert proc is not None
 
     with pytest.raises(Exception, match="bad initialize"):
-        client.initialize(cwd=".", model="dsagent")
+        client.initialize(provider="deepseek", cwd=".", model="dsagent")
 
     assert proc.wait(timeout=1) is not None
     assert client._proc is None
@@ -611,7 +611,7 @@ for line in sys.stdin:
 
     client = HarnessClient(HarnessConfig(launch_args_override=(sys.executable, str(script))))
     client.start()
-    client.initialize(cwd="/workspace", model="dsagent")
+    client.initialize(provider="deepseek", cwd="/workspace", model="dsagent")
     client.close()
     client.close()
 
@@ -634,7 +634,7 @@ sys.exit(42)
         )
     ) as client:
         with pytest.raises(Exception, match="fatal bridge exploded"):
-            client.initialize(cwd="/workspace", model="dsagent")
+            client.initialize(provider="deepseek", cwd="/workspace", model="dsagent")
 
 
 def test_client_serializes_concurrent_writes(tmp_path: Path) -> None:
@@ -665,7 +665,7 @@ with open(os.environ["SEEN"], "w") as seen:
             env={"SEEN": str(output)},
         )
     ) as client:
-        client.initialize(cwd="/workspace", model="dsagent")
+        client.initialize(provider="deepseek", cwd="/workspace", model="dsagent")
         threads = [
             threading.Thread(target=client.notify, args=(f"notice-{index}", {"index": index}))
             for index in range(50)
@@ -736,7 +736,7 @@ def test_client_default_launch_uses_bundled_runtime_and_injects_default_config(
         monkeypatch.setenv("DSH_CORDIS_CONFIG", ambient_config)
 
     with HarnessClient(HarnessConfig(env={"ENV_DUMP": str(env_dump)})) as client:
-        init = client.initialize(cwd="/workspace", model="deepseek-v4-pro")
+        init = client.initialize(provider="deepseek", cwd="/workspace", model="deepseek-v4-pro")
 
     assert init.serverInfo.name == "bundled-runtime"
     assert json.loads(env_dump.read_text())["DSH_CORDIS_CONFIG"] == str(default_config)
@@ -752,7 +752,7 @@ def test_client_respects_explicit_config_over_bundled_default(
     with HarnessClient(
         HarnessConfig(env={"ENV_DUMP": str(env_dump), "DSH_CORDIS_CONFIG": "./explicit.yml"})
     ) as client:
-        client.initialize(cwd="/workspace", model="deepseek-v4-pro")
+        client.initialize(provider="deepseek", cwd="/workspace", model="deepseek-v4-pro")
 
     assert json.loads(env_dump.read_text())["DSH_CORDIS_CONFIG"] == "./explicit.yml"
 
diff --git a/scripts/check-workspace-constraints.ts b/scripts/check-workspace-constraints.ts
index 4dac840da2..c10dd1396c 100644
--- a/scripts/check-workspace-constraints.ts
+++ b/scripts/check-workspace-constraints.ts
@@ -116,11 +116,31 @@ const dshWorkerPackageFiles = [
   'src',
 ] as const
 
+const packageFileExtras: Readonly<Record<string, readonly string[]>> = {
+  '@deepseek-ai/dsh-helper': ['lib/assets'],
+  '@deepseek-ai/dsh-scripts': [
+    'lib/dev/tsdown-config.js',
+    'lib/local-plugin-loader-hooks.js',
+    'lib/assets',
+  ],
+}
+
 function sameStringList(actual: readonly string[] | undefined, expected: readonly string[]): boolean {
   return !!actual && actual.length === expected.length && actual.every((value, index) => value === expected[index])
 }
 
 function expectedDshPackageFiles(manifest: PackageManifest): readonly string[] {
+  const extras = manifest.name ? packageFileExtras[manifest.name] ?? [] : []
+  if (extras.length > 0) {
+    return [
+      'lib/index.js',
+      ...manifest.bin ? ['lib/bin.js'] : [],
+      ...extras,
+      'lib/types/**/*.d.ts',
+      'lib/types/**/*.d.ts.map',
+      'src',
+    ]
+  }
   if (manifest.bin) return dshBinPackageFiles
   // A declared "./worker" subpath export sanctions the one extra runtime
   // bundle a worker-thread entry needs (and NodeNext/publint then validate
diff --git a/scripts/cordis-walk.ts b/scripts/cordis-walk.ts
new file mode 100644
index 0000000000..f4f045b06d
--- /dev/null
+++ b/scripts/cordis-walk.ts
@@ -0,0 +1,91 @@
+/**
+ * AST walkers for the Cordis catalog generator: locate the Cordis module merge
+ * in a source file, enumerate its `interface Events` members, and resolve the
+ * `interface Context` service keys to their service classes.
+ */
+
+import ts from 'typescript'
+import { parseJsDoc, pointer, rawJsDoc } from './jsdoc.ts'
+
+/** The body of the cordis module merge in `sf`: `declare module 'cordis'`
+ * (harness packages) or `declare module './context.ts'` (vendor core), or
+ * null when the file has neither. */
+export function cordisModuleBody(sf: ts.SourceFile): ts.ModuleBlock | null {
+  for (const stmt of sf.statements) {
+    if (!ts.isModuleDeclaration(stmt) || !ts.isStringLiteral(stmt.name)) continue
+    if (stmt.name.text !== 'cordis' && stmt.name.text !== './context.ts') continue
+    if (stmt.body && ts.isModuleBlock(stmt.body)) return stmt.body
+  }
+  return null
+}
+
+/** Every `interface Events` method member of a cordis module merge, with the
+ * event name resolved from its (possibly string-literal) property name. */
+export function eventMembers(body: ts.ModuleBlock, sf: ts.SourceFile): { name: string; member: ts.MethodSignature }[] {
+  const out: { name: string; member: ts.MethodSignature }[] = []
+  for (const stmt of body.statements) {
+    if (!ts.isInterfaceDeclaration(stmt) || stmt.name.text !== 'Events') continue
+    for (const member of stmt.members) {
+      if (!ts.isMethodSignature(member)) continue
+      const name = ts.isStringLiteral(member.name) ? member.name.text : member.name.getText(sf)
+      out.push({ name, member })
+    }
+  }
+  return out
+}
+
+/** The `ctx.<key> → type name` map declared by a merge's `interface Context`. */
+function contextKeyMap(body: ts.ModuleBlock, sf: ts.SourceFile): Map<string, string> {
+  const keyToType = new Map<string, string>()
+  for (const stmt of body.statements) {
+    if (!ts.isInterfaceDeclaration(stmt) || stmt.name.text !== 'Context') continue
+    for (const member of stmt.members) {
+      if (!ts.isPropertySignature(member) || !member.type) continue
+      keyToType.set(member.name.getText(sf), member.type.getText(sf))
+    }
+  }
+  return keyToType
+}
+
+/** One `ctx.<key>` service class resolved from a Context merge. */
+export interface ServiceClass {
+  key: string
+  type: string
+  cls: ts.ClassDeclaration
+  abstract: boolean
+  /** Class-level JSDoc prose (empty string when missing — also reported). */
+  doc: string
+}
+
+/**
+ * Resolve each `ctx.<key>` of a merge to the service class declared in the
+ * same file. A key whose type is not a class here (a Pick-mixin member, e.g.
+ * timer helpers) is skipped. A class without JSDoc prose is reported into
+ * `violations` (named `where` by the caller's gate).
+ *
+ * @param body — the cordis module merge body.
+ * @param sf — the source file containing the merge.
+ * @param rel — repo-relative path of `sf`, for violation pointers.
+ * @param violations — sink for JSDoc-completeness violations.
+ * @returns the resolved service classes, in Context-declaration order.
+ */
+export function serviceClasses(
+  body: ts.ModuleBlock,
+  sf: ts.SourceFile,
+  rel: string,
+  violations: string[],
+): ServiceClass[] {
+  const text = sf.getFullText()
+  const out: ServiceClass[] = []
+  for (const [key, type] of contextKeyMap(body, sf)) {
+    const cls = sf.statements.find(
+      (s): s is ts.ClassDeclaration => ts.isClassDeclaration(s) && s.name?.text === type,
+    )
+    if (!cls) continue // a Pick-mixin member, not a class here
+    const abstract = cls.modifiers?.some(m => m.kind === ts.SyntaxKind.AbstractKeyword) ?? false
+    const doc = parseJsDoc(rawJsDoc(text, cls)).doc
+    if (!doc) violations.push(`service ctx.${key} (${pointer(rel, sf, cls)}): class ${type} has no JSDoc.`)
+    out.push({ key, type, cls, abstract, doc })
+  }
+  return out
+}
diff --git a/scripts/doc-budgets.manifest.json b/scripts/doc-budgets.manifest.json
index 72d2d74533..3b99ad454a 100644
--- a/scripts/doc-budgets.manifest.json
+++ b/scripts/doc-budgets.manifest.json
@@ -1,11 +1,11 @@
 {
-  "AGENTS.md": 1370,
+  "AGENTS.md": 1500,
   "docs/AGENTS.md": 1100,
   "docs/architecture.md": 1790,
   "docs/cordis-primer.md": 600,
   "docs/defensive-patterns.md": 550,
-  "docs/testing.md": 800,
-  "examples/AGENTS.md": 200,
+  "docs/testing.md": 960,
+  "examples/AGENTS.md": 310,
   "packages/AGENTS.md": 290,
   "packages/README.md": 760
 }
diff --git a/scripts/doc-typecheck.ts b/scripts/doc-typecheck.ts
index 86266f89b2..b725cb29c9 100644
--- a/scripts/doc-typecheck.ts
+++ b/scripts/doc-typecheck.ts
@@ -1,13 +1,14 @@
 /**
- * Typecheck Markdown `ts` fences against workspace sources. `ignore-check`
- * fences are reported as opt-outs; generated catalog fragments and
- * `type-equiv` blocks are skipped here because their owning gates verify them.
+ * Typecheck Markdown `ts` fences against the workspace API. `ignore-check` fences are reported as
+ * opt-outs; generated catalog fragments and `type-equiv` blocks are skipped here because their
+ * owning gates verify them. A build-coordinated mode consumes existing declarations without emit.
  */
 
 import { execFileSync } from 'node:child_process'
 import { globSync, mkdtempSync, readFileSync, rmSync, writeFileSync } from 'node:fs'
 import { join, relative, resolve } from 'node:path'
 import ts from 'typescript'
+import { extractFences } from './md-fences.ts'
 
 const root = resolve(import.meta.dirname, '..')
 
@@ -27,60 +28,123 @@ interface Block {
   code: string
 }
 
+/** The info-string → kind table this gate tracks. */
+const KIND_BY_INFO: Record<string, BlockKind> = {
+  'ts': 'check',
+  'ts ignore-check': 'ignore',
+  'ts type-equiv': 'type-equiv',
+  'ts cordis-catalog': 'cordis-catalog',
+  'ts persistence-catalog': 'persistence-catalog',
+  'ts config-catalog': 'config-catalog',
+}
+
 /** Extract every recognized TypeScript fence from one Markdown file. */
 function extractBlocks(absPath: string): Block[] {
-  const text = readFileSync(absPath, 'utf8')
-  const lines = text.split('\n')
   const file = relative(root, absPath)
-  const blocks: Block[] = []
-  let open: { line: number; kind: BlockKind; body: string[] } | null = null
+  return extractFences(absPath, info => KIND_BY_INFO[info] ?? null)
+    .map(f => ({ file, line: f.line, kind: f.kind, code: f.code }))
+}
 
-  lines.forEach((raw, i) => {
-    const fence = /^```(\s*)(\S.*)?$/.exec(raw)
-    if (!fence) {
-      if (open) open.body.push(raw)
-      return
-    }
-    if (open) {
-      // closing fence
-      blocks.push({ file, line: open.line, kind: open.kind, code: open.body.join('\n') })
-      open = null
-      return
-    }
-    // Ignore non-TypeScript fences.
-    const info = (fence[2] ?? '').trim()
-    const kind: BlockKind | null =
-      info === 'ts' ? 'check'
-        : info === 'ts ignore-check' ? 'ignore'
-          : info === 'ts type-equiv' ? 'type-equiv'
-            : info === 'ts cordis-catalog' ? 'cordis-catalog'
-              : info === 'ts persistence-catalog' ? 'persistence-catalog'
-                : info === 'ts config-catalog' ? 'config-catalog'
-                  : null
-    if (kind) open = { line: i + 1, kind, body: [] }
+const configHost: ts.ParseConfigFileHost = {
+  ...ts.sys,
+  getCurrentDirectory: () => root,
+  onUnRecoverableConfigFileDiagnostic(diagnostic) {
+    throw new Error(ts.flattenDiagnosticMessageText(diagnostic.messageText, '\n'))
+  },
+}
+
+/** Load root settings and redirect workspace aliases to declarations from the coordinated build. */
+function builtTypeCompilerOptions(): ts.CompilerOptions {
+  const configPath = join(root, 'tsconfig.json')
+  const parsed = ts.getParsedCommandLineOfConfigFile(configPath, {}, configHost)
+  if (!parsed) throw new Error(`doc-typecheck: cannot parse ${configPath}`)
+  if (parsed.errors.length > 0) {
+    throw new Error(parsed.errors.map(error => ts.flattenDiagnosticMessageText(error.messageText, '\n')).join('\n'))
+  }
+  if (parsed.options.paths === undefined) throw new Error('doc-typecheck: root tsconfig has no workspace paths')
+  const paths = Object.fromEntries(Object.entries(parsed.options.paths).map(([specifier, candidates]) => [
+    specifier,
+    candidates.map((candidate) => {
+      if (!candidate.endsWith('/src')) {
+        throw new Error(`doc-typecheck: cannot map workspace source path to built declarations: ${candidate}`)
+      }
+      return `${candidate.slice(0, -'/src'.length)}/lib/types`
+    }),
+  ]))
+  const options: ts.CompilerOptions = {
+    ...parsed.options,
+    paths,
+    noEmit: true,
+    composite: false,
+    incremental: false,
+    declaration: false,
+    declarationMap: false,
+    sourceMap: false,
+    noUnusedLocals: false,
+    noUnusedParameters: false,
+  }
+  delete options.tsBuildInfoFile
+  return options
+}
+
+/** Compile Markdown blocks as virtual files against declarations from the coordinated build. */
+function compileBlocksAgainstBuiltTypes(blocks: Block[]): readonly ts.Diagnostic[] {
+  const options = builtTypeCompilerOptions()
+  const sources = new Map<string, string>()
+  for (const [index, block] of blocks.entries()) {
+    const fileName = resolve(root, '.doc-typecheck', `block-${index}.ts`)
+    sources.set(fileName, block.code.endsWith('\n') ? block.code : `${block.code}\n`)
+  }
+
+  const baseHost = ts.createCompilerHost(options, true)
+  const host: ts.CompilerHost = {
+    ...baseHost,
+    fileExists(fileName) {
+      return sources.has(resolve(fileName)) || baseHost.fileExists(fileName)
+    },
+    readFile(fileName) {
+      return sources.get(resolve(fileName)) ?? baseHost.readFile(fileName)
+    },
+    getSourceFile(fileName, languageVersion, onError, shouldCreateNewSourceFile) {
+      const source = sources.get(resolve(fileName))
+      if (source !== undefined) return ts.createSourceFile(fileName, source, languageVersion, true)
+      return baseHost.getSourceFile(fileName, languageVersion, onError, shouldCreateNewSourceFile)
+    },
+    writeFile() {
+      throw new Error('doc-typecheck: noEmit compilation attempted to write output')
+    },
+  }
+  const program = ts.createProgram([...sources.keys()], options, host)
+  return ts.getPreEmitDiagnostics(program)
+}
+
+/** Render compiler diagnostics with virtual block paths mapped back to Markdown. */
+function formatDiagnostics(diagnostics: readonly ts.Diagnostic[], blocks: Block[]): string {
+  const formatted = ts.formatDiagnostics(diagnostics, {
+    getCanonicalFileName: fileName => fileName,
+    getCurrentDirectory: () => root,
+    getNewLine: () => ts.sys.newLine,
   })
-  return blocks
+  return remapBlockPaths(formatted, blocks)
 }
 
 /** Reuse the repo typecheck graph references from a temp project one directory below root. */
 function workspaceReferences(): { path: string }[] {
   const file = join(root, 'tsconfig.json')
-  // Parse with TypeScript's own JSONC reader, not a hand-rolled comment strip:
-  // a regex strip mistakes the `/*/` in a wildcard path candidate
-  // (`./packages/core/*/src`) for a block comment and corrupts the map.
-  const result = ts.readConfigFile(file, p => readFileSync(p, 'utf8'))
+  // Parse with TypeScript's own JSONC reader: a regex comment stripper corrupts the `/*/` path
+  // candidate in the workspace wildcard.
+  const result = ts.readConfigFile(file, path => readFileSync(path, 'utf8'))
   if (result.error) {
     throw new Error(`doc-typecheck: cannot read ${file}: ${ts.flattenDiagnosticMessageText(result.error.messageText, '\n')}`)
   }
-  // `config` is typed `any` by the TS API; narrow it to the one field we read.
-  const { references } = result.config as { compilerOptions: { paths: Record<string, string[]> }; references: { path: string }[] }
-  return references.map(({ path }) => {
-    const relativeToTemp = path.startsWith('./') ? `../${path.slice(2)}` : `../${path}`
-    return { path: relativeToTemp }
-  })
+  // `config` is typed `any` by the TS API; narrow it to the one field read here.
+  const { references } = result.config as { references: { path: string }[] }
+  return references.map(({ path }) => ({
+    path: path.startsWith('./') ? `../${path.slice(2)}` : `../${path}`,
+  }))
 }
 
-/** The standalone tsconfig for the temp typecheck project. */
+/** The standalone temp project used when no coordinated build owns declaration freshness. */
 function tempTsconfig(): string {
   return JSON.stringify({
     extends: '../tsconfig.json',
@@ -94,6 +158,39 @@ function tempTsconfig(): string {
   })
 }
 
+/** Compile blocks through project references for the standalone command. */
+function compileBlocksStandalone(blocks: Block[]): string | undefined {
+  const tmp = mkdtempSync(join(root, '.doc-typecheck-'))
+  try {
+    writeFileSync(join(tmp, 'tsconfig.json'), tempTsconfig())
+    for (const [index, block] of blocks.entries()) {
+      writeFileSync(join(tmp, `block-${index}.ts`), block.code.endsWith('\n') ? block.code : `${block.code}\n`)
+    }
+    try {
+      // Invoke tsc's JS entry through Node instead of a platform-specific shell shim.
+      execFileSync(process.execPath, ['node_modules/typescript/bin/tsc', '-b', join(tmp, 'tsconfig.json')], {
+        cwd: root,
+        stdio: 'pipe',
+      })
+      return undefined
+    } catch (error: unknown) {
+      const failed = error as { stdout?: Buffer; stderr?: Buffer }
+      return remapBlockPaths(`${failed.stdout?.toString() ?? ''}${failed.stderr?.toString() ?? ''}`, blocks)
+    }
+  } finally {
+    rmSync(tmp, { recursive: true, force: true })
+  }
+}
+
+/** Map virtual or temporary block paths back to their owning Markdown fences. */
+function remapBlockPaths(output: string, blocks: Block[]): string {
+  return output.replace(/(?:[^\s:()]*[/\\])?block-(\d+)\.ts\((\d+),(\d+)\)/g, (_match, index: string, line: string, column: string) => {
+    const block = blocks[Number(index)]
+    if (!block) return `block-${index}.ts(${line},${column})`
+    return `${block.file} (block at line ${block.line}, +${line}:${column})`
+  })
+}
+
 const markdownGlobs = ['README.md', 'docs/**/*.md', 'packages/*/*.md', 'packages/*/*/*.md']
 
 const files: string[] = []
@@ -114,45 +211,24 @@ if (checked.length === 0) {
   process.exit(0)
 }
 
-const tmp = mkdtempSync(join(root, '.doc-typecheck-'))
-try {
-  writeFileSync(join(tmp, 'tsconfig.json'), tempTsconfig())
-  const fileForBlock = new Map<string, Block>()
-  checked.forEach((block, i) => {
-    const name = `block-${i}.ts`
-    writeFileSync(join(tmp, name), block.code.endsWith('\n') ? block.code : `${block.code}\n`)
-    fileForBlock.set(name, block)
-  })
-
-  try {
-    // tsc's JS entry via the current node, not the .bin shim: the extensionless
-    // shim is not spawnable on Windows (the CVE-2024-27980 class the sibling
-    // scripts hit), and the .cmd variant would need shell:true, which
-    // concatenates args UNESCAPED — a hazard for the temp project path. The JS
-    // entry behaves identically on every platform.
-    execFileSync(process.execPath, ['node_modules/typescript/bin/tsc', '-b', join(tmp, 'tsconfig.json')], { cwd: root, stdio: 'pipe' })
-  } catch (error: unknown) {
-    const failed = error as { stdout?: Buffer; stderr?: Buffer }
-    const out = `${failed.stdout?.toString() ?? ''}${failed.stderr?.toString() ?? ''}`
-    // Rewrite "block-N.ts(line,col)" to the real "file:fenceLine" for triage.
-    const remapped = out.replace(/(?:[^\s:()]*[/\\])?block-(\d+)\.ts\((\d+),(\d+)\)/g, (_m, idx: string, ln: string, col: string) => {
-      const block = fileForBlock.get(`block-${idx}.ts`)
-      if (!block) return `block-${idx}.ts(${ln},${col})`
-      return `${block.file} (block at line ${block.line}, +${ln}:${col})`
-    })
-    console.error('doc-typecheck: documentation code blocks failed to compile.\n')
-    console.error(remapped)
-    process.exit(1)
-  }
-
-  const ratio = ignored.length / ratioDenominator
-  const skipped = all.length - ratioDenominator
-  console.log(`doc-typecheck: ${checked.length} block(s) compiled, ${ignored.length} ignored (${(ratio * 100).toFixed(0)}% opt-out), ${skipped} type-equiv/catalog (checked elsewhere).`)
-  // Guard against the escape hatch becoming the norm.
-  if (ratioDenominator >= 4 && ratio > 0.5) {
-    console.error(`doc-typecheck: too many blocks opt out of checking (${ignored.length}/${ratioDenominator}). Make them compile or delete them.`)
-    process.exit(1)
-  }
-} finally {
-  rmSync(tmp, { recursive: true, force: true })
+const useBuiltTypes = process.env.DSH_DOC_TYPECHECK_USE_BUILD_OUTPUT === '1'
+const compilationError = useBuiltTypes
+  ? (() => {
+    const diagnostics = compileBlocksAgainstBuiltTypes(checked)
+    return diagnostics.length === 0 ? undefined : formatDiagnostics(diagnostics, checked)
+  })()
+  : compileBlocksStandalone(checked)
+if (compilationError !== undefined) {
+  console.error('doc-typecheck: documentation code blocks failed to compile.\n')
+  console.error(compilationError)
+  process.exit(1)
+}
+
+const ratio = ignored.length / ratioDenominator
+const skipped = all.length - ratioDenominator
+console.log(`doc-typecheck: ${checked.length} block(s) compiled, ${ignored.length} ignored (${(ratio * 100).toFixed(0)}% opt-out), ${skipped} type-equiv/catalog (checked elsewhere).`)
+// Guard against the escape hatch becoming the norm.
+if (ratioDenominator >= 4 && ratio > 0.5) {
+  console.error(`doc-typecheck: too many blocks opt out of checking (${ignored.length}/${ratioDenominator}). Make them compile or delete them.`)
+  process.exit(1)
 }
diff --git a/scripts/gen-cordis-catalog.ts b/scripts/gen-cordis-catalog.ts
index 05a308f1ac..bd9f3103bc 100644
--- a/scripts/gen-cordis-catalog.ts
+++ b/scripts/gen-cordis-catalog.ts
@@ -9,6 +9,7 @@ import { globSync, readFileSync, writeFileSync } from 'node:fs'
 import { resolve, sep } from 'node:path'
 import ts from 'typescript'
 import { checkParams, checkReturns, parseJsDoc, parseTags, pointer, rawJsDoc, reportViolations, type Mode } from './jsdoc.ts'
+import { cordisModuleBody, eventMembers, serviceClasses } from './cordis-walk.ts'
 
 const root = resolve(import.meta.dirname, '..')
 const OUT_EVENTS = 'docs/cordis-catalog/events.md'
@@ -37,6 +38,7 @@ export const LINK_MAP: Record<string, string> = {
   TurnEndReason: 'session.md',
   ToolDefinition: 'tools.md',
   ToolExecution: 'tools.md',
+  ToolExecutionMode: 'tools.md',
   ToolExecutionInput: 'tools.md',
   ToolExecutionResult: 'tools.md',
   ToolExecutionToken: 'tools.md',
@@ -46,8 +48,6 @@ export const LINK_MAP: Record<string, string> = {
   BashExecRequest: 'bash.md',
   BashExecSpec: 'bash.md',
   BashRunResult: 'bash.md',
-  BashTask: 'bash.md',
-  BashTaskRead: 'bash.md',
   ConfinedArgv: 'sandbox.md',
   SandboxMode: 'sandbox.md',
   SandboxPolicy: 'sandbox.md',
@@ -104,15 +104,7 @@ interface InheritedEntry {
   source: string
 }
 
-/** Find the `declare module 'cordis'` body in a source file, or null. */
-function cordisModuleBody(sf: ts.SourceFile): ts.ModuleBlock | null {
-  for (const stmt of sf.statements) {
-    if (ts.isModuleDeclaration(stmt) && ts.isStringLiteral(stmt.name) && stmt.name.text === 'cordis') {
-      if (stmt.body && ts.isModuleBlock(stmt.body)) return stmt.body
-    }
-  }
-  return null
-}
+// cordisModuleBody / eventMembers / serviceClasses live in cordis-walk.ts.
 
 /** The signature text of a method-signature member (everything but a body). */
 function memberSignature(member: ts.TypeElement | ts.ClassElement, sf: ts.SourceFile): string {
@@ -136,38 +128,33 @@ export function collectEvents(scanRoot: string = root): EventEntry[] {
     const sf = ts.createSourceFile(abs, text, ts.ScriptTarget.Latest, true)
     const body = cordisModuleBody(sf)
     if (!body) continue
-    for (const stmt of body.statements) {
-      if (!ts.isInterfaceDeclaration(stmt) || stmt.name.text !== 'Events') continue
-      for (const member of stmt.members) {
-        if (!ts.isMethodSignature(member)) continue
-        const name = ts.isStringLiteral(member.name) ? member.name.text : member.name.getText(sf)
-        const signature = memberSignature(member, sf)
-        const raw = rawJsDoc(text, member)
-        const { doc, mode } = parseJsDoc(raw)
-        const src = pointer(rel, sf, member)
-        const where = `event '${name}' (${src})`
-        if (!mode) {
-          violations.push(`${where} is missing an @mode tag. Add '@mode emit|waterfall|parallel|serial' to its JSDoc (see AGENTS.md).`)
-        }
-        // Conclusive structural check: a trailing `next: () => …` parameter is a
-        // waterfall. (emit vs parallel vs serial is not structurally
-        // distinguishable, so it is trusted from the tag.)
-        const last = member.parameters.at(-1)
-        const hasNext = !!last && last.name.getText(sf) === 'next'
-        if (mode && hasNext && mode !== 'waterfall') {
-          violations.push(`${where} has a trailing 'next' parameter (structurally a waterfall) but is tagged '@mode ${mode}'. Fix the tag or the signature.`)
-        }
-        if (mode && !hasNext && mode === 'waterfall') {
-          violations.push(`${where} is tagged '@mode waterfall' but has no trailing 'next' parameter. A waterfall delegates via next().`)
-        }
-        if (!doc) violations.push(`${where} has no description prose. Say what happened / what a listener may do, above the block tags.`)
-        // Payload parameters need a non-empty @param. The `this` receiver is not
-        // payload, and a waterfall's trailing `next` is covered by its mode.
-        const { params } = parseTags(raw)
-        checkParams(where, 'event', member.parameters, params, sf,
-          p => (ts.isIdentifier(p.name) && p.name.text === 'this') || (hasNext && p === last), violations)
-        if (mode) entries.push({ name, scope: name.split('/')[0] ?? name, signature, mode, doc, source: src })
+    for (const { name, member } of eventMembers(body, sf)) {
+      const signature = memberSignature(member, sf)
+      const raw = rawJsDoc(text, member)
+      const { doc, mode } = parseJsDoc(raw)
+      const src = pointer(rel, sf, member)
+      const where = `event '${name}' (${src})`
+      if (!mode) {
+        violations.push(`${where} is missing an @mode tag. Add '@mode emit|waterfall|parallel|serial' to its JSDoc (see AGENTS.md).`)
       }
+      // Conclusive structural check: a trailing `next: () => …` parameter is a
+      // waterfall. (emit vs parallel vs serial is not structurally
+      // distinguishable, so it is trusted from the tag.)
+      const last = member.parameters.at(-1)
+      const hasNext = !!last && last.name.getText(sf) === 'next'
+      if (mode && hasNext && mode !== 'waterfall') {
+        violations.push(`${where} has a trailing 'next' parameter (structurally a waterfall) but is tagged '@mode ${mode}'. Fix the tag or the signature.`)
+      }
+      if (mode && !hasNext && mode === 'waterfall') {
+        violations.push(`${where} is tagged '@mode waterfall' but has no trailing 'next' parameter. A waterfall delegates via next().`)
+      }
+      if (!doc) violations.push(`${where} has no description prose. Say what happened / what a listener may do, above the block tags.`)
+      // Payload parameters need a non-empty @param. The `this` receiver is not
+      // payload, and a waterfall's trailing `next` is covered by its mode.
+      const { params } = parseTags(raw)
+      checkParams(where, 'event', member.parameters, params, sf,
+        p => (ts.isIdentifier(p.name) && p.name.text === 'this') || (hasNext && p === last), violations)
+      if (mode) entries.push({ name, scope: name.split('/')[0] ?? name, signature, mode, doc, source: src })
     }
   }
   reportViolations('gen-cordis-catalog', violations)
@@ -190,26 +177,8 @@ export function collectServices(scanRoot: string = root): ServiceEntry[] {
     const sf = ts.createSourceFile(abs, text, ts.ScriptTarget.Latest, true)
     const body = cordisModuleBody(sf)
     if (!body) continue
-    // The ctx key → type mapping(s) declared in this file's interface Context.
-    const keyToType = new Map<string, string>()
-    for (const stmt of body.statements) {
-      if (!ts.isInterfaceDeclaration(stmt) || stmt.name.text !== 'Context') continue
-      for (const member of stmt.members) {
-        if (!ts.isPropertySignature(member) || !member.type) continue
-        const key = member.name.getText(sf)
-        keyToType.set(key, member.type.getText(sf))
-      }
-    }
-    if (keyToType.size === 0) continue
-    // Find each service class declared in the same file and emit an entry.
-    for (const [key, type] of keyToType) {
-      const cls = sf.statements.find(
-        (s): s is ts.ClassDeclaration => ts.isClassDeclaration(s) && s.name?.text === type,
-      )
-      if (!cls) continue // a Pick-mixin member (e.g. timer helpers), not a class here
-      const abstract = cls.modifiers?.some(m => m.kind === ts.SyntaxKind.AbstractKeyword) ?? false
-      const clsDoc = parseJsDoc(rawJsDoc(text, cls)).doc
-      if (!clsDoc) violations.push(`service ctx.${key} (${pointer(rel, sf, cls)}): class ${type} has no JSDoc.`)
+    // Resolve each ctx key to its service class (shared walk) and emit an entry.
+    for (const { key, type, cls, abstract, doc: clsDoc } of serviceClasses(body, sf, rel, violations)) {
       const methods: string[] = []
       for (const member of cls.members) {
         if (!ts.isMethodDeclaration(member)) continue
@@ -260,14 +229,14 @@ export function collectServices(scanRoot: string = root): ServiceEntry[] {
  * sibling check is N/A; keep them current on a vendor bump.
  */
 const INHERITED_EVENTS: InheritedEntry[] = [
-  { name: 'internal/plugin', summary: 'A plugin fiber was created.', source: 'vendor/cordis/src/events.ts:197' },
-  { name: 'internal/status', summary: 'A fiber changed lifecycle state.', source: 'vendor/cordis/src/events.ts:198' },
-  { name: 'internal/service', summary: 'Interception hook for a service binding (no core producer).', source: 'vendor/cordis/src/events.ts:199' },
-  { name: 'internal/update', summary: 'Waterfall: a fiber config update is being applied.', source: 'vendor/cordis/src/events.ts:200' },
-  { name: 'internal/get', summary: 'Waterfall: a service is being read from the store.', source: 'vendor/cordis/src/events.ts:201' },
-  { name: 'internal/set', summary: 'Waterfall: a service is being written to the store.', source: 'vendor/cordis/src/events.ts:202' },
-  { name: 'internal/listener', summary: 'A listener was registered.', source: 'vendor/cordis/src/events.ts:203' },
-  { name: 'internal/dispatch', summary: 'An event is being dispatched to listeners.', source: 'vendor/cordis/src/events.ts:204' },
+  { name: 'internal/plugin', summary: 'A plugin fiber was created.', source: 'vendor/cordis/src/events.ts:328' },
+  { name: 'internal/status', summary: 'A fiber changed lifecycle state.', source: 'vendor/cordis/src/events.ts:330' },
+  { name: 'internal/service', summary: 'Interception hook for a service binding (no core producer).', source: 'vendor/cordis/src/events.ts:332' },
+  { name: 'internal/update', summary: 'Waterfall: a fiber config update is being applied.', source: 'vendor/cordis/src/events.ts:334' },
+  { name: 'internal/get', summary: 'Waterfall: a service is being read from the store.', source: 'vendor/cordis/src/events.ts:336' },
+  { name: 'internal/set', summary: 'Waterfall: a service is being written to the store.', source: 'vendor/cordis/src/events.ts:338' },
+  { name: 'internal/listener', summary: 'A listener was registered.', source: 'vendor/cordis/src/events.ts:340' },
+  { name: 'internal/dispatch', summary: 'An event is being dispatched to listeners.', source: 'vendor/cordis/src/events.ts:342' },
   { name: 'hmr/change', summary: 'A watched source file changed on disk.', source: 'vendor/hmr/src/index.ts:20' },
   { name: 'hmr/reload', summary: 'Plugins are being reloaded after a change.', source: 'vendor/hmr/src/index.ts:21' },
   { name: 'exit', summary: 'The process is exiting on a signal.', source: 'vendor/loader/src/index.ts:23' },
@@ -278,12 +247,12 @@ const INHERITED_EVENTS: InheritedEntry[] = [
 ]
 
 export const INHERITED_SERVICES: InheritedEntry[] = [
-  { name: 'ctx.on / ctx.once', summary: 'Register an event listener (disposable).', source: 'vendor/cordis/src/events.ts:29' },
-  { name: 'ctx.emit / ctx.parallel / ctx.serial / ctx.bail / ctx.waterfall', summary: 'Dispatch an event (sync / awaited / first-bail / veto-chain).', source: 'vendor/cordis/src/events.ts:29' },
-  { name: 'ctx.plugin / ctx.inject', summary: 'Load a plugin / declare required services.', source: 'vendor/cordis/src/registry.ts:144' },
+  { name: 'ctx.on / ctx.once', summary: 'Register an event listener (disposable).', source: 'vendor/cordis/src/events.ts:34' },
+  { name: 'ctx.emit / ctx.parallel / ctx.serial / ctx.bail / ctx.waterfall', summary: 'Dispatch an event (sync / awaited / first-bail / veto-chain).', source: 'vendor/cordis/src/events.ts:34' },
+  { name: 'ctx.plugin / ctx.inject', summary: 'Load a plugin / declare required services.', source: 'vendor/cordis/src/registry.ts:164' },
   { name: 'ctx.effect', summary: 'Register a disposable side effect tied to the fiber.', source: 'vendor/cordis/src/fiber.ts:9' },
   { name: 'ctx.get / ctx.set / ctx.provide / ctx.accessor / ctx.mixin', summary: 'Low-level service-store access and binding.', source: 'vendor/cordis/src/reflect.ts:7' },
-  { name: 'ctx.extend / ctx.isolate / ctx.intercept', summary: 'Derive a child context (scoped services / isolation / interception).', source: 'vendor/cordis/src/context.ts:35' },
+  { name: 'ctx.extend / ctx.isolate / ctx.intercept', summary: 'Derive a child context (scoped services / isolation / interception).', source: 'vendor/cordis/src/context.ts:42' },
   { name: 'ctx.root / ctx.scope / ctx.fiber / ctx.registry / ctx.reflect / ctx.events / ctx.logger', summary: 'Ambient handles onto the running context graph.', source: 'vendor/cordis/src/context.ts:16' },
   { name: 'ctx.timer (+ interval / timeout / throttle / debounce / setTimeout / setInterval)', summary: 'Disposable timer helpers. The `timer` key is provided at runtime; the six helpers are mixed onto ctx directly (declared via Pick).', source: 'vendor/timer/src/index.ts:4' },
   { name: 'ctx.loader', summary: 'The config Loader that booted the app (present under the loader).', source: 'vendor/loader/src/index.ts:30' },
diff --git a/scripts/gen-doc-graphs.ts b/scripts/gen-doc-graphs.ts
index 0e1774c2e0..f6aaa6c2d9 100644
--- a/scripts/gen-doc-graphs.ts
+++ b/scripts/gen-doc-graphs.ts
@@ -64,7 +64,10 @@ const GROUP_ORDER = [
   'skill',
   'compact',
   'subagent',
+  'tasks',
+  'workflow',
   'web',
+  'spill',
   'todo',
   'cordis',
   'hooks',
@@ -84,6 +87,14 @@ const SERVICE_ROLES: ServiceRole[] = [
     consumers: ['agent-loop', 'compact-basic'],
     note: 'Adapters register provider implementations; the loop and compaction call the provider-neutral stream service.',
   },
+  {
+    key: 'tokenMeter',
+    pkg: 'token-meter',
+    title: 'Replay token measurement',
+    mode: 'core',
+    consumers: ['compact-basic'],
+    note: 'Owns isolated per-session replay folds; pressure consumers share immutable revisioned measurements.',
+  },
   {
     key: 'sessions',
     pkg: 'session',
@@ -98,15 +109,15 @@ const SERVICE_ROLES: ServiceRole[] = [
     title: 'Durable session persistence seam',
     mode: 'seam',
     implementations: ['session-persistence-jsonl', 'session-persistence-sqlite'],
-    consumers: ['agent-loop', 'acp', 'session-query'],
+    consumers: ['agent-loop', 'tool-bash', 'hooks-claude', 'hooks-codex', 'acp', 'session-query'],
     note: 'Backends persist the same SessionEvent vocabulary; apps choose a backend at composition time.',
   },
   {
     key: 'sessionQuery',
     pkg: 'session-query',
-    title: 'Exact session-history reads',
+    title: 'Exact session-history reads and traces',
     mode: 'seam',
-    note: 'Resolves live and optional persisted logs into one logical corpus for exact reads.',
+    note: 'Resolves live and optional persisted logs into one logical corpus for exact reads and relationship traces.',
   },
   {
     key: 'systemPrompt',
@@ -167,6 +178,13 @@ const SERVICE_ROLES: ServiceRole[] = [
     consumers: ['tool-bash', 'hooks-claude', 'hooks-codex'],
     note: 'The model-facing bash tools and hook bridges consume this seam; sandboxed or remote executors replace bash-local without touching them.',
   },
+  {
+    key: 'bashEnv',
+    pkg: 'tool-bash',
+    title: 'Managed bash environment registry',
+    mode: 'core',
+    note: 'Plugins declare effect-scoped DSH_* facts; tool-bash collects one trusted snapshot per execution and the executor rebuilds the namespace.',
+  },
   {
     key: 'sandbox',
     pkg: 'sandbox',
@@ -231,6 +249,14 @@ const SERVICE_ROLES: ServiceRole[] = [
     consumers: ['tool-subagent'],
     note: 'Providers implement transports; tool-subagent exposes one configured provider as a model-facing tool name.',
   },
+  {
+    key: 'tasks',
+    pkg: 'tasks',
+    title: 'Background task registry',
+    mode: 'core',
+    consumers: ['tool-bash', 'tool-subagent', 'tool-tasks'],
+    note: 'Producers (tool-bash background commands, tool-subagent background delegations) register running work; tool-tasks is the model-facing control surface that reads, lists, and kills it.',
+  },
   {
     key: 'web',
     pkg: 'web',
@@ -240,6 +266,15 @@ const SERVICE_ROLES: ServiceRole[] = [
     consumers: ['tool-web'],
     note: 'Search and fetch providers register into one ctx.web seam; tool-web owns the stable model-facing names.',
   },
+  {
+    key: 'spillStore',
+    pkg: 'spill',
+    title: 'Spill storage seam',
+    mode: 'seam',
+    implementations: ['spill-local'],
+    consumers: ['spill-policy'],
+    note: 'The backend saves oversized tool text and returns a model-facing locator plus retrieval hint; spill-policy is the tools/post-execute consumer that decides when to spill.',
+  },
   {
     key: 'workflows',
     pkg: 'workflow',
@@ -812,10 +847,19 @@ function renderLifecycle(): string {
     `  Session-->>SDK: ${mermaidCode('session/event')} ${mermaidCode('assistant/chunk')}*`,
     `  Driver->>Hooks: ${mermaidCode('agent/step-result')} waterfall`,
     `  Driver->>Session: ${mermaidCode('assistant/message')}`,
-    `  Driver->>Session: ${mermaidCode('tool/call')}`,
-    '  Driver->>Tools: execute through pre and post waterfalls',
-    '  Tools-->>Session: tool-owned events when applicable',
-    `  Driver->>Session: ${mermaidCode('tool/result')} and ${mermaidCode('step/end')}`,
+    '  Driver->>Tools: classify pending call by executionMode',
+    '  loop barriers and bounded rolling pool, reclassify before start',
+    '    opt call starts',
+    `      Driver->>Session: ${mermaidCode('tool/call')}`,
+    '      Driver->>Tools: ordered pre, concurrent execute',
+    '      Tools-->>Session: tool-owned events when applicable',
+    '    end',
+    '    opt next model-order result ready',
+    '      Driver->>Tools: ordered post',
+    `      Driver->>Session: ${mermaidCode('tool/result')}`,
+    '    end',
+    '  end',
+    `  Driver->>Session: ${mermaidCode('step/end')}`,
     `  Driver->>Hooks: ${mermaidCode('agent/turn-continuation')} waterfall`,
     `  Driver->>Hooks: ${mermaidCode('agent/turn-stop')} serial terminal checkpoint`,
     `  Driver->>Session: ${mermaidCode('turn/end')}`,
@@ -823,6 +867,8 @@ function renderLifecycle(): string {
     `  Driver-->>SDK: ${mermaidCode('agent/status')} idle`,
     '```',
     '',
+    'The `assistant/message` edge records every successful provider call, including content-less and `max-tokens` finishes. Empty content stays out of derived history while the durable anchor retains usage and exact chunk provenance, including an explicit empty source set.',
+    '',
     'SDK users that need replayable transcript data should consume `session/event`; `agent/*` is the live coordination surface for queue/status, prompt interception, request shaping, steering, continuation, and errors.',
     '',
     ...maintenanceFooter(maintenance),
@@ -850,9 +896,9 @@ function renderToolPipeline(): string {
     `  owned["Tool-owned session events<br/>${mermaidCode('todo/write')}, ${mermaidCode('fs/observed')}, ${mermaidCode('hook/invoked')}, ${mermaidCode('hook/result')}, ${mermaidCode('tool/code-dispatch')}"]`,
     `  post["${mermaidCode('tools/post-execute')} waterfall<br/>accept, block, replace, add context"]`,
     `  final["${mermaidCode('tools/result')} synchronous notification<br/>frozen authoritative outcome"]`,
-    '  context["Buffered additionalContext<br/>context/message after all tool results"]',
+    '  context["Active-batch additionalContexts FIFO<br/>context/message after recorded tool results"]',
     `  toolResult["Session event: ${mermaidCode('tool/result')}<br/>single model-facing outcome"]`,
-    '  allResults["All calls in the step settled<br/>and tool/result events recorded"]',
+    '  allResults["Tool batch settled<br/>recorded tool/result events complete"]',
     '  presentResult["UI completed card<br/>presentResult(args, result)"]',
     '  model --> toolCall',
     '  toolCall --> presentCall',
@@ -878,7 +924,7 @@ function renderToolPipeline(): string {
     '  allResults --> context',
     '```',
     '',
-    'Filesystem read-before-edit checks stay below `tool-fs` on `fs/*` events. Generic pre/post waterfalls host hooks and approval policy; `ctx.approval` resolves asks before monotonic guards, and owner policy that must not be reordered remains a registered guard. Around-dispatch concerns such as timeouts wrap `tools/execute`, while `tools/result` observes the immutable outcome after transforms, lossless-JSON validation, and outer error normalization. This lets hooks span tool families without coupling the tools to one policy service. Code Mode sends both the reserved `run_code` transport and its serialized sub-calls through the pipeline; sub-calls carry the parent token, log `tool/code-dispatch`, surface denials as binding rejections, and omit `additionalContext` to preserve call/result adjacency.',
+    'Filesystem read-before-edit checks stay below `tool-fs` on `fs/*` events. Generic pre/post waterfalls host hooks and approval policy; `ctx.approval` resolves asks before monotonic guards, and owner policy that must not be reordered remains a registered guard. Around-dispatch concerns such as timeouts wrap `tools/execute`, while `tools/result` observes the immutable outcome after transforms, lossless-JSON validation, and outer error normalization. This lets hooks span tool families without coupling the tools to one policy service. Code Mode sends both the reserved `run_code` transport and its serialized sub-calls through the pipeline; sub-calls carry the parent token, log `tool/code-dispatch`, surface denials as binding rejections, and omit `additionalContexts` to preserve call/result adjacency.',
     '',
     ...maintenanceFooter(maintenance),
   ].join('\n')
diff --git a/scripts/gen-module-graph.ts b/scripts/gen-module-graph.ts
index 6ae5dd3044..520375c9ae 100644
--- a/scripts/gen-module-graph.ts
+++ b/scripts/gen-module-graph.ts
@@ -27,6 +27,7 @@ const GROUP_ORDER = [
   'compact',
   'subagent',
   'web',
+  'spill',
   'timeout',
   'todo',
   'cordis',
diff --git a/scripts/gen-persistence-catalog.ts b/scripts/gen-persistence-catalog.ts
index 9469081607..8762b8529e 100644
--- a/scripts/gen-persistence-catalog.ts
+++ b/scripts/gen-persistence-catalog.ts
@@ -1,7 +1,7 @@
 /**
  * Generate `docs/persistence-catalog.md` from every `SessionEventMap` merge and
- * the owning `SurfaceEventType` union. This is the durable-record vocabulary,
- * not the live Cordis bus. Event declarations must be unique, explicitly typed,
+ * the owning event-envelope types. This is the durable-record vocabulary, not
+ * the live Cordis bus. Event declarations must be unique, explicitly typed,
  * documented, inheritance-free, and free of Cordis-only `@mode` tags; every
  * surface-union member must resolve to one. `--check` verifies the artifact.
  */
@@ -14,13 +14,23 @@ import { parseJsDoc, pointer, rawJsDoc, reportViolations } from './jsdoc.ts'
 const root = resolve(import.meta.dirname, '..')
 const OUT = 'docs/persistence-catalog.md'
 
-/** The fenced-block info string for generated payload blocks (skipped by
- * doc-typecheck, since a bare payload fragment is not standalone-compilable). */
+/** The fenced-block info string for generated declaration blocks (skipped by
+ * doc-typecheck, since their imported types are not standalone-compilable). */
 const FENCE = 'ts persistence-catalog'
 
 /** The package whose module id plugin merges augment (`declare module '…'`). */
 const SESSION_MODULE = '@deepseek-ai/dsh-session'
 
+/** Event-envelope declarations rendered before the per-event vocabulary. */
+const EVENT_ENVELOPE_TYPE_NAMES = [
+  'SessionEventType',
+  'SurfaceEventType',
+  'SurfaceOp',
+  'SessionEvent',
+] as const
+
+type EventEnvelopeTypeName = typeof EVENT_ENVELOPE_TYPE_NAMES[number]
+
 /** Primary core-data-structures page for linked payload types. */
 const LINK_MAP: Record<string, string> = {
   CallId: 'core.md',
@@ -41,6 +51,8 @@ export interface LogEventEntry {
   scope: string
   /** Payload type text (the member's type annotation, whitespace-collapsed). */
   payload: string
+  /** Source member declaration and complete JSDoc, dedented from its container. */
+  declaration: string
   /** Description prose (the member's JSDoc), one line per paragraph. */
   doc: string
   /** Source pointer `packages/…/file.ts:line` of the declaration. */
@@ -53,6 +65,16 @@ export interface AnnotatedLogEventEntry extends LogEventEntry {
   surface: boolean
 }
 
+/** One owning event-envelope declaration pasted into the generated catalog. */
+export interface EventEnvelopeTypeEntry {
+  /** Exported declaration name. */
+  name: EventEnvelopeTypeName
+  /** Verbatim type declaration, including its complete leading JSDoc. */
+  declaration: string
+  /** Source pointer `packages/…/file.ts:line` of the declaration. */
+  source: string
+}
+
 const printer = ts.createPrinter({ removeComments: true })
 
 /**
@@ -67,6 +89,24 @@ function payloadText(type: ts.TypeNode, sf: ts.SourceFile): string {
     .trim()
 }
 
+/**
+ * Copy a declaration from its leading JSDoc through its closing token while
+ * removing only the indentation imposed by its containing interface/module.
+ */
+function declarationText(text: string, sf: ts.SourceFile, node: ts.Node): string {
+  const raw = rawJsDoc(text, node)
+  const nodeStart = node.getStart(sf)
+  const start = raw ? text.lastIndexOf(raw, nodeStart) : nodeStart
+  const { line } = sf.getLineAndCharacterOfPosition(start)
+  const lineStart = sf.getPositionOfLineAndCharacter(line, 0)
+  const indent = text.slice(lineStart, start)
+  return text.slice(lineStart, node.end)
+    .split('\n')
+    .map(lineText => lineText.startsWith(indent) ? lineText.slice(indent.length) : lineText)
+    .join('\n')
+    .trimEnd()
+}
+
 /**
  * Every `interface SessionEventMap` declaration in a source file: the owning
  * top-level declaration (in `@deepseek-ai/dsh-session`) and any declaration
@@ -177,7 +217,8 @@ export function collectLogEvents(scanRoot: string = root): LogEventEntry[] {
         if (!doc) {
           violations.push(`${where} has no description prose. Say what the event records and what its payload means — the JSDoc becomes the catalog entry.`)
         }
-        entries.push({ name, scope: name.split('/')[0] ?? name, payload, doc, source: src })
+        const declaration = declarationText(text, sf, member)
+        entries.push({ name, scope: name.split('/')[0] ?? name, payload, declaration, doc, source: src })
       }
     }
   }
@@ -185,6 +226,51 @@ export function collectLogEvents(scanRoot: string = root): LogEventEntry[] {
   return entries
 }
 
+/**
+ * Collect the exported declarations that compose the persisted event envelope,
+ * preserving their source JSDoc and declaration text.
+ */
+export function collectEventEnvelopeTypes(scanRoot: string = root): EventEnvelopeTypeEntry[] {
+  const found = new Map<EventEnvelopeTypeName, EventEnvelopeTypeEntry>()
+  const violations: string[] = []
+  const wanted = new Set<string>(EVENT_ENVELOPE_TYPE_NAMES)
+  for (const rel of globSync('packages/*/*/src/**/*.ts', { cwd: scanRoot }).map(s => s.split(sep).join('/')).sort()) {
+    const abs = resolve(scanRoot, rel)
+    const text = readFileSync(abs, 'utf8')
+    if (!EVENT_ENVELOPE_TYPE_NAMES.some(name => text.includes(name))) continue
+    if (packageNameFor(rel, scanRoot) !== SESSION_MODULE) continue
+    const sf = ts.createSourceFile(abs, text, ts.ScriptTarget.Latest, true)
+    for (const stmt of sf.statements) {
+      if (!ts.isTypeAliasDeclaration(stmt) || !wanted.has(stmt.name.text)) continue
+      const name = stmt.name.text as EventEnvelopeTypeName
+      const src = pointer(rel, sf, stmt)
+      const where = `event-envelope type '${name}' (${src})`
+      const prior = found.get(name)
+      if (prior) {
+        violations.push(`${where} is already declared at ${prior.source}; the persisted envelope type has exactly one owner.`)
+        continue
+      }
+      if (!(stmt.modifiers?.some(m => m.kind === ts.SyntaxKind.ExportKeyword) ?? false)) {
+        violations.push(`${where} is not exported.`)
+      }
+      const { doc, hasMode } = parseJsDoc(rawJsDoc(text, stmt))
+      if (hasMode) violations.push(`${where} carries an @mode tag, but a persisted type has no dispatch mode.`)
+      if (!doc) violations.push(`${where} has no description prose. The full JSDoc is part of the generated catalog.`)
+      found.set(name, { name, declaration: declarationText(text, sf, stmt), source: src })
+    }
+  }
+  const missing = EVENT_ENVELOPE_TYPE_NAMES.filter(name => !found.has(name))
+  if (missing.length > 0) {
+    violations.push(`missing event-envelope declaration(s): ${missing.join(', ')}.`)
+  }
+  reportViolations('gen-persistence-catalog', violations)
+  return EVENT_ENVELOPE_TYPE_NAMES.map((name) => {
+    const entry = found.get(name)
+    if (!entry) throw new Error(`gen-persistence-catalog: missing checked event-envelope declaration '${name}'.`)
+    return entry
+  })
+}
+
 /**
  * Parse the `SurfaceEventType` union — the surface-eligible subset of event
  * types — from source. Hard-errors when the alias is missing, declared more
@@ -246,8 +332,7 @@ function typeLinks(payload: string): string {
 /** Render one log event entry. */
 function renderEvent(e: AnnotatedLogEventEntry): string[] {
   const out = [`#### \`${e.name}\` — ${e.surface ? 'surface' : 'log-only'}`, '']
-  if (e.doc) out.push(e.doc, '')
-  out.push('```' + FENCE, `'${e.name}': ${e.payload}`, '```', '')
+  out.push('```' + FENCE, e.declaration, '```', '')
   const links = typeLinks(e.payload)
   if (links) out.push(links, '')
   out.push(`Source: [\`${e.source}\`](../${e.source.split(':')[0]})`, '')
@@ -255,18 +340,26 @@ function renderEvent(e: AnnotatedLogEventEntry): string[] {
 }
 
 /** Render the full catalog (pure, deterministic given the collected inputs). */
-export function render(events: AnnotatedLogEventEntry[]): string {
+export function render(events: AnnotatedLogEventEntry[], envelopeTypes: EventEnvelopeTypeEntry[]): string {
   const lines: string[] = [
     '<!-- Generated by scripts/gen-persistence-catalog.ts — do not edit by hand.',
     '     Run `pnpm run gen-persistence-catalog` to regenerate. -->',
     '',
-    '# Persistence Log Event Catalog',
+    '# Session Persistence Event Catalog',
     '',
-    'Every event type that can appear in a session\'s durable event log: each member of the merge-extensible `SessionEventMap` — the owning vocabulary in `@deepseek-ai/dsh-session` plus every plugin declaration merge in this repo — with the payload it carries, its surface badge, and the declaration it comes from. It complements [session.md](core-data-structures/session.md) (the `SessionEvent` envelope, surface list, and `deriveMessages()` projection), [persistence.md](core-data-structures/persistence.md) (how the log is made durable), and the [cordis events catalog](cordis-catalog/events.md) (the live bus wiring — a log event is NOT a cordis event; it reaches listeners via the single `session/event` emit).',
+    'Every event type that can appear in a session\'s durable event log: the complete persisted `SessionEvent` envelope and each member of the merge-extensible `SessionEventMap` — the owning vocabulary in `@deepseek-ai/dsh-session` plus every plugin declaration merge in this repo — with source JSDoc, full payload declaration, surface badge, and declaration site. It complements [session.md](core-data-structures/session.md) (surface ordering and the `deriveMessages()` projection), [persistence.md](core-data-structures/persistence.md) (how the log is made durable), and the [cordis events catalog](cordis-catalog/events.md) (the live bus wiring — a log event is NOT a cordis event; it reaches listeners via the single `session/event` emit).',
     '',
-    'This file is GENERATED from source (`scripts/gen-persistence-catalog.ts`) and verified fresh by `pnpm run verify-persistence-catalog` (part of `doc-sync`) — do not edit it by hand. Payload blocks use a `ts persistence-catalog` fence (skipped by doc-typecheck, since a bare payload fragment is not standalone-compilable). Type names in a payload link to the page that documents them. See [the persistence-log-catalog RFC](rfc/implemented/process/2026-07-04-persistence-log-catalog.md).',
+    'This file is GENERATED from source (`scripts/gen-persistence-catalog.ts`) and verified fresh by `pnpm run verify-persistence-catalog` (part of `doc-sync`) — do not edit it by hand. Declaration blocks retain the source declaration and nested property JSDoc, removing only the indentation imposed by a containing interface/module, and use a `ts persistence-catalog` fence (skipped by doc-typecheck because declarations reference types from their owning modules). Type names in a payload link to the page that documents them. See [the persistence-log-catalog RFC](rfc/implemented/process/2026-07-04-persistence-log-catalog.md).',
     '',
-    'The on-disk envelope around every payload is `SessionEvent` — `type`, monotonic `seq`, epoch-ms `time`, the `data` documented here, plus `surfaceOp`/`sourceEventSeqs` on **surface** events only ([envelope](core-data-structures/session.md#sessioneventt--one-log-entry)). **surface** marks a `SurfaceEventType` member: it produces an LLM message and declares how it joins the surface list. **log-only** marks everything else: durable, replayable record with no derived-history contribution. Every payload is JSON-serializable (enforced at `Session.append`), and the whole format is pinned at `SESSION_FORMAT_VERSION = 0` — pre-release, no compatibility implied ([the version stance](core-data-structures/persistence.md)). Scope: the packages in this repo; a downstream plugin can merge further event types, which are outside this catalog by construction.',
+    'The envelope declarations below compose each event\'s `type`, monotonic `seq`, epoch-ms `time`, `data`, and the conditional `surfaceOp`/`sourceEventSeqs` fields. **surface** marks a `SurfaceEventType` member: it produces an LLM message and declares how it joins the surface list. **log-only** marks everything else: a durable, replayable record with no derived-history contribution. Every payload is JSON-serializable (enforced at `Session.append`), and the whole format is pinned at `SESSION_FORMAT_VERSION = 0` — pre-release, no compatibility implied ([the version stance](core-data-structures/persistence.md)). Scope: the packages in this repo; a downstream plugin can merge further event types, which are outside this catalog by construction.',
+    '',
+    '## Event envelope',
+    '',
+    '```' + FENCE,
+    envelopeTypes.map(entry => entry.declaration).join('\n\n'),
+    '```',
+    '',
+    `Sources: ${envelopeTypes.map(entry => `[\`${entry.source}\`](../${entry.source.split(':')[0]})`).join(' · ')}`,
     '',
     '## Events',
     '',
@@ -285,7 +378,7 @@ export function render(events: AnnotatedLogEventEntry[]): string {
  * is stale. Guarded behind an entry-point check so importing this module for
  * tests neither regenerates the committed file nor calls process.exit. */
 function main(): void {
-  const content = render(annotateSurface(collectLogEvents(), collectSurfaceEventTypes()))
+  const content = render(annotateSurface(collectLogEvents(), collectSurfaceEventTypes()), collectEventEnvelopeTypes())
   if (process.argv.includes('--check')) {
     let committed: string | null = null
     try {
diff --git a/scripts/gen-tool-catalog.ts b/scripts/gen-tool-catalog.ts
index ab6b01c10d..f178f70e9f 100644
--- a/scripts/gen-tool-catalog.ts
+++ b/scripts/gen-tool-catalog.ts
@@ -22,11 +22,14 @@ import SubagentService from '@deepseek-ai/dsh-subagent'
 import * as SubagentMock from '@deepseek-ai/dsh-subagent-mock'
 import SkillService from '@deepseek-ai/dsh-skill'
 import * as SkillLocal from '@deepseek-ai/dsh-skill-local'
+import TaskService from '@deepseek-ai/dsh-tasks'
 import * as ToolAskUser from '@deepseek-ai/dsh-tool-ask-user'
 import * as ToolBash from '@deepseek-ai/dsh-tool-bash'
 import * as ToolCordis from '@deepseek-ai/dsh-tool-cordis'
 import * as ToolFs from '@deepseek-ai/dsh-tool-fs'
+import * as ToolFsSearch from '@deepseek-ai/dsh-tool-fs-search'
 import * as ToolSkill from '@deepseek-ai/dsh-tool-skill'
+import * as ToolTasks from '@deepseek-ai/dsh-tool-tasks'
 import * as ToolTodo from '@deepseek-ai/dsh-tool-todo'
 import * as ToolSubagent from '@deepseek-ai/dsh-tool-subagent'
 import * as ToolWeb from '@deepseek-ai/dsh-tool-web'
@@ -111,14 +114,14 @@ const TOOL_PACKAGES: ToolPackage[] = [
     pkg: '@deepseek-ai/dsh-tool-bash',
     dir: 'tool-bash',
     source: 'packages/bash/tool-bash/src/index.ts',
-    requires: ['ctx.tools', 'ctx.bash'],
-    writes: ['tool/call', 'tool/result', 'context/message via agent.inject() for background completion notices'],
+    requires: ['ctx.tools', 'ctx.bash', 'ctx.tasks at call time for run_in_background'],
+    writes: ['tool/call', 'tool/result'],
     async mount(ctx) {
       await ctx.plugin(LocalBashExecutor)
       await ctx.plugin(ToolBash)
     },
     note:
-      'The bash/bash_output/bash_kill tools are model-facing consumers of the bash executor seam.',
+      'The bash tool is the model-facing consumer of the bash executor seam. A `run_in_background` run registers with the generic `ctx.tasks` runtime and is collected/stopped through the `task_*` tools from `@deepseek-ai/dsh-tool-tasks`; the `enableRunInBackground` config (default true) removes the parameter entirely when disabled.',
   },
   {
     pkg: '@deepseek-ai/dsh-tool-cordis',
@@ -130,7 +133,7 @@ const TOOL_PACKAGES: ToolPackage[] = [
       await ctx.plugin(ToolCordis)
     },
     note:
-      'Ships in examples/cordis-agent only (a deliberate opt-in — mounted code gets the real ctx, see docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md). Plugins the model mounts may register ADDITIONAL model-visible tools at runtime; the request-header ToolsDelta logs those tool-set changes.',
+      'Ships in examples/cordis-agent only (a deliberate opt-in — mounted code gets the real ctx, see docs/rfc/implemented/feature/2026-07-08-self-referential-cordis-toolset.md). Plugins the model mounts may register ADDITIONAL model-visible tools at runtime; a full changed request header logs those tool-set changes.',
   },
   {
     pkg: '@deepseek-ai/dsh-tool-fs',
@@ -147,6 +150,23 @@ const TOOL_PACKAGES: ToolPackage[] = [
     note:
       'The read-before-write/edit policy is added by `@deepseek-ai/dsh-fs-policy` (an `fs/*` event-gate plugin, no schema change); a deployment that loads these tools is expected to also load it. The tool schemas above are identical with or without the policy plugin.',
   },
+  {
+    pkg: '@deepseek-ai/dsh-tool-fs-search',
+    dir: 'tool-fs-search',
+    source: 'packages/fs/tool-fs-search/src/index.ts',
+    requires: ['ctx.tools', 'ctx.bash', 'ctx.systemPrompt'],
+    writes: ['tool/call', 'tool/result'],
+    async mount(ctx) {
+      // The tools inject `bash` (search executes fixed `rg` commands through
+      // the executor seam, not ctx.fs); boot the local executor to satisfy it.
+      // `ctx.spillStore` is optional (read via ctx.get) and does not affect the
+      // schemas, so no spill backend is mounted.
+      await ctx.plugin(LocalBashExecutor)
+      await ctx.plugin(ToolFsSearch)
+    },
+    note:
+      'glob and grep are bash-backed discovery tools: they run fixed ripgrep commands through ctx.bash as ordinary foreground calls (never background tasks). Capped results save the complete formatted list through the optional ctx.spillStore backend; returned locators are follow-up-readable/searchable when the backend exposes local paths in co-located deployments.',
+  },
   {
     pkg: '@deepseek-ai/dsh-tool-skill',
     dir: 'tool-skill',
@@ -178,6 +198,19 @@ const TOOL_PACKAGES: ToolPackage[] = [
     note:
       'The registered tool name is the load-time `toolName` config (default `subagent`); the schema above is that default. The shipped example agents load this package once per subagent backend, so the model additionally sees `subagent_fork` (bound to the fork backend) with an identical schema — see `examples/coding-agent/cordis.yml` and `examples/acp-agent/cordis.yml`.',
   },
+  {
+    pkg: '@deepseek-ai/dsh-tool-tasks',
+    dir: 'tool-tasks',
+    source: 'packages/tasks/tool-tasks/src/index.ts',
+    requires: ['ctx.tools', 'ctx.tasks', 'ctx.systemPrompt'],
+    writes: ['tool/call', 'tool/result', 'context/message via agent.inject() for background completion notices'],
+    async mount(ctx) {
+      await ctx.plugin(TaskService)
+      await ctx.plugin(ToolTasks)
+    },
+    note:
+      'The kind-agnostic background-task control surface: a background bash command and a background subagent are read, listed, and killed through the same three tools. Loading the plugin attaches the control surface that arms producers\' `ctx.tasks.start()`.',
+  },
   {
     pkg: '@deepseek-ai/dsh-tool-todo',
     dir: 'tool-todo',
diff --git a/scripts/md-fences.ts b/scripts/md-fences.ts
new file mode 100644
index 0000000000..ad97164369
--- /dev/null
+++ b/scripts/md-fences.ts
@@ -0,0 +1,55 @@
+/**
+ * Shared fenced-code-block extractor for the Markdown doc gates
+ * (currently `doc-typecheck.ts`; future Markdown gates can share it). One scanner, per-gate
+ * classification: each gate maps a fence info string (` ```ts `,
+ * ` ```yaml ignore-check `, …) to its own kind tag and receives every
+ * classified block with its 1-based opening-fence line.
+ */
+
+import { readFileSync } from 'node:fs'
+
+/** One extracted fenced block, classified by the caller's `classify`. */
+export interface Fence<K> {
+  /** 1-based line of the opening fence. */
+  line: number
+  kind: K
+  code: string
+}
+
+/**
+ * Extract every fenced block of `absPath` whose info string `classify` maps
+ * to a kind. Blocks classified `null` are skipped (their bodies are still
+ * consumed, so an unrelated fence can never leak into a tracked one).
+ *
+ * @param absPath — absolute path of the Markdown file.
+ * @param classify — info string (trimmed, e.g. `ts ignore-check`) → kind, or
+ *   null for fences this gate does not track.
+ * @returns the classified blocks in document order.
+ */
+export function extractFences<K>(absPath: string, classify: (info: string) => K | null): Fence<K>[] {
+  const lines = readFileSync(absPath, 'utf8').split('\n')
+  const blocks: Fence<K>[] = []
+  let open: { line: number; kind: K; body: string[] } | null = null
+  let skipping = false
+
+  lines.forEach((raw, i) => {
+    const fence = /^```(\s*)(\S.*)?$/.exec(raw)
+    if (!fence) {
+      if (open) open.body.push(raw)
+      return
+    }
+    if (open) {
+      blocks.push({ line: open.line, kind: open.kind, code: open.body.join('\n') })
+      open = null
+      return
+    }
+    if (skipping) {
+      skipping = false
+      return
+    }
+    const kind = classify((fence[2] ?? '').trim())
+    if (kind !== null) open = { line: i + 1, kind, body: [] }
+    else skipping = true
+  })
+  return blocks
+}
diff --git a/scripts/run-gates.ts b/scripts/run-gates.ts
index 95571f4f0a..9eb82b5e15 100644
--- a/scripts/run-gates.ts
+++ b/scripts/run-gates.ts
@@ -24,6 +24,7 @@ type GateStatus = 'pending' | 'running' | 'passed' | 'failed' | 'skipped'
 interface Gate {
   id: string
   label: string
+  displayCommand: string
   command: string
   args: string[]
   needs?: string[]
@@ -38,22 +39,39 @@ interface GateResult {
   durationMs: number
   stdout: string
   stderr: string
+  output: GateOutputChunk[]
   exitCode: number | null
   error?: string
 }
 
+interface GateOutputChunk {
+  stream: 'stdout' | 'stderr'
+  text: string
+}
+
 interface RunningGate {
   gate: Gate
   promise: Promise<GateResult>
 }
 
+interface ConcurrencyDefault {
+  workers: number
+  source: string
+}
+
 const root = resolve(import.meta.dirname, '..')
 const mode = parseMode(process.argv[2])
 const gates = gatesForMode(mode)
-const maxConcurrency = concurrencyFromEnv('DSH_GATE_CONCURRENCY', defaultConcurrency(gates.length))
+const concurrencyDefault = defaultConcurrency(mode, gates.length)
+const concurrencyOverride = process.env.DSH_GATE_CONCURRENCY
+const maxConcurrency = concurrencyFromEnv('DSH_GATE_CONCURRENCY', concurrencyDefault.workers)
+const verbose = process.env.DSH_GATE_VERBOSE === '1'
 const startedAt = performance.now()
 
-console.log(`run-gates: ${mode} running ${gates.length} gate(s) with ${maxConcurrency} worker(s).`)
+const concurrencySource = concurrencyOverride === undefined || concurrencyOverride === ''
+  ? concurrencyDefault.source
+  : '$DSH_GATE_CONCURRENCY'
+console.log(`run-gates: ${mode} running ${gates.length} gate(s) with ${maxConcurrency} worker(s) from ${concurrencySource}.`)
 
 const results = await runGates(gates, maxConcurrency)
 printSummary(results, performance.now() - startedAt)
@@ -78,8 +96,15 @@ function parseMode(raw: string | undefined): Mode {
   }
 }
 
-function defaultConcurrency(total: number): number {
-  return Math.min(total, Math.max(4, availableParallelism()))
+function defaultConcurrency(selectedMode: Mode, total: number): ConcurrencyDefault {
+  const available = availableParallelism()
+  const modeLimit = selectedMode === 'pre-push' ? Math.min(4, available) : available
+  return {
+    workers: Math.min(total, modeLimit),
+    source: selectedMode === 'pre-push'
+      ? `${available} available CPU(s), pre-push cap 4`
+      : `${available} available CPU(s)`,
+  }
 }
 
 function concurrencyFromEnv(name: string, fallback: number): number {
@@ -96,6 +121,7 @@ function pnpmScript(id: string, script: string, options: Partial<Gate> = {}): Ga
   return {
     id,
     label: options.label ?? script,
+    displayCommand: `pnpm run ${script}`,
     ...pnpmInvocation(['run', script]),
     ...options,
   }
@@ -105,6 +131,7 @@ function pnpmExec(id: string, args: string[], options: Partial<Gate> = {}): Gate
   return {
     id,
     label: options.label ?? `pnpm exec ${args.join(' ')}`,
+    displayCommand: `pnpm exec ${args.join(' ')}`,
     ...pnpmInvocation(['exec', ...args]),
     ...options,
   }
@@ -136,11 +163,13 @@ function gatesForMode(selected: Mode): Gate[] {
       ]
     case 'ci-coverage':
       return [
+        pnpmScript('build', 'build'),
         coverageGate(),
       ]
     case 'ci-snapshot':
       return [
-        pnpmScript('snapshot', 'test:snapshot'),
+        pnpmScript('build', 'build'),
+        snapshotGate(),
       ]
     case 'ci-artifacts':
       return ciArtifactGates()
@@ -159,10 +188,13 @@ function gatesForMode(selected: Mode): Gate[] {
         pnpmScript('cordis-config', 'verify-cordis-config', { label: 'Cordis config' }),
         pnpmScript('test', 'test'),
         pnpmScript('duplication', 'duplication'),
-        pnpmScript('snapshot', 'test:snapshot'),
+        snapshotGate(),
         pnpmScript('build', 'build'),
         ...hygieneLeafGates({ artifactNeeds: ['build'] }),
-        ...docSyncLeafGates(),
+        ...docSyncLeafGates({
+          docTypecheckNeeds: ['build'],
+          docTypecheckEnv: { DSH_DOC_TYPECHECK_USE_BUILD_OUTPUT: '1' },
+        }),
         pnpmScript('module-graph', 'verify-module-graph', { label: 'module graph' }),
       ]
   }
@@ -177,11 +209,12 @@ function ciPrimaryGates(): Gate[] {
     lintGate(),
     pnpmScript('duplication', 'duplication'),
     coverageGate(),
-    pnpmScript('snapshot', 'test:snapshot'),
+    snapshotGate(),
     demoSmokeGate({ needs: ['lint'] }),
     ...docSyncLeafGates(),
     pnpmScript('module-graph', 'verify-module-graph', { label: 'module graph' }),
     pnpmScript('knip', 'knip'),
+    pnpmScript('website-build', 'website:build', { label: 'website build' }),
     pnpmScript('build', 'build', { needs: ['typecheck'] }),
     pnpmScript('publint', 'publint', { needs: ['build'] }),
     pnpmScript('node-next-types', 'verify-node-next-types', {
@@ -201,6 +234,7 @@ function ciStaticGates(): Gate[] {
     ...docSyncLeafGates(),
     pnpmScript('module-graph', 'verify-module-graph', { label: 'module graph' }),
     pnpmScript('knip', 'knip'),
+    pnpmScript('website-build', 'website:build', { label: 'website build' }),
   ]
 }
 
@@ -249,6 +283,18 @@ function coverageGate(): Gate {
     ...positiveIntArg('DSH_COVERAGE_MAX_WORKERS', '--maxWorkers'),
   ], {
     label: 'test:coverage',
+    env: { DSH_EXAMPLE_MODE: 'lib' },
+    needs: ['build'],
+  })
+}
+
+// The snapshot suite boots the example bins in `lib` mode (built artifact under plain Node,
+// plugins via real exports) — CI and pre-push already build, so they exercise what ships rather
+// than the tsx/source path dev uses. It therefore waits on `build`.
+function snapshotGate(): Gate {
+  return pnpmScript('snapshot', 'test:snapshot', {
+    env: { DSH_EXAMPLE_MODE: 'lib' },
+    needs: ['build'],
   })
 }
 
@@ -275,9 +321,15 @@ function hygieneLeafGates(options: { artifactNeeds?: string[] } = {}): Gate[] {
   ]
 }
 
-function docSyncLeafGates(): Gate[] {
+function docSyncLeafGates(options: {
+  docTypecheckNeeds?: string[]
+  docTypecheckEnv?: Record<string, string | undefined>
+} = {}): Gate[] {
+  const docTypecheckOptions: Partial<Gate> = {}
+  if (options.docTypecheckNeeds !== undefined) docTypecheckOptions.needs = options.docTypecheckNeeds
+  if (options.docTypecheckEnv !== undefined) docTypecheckOptions.env = options.docTypecheckEnv
   return [
-    pnpmScript('doc-typecheck', 'doc-typecheck'),
+    pnpmScript('doc-typecheck', 'doc-typecheck', docTypecheckOptions),
     pnpmScript('cordis-catalog', 'verify-cordis-catalog', { label: 'cordis catalog' }),
     pnpmScript('export-jsdoc', 'verify-export-jsdoc', { label: 'export jsdoc' }),
     pnpmScript('tool-catalog', 'verify-tool-catalog', { label: 'tool catalog' }),
@@ -294,6 +346,7 @@ function docSyncLeafGates(): Gate[] {
     pnpmScript('rfc-classification', 'verify-rfc-classification', { label: 'rfc classification' }),
     pnpmScript('rfc-format', 'verify-rfc-format', { label: 'rfc format' }),
     pnpmScript('type-equivalence', 'verify-type-equiv', { label: 'type equivalence' }),
+    pnpmScript('translation-prompt', 'verify-translation-prompt', { label: 'translation prompt' }),
     pnpmScript('translation-pairing', 'verify-translation-pairing', { label: 'translation pairing' }),
     pnpmScript('doc-budgets', 'verify-doc-budgets', { label: 'doc budgets' }),
     pnpmScript('docs-site', 'docs:check', { label: 'documentation site' }),
@@ -306,6 +359,7 @@ function demoSmokeGate(options: { needs?: string[] } = {}): Gate {
   return {
     id: 'demo-smoke',
     label: 'demo smoke',
+    displayCommand: 'pnpm run demo:echo',
     ...pnpmInvocation(['run', 'demo:echo']),
     input: 'echo ci smoke\n',
     ...dependencyOptions,
@@ -382,6 +436,7 @@ async function runGates(allGates: Gate[], maxActive: number): Promise<GateResult
           durationMs: 0,
           stdout: '',
           stderr: '',
+          output: [],
           exitCode: null,
           error: `dependency failed or skipped: ${failedDeps.join(', ')}`,
         }
@@ -416,8 +471,10 @@ async function runGate(gate: Gate): Promise<GateResult> {
   const started = performance.now()
   let stdout = ''
   let stderr = ''
+  const output: GateOutputChunk[] = []
+  let spawnError: string | undefined
 
-  const exitCode = await new Promise<number | null>((resolveExit, reject) => {
+  const exitCode = await new Promise<number | null>((resolveExit) => {
     const child = spawn(gate.command, gate.args, {
       cwd: root,
       env: { ...process.env, ...gate.env },
@@ -425,19 +482,28 @@ async function runGate(gate: Gate): Promise<GateResult> {
     })
     child.stdout.setEncoding('utf8')
     child.stderr.setEncoding('utf8')
-    child.stdout.on('data', (chunk: string) => { stdout += chunk })
-    child.stderr.on('data', (chunk: string) => { stderr += chunk })
-    child.on('error', reject)
+    child.stdout.on('data', (chunk: string) => {
+      stdout += chunk
+      output.push({ stream: 'stdout', text: chunk })
+    })
+    child.stderr.on('data', (chunk: string) => {
+      stderr += chunk
+      output.push({ stream: 'stderr', text: chunk })
+    })
+    child.on('error', (error) => {
+      spawnError = `failed to start command: ${error.message}`
+      resolveExit(null)
+    })
     child.on('close', resolveExit)
     if (gate.input !== undefined) child.stdin.end(gate.input)
     else child.stdin.end()
   })
 
-  let status: GateStatus = exitCode === 0 ? 'passed' : 'failed'
-  let error: string | undefined
+  let status: GateStatus = exitCode === 0 && spawnError === undefined ? 'passed' : 'failed'
+  let error = spawnError
   if (status === 'passed' && gate.verify !== undefined) {
     try {
-      await gate.verify({ gate, status, durationMs: performance.now() - started, stdout, stderr, exitCode })
+      await gate.verify({ gate, status, durationMs: performance.now() - started, stdout, stderr, output, exitCode })
     } catch (verifyError: unknown) {
       status = 'failed'
       error = verifyError instanceof Error ? verifyError.message : String(verifyError)
@@ -450,6 +516,7 @@ async function runGate(gate: Gate): Promise<GateResult> {
     durationMs: performance.now() - started,
     stdout,
     stderr,
+    output,
     exitCode,
   }
   if (error !== undefined) result.error = error
@@ -458,9 +525,16 @@ async function runGate(gate: Gate): Promise<GateResult> {
 
 function printResult(result: GateResult): void {
   const seconds = (result.durationMs / 1000).toFixed(2)
-  console.log(`\n== ${result.status.toUpperCase()} ${result.gate.label} (${seconds}s) ==`)
-  process.stdout.write(result.stdout)
-  process.stderr.write(result.stderr)
+  if (result.status === 'passed' && !verbose) {
+    console.log(`run-gates: PASS ${result.gate.label} (${seconds}s)`)
+    return
+  }
+
+  const heading = `${result.status.toUpperCase()} ${result.gate.label} (${seconds}s)`
+  const writeHeading = result.status === 'passed' ? console.log : console.error
+  writeHeading(`\n== ${heading} ==`)
+  if (result.status !== 'passed') console.error(`command: ${result.gate.displayCommand}`)
+  printOutput(result.output)
   if (result.error !== undefined) console.error(result.error)
 }
 
@@ -470,4 +544,22 @@ function printSummary(results: GateResult[], durationMs: number): void {
   const skipped = results.filter(result => result.status === 'skipped').length
   const seconds = (durationMs / 1000).toFixed(2)
   console.log(`\nrun-gates: ${passed} passed, ${failed} failed, ${skipped} skipped in ${seconds}s.`)
+
+  const unsuccessful = results.filter(result => result.status === 'failed' || result.status === 'skipped')
+  if (unsuccessful.length === 0) return
+
+  console.error('run-gates: unsuccessful gates:')
+  for (const result of unsuccessful) {
+    const duration = (result.durationMs / 1000).toFixed(2)
+    const reason = result.error ?? (result.exitCode === null ? 'no exit code' : `exit ${result.exitCode}`)
+    console.error(`  - ${result.status.toUpperCase()} ${result.gate.label} (${duration}s, ${reason})`)
+    console.error(`    ${result.gate.displayCommand}`)
+  }
+}
+
+function printOutput(output: GateOutputChunk[]): void {
+  for (const chunk of output) {
+    if (chunk.stream === 'stdout') process.stdout.write(chunk.text)
+    else process.stderr.write(chunk.text)
+  }
 }
diff --git a/scripts/smoke-python-runtime.py b/scripts/smoke-python-runtime.py
index be9d36a6cd..643716515f 100644
--- a/scripts/smoke-python-runtime.py
+++ b/scripts/smoke-python-runtime.py
@@ -57,6 +57,7 @@ CUSTOM_CORDIS = """\
 - id: agent-core
   name: '@deepseek-ai/dsh-agent-spine-demo'
   config:
+    workspaceContext: false
     tools:
       mode: both
 - id: sessions
@@ -379,6 +380,7 @@ def smoke_sdk_default(base_url: str) -> None:
         root = Path(temporary).resolve()
         sessions = root / "sessions"
         with DeepSeekHarness(
+            provider="deepseek",
             model="smoke-model",
             cwd=str(root),
             session_root=str(sessions),
@@ -401,6 +403,7 @@ def smoke_sdk_custom(base_url: str, executable: Path) -> None:
         cordis = root / "cordis.yml"
         cordis.write_text(CUSTOM_CORDIS)
         with DeepSeekHarness(
+            provider="deepseek",
             model="smoke-model",
             cwd=str(root),
             session_root=str(sessions),
@@ -432,6 +435,7 @@ def smoke_sdk_snapshot(base_url: str, executable: Path, update_snapshots: bool)
         cordis = root / "cordis.yml"
         cordis.write_text(CUSTOM_CORDIS)
         with DeepSeekHarness(
+            provider="deepseek",
             model="smoke-model",
             cwd=str(root),
             session_root=str(sessions),
@@ -481,7 +485,7 @@ def smoke_direct(base_url: str, executable: Path) -> None:
         }
         peer = RuntimePeer([str(executable)], root, environment)
         try:
-            peer.send({"jsonrpc": "2.0", "id": "initialize", "method": "initialize", "params": {"cwd": str(root), "model": "smoke-model"}})
+            peer.send({"jsonrpc": "2.0", "id": "initialize", "method": "initialize", "params": {"cwd": str(root), "provider": "deepseek", "model": "smoke-model"}})
             peer.read_until(lambda message: message.get("id") == "initialize")
             peer.send({
                 "jsonrpc": "2.0",
@@ -703,7 +707,7 @@ def normalize_snapshot_value(
 
 
 def scrub_snapshot_header(value: dict[object, object]) -> None:
-    """Tokenize request-header bulk while retaining delta tool names."""
+    """Tokenize full request-header bulk while retaining tool names."""
     data = value.get("data")
     if not isinstance(data, dict):
         return
@@ -721,29 +725,6 @@ def scrub_snapshot_header(value: dict[object, object]) -> None:
             ]
         if isinstance(header.get("messagePrefix"), list):
             header["messagePrefix"] = ["{{messagePrefix}}" for _ in header["messagePrefix"]]
-        return
-    if value.get("type") != "request/header-delta":
-        return
-    system = data.get("system")
-    if isinstance(system, dict) and isinstance(system.get("insert"), list):
-        system["insert"] = ["{{system}}" for _ in system["insert"]]
-    tools = data.get("tools")
-    if isinstance(tools, dict):
-        for key in ("added", "changed"):
-            if isinstance(tools.get(key), list):
-                tools[key] = [scrub_snapshot_tool_schema(tool) for tool in tools[key]]
-    if isinstance(data.get("messagePrefix"), list):
-        data["messagePrefix"] = ["{{messagePrefix}}" for _ in data["messagePrefix"]]
-
-
-def scrub_snapshot_tool_schema(value: object) -> object:
-    """Keep a changed tool's name while tokenizing its schema bulk."""
-    if not isinstance(value, dict):
-        return value
-    return {
-        key: item if key == "name" else "{{tools}}"
-        for key, item in value.items()
-    }
 
 
 def render_jsonl(records: list[object]) -> str:
diff --git a/scripts/snapshots/python-sdk-single-exe/advanced/result.json b/scripts/snapshots/python-sdk-single-exe/advanced/result.json
index ca89b2d9a9..499ddafdf8 100644
--- a/scripts/snapshots/python-sdk-single-exe/advanced/result.json
+++ b/scripts/snapshots/python-sdk-single-exe/advanced/result.json
@@ -253,7 +253,7 @@
             "workflow"
           ]
         },
-        "reason": "fallback"
+        "reason": "change"
       }
     },
     {
@@ -915,22 +915,29 @@
       }
     },
     {
-      "type": "request/header-delta",
+      "type": "request/header",
       "seq": 56,
       "time": 0,
       "data": {
-        "system": {
-          "keepStart": 62,
-          "keepEnd": 34,
-          "insert": []
+        "header": {
+          "config": {
+            "model": "smoke-model"
+          },
+          "system": "{{system}}",
+          "tools": [
+            "bash",
+            "bash_kill",
+            "bash_output",
+            "cordis_inspect",
+            "cordis_mount",
+            "cordis_unmount",
+            "run_code",
+            "skill",
+            "subagent",
+            "workflow"
+          ]
         },
-        "tools": {
-          "added": [],
-          "removed": [
-            "snapshot_double"
-          ],
-          "changed": []
-        }
+        "reason": "change"
       }
     },
     {
@@ -1396,7 +1403,7 @@
                 "workflow"
               ]
             },
-            "reason": "fallback"
+            "reason": "change"
           }
         }
       }
@@ -2358,22 +2365,29 @@
       "payload": {
         "sessionId": "{{parent}}",
         "event": {
-          "type": "request/header-delta",
+          "type": "request/header",
           "seq": 56,
           "time": 0,
           "data": {
-            "system": {
-              "keepStart": 62,
-              "keepEnd": 34,
-              "insert": []
+            "header": {
+              "config": {
+                "model": "smoke-model"
+              },
+              "system": "{{system}}",
+              "tools": [
+                "bash",
+                "bash_kill",
+                "bash_output",
+                "cordis_inspect",
+                "cordis_mount",
+                "cordis_unmount",
+                "run_code",
+                "skill",
+                "subagent",
+                "workflow"
+              ]
             },
-            "tools": {
-              "added": [],
-              "removed": [
-                "snapshot_double"
-              ],
-              "changed": []
-            }
+            "reason": "change"
           }
         }
       }
diff --git a/scripts/snapshots/python-sdk-single-exe/advanced/session.jsonl b/scripts/snapshots/python-sdk-single-exe/advanced/session.jsonl
index 82fb190727..3d09295028 100644
--- a/scripts/snapshots/python-sdk-single-exe/advanced/session.jsonl
+++ b/scripts/snapshots/python-sdk-single-exe/advanced/session.jsonl
@@ -13,7 +13,7 @@
 {"type":"tool/result","seq":11,"time":0,"data":{"turn":1,"step":1,"callId":"advanced-mount","content":[{"type":"text","text":"mounted dyn-1 (plugin \"<anonymous>\", state: active)"}],"isError":false},"sourceEventSeqs":[10],"surfaceOp":"append"}
 {"type":"step/end","seq":12,"time":0,"data":{"turn":1,"step":1}}
 {"type":"step/start","seq":13,"time":0,"data":{"turn":1,"step":2}}
-{"type":"request/header","seq":14,"time":0,"data":{"header":{"config":{"model":"smoke-model"},"system":"{{system}}","tools":["bash","bash_kill","bash_output","cordis_inspect","cordis_mount","cordis_unmount","run_code","skill","snapshot_double","subagent","workflow"]},"reason":"fallback"}}
+{"type":"request/header","seq":14,"time":0,"data":{"header":{"config":{"model":"smoke-model"},"system":"{{system}}","tools":["bash","bash_kill","bash_output","cordis_inspect","cordis_mount","cordis_unmount","run_code","skill","snapshot_double","subagent","workflow"]},"reason":"change"}}
 {"type":"assistant/chunk","seq":15,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}
 {"type":"assistant/chunk","seq":16,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"tool-call-delta","index":0,"id":"advanced-code","name":"run_code","argumentsDelta":"{\"code\": \"return await tools.snapshot_double({ value: 21 })\"}"}}}
 {"type":"assistant/chunk","seq":17,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\": \"return await tools.snapshot_double({ value: 21 })\"}"}}}}
@@ -55,7 +55,7 @@
 {"type":"tool/result","seq":53,"time":0,"data":{"turn":1,"step":5,"callId":"advanced-unmount","content":[{"type":"text","text":"unmounted dyn-1 (plugin \"<anonymous>\")"}],"isError":false},"sourceEventSeqs":[52],"surfaceOp":"append"}
 {"type":"step/end","seq":54,"time":0,"data":{"turn":1,"step":5}}
 {"type":"step/start","seq":55,"time":0,"data":{"turn":1,"step":6}}
-{"type":"request/header-delta","seq":56,"time":0,"data":{"system":{"keepStart":62,"keepEnd":34,"insert":[]},"tools":{"added":[],"removed":["snapshot_double"],"changed":[]}}}
+{"type":"request/header","seq":56,"time":0,"data":{"header":{"config":{"model":"smoke-model"},"system":"{{system}}","tools":["bash","bash_kill","bash_output","cordis_inspect","cordis_mount","cordis_unmount","run_code","skill","subagent","workflow"]},"reason":"change"}}
 {"type":"assistant/chunk","seq":57,"time":0,"data":{"turn":1,"step":6,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}
 {"type":"assistant/chunk","seq":58,"time":0,"data":{"turn":1,"step":6,"chunk":{"type":"text-delta","index":0,"text":"ADVANCED_EXECUTABLE_OK"}}}
 {"type":"assistant/chunk","seq":59,"time":0,"data":{"turn":1,"step":6,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"ADVANCED_EXECUTABLE_OK"}}}}
diff --git a/scripts/translation-pairing.spec.ts b/scripts/translation-pairing.spec.ts
new file mode 100644
index 0000000000..39051969fc
--- /dev/null
+++ b/scripts/translation-pairing.spec.ts
@@ -0,0 +1,95 @@
+/** Regression tests for the bilingual cutoff and structural signature. */
+
+import { describe, expect, it } from 'vitest'
+import {
+  datedDocumentDate,
+  isIsoDate,
+  parseTranslationMarkdown,
+  parseTranslationPairingManifest,
+  requiresPairByDate,
+  translationStructureDiff,
+  translationStructureSignature,
+} from './translation-pairing.ts'
+
+function signature(markdown: string) {
+  return translationStructureSignature(parseTranslationMarkdown(markdown), 'counterpart.zh.md')
+}
+
+describe('translation pairing manifest', () => {
+  it('accepts a real ISO cutoff and string-array fields', () => {
+    expect(parseTranslationPairingManifest(JSON.stringify({
+      requiredSince: '2026-07-14',
+      required: ['README.md'],
+      excluded: ['docs/generated/'],
+    }))).toEqual({
+      requiredSince: '2026-07-14',
+      required: ['README.md'],
+      excluded: ['docs/generated/'],
+    })
+  })
+
+  it.each(['2026-7-14', '2026-02-29', '2026-13-01', 'not-a-date'])('rejects invalid cutoff %s', (cutoff) => {
+    expect(isIsoDate(cutoff)).toBe(false)
+    expect(() => parseTranslationPairingManifest(JSON.stringify({
+      requiredSince: cutoff,
+      required: [],
+      excluded: [],
+    }))).toThrow('requiredSince must be a valid YYYY-MM-DD date')
+  })
+
+  it('rejects non-string manifest arrays', () => {
+    expect(() => parseTranslationPairingManifest(JSON.stringify({
+      requiredSince: '2026-07-14',
+      required: [42],
+      excluded: [],
+    }))).toThrow('required must be an array of strings')
+  })
+})
+
+describe('date-based pairing frontier', () => {
+  const cutoff = '2026-07-14'
+
+  it('enforces the cutoff day and every later day, but not the preceding day', () => {
+    expect(requiresPairByDate('docs/rfc/2026-07-13-before.md', cutoff)).toBe(false)
+    expect(requiresPairByDate('docs/rfc/2026-07-14-at-cutoff.md', cutoff)).toBe(true)
+    expect(requiresPairByDate('docs/rfc/2026-07-15-after.md', cutoff)).toBe(true)
+  })
+
+  it('matches only a date at the start of the basename', () => {
+    expect(datedDocumentDate('docs/rfc/2026-07-14-proposal.md')).toBe('2026-07-14')
+    expect(datedDocumentDate('docs/release-notes-2026-07-14-alpha.md')).toBeUndefined()
+    expect(requiresPairByDate('docs/release-notes-2026-07-14-alpha.md', cutoff)).toBe(false)
+  })
+})
+
+describe('translation structural signature', () => {
+  it('accepts matching list kinds, starts, and item counts', () => {
+    const source = signature('3. One\n4. Two\n\n- A\n- B\n')
+    const counterpart = signature('3. 一\n4. 二\n\n- 甲\n- 乙\n')
+    expect(translationStructureDiff(source, counterpart)).toEqual([])
+  })
+
+  it('rejects an altered ordered-list start', () => {
+    const source = signature('3. One\n4. Two\n\n- A\n- B\n')
+    const counterpart = signature('1. 一\n2. 二\n\n- 甲\n- 乙\n')
+    expect(translationStructureDiff(source, counterpart)).toEqual([
+      'list (kind, start, item count) #1 diverges between the pair: "ordered:start=3:items=2" vs "ordered:start=1:items=2"',
+    ])
+  })
+
+  it('rejects a missing list item', () => {
+    const source = signature('- A\n- B\n')
+    const counterpart = signature('- 甲\n')
+    expect(translationStructureDiff(source, counterpart)).toEqual([
+      'list (kind, start, item count) #1 diverges between the pair: "bullet:items=2" vs "bullet:items=1"',
+    ])
+  })
+
+  it('rejects altered table row or column counts', () => {
+    const source = signature('| A | B |\n|---|---|\n| 1 | 2 |\n| 3 | 4 |\n')
+    const counterpart = signature('| 甲 | 乙 |\n|---|---|\n| 一 | 二 |\n')
+    expect(translationStructureDiff(source, counterpart)).toEqual([
+      'table (row x column count) #1 diverges between the pair: "3x2" vs "2x2"',
+    ])
+  })
+})
diff --git a/scripts/translation-pairing.ts b/scripts/translation-pairing.ts
new file mode 100644
index 0000000000..b7238fa3aa
--- /dev/null
+++ b/scripts/translation-pairing.ts
@@ -0,0 +1,164 @@
+/**
+ * Pure parsing and structural helpers for the bilingual-document pairing
+ * gate. Kept separate from the CLI so cutoff and signature behavior can be
+ * regression-tested without reading or mutating the repository tree.
+ */
+
+import { fromMarkdown } from 'mdast-util-from-markdown'
+import { gfmFromMarkdown } from 'mdast-util-gfm'
+import { gfm } from 'micromark-extension-gfm'
+import type { Nodes } from 'mdast'
+
+/** Validated shape of `scripts/translation-pairing.manifest.json`. */
+export interface TranslationPairingManifest {
+  required: string[]
+  excluded: string[]
+  /** Date-named documents on or after this day must merge bilingual. */
+  requiredSince: string
+}
+
+const ISO_DATE = /^\d{4}-\d{2}-\d{2}$/
+const DATED_DOCUMENT = /(?:^|\/)(\d{4}-\d{2}-\d{2})-[^/]*\.md$/
+
+/** Whether a string names one real calendar day in canonical ISO form. */
+export function isIsoDate(value: string): boolean {
+  if (!ISO_DATE.test(value)) return false
+  const date = new Date(`${value}T00:00:00.000Z`)
+  return !Number.isNaN(date.getTime()) && date.toISOString().slice(0, 10) === value
+}
+
+/** Read one manifest string-array field or fail before enforcement starts. */
+function stringArrayField(record: Record<string, unknown>, field: 'required' | 'excluded'): string[] {
+  const value = record[field]
+  if (!Array.isArray(value)) {
+    throw new Error(`translation-pairing.manifest.json: ${field} must be an array of strings`)
+  }
+  const entries: unknown[] = value
+  if (!entries.every((entry): entry is string => typeof entry === 'string')) {
+    throw new Error(`translation-pairing.manifest.json: ${field} must be an array of strings`)
+  }
+  return entries
+}
+
+/** Parse and validate the checked-in bilingual manifest. */
+export function parseTranslationPairingManifest(content: string): TranslationPairingManifest {
+  const value: unknown = JSON.parse(content)
+  if (typeof value !== 'object' || value === null || Array.isArray(value)) {
+    throw new Error('translation-pairing.manifest.json: expected an object')
+  }
+  const record = value as Record<string, unknown>
+  const requiredSince = record.requiredSince
+  if (typeof requiredSince !== 'string' || !isIsoDate(requiredSince)) {
+    throw new Error(`translation-pairing.manifest.json: requiredSince must be a valid YYYY-MM-DD date; got ${JSON.stringify(requiredSince)}`)
+  }
+  return {
+    required: stringArrayField(record, 'required'),
+    excluded: stringArrayField(record, 'excluded'),
+    requiredSince,
+  }
+}
+
+/** Return the leading date of a `yyyy-mm-dd-*.md` basename, if present. */
+export function datedDocumentDate(file: string): string | undefined {
+  return DATED_DOCUMENT.exec(file)?.[1]
+}
+
+/** Whether a date-named document falls on or after the pairing cutoff. */
+export function requiresPairByDate(file: string, requiredSince: string): boolean {
+  const date = datedDocumentDate(file)
+  return date !== undefined && date >= requiredSince
+}
+
+/** The structural surface compared between the two sides of a pair. */
+export interface TranslationStructureSignature {
+  /** Heading depths in document order (h2 -> 2). */
+  headings: number[]
+  /** Fenced code blocks verbatim: info string plus content, in order. */
+  code: string[]
+  /** Row and column count of each table, in order. */
+  tables: string[]
+  /** Kind, ordered-list start, and direct item count of each list, in order. */
+  lists: string[]
+  /** Every link target in order; the language switcher is excluded. */
+  links: string[]
+}
+
+/** Parse Markdown with the same GFM extensions used by the pairing gate. */
+export function parseTranslationMarkdown(content: string): Nodes {
+  return fromMarkdown(content, { extensions: [gfm()], mdastExtensions: [gfmFromMarkdown()] })
+}
+
+/** Whether the tree contains a link to exactly `target`. */
+export function linksTo(tree: Nodes, target: string): boolean {
+  let found = false
+  const visit = (node: Nodes): void => {
+    if (node.type === 'link' && node.url === target) found = true
+    if ('children' in node) for (const child of node.children) visit(child)
+  }
+  visit(tree)
+  return found
+}
+
+/** Collect the ordered structural signature, skipping one switcher target. */
+export function translationStructureSignature(tree: Nodes, switcherTarget: string): TranslationStructureSignature {
+  const sig: TranslationStructureSignature = { headings: [], code: [], tables: [], lists: [], links: [] }
+  const visit = (node: Nodes): void => {
+    switch (node.type) {
+      case 'heading':
+        sig.headings.push(node.depth)
+        break
+      case 'code':
+        sig.code.push(`\`\`\`${node.lang ?? ''}${node.meta ? ` ${node.meta}` : ''}\n${node.value}`)
+        break
+      case 'table':
+        sig.tables.push(`${node.children.length}x${node.children[0]?.children.length ?? 0}`)
+        break
+      case 'list':
+        sig.lists.push(node.ordered
+          ? `ordered:start=${node.start ?? 1}:items=${node.children.length}`
+          : `bullet:items=${node.children.length}`)
+        break
+      case 'link':
+        if (node.url !== switcherTarget) sig.links.push(node.url)
+        break
+      default:
+        // Every other node kind is prose or a container, not part of the signature.
+        break
+    }
+    if ('children' in node) for (const child of node.children) visit(child)
+  }
+  visit(tree)
+  return sig
+}
+
+/** Render a signature element for an error message, truncated for readability. */
+function show(value: string | number | undefined): string {
+  if (value === undefined) return 'nothing'
+  const text = JSON.stringify(value)
+  return text.length > 72 ? `${text.slice(0, 72)}…` : text
+}
+
+/** Return the first divergence for each structural field; empty means equal. */
+export function translationStructureDiff(
+  source: TranslationStructureSignature,
+  zh: TranslationStructureSignature,
+): string[] {
+  const out: string[] = []
+  const fields: [string, (string | number)[], (string | number)[]][] = [
+    ['heading (depth)', source.headings, zh.headings],
+    ['code block', source.code, zh.code],
+    ['table (row x column count)', source.tables, zh.tables],
+    ['list (kind, start, item count)', source.lists, zh.lists],
+    ['link target', source.links, zh.links],
+  ]
+  for (const [field, sourceValues, zhValues] of fields) {
+    const length = Math.max(sourceValues.length, zhValues.length)
+    for (let index = 0; index < length; index++) {
+      if (sourceValues[index] !== zhValues[index]) {
+        out.push(`${field} #${index + 1} diverges between the pair: ${show(sourceValues[index])} vs ${show(zhValues[index])}`)
+        break
+      }
+    }
+  }
+  return out
+}
diff --git a/scripts/translation-prompt.spec.ts b/scripts/translation-prompt.spec.ts
new file mode 100644
index 0000000000..93754a79a1
--- /dev/null
+++ b/scripts/translation-prompt.spec.ts
@@ -0,0 +1,76 @@
+/** Regression tests for the executable translation prompt contract. */
+
+import { describe, expect, it } from 'vitest'
+import {
+  parseTranslationResponse,
+  renderTranslationPrompt,
+  renderTranslationResponse,
+} from './translation-prompt.ts'
+
+const document = `# Wrapper
+
+## 模板正文
+
+\`\`\`\`text
+{{source_lang}} to {{target_lang}}
+{{translation_rules}}
+{{terminology}}
+[English]({{source_filename}}) | [中文]({{source_filename_zh}})
+\`\`\`\`
+`
+
+describe('translation prompt rendering', () => {
+  it('renders every supported placeholder without recursively rewriting injected rules', () => {
+    const rendered = renderTranslationPrompt(document, {
+      sourceLanguage: 'English',
+      sourceFilename: 'guide.md',
+      translationRules: 'A literal {{source_lang}} in injected rules.',
+      terminology: '| English | 中文 |',
+    })
+    expect(rendered).toContain('English to Chinese')
+    expect(rendered).toContain('A literal {{source_lang}} in injected rules.')
+    expect(rendered).toContain('[English](guide.md) | [中文](guide.zh.md)')
+  })
+
+  it('rejects a filename whose suffix contradicts the source language', () => {
+    expect(() => renderTranslationPrompt(document, {
+      sourceLanguage: 'Chinese',
+      sourceFilename: 'guide.md',
+      translationRules: 'rules',
+      terminology: 'terms',
+    })).toThrow('does not match source language Chinese')
+  })
+
+  it('rejects malformed template placeholders before injecting rule contents', () => {
+    expect(() => renderTranslationPrompt(document.replace('{{source_lang}}', '{{source-lang}}'), {
+      sourceLanguage: 'English',
+      sourceFilename: 'guide.md',
+      translationRules: 'A literal {{source_lang}} in injected rules.',
+      terminology: '| English | 中文 |',
+    })).toThrow('template contains malformed placeholder syntax')
+  })
+})
+
+describe('translation response XML', () => {
+  it('round-trips Markdown and the CDATA terminator', () => {
+    const response = {
+      translation: '# Draft\n\nA ]]> marker.',
+      review: '- [Tone] Fixed.',
+      final: '# Final\n\nA ]]> marker.',
+    }
+    expect(parseTranslationResponse(renderTranslationResponse(response))).toEqual(response)
+  })
+
+  it('rejects missing, reordered, nested, attributed, or non-CDATA children', () => {
+    expect(() => parseTranslationResponse('<dsh-translation-response version="1"/>')).toThrow('translation, review, and final')
+    expect(() => parseTranslationResponse('<dsh-translation-response version="1"><review><![CDATA[x]]></review></dsh-translation-response>'))
+      .toThrow('expected translation, got review')
+    expect(() => parseTranslationResponse(renderTranslationResponse({ translation: 'x', review: 'y', final: 'z' })
+      .replace('<translation><![CDATA[x]]></translation>', '<translation><b><![CDATA[x]]></b></translation>')))
+      .toThrow('nested element b is not allowed')
+    expect(() => parseTranslationResponse(renderTranslationResponse({ translation: 'x', review: 'y', final: 'z' }).replace('<review>', '<review lang="en">')))
+      .toThrow('review must not have attributes')
+    expect(() => parseTranslationResponse(renderTranslationResponse({ translation: 'x', review: 'y', final: 'z' }).replace('<![CDATA[x]]>', 'x')))
+      .toThrow('all response field content must be inside CDATA')
+  })
+})
diff --git a/scripts/translation-prompt.ts b/scripts/translation-prompt.ts
new file mode 100644
index 0000000000..e30c962498
--- /dev/null
+++ b/scripts/translation-prompt.ts
@@ -0,0 +1,171 @@
+/**
+ * Executable renderer and strict response parser for the committed
+ * documentation-translation prompt contract.
+ */
+
+import { basename } from 'node:path'
+import { SaxesParser } from 'saxes'
+
+/** Placeholder names supported by the committed translation prompt. */
+export const TRANSLATION_PROMPT_PLACEHOLDERS = [
+  'source_lang',
+  'target_lang',
+  'translation_rules',
+  'terminology',
+  'source_filename',
+  'source_filename_zh',
+] as const
+
+type TranslationPromptPlaceholder = (typeof TRANSLATION_PROMPT_PLACEHOLDERS)[number]
+
+/** Languages accepted by the bidirectional prompt. */
+type TranslationLanguage = 'English' | 'Chinese'
+
+/** Inputs that vary for one rendered translation request. */
+export interface TranslationPromptInput {
+  sourceLanguage: TranslationLanguage
+  /** Source basename, including `.md` or `.zh.md`. */
+  sourceFilename: string
+  /** Complete current `translation-rules.md` contents. */
+  translationRules: string
+  /** Complete current `terminology.md` contents. */
+  terminology: string
+}
+
+/** Parsed contents of the three-element XML response. */
+export interface TranslationResponse {
+  translation: string
+  review: string
+  final: string
+}
+
+const PLACEHOLDER = /{{([a-z_]+)}}/g
+const TEMPLATE_OPEN = '## 模板正文\n\n````text\n'
+const TEMPLATE_CLOSE = '\n````'
+const RESPONSE_CHILDREN = ['translation', 'review', 'final'] as const
+
+/** Extract the machine-consumed text fence from `translation-prompt.md`. */
+function extractTranslationPrompt(document: string): string {
+  const start = document.indexOf(TEMPLATE_OPEN)
+  if (start === -1) throw new Error('translation prompt: missing `## 模板正文` text fence')
+  const contentStart = start + TEMPLATE_OPEN.length
+  const end = document.indexOf(TEMPLATE_CLOSE, contentStart)
+  if (end === -1) throw new Error('translation prompt: missing closing four-backtick fence')
+  return document.slice(contentStart, end)
+}
+
+/** Read the placeholder names documented in the prompt's contract table. */
+export function documentedTranslationPromptPlaceholders(document: string): string[] {
+  const preambleEnd = document.indexOf(TEMPLATE_OPEN)
+  if (preambleEnd === -1) throw new Error('translation prompt: missing template body')
+  return [...document.slice(0, preambleEnd).matchAll(/^\| `{{([a-z_]+)}}` \|/gm)].map(match => match[1] ?? '')
+}
+
+/** Render one system prompt from the checked-in template and canonical rules. */
+export function renderTranslationPrompt(document: string, input: TranslationPromptInput): string {
+  if (basename(input.sourceFilename) !== input.sourceFilename) {
+    throw new Error(`translation prompt: sourceFilename must be a basename; got ${JSON.stringify(input.sourceFilename)}`)
+  }
+  const sourceIsChinese = input.sourceFilename.endsWith('.zh.md')
+  if (input.sourceLanguage === 'Chinese' ? !sourceIsChinese : sourceIsChinese || !input.sourceFilename.endsWith('.md')) {
+    throw new Error(`translation prompt: ${input.sourceFilename} does not match source language ${input.sourceLanguage}`)
+  }
+
+  const targetLanguage: TranslationLanguage = input.sourceLanguage === 'English' ? 'Chinese' : 'English'
+  const sourceFilenameZh = sourceIsChinese ? input.sourceFilename : input.sourceFilename.replace(/\.md$/, '.zh.md')
+  const values: Record<TranslationPromptPlaceholder, string> = {
+    source_lang: input.sourceLanguage,
+    target_lang: targetLanguage,
+    translation_rules: input.translationRules,
+    terminology: input.terminology,
+    source_filename: input.sourceFilename,
+    source_filename_zh: sourceFilenameZh,
+  }
+  const template = extractTranslationPrompt(document)
+  const placeholderFreeTemplate = template.replace(PLACEHOLDER, '')
+  if (placeholderFreeTemplate.includes('{{') || placeholderFreeTemplate.includes('}}')) {
+    throw new Error('translation prompt: template contains malformed placeholder syntax')
+  }
+  const names = [...template.matchAll(PLACEHOLDER)].map(match => match[1] ?? '')
+  const unknown = names.filter(name => !TRANSLATION_PROMPT_PLACEHOLDERS.includes(name as TranslationPromptPlaceholder))
+  if (unknown.length > 0) throw new Error(`translation prompt: unsupported placeholder(s): ${[...new Set(unknown)].join(', ')}`)
+  const missing = TRANSLATION_PROMPT_PLACEHOLDERS.filter(name => !names.includes(name))
+  if (missing.length > 0) throw new Error(`translation prompt: template does not use required placeholder(s): ${missing.join(', ')}`)
+
+  return template.replace(PLACEHOLDER, (_token, name: string) => values[name as TranslationPromptPlaceholder])
+}
+
+/** Escape one value so it remains byte-identical inside an XML CDATA field. */
+function escapeTranslationCdata(value: string): string {
+  return value.replaceAll(']]>', ']]]]><![CDATA[>')
+}
+
+/** Serialize a response using the exact XML wire contract in the prompt. */
+export function renderTranslationResponse(response: TranslationResponse): string {
+  return [
+    '<dsh-translation-response version="1">',
+    `<translation><![CDATA[${escapeTranslationCdata(response.translation)}]]></translation>`,
+    `<review><![CDATA[${escapeTranslationCdata(response.review)}]]></review>`,
+    `<final><![CDATA[${escapeTranslationCdata(response.final)}]]></final>`,
+    '</dsh-translation-response>',
+  ].join('\n')
+}
+
+/** Parse and validate the exact XML response shape emitted by the model. */
+export function parseTranslationResponse(xml: string): TranslationResponse {
+  const values: TranslationResponse = { translation: '', review: '', final: '' }
+  const stack: string[] = []
+  const cdataFields = new Set<string>()
+  let rootSeen = false
+  let childIndex = 0
+  const fail = (message: string): never => {
+    throw new Error(`translation response: ${message}`)
+  }
+  const parser = new SaxesParser({ xmlns: false })
+
+  parser.on('opentag', (tag) => {
+    if (stack.length === 0) {
+      if (rootSeen) fail('contains more than one root element')
+      if (tag.name !== 'dsh-translation-response') fail(`expected dsh-translation-response root, got ${tag.name}`)
+      const attributes = Object.keys(tag.attributes)
+      if (attributes.length !== 1 || tag.attributes.version !== '1') fail('root must have only version="1"')
+      rootSeen = true
+    } else if (stack.length === 1) {
+      const expected = RESPONSE_CHILDREN[childIndex]
+      if (tag.name !== expected) fail(`expected ${expected ?? 'no more children'}, got ${tag.name}`)
+      if (Object.keys(tag.attributes).length !== 0) fail(`${tag.name} must not have attributes`)
+      childIndex++
+    } else {
+      fail(`nested element ${tag.name} is not allowed`)
+    }
+    stack.push(tag.name)
+  })
+  parser.on('text', (value) => {
+    if (stack.length <= 1 && value.trim() === '') return
+    fail('all response field content must be inside CDATA')
+  })
+  parser.on('cdata', (value) => {
+    const field = stack.at(-1)
+    if (field === undefined || !RESPONSE_CHILDREN.includes(field as (typeof RESPONSE_CHILDREN)[number])) {
+      fail('CDATA is allowed only inside translation, review, or final')
+    }
+    const key = field as (typeof RESPONSE_CHILDREN)[number]
+    values[key] += value
+    cdataFields.add(key)
+  })
+  parser.on('closetag', (tag) => {
+    const expected = stack.pop()
+    if (expected !== tag.name) fail(`closing ${tag.name} does not match ${expected ?? 'nothing'}`)
+  })
+  parser.on('comment', () => fail('comments are not allowed'))
+  parser.on('doctype', () => fail('doctypes are not allowed'))
+  parser.on('processinginstruction', () => fail('processing instructions are not allowed'))
+  parser.on('error', error => fail(`invalid XML: ${error.message}`))
+  parser.write(xml).close()
+
+  if (childIndex !== RESPONSE_CHILDREN.length) fail('translation, review, and final must each appear exactly once and in order')
+  for (const field of RESPONSE_CHILDREN) {
+    if (!cdataFields.has(field)) fail(`${field} must contain a CDATA section`)
+  }
+  return values
+}
diff --git a/scripts/type-equiv.manifest.json b/scripts/type-equiv.manifest.json
index 51d98c659e..c16245f7cf 100644
--- a/scripts/type-equiv.manifest.json
+++ b/scripts/type-equiv.manifest.json
@@ -3,13 +3,17 @@
   "entries": [
     { "doc": "docs/core-data-structures/core.md", "symbol": "Branded", "source": "packages/util/brand/src/index.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "ContentBlockMap", "source": "packages/llm/llm/src/types.ts" },
+    { "doc": "docs/core-data-structures/core.md", "symbol": "AssistantProvenance", "source": "packages/llm/llm/src/types.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "Message", "source": "packages/llm/llm/src/types.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "MessageSourceMap", "source": "packages/llm/llm/src/types.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "FinishReasonMap", "source": "packages/llm/llm/src/types.ts" },
+    { "doc": "docs/core-data-structures/core.md", "symbol": "LlmProviderInfo", "source": "packages/llm/llm/src/types.ts" },
+    { "doc": "docs/core-data-structures/core.md", "symbol": "LlmModelInfo", "source": "packages/llm/llm/src/types.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "GenerateOptions", "source": "packages/llm/llm/src/types.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "ToolSchema", "source": "packages/llm/llm/src/types.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "LlmCallConfig", "source": "packages/llm/llm/src/call-config.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "SessionEvent", "source": "packages/core/session/src/types.ts" },
+    { "doc": "docs/core-data-structures/core.md", "symbol": "InjectOptions", "source": "packages/core/agent/src/types.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "Agent", "source": "packages/core/agent/src/types.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "HookContext", "source": "packages/core/agent/src/types.ts" },
     { "doc": "docs/core-data-structures/core.md", "symbol": "PromptDecision", "source": "packages/core/agent/src/types.ts" },
@@ -30,6 +34,10 @@
     { "doc": "docs/core-data-structures/llm-streaming.md", "symbol": "ContentBlockMap", "source": "packages/llm/llm/src/types.ts" },
     { "doc": "docs/core-data-structures/llm-streaming.md", "symbol": "AppIdentity", "source": "packages/llm/llm/src/attribution.ts" },
 
+    { "doc": "docs/core-data-structures/token-meter.md", "symbol": "TokenMeasurement", "source": "packages/llm/token-meter/src/types.ts" },
+    { "doc": "docs/core-data-structures/token-meter.md", "symbol": "TokenSurfaceNode", "source": "packages/llm/token-meter/src/types.ts" },
+
+    { "doc": "docs/core-data-structures/session.md", "symbol": "ContextEnvelope", "source": "packages/core/session/src/types.ts" },
     { "doc": "docs/core-data-structures/session.md", "symbol": "SessionEventMap", "source": "packages/core/session/src/types.ts" },
     { "doc": "docs/core-data-structures/session.md", "symbol": "EpochHeader", "source": "packages/core/session/src/types.ts" },
     { "doc": "docs/core-data-structures/session.md", "symbol": "TodoItem", "source": "packages/core/session/src/types.ts" },
@@ -39,19 +47,23 @@
     { "doc": "docs/core-data-structures/session.md", "symbol": "SurfaceEventType", "source": "packages/core/session/src/types.ts" },
     { "doc": "docs/core-data-structures/session.md", "symbol": "SurfaceOp", "source": "packages/core/session/src/types.ts" },
     { "doc": "docs/core-data-structures/session.md", "symbol": "SurfaceIntent", "source": "packages/core/session/src/types.ts" },
-    { "doc": "docs/core-data-structures/session.md", "symbol": "SurfaceNode", "source": "packages/core/session/src/surface.ts" },
     { "doc": "docs/core-data-structures/session.md", "symbol": "SurfaceFoldReplacement", "source": "packages/core/session/src/surface.ts" },
     { "doc": "docs/core-data-structures/session.md", "symbol": "SurfaceFoldResult", "source": "packages/core/session/src/surface.ts" },
 
     { "doc": "docs/core-data-structures/persistence.md", "symbol": "SessionHeader", "source": "packages/core/session/src/types.ts" },
     { "doc": "docs/core-data-structures/persistence.md", "symbol": "CreateSessionOptions", "source": "packages/core/session/src/types.ts" },
+    { "doc": "docs/core-data-structures/persistence.md", "symbol": "SessionLocation", "source": "packages/session-persistence/session-persistence/src/index.ts" },
 
     { "doc": "docs/core-data-structures/session-query.md", "symbol": "SessionEventSurface", "source": "packages/session-query/session-query/src/types.ts" },
     { "doc": "docs/core-data-structures/session-query.md", "symbol": "SessionRecord", "source": "packages/session-query/session-query/src/types.ts" },
     { "doc": "docs/core-data-structures/session-query.md", "symbol": "SessionEventRecord", "source": "packages/session-query/session-query/src/types.ts" },
+    { "doc": "docs/core-data-structures/session-query.md", "symbol": "SessionLineageNode", "source": "packages/session-query/session-query/src/types.ts" },
+    { "doc": "docs/core-data-structures/session-query.md", "symbol": "SessionLineageTrace", "source": "packages/session-query/session-query/src/types.ts" },
     { "doc": "docs/core-data-structures/session-query.md", "symbol": "SessionQueryErrorCode", "source": "packages/session-query/session-query/src/config.ts" },
     { "doc": "docs/core-data-structures/session-query.md", "symbol": "SessionEventReadRequest", "source": "packages/session-query/session-query/src/types.ts" },
     { "doc": "docs/core-data-structures/session-query.md", "symbol": "SessionEventWindow", "source": "packages/session-query/session-query/src/types.ts" },
+    { "doc": "docs/core-data-structures/session-query.md", "symbol": "SessionEventTraceRequest", "source": "packages/session-query/session-query/src/types.ts" },
+    { "doc": "docs/core-data-structures/session-query.md", "symbol": "SessionEventTrace", "source": "packages/session-query/session-query/src/types.ts" },
 
     { "doc": "docs/core-data-structures/tools.md", "symbol": "ToolDefinition", "source": "packages/core/tools/src/index.ts" },
     { "doc": "docs/core-data-structures/tools.md", "symbol": "SchemaProp", "source": "packages/core/tools/src/schema.ts" },
@@ -60,6 +72,8 @@
     { "doc": "docs/core-data-structures/tools.md", "symbol": "ToolExecutionToken", "source": "packages/core/tools/src/index.ts" },
     { "doc": "docs/core-data-structures/tools.md", "symbol": "ToolExecutionInput", "source": "packages/core/tools/src/index.ts" },
     { "doc": "docs/core-data-structures/tools.md", "symbol": "ToolExecution", "source": "packages/core/tools/src/index.ts" },
+    { "doc": "docs/core-data-structures/tools.md", "symbol": "ToolExecutionMode", "source": "packages/core/tools/src/index.ts" },
+    { "doc": "docs/core-data-structures/tools.md", "symbol": "ToolRunContext", "source": "packages/core/tools/src/index.ts" },
     { "doc": "docs/core-data-structures/tools.md", "symbol": "ToolGuard", "source": "packages/core/tools/src/index.ts" },
     { "doc": "docs/core-data-structures/tools.md", "symbol": "ToolRestriction", "source": "packages/core/tools/src/index.ts" },
     { "doc": "docs/core-data-structures/tools.md", "symbol": "ToolExecutionResult", "source": "packages/core/tools/src/index.ts" },
@@ -83,13 +97,22 @@
     { "doc": "docs/core-data-structures/approval.md", "symbol": "ApprovalPolicy", "source": "packages/ui/user-approval/src/index.ts" },
     { "doc": "docs/core-data-structures/approval.md", "symbol": "ApprovalRequest", "source": "packages/ui/user-approval/src/index.ts" },
 
+    { "doc": "docs/core-data-structures/bash.md", "symbol": "DshEnvironmentKey", "source": "packages/bash/bash/src/types.ts" },
+    { "doc": "docs/core-data-structures/bash.md", "symbol": "DshEnvironment", "source": "packages/bash/bash/src/types.ts" },
     { "doc": "docs/core-data-structures/bash.md", "symbol": "BashExecRequest", "source": "packages/bash/bash/src/types.ts" },
     { "doc": "docs/core-data-structures/bash.md", "symbol": "BashExecSpec", "source": "packages/bash/bash/src/types.ts" },
     { "doc": "docs/core-data-structures/bash.md", "symbol": "BashRunResult", "source": "packages/bash/bash/src/types.ts" },
     { "doc": "docs/core-data-structures/bash.md", "symbol": "BashSandboxInfo", "source": "packages/bash/bash/src/types.ts" },
     { "doc": "docs/core-data-structures/bash.md", "symbol": "CollectedOutput", "source": "packages/bash/bash/src/types.ts" },
-    { "doc": "docs/core-data-structures/bash.md", "symbol": "BashTask", "source": "packages/bash/bash/src/types.ts" },
-    { "doc": "docs/core-data-structures/bash.md", "symbol": "BashTaskRead", "source": "packages/bash/bash/src/types.ts" },
+    { "doc": "docs/core-data-structures/bash.md", "symbol": "BashProcess", "source": "packages/bash/bash/src/types.ts" },
+    { "doc": "docs/core-data-structures/bash.md", "symbol": "BashProcessRead", "source": "packages/bash/bash/src/types.ts" },
+
+    { "doc": "docs/core-data-structures/tasks.md", "symbol": "TaskKindMap", "source": "packages/tasks/tasks/src/types.ts" },
+    { "doc": "docs/core-data-structures/tasks.md", "symbol": "TaskStart", "source": "packages/tasks/tasks/src/types.ts" },
+    { "doc": "docs/core-data-structures/tasks.md", "symbol": "TaskHooks", "source": "packages/tasks/tasks/src/types.ts" },
+    { "doc": "docs/core-data-structures/tasks.md", "symbol": "TaskOutcome", "source": "packages/tasks/tasks/src/types.ts" },
+    { "doc": "docs/core-data-structures/tasks.md", "symbol": "TaskSnapshot", "source": "packages/tasks/tasks/src/types.ts" },
+    { "doc": "docs/core-data-structures/tasks.md", "symbol": "TaskRead", "source": "packages/tasks/tasks/src/types.ts" },
 
     { "doc": "docs/core-data-structures/sandbox.md", "symbol": "SandboxMode", "source": "packages/sandbox/sandbox/src/index.ts" },
     { "doc": "docs/core-data-structures/sandbox.md", "symbol": "ConfinedSandboxMode", "source": "packages/sandbox/sandbox/src/index.ts" },
@@ -107,6 +130,7 @@
     { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsTargetKey", "source": "packages/fs/fs/src/types.ts" },
     { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsVersion", "source": "packages/fs/fs/src/types.ts" },
     { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsInfo", "source": "packages/fs/fs/src/types.ts" },
+    { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsPathInfo", "source": "packages/fs/fs/src/types.ts" },
     { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsDirEntry", "source": "packages/fs/fs/src/types.ts" },
     { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsWriteIntent", "source": "packages/fs/fs/src/types.ts" },
     { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsWriteOutcome", "source": "packages/fs/fs/src/types.ts" },
@@ -142,6 +166,12 @@
     { "doc": "docs/core-data-structures/web.md", "symbol": "WebFetchResult", "source": "packages/web/web/src/types.ts" },
     { "doc": "docs/core-data-structures/web.md", "symbol": "WebFetchBody", "source": "packages/web/web/src/types.ts" },
 
+    { "doc": "docs/core-data-structures/spill.md", "symbol": "SaveTextSpill", "source": "packages/spill/spill/src/types.ts" },
+    { "doc": "docs/core-data-structures/spill.md", "symbol": "SpillOwner", "source": "packages/spill/spill/src/types.ts" },
+    { "doc": "docs/core-data-structures/spill.md", "symbol": "SpillSource", "source": "packages/spill/spill/src/types.ts" },
+    { "doc": "docs/core-data-structures/spill.md", "symbol": "SpillRef", "source": "packages/spill/spill/src/types.ts" },
+    { "doc": "docs/core-data-structures/spill.md", "symbol": "SpillLocator", "source": "packages/spill/spill/src/types.ts" },
+
     { "doc": "docs/core-data-structures/workflow.md", "symbol": "WorkflowStartRequest", "source": "packages/workflow/workflow/src/types.ts" },
     { "doc": "docs/core-data-structures/workflow.md", "symbol": "WorkflowMeta", "source": "packages/workflow/workflow/src/types.ts" },
     { "doc": "docs/core-data-structures/workflow.md", "symbol": "WorkflowResult", "source": "packages/workflow/workflow/src/types.ts" },
diff --git a/scripts/verify-cordis-config.ts b/scripts/verify-cordis-config.ts
index 131bcf0e58..88615ee57a 100644
--- a/scripts/verify-cordis-config.ts
+++ b/scripts/verify-cordis-config.ts
@@ -1,18 +1,32 @@
 /**
- * Reject JavaScript expressions in Cordis Loader entry metadata.
+ * Validate Cordis Loader entry metadata and example package resolution.
  *
  * The Loader interpolates only a plugin entry's `config`; expression objects in
  * fields such as `disabled` remain truthy data and silently change composition.
+ * Example configs run from built packages, so every named package must resolve
+ * from the examples workspace and every local package must be in the root
+ * TypeScript project graph.
  */
 
 import { globSync, readFileSync } from 'node:fs'
-import { resolve } from 'node:path'
+import { dirname, relative, resolve } from 'node:path'
 import * as yaml from 'js-yaml'
+import ts from 'typescript'
 
 interface JsExpr {
   __jsExpr: string
 }
 
+interface PackageManifest {
+  name?: string
+  dependencies?: Record<string, string>
+}
+
+interface PluginReference {
+  file: string
+  name: string
+}
+
 const root = resolve(import.meta.dirname, '..')
 const metadataFields = ['id', 'name', 'group', 'disabled', 'inject', 'intercept', 'isolate'] as const
 const jsExprType = new yaml.Type('tag:yaml.org,2002:js', {
@@ -30,6 +44,7 @@ const files = globSync(['**/*cordis*.yml', '**/*cordis*.yaml'], {
   exclude: ['.claude/**', 'node_modules/**', 'vendor/**'],
 }).sort()
 const errors: string[] = []
+const examplePluginReferences: PluginReference[] = []
 
 for (const file of files) {
   const document: unknown = yaml.load(readFileSync(resolve(root, file), 'utf8'), { schema })
@@ -42,8 +57,10 @@ for (const file of files) {
   }
 }
 
+errors.push(...validateExampleResolution())
+
 if (errors.length > 0) {
-  console.error('verify-cordis-config: Loader entry metadata is static; move !!js under plugin config or select an explicit overlay.')
+  console.error('verify-cordis-config: invalid Loader metadata or example package resolution:')
   for (const error of errors) console.error(`- ${error}`)
   process.exitCode = 1
 } else {
@@ -55,6 +72,7 @@ function validateEntry(value: unknown, file: string, path: string): void {
     errors.push(`${file}${path}: entry must be an object`)
     return
   }
+  recordExamplePlugin(value, file)
   validateMetadata(value, file, path)
   if ((value.group === true || value.name === '@cordisjs/plugin-group') && isUnknownArray(value.config)) {
     for (let index = 0; index < value.config.length; index++) {
@@ -68,6 +86,7 @@ function validateEntry(value: unknown, file: string, path: string): void {
     const patch = config.patches[index]
     const patchPath = `${path}.config.patches[${index}]`
     if (!isRecord(patch)) continue
+    recordExamplePlugin(patch, file)
     validateMetadata(patch, file, patchPath)
     if (!isUnknownArray(patch.insert)) continue
     for (let insertIndex = 0; insertIndex < patch.insert.length; insertIndex++) {
@@ -76,6 +95,83 @@ function validateEntry(value: unknown, file: string, path: string): void {
   }
 }
 
+function recordExamplePlugin(entry: Record<string, unknown>, file: string): void {
+  if (file.startsWith('examples/') && typeof entry.name === 'string') {
+    examplePluginReferences.push({ file, name: entry.name })
+  }
+}
+
+function validateExampleResolution(): string[] {
+  const violations: string[] = []
+  const exampleManifest = readManifest('examples/package.json')
+  const dependencies = exampleManifest.dependencies ?? {}
+  const localPackages = localPackageDirectories()
+  const rootReferences = rootProjectReferences()
+  const requiredPackages = new Map<string, Set<string>>()
+
+  for (const reference of examplePluginReferences) {
+    const packageName = packageNameFromSpecifier(reference.name)
+    if (packageName === undefined) continue
+    const locations = requiredPackages.get(packageName) ?? new Set<string>()
+    locations.add(reference.file)
+    requiredPackages.set(packageName, locations)
+  }
+
+  for (const [packageName, locations] of requiredPackages) {
+    if (!(packageName in dependencies)) {
+      violations.push(`${[...locations].join(', ')}: ${packageName} must be declared in examples/package.json dependencies`)
+    }
+  }
+
+  const localExamplePackages = new Set([
+    ...Object.keys(dependencies),
+    ...requiredPackages.keys(),
+  ])
+  for (const packageName of localExamplePackages) {
+    const packageDirectory = localPackages.get(packageName)
+    if (packageDirectory === undefined || rootReferences.has(packageDirectory)) continue
+    const repoPath = relative(root, packageDirectory).replaceAll('\\', '/')
+    violations.push(`tsconfig.json: missing project reference for ${packageName} (${repoPath})`)
+  }
+
+  return violations
+}
+
+function readManifest(path: string): PackageManifest {
+  return JSON.parse(readFileSync(resolve(root, path), 'utf8')) as PackageManifest
+}
+
+function localPackageDirectories(): Map<string, string> {
+  const manifests = globSync(['packages/*/*/package.json', 'vendor/*/package.json'], { cwd: root })
+  const packages = new Map<string, string>()
+  for (const manifestPath of manifests) {
+    const manifest = readManifest(manifestPath)
+    if (manifest.name !== undefined) packages.set(manifest.name, resolve(root, dirname(manifestPath)))
+  }
+  return packages
+}
+
+function rootProjectReferences(): Set<string> {
+  const config = ts.readConfigFile(resolve(root, 'tsconfig.json'), path => ts.sys.readFile(path))
+  if (config.error !== undefined) {
+    throw new Error(ts.flattenDiagnosticMessageText(config.error.messageText, '\n'))
+  }
+  const references = (config.config as { references?: Array<{ path?: unknown }> }).references ?? []
+  return new Set(references.flatMap((reference) => {
+    if (typeof reference.path !== 'string') return []
+    return [resolve(root, reference.path)]
+  }))
+}
+
+function packageNameFromSpecifier(specifier: string): string | undefined {
+  if (specifier.startsWith('.') || specifier.startsWith('/') || specifier.startsWith('file:')) return undefined
+  const segments = specifier.split('/')
+  if (specifier.startsWith('@')) {
+    return segments.length >= 2 ? `${segments[0]}/${segments[1]}` : undefined
+  }
+  return segments[0] || undefined
+}
+
 function validateMetadata(entry: Record<string, unknown>, file: string, path: string): void {
   for (const field of metadataFields) {
     if (!(field in entry)) continue
diff --git a/scripts/verify-export-jsdoc.ts b/scripts/verify-export-jsdoc.ts
index 44729e73f0..8ff42fa876 100644
--- a/scripts/verify-export-jsdoc.ts
+++ b/scripts/verify-export-jsdoc.ts
@@ -7,8 +7,8 @@
  * re-exports keep their docs at the declaring contract. Unknown forms fail closed.
  */
 
-import { existsSync, globSync } from 'node:fs'
-import { resolve } from 'node:path'
+import { existsSync, globSync, readFileSync } from 'node:fs'
+import { relative, resolve, sep } from 'node:path'
 import ts from 'typescript'
 import { checkParams, checkReturns, parseJsDoc, parseTags, pointer, rawJsDoc } from './jsdoc.ts'
 
@@ -389,7 +389,13 @@ function checkDecl(
  * @param w - the walk state violations append to.
  * @param ambient - whether this scope is ambient (`declare` namespace or a declaration file), where members export implicitly.
  */
-function checkScope(statements: readonly ts.Statement[], prefix: string, w: Walk, ambient: boolean): void {
+function checkScope(
+  statements: readonly ts.Statement[],
+  prefix: string,
+  w: Walk,
+  ambient: boolean,
+  allowedNames?: ReadonlySet<string>,
+): void {
   const byName = new Map<string, ts.Statement[]>()
   const overloadSigs = new Set<string>()
   const add = (name: string, stmt: ts.Statement): void => {
@@ -455,7 +461,20 @@ function checkScope(statements: readonly ts.Statement[], prefix: string, w: Walk
       }
       continue
     }
-    if (isExported(stmt) || (ambient && !ts.isImportDeclaration(stmt))) request(stmt, null)
+    if (isExported(stmt) || (ambient && !ts.isImportDeclaration(stmt))) {
+      if (allowedNames === undefined) {
+        request(stmt, null)
+      } else if (ts.isVariableStatement(stmt)) {
+        for (const declaration of stmt.declarationList.declarations) {
+          if (ts.isIdentifier(declaration.name) && allowedNames.has(declaration.name.text)) {
+            request(stmt, declaration.name.text)
+          }
+        }
+      } else {
+        const name = declarationName(stmt) ?? 'default'
+        if (allowedNames.has(name)) request(stmt, null)
+      }
+    }
   }
   for (const stmt of statements) {
     const only = requested.get(stmt)
@@ -463,6 +482,67 @@ function checkScope(statements: readonly ts.Statement[], prefix: string, w: Walk
   }
 }
 
+function exportedTargets(value: unknown): string[] {
+  if (typeof value === 'string') return [value]
+  if (!value || typeof value !== 'object') return []
+  return Object.values(value).flatMap(exportedTargets)
+}
+
+function sourceEntry(target: string): string | undefined {
+  if (target.startsWith('./lib/types/') && target.endsWith('.d.ts')) {
+    return `src/${target.slice('./lib/types/'.length, -'.d.ts'.length)}.ts`
+  }
+  if (target.startsWith('./lib/') && target.endsWith('.js')) {
+    return `src/${target.slice('./lib/'.length, -'.js'.length)}.ts`
+  }
+  return undefined
+}
+
+function declarationName(declaration: ts.Node): string | undefined {
+  const name = (declaration as ts.NamedDeclaration).name
+  if (name && ts.isIdentifier(name)) return name.text
+  return undefined
+}
+
+/** Resolve the declarations reachable through packages that do not export src/*. */
+function restrictedPublicNames(
+  scanRoot: string,
+  rels: readonly string[],
+  program: ts.Program,
+  checker: ts.TypeChecker,
+): { restrictedPackages: Set<string>; namesByFile: Map<string, Set<string>> } {
+  const restrictedPackages = new Set<string>()
+  const namesByFile = new Map<string, Set<string>>()
+  const packages = new Set(rels.map(rel => rel.split('/').slice(0, 3).join('/')))
+  for (const packageDir of packages) {
+    const manifestPath = resolve(scanRoot, packageDir, 'package.json')
+    if (!existsSync(manifestPath)) continue
+    const manifest = JSON.parse(readFileSync(manifestPath, 'utf8')) as { exports?: Record<string, unknown> }
+    if (!manifest.exports || manifest.exports['./src/*'] !== undefined) continue
+    restrictedPackages.add(packageDir)
+    const entries = new Set(Object.values(manifest.exports).flatMap(exportedTargets).flatMap((target) => {
+      const entry = sourceEntry(target)
+      return entry ? [`${packageDir}/${entry}`] : []
+    }))
+    for (const entry of entries) {
+      const source = program.getSourceFile(resolve(scanRoot, entry))
+      const moduleSymbol = source && checker.getSymbolAtLocation(source)
+      if (!source || !moduleSymbol) continue
+      for (const exported of checker.getExportsOfModule(moduleSymbol)) {
+        const target = (exported.flags & ts.SymbolFlags.Alias) !== 0 ? checker.getAliasedSymbol(exported) : exported
+        for (const declaration of target.declarations ?? []) {
+          const name = declarationName(declaration)
+          const file = declaration.getSourceFile().fileName
+          const rel = relative(scanRoot, file).split(sep).join('/')
+          if (!name || !rel.startsWith(`${packageDir}/src/`)) continue
+          namesByFile.set(rel, new Set([...(namesByFile.get(rel) ?? []), name]))
+        }
+      }
+    }
+  }
+  return { restrictedPackages, namesByFile }
+}
+
 /**
  * Compiler options for the walk's program.
  *
@@ -495,15 +575,26 @@ function loadCompilerOptions(scanRoot: string): ts.CompilerOptions {
  */
 export function collectExportJsdocViolations(scanRoot: string = root): string[] {
   const violations: string[] = []
-  const rels = globSync('packages/*/*/src/**/*.ts', { cwd: scanRoot }).sort()
+  const rels = globSync('packages/*/*/src/**/*.ts', { cwd: scanRoot })
+    .map(path => path.split(sep).join('/'))
+    .sort()
   const program = ts.createProgram(rels.map(rel => resolve(scanRoot, rel)), loadCompilerOptions(scanRoot))
   const checker = program.getTypeChecker()
+  const { restrictedPackages, namesByFile } = restrictedPublicNames(scanRoot, rels, program, checker)
   for (const rel of rels) {
     const sf = program.getSourceFile(resolve(scanRoot, rel))
     if (!sf) continue // program root files always resolve; guard for narrowing
     // A script-style declaration file (no imports/exports) is one big ambient
     // scope; a module-style .d.ts still honors explicit export modifiers.
-    checkScope(sf.statements, '', { rel, sf, text: sf.text, checker, violations }, sf.isDeclarationFile && !ts.isExternalModule(sf))
+    const packageDir = rel.split('/').slice(0, 3).join('/')
+    const allowedNames = restrictedPackages.has(packageDir) ? namesByFile.get(rel) ?? new Set<string>() : undefined
+    checkScope(
+      sf.statements,
+      '',
+      { rel, sf, text: sf.text, checker, violations },
+      sf.isDeclarationFile && !ts.isExternalModule(sf),
+      allowedNames,
+    )
   }
   return violations
 }
diff --git a/scripts/verify-package-readme-model-experience.ts b/scripts/verify-package-readme-model-experience.ts
index 5c128954f6..75fb2b75ad 100644
--- a/scripts/verify-package-readme-model-experience.ts
+++ b/scripts/verify-package-readme-model-experience.ts
@@ -30,6 +30,7 @@ interface SentenceContract {
 const NO_MODEL_EXPERIENCE_SECTION: Readonly<Record<string, string>> = {
   'packages/core/scope': 'The package is a model-agnostic registration and lifecycle primitive; model-facing consumers own any context selection.',
   'packages/util/brand': 'The package is a type-only primitive erased at compile time.',
+  'packages/util/paths': 'The package only resolves harness-owned host paths; model-facing consumers own any rendered use.',
 }
 
 /**
@@ -47,23 +48,33 @@ const SENTENCE_MODEL_EXPERIENCE: Readonly<Record<string, SentenceContract>> = {
   'packages/fs/fs-local': { kind: 'indirect', reason: 'The provider backend delegates model rendering to dsh-tool-fs.' },
   'packages/hooks/hook-protocol': { kind: 'indirect', reason: 'Only the hook bridge plugins render decoded hook output to a model.' },
   'packages/llm/llm': { kind: 'none', reason: 'The adapter registry forwards already-assembled requests unchanged.' },
+  'packages/llm/token-meter': { kind: 'indirect', reason: 'The measurement service leaves model-visible changes to its consumers.' },
   'packages/sandbox/sandbox-local': { kind: 'indirect', reason: 'The provider backend delegates model rendering to dsh-bash-sandbox and dsh-tool-bash.' },
+  'packages/sdk/create-sdk': { kind: 'indirect', reason: 'The initializer only writes project files; selected runtime plugins provide the generated project model surface.' },
+  'packages/sdk/helper': { kind: 'none', reason: 'The project domain edits files and registers no live agent or model surface.' },
+  'packages/sdk/scripts': { kind: 'indirect', reason: 'The launcher delegates model context to the loaded project plugin tree.' },
   'packages/session-query/session-query': { kind: 'none', reason: 'The trusted query service exposes cloned records only to callers and registers no model surface.' },
   'packages/skill/skill': { kind: 'indirect', reason: 'The provider registry delegates model rendering to dsh-tool-skill.' },
   'packages/skill/skill-local': { kind: 'indirect', reason: 'The provider backend delegates model rendering to dsh-tool-skill.' },
+  'packages/spill/spill': { kind: 'indirect', reason: 'The storage seam delegates model rendering to spill consumers.' },
+  'packages/spill/spill-local': { kind: 'indirect', reason: 'The storage backend delegates model rendering to spill consumers.' },
   'packages/subagent/subagent': { kind: 'indirect', reason: 'The provider registry delegates parent-model rendering to dsh-tool-subagent.' },
   'packages/subagent/subagent-subprocess': { kind: 'indirect', reason: 'Only process-based subagent backends compose a child model request.' },
   'packages/support/acp-snapshot': { kind: 'none', reason: 'The test harness observes and normalizes transcripts without changing live requests.' },
+  'packages/support/agent-loop-testkit': { kind: 'none', reason: 'The test helper mounts services but neither drives nor modifies model requests.' },
   'packages/support/invariants': { kind: 'none', reason: 'The observer validates requests but never rewrites their context.' },
   'packages/support/loader-smoke': { kind: 'none', reason: 'The test harness observes child-process streams without changing live requests.' },
   'packages/support/llm-replay': { kind: 'none', reason: 'The keyless adapter invokes no provider model.' },
   'packages/support/subagent-mock': { kind: 'indirect', reason: 'Only dsh-tool-subagent renders its configured test outcome.' },
+  'packages/tasks/tasks': { kind: 'indirect', reason: 'Producer and control-surface plugins own all model rendering over the task registry.' },
   'packages/examples/acp-demo': { kind: 'indirect', reason: 'The app bundle delegates request composition to dsh-agent-spine-demo and dsh-acp.' },
   'packages/ui/app-boot': { kind: 'indirect', reason: 'Only the loaded plugin tree contributes model context.' },
   'packages/examples/jsonrpc-demo': { kind: 'indirect', reason: 'Only the externally configured plugin tree contributes model context.' },
   'packages/ui/permission': { kind: 'indirect', reason: 'The service writes mechanism events rendered by dsh-user-approval and dsh-tool-bash.' },
   'packages/ui/user-interaction': { kind: 'indirect', reason: 'Model-facing consumers render provider answers and seam errors.' },
+  'packages/util/home': { kind: 'indirect', reason: 'Only dsh-tool-bash exposes the resolved home to model commands.' },
   'packages/util/timeout': { kind: 'indirect', reason: 'Only timeout consumers render timeout outcomes.' },
+  'packages/util/retention': { kind: 'indirect', reason: 'Only retention consumers render retained content and omission metadata.' },
   'packages/web/web': { kind: 'indirect', reason: 'The provider registry delegates model rendering to dsh-tool-web.' },
   'packages/web/web-fetch-local': { kind: 'indirect', reason: 'The provider backend delegates model rendering to dsh-tool-web.' },
   'packages/web/web-search-exa': { kind: 'indirect', reason: 'The provider backend delegates model rendering to dsh-tool-web.' },
diff --git a/scripts/verify-translation-pairing.ts b/scripts/verify-translation-pairing.ts
index 4aaf7de79b..ced55d5be1 100644
--- a/scripts/verify-translation-pairing.ts
+++ b/scripts/verify-translation-pairing.ts
@@ -10,10 +10,15 @@
 import { createHash } from 'node:crypto'
 import { existsSync, globSync, readFileSync, writeFileSync } from 'node:fs'
 import { basename, join, resolve, sep } from 'node:path'
-import { fromMarkdown } from 'mdast-util-from-markdown'
-import { gfmFromMarkdown } from 'mdast-util-gfm'
-import { gfm } from 'micromark-extension-gfm'
-import type { Nodes } from 'mdast'
+import {
+  datedDocumentDate,
+  linksTo,
+  parseTranslationMarkdown,
+  parseTranslationPairingManifest,
+  requiresPairByDate,
+  translationStructureDiff,
+  translationStructureSignature,
+} from './translation-pairing.ts'
 
 const root = resolve(import.meta.dirname, '..')
 const listMode = process.argv.includes('--list')
@@ -22,14 +27,7 @@ const writeMode = process.argv.includes('--write')
 /** Scope of the bilingual contract: the root README, the docs tree, and the Python SDK tree. */
 const SCOPE_PATTERNS = ['README.md', 'README.zh.md', 'README.i18n.yaml', 'docs/**/*.md', 'docs/**/*.i18n.yaml', 'python/**/*.md', 'python/**/*.i18n.yaml']
 
-/** The enforcement frontier and the never-paired set (docs/i18n/README.md § Scope). */
-interface Manifest {
-  required: string[]
-  excluded: string[]
-  /** Date-named documents (yyyy-mm-dd-*.md, i.e. RFCs) dated on/after this day must merge bilingual. */
-  requiredSince: string
-}
-const manifest = JSON.parse(readFileSync(join(root, 'scripts/translation-pairing.manifest.json'), 'utf8')) as Manifest
+const manifest = parseTranslationPairingManifest(readFileSync(join(root, 'scripts/translation-pairing.manifest.json'), 'utf8'))
 
 /**
  * An excluded entry ending in `/` excludes the whole directory. The trailing
@@ -81,98 +79,6 @@ function renderMeta(source: string, sourceHash: string, zh: string, zhHash: stri
   ].join('\n')
 }
 
-/**
- * The structural signature the two sides must share, as ordered sequences so
- * a swap or a level change is caught, not just a count change. Prose is
- * deliberately absent: the gate checks shape, never wording.
- */
-interface Signature {
-  /** Heading depths in document order (h2 → 2). */
-  headings: number[]
-  /** Fenced code blocks verbatim: info string + content, in order. */
-  code: string[]
-  /** Column count of each table, in order. */
-  tables: number[]
-  /** Each list's kind (ordered vs bullet), in order. */
-  lists: string[]
-  /** Every link target in order, the language switcher's excluded. */
-  links: string[]
-}
-
-/** Whether the tree contains a link to exactly `target` (the switcher check). */
-function linksTo(tree: Nodes, target: string): boolean {
-  let found = false
-  const visit = (node: Nodes): void => {
-    if (node.type === 'link' && node.url === target) found = true
-    if ('children' in node) for (const child of node.children) visit(child)
-  }
-  visit(tree)
-  return found
-}
-
-/** Collect the structural signature, skipping links to `switcherTarget`. */
-function signatureOf(tree: Nodes, switcherTarget: string): Signature {
-  const sig: Signature = { headings: [], code: [], tables: [], lists: [], links: [] }
-  const visit = (node: Nodes): void => {
-    switch (node.type) {
-      case 'heading':
-        sig.headings.push(node.depth)
-        break
-      case 'code':
-        sig.code.push(`\`\`\`${node.lang ?? ''}${node.meta ? ` ${node.meta}` : ''}\n${node.value}`)
-        break
-      case 'table':
-        sig.tables.push(node.children[0]?.children.length ?? 0)
-        break
-      case 'list':
-        sig.lists.push(node.ordered ? 'ordered' : 'bullet')
-        break
-      case 'link':
-        if (node.url !== switcherTarget) sig.links.push(node.url)
-        break
-      default:
-        // Every other node kind is prose or container — not part of the signature.
-        break
-    }
-    if ('children' in node) for (const child of node.children) visit(child)
-  }
-  visit(tree)
-  return sig
-}
-
-/** Render a signature element for an error message, truncated for readability. */
-function show(value: string | number | undefined): string {
-  if (value === undefined) return 'nothing'
-  const text = JSON.stringify(value)
-  return text.length > 72 ? `${text.slice(0, 72)}…` : text
-}
-
-/** First divergence between two signatures, as messages; empty when identical. */
-function signatureDiff(source: Signature, zh: Signature): string[] {
-  const out: string[] = []
-  const fields: [string, (string | number)[], (string | number)[]][] = [
-    ['heading (depth)', source.headings, zh.headings],
-    ['code block', source.code, zh.code],
-    ['table (column count)', source.tables, zh.tables],
-    ['list (kind)', source.lists, zh.lists],
-    ['link target', source.links, zh.links],
-  ]
-  for (const [field, s, z] of fields) {
-    const length = Math.max(s.length, z.length)
-    for (let i = 0; i < length; i++) {
-      if (s[i] !== z[i]) {
-        out.push(`${field} #${i + 1} diverges between the pair: ${show(s[i])} vs ${show(z[i])}`)
-        break
-      }
-    }
-  }
-  return out
-}
-
-function parse(content: string): Nodes {
-  return fromMarkdown(content, { extensions: [gfm()], mdastExtensions: [gfmFromMarkdown()] })
-}
-
 // Enumerate the scope once.
 const files = new Set<string>()
 for (const pattern of SCOPE_PATTERNS) {
@@ -218,14 +124,13 @@ for (const req of manifest.required) {
 // 2. Date-named documents (RFCs) dated on/after the requiredSince cutoff merge
 // bilingual: a new RFC lands with its pair or not at all. Deterministic from
 // the filename alone — no git history, so it holds on shallow CI checkouts.
-const DATED = /(?:^|\/)(\d{4}-\d{2}-\d{2})-[^/]*\.md$/
 for (const source of sources) {
   if (isExcluded(source)) continue
-  const dated = DATED.exec(source)
-  if (!dated?.[1] || dated[1] < manifest.requiredSince) continue
+  const date = datedDocumentDate(source)
+  if (!requiresPairByDate(source, manifest.requiredSince) || date === undefined) continue
   const { zh } = pairPaths(source)
   if (!existsSync(join(root, zh))) {
-    errors.push(`${source}: dated ${dated[1]} — documents dated on/after ${manifest.requiredSince} merge bilingual (docs/i18n/README.md); add the counterpart and record the pair`)
+    errors.push(`${source}: dated ${date} — documents dated on/after ${manifest.requiredSince} merge bilingual (docs/i18n/README.md); add the counterpart and record the pair`)
     state.set(source, 'missing')
   }
 }
@@ -273,15 +178,18 @@ for (const source of [...pairAnchors].sort()) {
     continue
   }
 
-  const sourceTree = parse(sourceContent.toString('utf8'))
-  const zhTree = parse(zhContent.toString('utf8'))
+  const sourceTree = parseTranslationMarkdown(sourceContent.toString('utf8'))
+  const zhTree = parseTranslationMarkdown(zhContent.toString('utf8'))
   if (!linksTo(zhTree, basename(source))) {
     errors.push(`${zh}: missing language switcher — no link to ${basename(source)}`)
   }
   if (!linksTo(sourceTree, basename(zh))) {
     errors.push(`${source}: missing language switcher — no link back to ${basename(zh)}`)
   }
-  for (const divergence of signatureDiff(signatureOf(sourceTree, basename(zh)), signatureOf(zhTree, basename(source)))) {
+  for (const divergence of translationStructureDiff(
+    translationStructureSignature(sourceTree, basename(zh)),
+    translationStructureSignature(zhTree, basename(source)),
+  )) {
     errors.push(`${source} ↔ ${zh}: ${divergence}`)
   }
   if (!state.has(source)) state.set(source, 'ok')
@@ -297,8 +205,7 @@ if (listMode) {
   const rows = [...state.entries()].sort((a, b) => order[a[1]] - order[b[1]] || a[0].localeCompare(b[0]))
   for (const [file, status] of rows) {
     const required = manifest.required.includes(file)
-    const date = DATED.exec(file)?.[1]
-    const tag = required ? '  (required)' : date && date >= manifest.requiredSince ? '  (required by date)' : '  (backlog)'
+    const tag = required ? '  (required)' : requiresPairByDate(file, manifest.requiredSince) ? '  (required by date)' : '  (backlog)'
     console.log(`${status.padEnd(11)} ${file}${status === 'missing' ? tag : ''}`)
   }
   const counts = { 'ok': 0, 'out-of-sync': 0, 'missing': 0 }
diff --git a/scripts/verify-translation-prompt.ts b/scripts/verify-translation-prompt.ts
new file mode 100644
index 0000000000..66d83e47ad
--- /dev/null
+++ b/scripts/verify-translation-prompt.ts
@@ -0,0 +1,56 @@
+/** Verify that the committed translation prompt renders and parses as documented. */
+
+import { readFileSync } from 'node:fs'
+import { join, resolve } from 'node:path'
+import {
+  documentedTranslationPromptPlaceholders,
+  parseTranslationResponse,
+  renderTranslationPrompt,
+  renderTranslationResponse,
+  TRANSLATION_PROMPT_PLACEHOLDERS,
+} from './translation-prompt.ts'
+
+const root = resolve(import.meta.dirname, '..')
+
+function read(path: string): string {
+  return readFileSync(join(root, path), 'utf8')
+}
+
+try {
+  const document = read('docs/i18n/translation-prompt.md')
+  const translationRules = read('docs/i18n/translation-rules.md')
+  const terminology = read('docs/i18n/terminology.md')
+  const documented = documentedTranslationPromptPlaceholders(document)
+  if (documented.join('\n') !== TRANSLATION_PROMPT_PLACEHOLDERS.join('\n')) {
+    throw new Error(`placeholder table must list exactly: ${TRANSLATION_PROMPT_PLACEHOLDERS.join(', ')}`)
+  }
+
+  const englishSource = renderTranslationPrompt(document, {
+    sourceLanguage: 'English',
+    sourceFilename: 'example.md',
+    translationRules,
+    terminology,
+  })
+  const chineseSource = renderTranslationPrompt(document, {
+    sourceLanguage: 'Chinese',
+    sourceFilename: 'example.zh.md',
+    translationRules,
+    terminology,
+  })
+  if (!englishSource.includes('[English](example.md) | 中文')) throw new Error('English-source render does not carry the Chinese switcher instruction')
+  if (!chineseSource.includes('English | [中文](example.zh.md)')) throw new Error('Chinese-source render does not carry the English switcher instruction')
+
+  const example = /```xml\n([\s\S]*?)\n```/.exec(englishSource)?.[1]
+  if (example === undefined) throw new Error('rendered prompt has no XML response example')
+  parseTranslationResponse(example)
+
+  const roundTrip = { translation: 'first ]]> pass', review: '- [None] No corrections.', final: 'final ]]> text' }
+  const parsed = parseTranslationResponse(renderTranslationResponse(roundTrip))
+  if (JSON.stringify(parsed) !== JSON.stringify(roundTrip)) throw new Error('CDATA split rule does not round-trip response content')
+
+  console.log('verify-translation-prompt: both directions render and the XML response contract parses.')
+} catch (error) {
+  const message = error instanceof Error ? error.message : String(error)
+  console.error(`verify-translation-prompt: ${message}`)
+  process.exit(1)
+}
diff --git a/tsconfig.base.json b/tsconfig.base.json
index 74696dc2d9..53f69b2cf5 100644
--- a/tsconfig.base.json
+++ b/tsconfig.base.json
@@ -41,6 +41,7 @@
       // explicit — TS project references have no wildcard form.
       "@deepseek-ai/dsh-*": [
         "./packages/core/*/src",
+        "./packages/prompt/*/src",
         "./packages/llm/*/src",
         "./packages/bash/*/src",
         "./packages/code-runtime/*/src",
@@ -50,8 +51,10 @@
         "./packages/context/*/src",
         "./packages/guard/*/src",
         "./packages/subagent/*/src",
+        "./packages/tasks/*/src",
         "./packages/workflow/*/src",
         "./packages/web/*/src",
+        "./packages/spill/*/src",
         "./packages/timeout/*/src",
         "./packages/todo/*/src",
         "./packages/cordis/*/src",
@@ -59,6 +62,7 @@
         "./packages/hooks/*/src",
         "./packages/session-persistence/*/src",
         "./packages/session-query/*/src",
+        "./packages/sdk/*/src",
         "./packages/ui/*/src",
         "./packages/examples/*/src",
         "./packages/util/*/src",
diff --git a/tsconfig.build.json b/tsconfig.build.json
index 05725fe0ae..dd13e018c3 100644
--- a/tsconfig.build.json
+++ b/tsconfig.build.json
@@ -11,8 +11,12 @@
     { "path": "./vendor/hmr" },
     { "path": "./vendor/logger-console" },
     { "path": "./packages/util/brand" },
+    { "path": "./packages/util/home" },
+    { "path": "./packages/util/paths" },
     { "path": "./packages/util/timeout" },
+    { "path": "./packages/util/retention" },
     { "path": "./packages/llm/llm" },
+    { "path": "./packages/llm/token-meter" },
     { "path": "./packages/core/session" },
     { "path": "./packages/core/scope" },
     { "path": "./packages/session-persistence/session-persistence" },
@@ -30,6 +34,7 @@
     { "path": "./packages/skill/skill-local" },
     { "path": "./packages/skill/tool-skill" },
     { "path": "./packages/ui/tool-ask-user" },
+    { "path": "./packages/context/workspace-context" },
     { "path": "./packages/core/agent-loop" },
     { "path": "./packages/examples/agent-spine-demo" },
     { "path": "./packages/bash/bash" },
@@ -48,14 +53,19 @@
     { "path": "./packages/fs/fs-local" },
     { "path": "./packages/fs/fs-policy" },
     { "path": "./packages/fs/tool-fs" },
+    { "path": "./packages/fs/tool-fs-search" },
     { "path": "./packages/web/web" },
     { "path": "./packages/web/web-search-exa" },
     { "path": "./packages/web/web-search-perplexity" },
     { "path": "./packages/web/web-search-deepseek" },
     { "path": "./packages/web/web-fetch-local" },
     { "path": "./packages/web/tool-web" },
+    { "path": "./packages/spill/spill" },
+    { "path": "./packages/spill/spill-local" },
+    { "path": "./packages/spill/spill-policy" },
     { "path": "./packages/timeout/timeout-policy" },
     { "path": "./packages/support/invariants" },
+    { "path": "./packages/support/agent-loop-testkit" },
     { "path": "./packages/ui/acp" },
     { "path": "./packages/examples/acp-demo" },
     { "path": "./packages/ui/app-boot" },
@@ -74,6 +84,8 @@
     { "path": "./packages/subagent/subagent-spawn" },
     { "path": "./packages/subagent/subagent-fork" },
     { "path": "./packages/subagent/subagent-acp" },
+    { "path": "./packages/tasks/tasks" },
+    { "path": "./packages/tasks/tool-tasks" },
     { "path": "./packages/workflow/workflow" },
     { "path": "./packages/workflow/workflow-workerthread" },
     { "path": "./packages/workflow/tool-workflow" },
@@ -83,6 +95,9 @@
     { "path": "./packages/hooks/hook-protocol" },
     { "path": "./packages/hooks/hooks-claude" },
     { "path": "./packages/hooks/hooks-codex" },
-    { "path": "./packages/mcp/mcp-client" }
+    { "path": "./packages/mcp/mcp-client" },
+    { "path": "./packages/sdk/helper" },
+    { "path": "./packages/sdk/scripts" },
+    { "path": "./packages/sdk/create-sdk" }
   ]
 }
diff --git a/tsconfig.json b/tsconfig.json
index 5720ca21ee..d64188e420 100644
--- a/tsconfig.json
+++ b/tsconfig.json
@@ -24,8 +24,12 @@
     { "path": "./vendor/hmr" },
     { "path": "./vendor/logger-console" },
     { "path": "./packages/util/brand" },
+    { "path": "./packages/util/home" },
+    { "path": "./packages/util/paths" },
     { "path": "./packages/util/timeout" },
+    { "path": "./packages/util/retention" },
     { "path": "./packages/llm/llm" },
+    { "path": "./packages/llm/token-meter" },
     { "path": "./packages/core/session" },
     { "path": "./packages/core/scope" },
     { "path": "./packages/session-persistence/session-persistence" },
@@ -43,6 +47,7 @@
     { "path": "./packages/skill/skill-local" },
     { "path": "./packages/skill/tool-skill" },
     { "path": "./packages/ui/tool-ask-user" },
+    { "path": "./packages/context/workspace-context" },
     { "path": "./packages/core/agent-loop" },
     { "path": "./packages/examples/agent-spine-demo" },
     { "path": "./packages/bash/bash" },
@@ -59,6 +64,7 @@
     { "path": "./packages/fs/fs-local" },
     { "path": "./packages/fs/fs-policy" },
     { "path": "./packages/fs/tool-fs" },
+    { "path": "./packages/fs/tool-fs-search" },
     { "path": "./packages/compact/compact" },
     { "path": "./packages/compact/compact-basic" },
     { "path": "./packages/web/web" },
@@ -67,8 +73,12 @@
     { "path": "./packages/web/web-search-deepseek" },
     { "path": "./packages/web/web-fetch-local" },
     { "path": "./packages/web/tool-web" },
+    { "path": "./packages/spill/spill" },
+    { "path": "./packages/spill/spill-local" },
+    { "path": "./packages/spill/spill-policy" },
     { "path": "./packages/timeout/timeout-policy" },
     { "path": "./packages/support/invariants" },
+    { "path": "./packages/support/agent-loop-testkit" },
     { "path": "./packages/ui/acp" },
     { "path": "./packages/examples/acp-demo" },
     { "path": "./packages/ui/app-boot" },
@@ -87,6 +97,8 @@
     { "path": "./packages/subagent/subagent-spawn" },
     { "path": "./packages/subagent/subagent-fork" },
     { "path": "./packages/subagent/subagent-acp" },
+    { "path": "./packages/tasks/tasks" },
+    { "path": "./packages/tasks/tool-tasks" },
     { "path": "./packages/workflow/workflow" },
     { "path": "./packages/workflow/workflow-workerthread" },
     { "path": "./packages/workflow/tool-workflow" },
@@ -96,6 +108,9 @@
     { "path": "./packages/hooks/hook-protocol" },
     { "path": "./packages/hooks/hooks-claude" },
     { "path": "./packages/hooks/hooks-codex" },
-    { "path": "./packages/mcp/mcp-client" }
+    { "path": "./packages/mcp/mcp-client" },
+    { "path": "./packages/sdk/helper" },
+    { "path": "./packages/sdk/scripts" },
+    { "path": "./packages/sdk/create-sdk" }
   ]
 }
diff --git a/vendor/README.md b/vendor/README.md
index 9bf226fb4d..ae43760ceb 100644
--- a/vendor/README.md
+++ b/vendor/README.md
@@ -36,6 +36,7 @@ Keep this log exhaustive — every divergence from upstream must be listed.
 4. **Vendored TypeScript source internal specifiers**: changed local relative imports/exports from upstream's specifier shape to explicit `.ts` specifiers so TypeScript rewrites emitted JS to `.js` while declarations keep explicit, NodeNext-safe `.ts` specifiers. This includes `loader/src/config/isolate.ts` using `declare module './entry.ts'`.
 5. **`schemastery/tsdown.config.ts` and `logger-console/tsdown.config.ts`**: ours, not upstream files — per-package build-shape overrides (dual ESM+CJS output; separate node/browser entries) for the repo-root tsdown build. They read the JS emitted under `lib/types` and then write the publish runtime entries under `lib/`. Like the regenerated tsconfigs, they are not part of the upstream sync surface.
 6. **`cordis/src/fiber.ts` lifecycle hardening**: locally closes three reentrant disposal gaps. An effect's owner-list wrapper is registered before its setup body runs, so an unload begun from inside setup awaits setup and every collected cleanup; synchronous setup failure removes the wrapper and rolls back collected cleanup. Async cleanup stays owner-visible until quiescence, and Cordis's internal effect composition joins an already-running cleanup while repeated public disposer calls retain their upstream single-shot result. Effect creation is rejected while the owner is `UNLOADING` (while `PENDING` and `LOADING` remain legal), preventing cleanup-time registrations from escaping the unload snapshot. Child fibers register and receive their parent-owned disposer before `internal/plugin` publication, resolve dependency declarations added by that notification before activation, drain effects attached while pending, skip plugin execution when reentrant disposal invalidates the load epoch before its first checkpoint, and contain teardown-notification failures per observer so one callback cannot starve peers or interrupt ownership cleanup.
+7. **`cordis/src/*.ts` JSDoc enrichment**: added `@param`/`@returns` tags and contract documentation (disposal semantics, waterfall veto, bail conditions, error cases) across the public plugin-author surface — `Context` (class, statics, and the `Context` interface properties incl. `root`), `EventsService`, `Fiber`, `RegistryService`, `ReflectService`, `Service`, `LoggerService` and their `declare module './context.ts'` overloads. Comment-only; no code changes. Motivation: the website API-reference generator renders these docs and hard-errors on undocumented members. Retire this entry when the enrichment is upstreamed to the fork.
 
 ## Sync procedure
 
diff --git a/vendor/cordis/src/context.ts b/vendor/cordis/src/context.ts
index 8b21c464b2..514b1e062a 100644
--- a/vendor/cordis/src/context.ts
+++ b/vendor/cordis/src/context.ts
@@ -14,14 +14,21 @@ import { Fiber } from './fiber.ts'
  * be read from `ctx`.
  */
 export interface Context {
+  /** Isolation map: service name → scope label. Lookups for a name resolve within its label. */
   [symbols.isolate]: Dict<symbol>
+  /** Intercept map: service name → config merged into that service's per-plugin config. */
   [symbols.intercept]: Dict
-  /** @experimental */
+  /** The root context of the application (every child context shares it). @experimental */
   root: this
+  /** Base URL used to resolve relative plugin/module specifiers, if the runtime sets one. */
   baseUrl?: string
+  /** The event bus. Its methods are also mixed onto `ctx` (`ctx.on`, `ctx.emit`, ...). */
   events: EventsService
+  /** The logging service. Call `ctx.logger(name)` for a named logger. */
   logger: LoggerService
+  /** The reflection layer backing the context proxy (`ctx.get`, `ctx.provide`, ...). */
   reflect: ReflectService
+  /** The plugin registry. Its methods are mixed onto `ctx` (`ctx.plugin`, `ctx.inject`). */
   registry: RegistryService
 }
 
@@ -33,12 +40,24 @@ export interface Context {
  * contexts without mutating their parent.
  */
 export class Context {
+  /** Symbol key under which a disposer exposes its {@link EffectMeta} diagnostics tree. */
   static readonly effect: unique symbol = symbols.effect
+  /** Symbol key for a context's listener filter, consulted on every event dispatch. */
   static readonly filter: unique symbol = symbols.filter
+  /** Symbol key of the isolation map (see the `Context[symbols.isolate]` property). */
   static readonly isolate: unique symbol = symbols.isolate
+  /** Symbol key of the intercept map (see the `Context[symbols.intercept]` property). */
   static readonly intercept: unique symbol = symbols.intercept
 
-  /** Returns true for Cordis context proxies and context prototypes. */
+  /**
+   * Returns true for Cordis context proxies and context prototypes.
+   *
+   * Works across realms and across multiple copies of cordis, because the
+   * brand is keyed by a global symbol rather than by `instanceof`.
+   *
+   * @param value — the value to test.
+   * @returns `true` if `value` is a Cordis context, narrowing its type.
+   */
   static is(value: any): value is Context {
     return !!value?.[Context.is as any]
   }
@@ -68,7 +87,15 @@ export class Context {
     return `Context <${this.fiber.name}>`
   }
 
-  /** Create a child context with extra metadata on top of the current scope. */
+  /**
+   * Create a child context with extra metadata on top of the current scope.
+   *
+   * The child prototypally inherits every property of this context; own
+   * properties of `meta` shadow the inherited ones. The parent is not mutated.
+   *
+   * @param meta — own properties (including symbol keys) to define on the child.
+   * @returns a child context inheriting from this one.
+   */
   extend(meta = {}): this {
     const shadow = Reflect.getOwnPropertyDescriptor(this, symbols.shadow)?.value
     const self = Object.create(getTraceable(this, this))
@@ -79,14 +106,36 @@ export class Context {
     return Object.assign(Object.create(self), { [symbols.shadow]: shadow })
   }
 
-  /** Create a child context with an independent service scope for `name`. */
+  /**
+   * Create a child context with an independent service scope for `name`.
+   *
+   * Below the returned context, reads and writes of the service `name`
+   * resolve against the new label instead of the parent's, so a different
+   * implementation can be provided without affecting the parent scope.
+   * Passing the same `label` to two `isolate()` calls joins their scopes.
+   *
+   * @param name — the service name to isolate.
+   * @param label — scope label to join; defaults to a fresh unique symbol.
+   * @returns a child context whose `name` service resolves in the new scope.
+   */
   isolate(name: string, label?: symbol) {
     const shadow = Object.create(this[symbols.isolate])
     shadow[name] = label ?? Symbol(name)
     return this.extend({ [symbols.isolate]: shadow })
   }
 
-  /** Add service-specific intercept config for plugins started below this context. */
+  /**
+   * Add service-specific intercept config for plugins started below this
+   * context.
+   *
+   * Plugins loaded under the returned context see `config` merged into the
+   * service's resolved config (ancestor entries first; see
+   * `Service[symbols.resolveConfig]`). The parent context is not affected.
+   *
+   * @param name — the service name whose config to intercept.
+   * @param config — the intercept config to merge for that service.
+   * @returns a child context carrying the additional intercept entry.
+   */
   intercept<K extends InjectKey>(name: K, config: Context[K] extends { [symbols.config]: infer T } ? T : never): this
   intercept(name: string, config: any): this
   intercept(name: string, config: any) {
diff --git a/vendor/cordis/src/events.ts b/vendor/cordis/src/events.ts
index 842a780cdb..09589d69da 100644
--- a/vendor/cordis/src/events.ts
+++ b/vendor/cordis/src/events.ts
@@ -3,7 +3,12 @@ import { Context } from './context.ts'
 import { Fiber, FiberState } from './fiber.ts'
 import { DisposableList, symbols } from './utils.ts'
 
-/** Return whether an event result should stop a bail-style dispatch. */
+/**
+ * Return whether an event result should stop a bail-style dispatch.
+ *
+ * @param value — a listener's return value.
+ * @returns `true` unless `value` is `null`, `false`, or `undefined`.
+ */
 export function isBailed(value: any) {
   return value !== null && value !== false && value !== undefined
 }
@@ -28,17 +33,75 @@ export type DispatchMode = 'emit' | 'parallel' | 'serial' | 'bail' | 'waterfall'
 declare module './context.ts' {
   export interface Context {
     /* eslint-disable max-len */
+    /**
+     * Dispatch an event, running all listeners concurrently.
+     *
+     * @param name — the event name.
+     * @param args — arguments passed to every listener.
+     * @returns a promise resolving once every listener has settled.
+     */
     parallel<K extends keyof Events>(name: K, ...args: Parameters<Events[K]>): Promise<void>
+    /** Same as above, with an explicit `this` for listeners (also used for filtering). */
     parallel<K extends keyof Events>(thisArg: NoInfer<ThisType<Events[K]>>, name: K, ...args: Parameters<Events[K]>): Promise<void>
+    /**
+     * Dispatch an event synchronously, ignoring listener return values.
+     *
+     * @param name — the event name.
+     * @param args — arguments passed to every listener.
+     */
     emit<K extends keyof Events>(name: K, ...args: Parameters<Events[K]>): void
+    /** Same as above, with an explicit `this` for listeners (also used for filtering). */
     emit<K extends keyof Events>(thisArg: NoInfer<ThisType<Events[K]>>, name: K, ...args: Parameters<Events[K]>): void
+    /**
+     * Dispatch an event, awaiting listeners in order until one bails.
+     *
+     * @param name — the event name.
+     * @param args — arguments passed to each listener.
+     * @returns the first bail value (non-null, non-false, non-undefined), if any.
+     */
     serial<K extends keyof Events>(name: K, ...args: Parameters<Events[K]>): Promisify<ReturnType<Events[K]>>
+    /** Same as above, with an explicit `this` for listeners (also used for filtering). */
     serial<K extends keyof Events>(thisArg: NoInfer<ThisType<Events[K]>>, name: K, ...args: Parameters<Events[K]>): Promisify<ReturnType<Events[K]>>
+    /**
+     * Dispatch an event, calling listeners in order until one bails.
+     *
+     * @param name — the event name.
+     * @param args — arguments passed to each listener.
+     * @returns the first bail value (non-null, non-false, non-undefined), if any.
+     */
     bail<K extends keyof Events>(name: K, ...args: Parameters<Events[K]>): ReturnType<Events[K]>
+    /** Same as above, with an explicit `this` for listeners (also used for filtering). */
     bail<K extends keyof Events>(thisArg: NoInfer<ThisType<Events[K]>>, name: K, ...args: Parameters<Events[K]>): ReturnType<Events[K]>
+    /**
+     * Dispatch an event whose last argument is a `next` continuation.
+     *
+     * Each listener wraps the rest of the chain: calling `next()` invokes the
+     * next listener (finally the built-in behavior); not calling it vetoes.
+     *
+     * @param name — the event name.
+     * @param args — listener arguments; the final one is the innermost `next`.
+     * @returns the outermost listener's return value.
+     */
     waterfall<K extends keyof Events>(name: K, ...args: Parameters<Events[K]>): ReturnType<Events[K]>
+    /** Same as above, with an explicit `this` for listeners (also used for filtering). */
     waterfall<K extends keyof Events>(thisArg: NoInfer<ThisType<Events[K]>>, name: K, ...args: Parameters<Events[K]>): ReturnType<Events[K]>
+    /**
+     * Register an event listener owned by the current fiber.
+     *
+     * @param name — the event name to listen for.
+     * @param listener — called with the dispatch arguments.
+     * @param options — listener options; a boolean is shorthand for `prepend`.
+     * @returns a disposer removing the listener; `true` if it was still registered.
+     */
     on<K extends keyof Events>(name: K, listener: Events[K], options?: boolean | EventOptions): () => boolean
+    /**
+     * Same as `on()`, but the listener disposes itself after its first call.
+     *
+     * @param name — the event name to listen for.
+     * @param listener — called at most once with the dispatch arguments.
+     * @param options — listener options; a boolean is shorthand for `prepend`.
+     * @returns a disposer removing the listener; `true` if it was still registered.
+     */
     once<K extends keyof Events>(name: K, listener: Events[K], options?: boolean | EventOptions): () => boolean
     /* eslint-enable max-len */
   }
@@ -91,7 +154,13 @@ export class EventsService {
     }, { global: true, prepend: true })
   }
 
-  /** Resolve listeners for one dispatch and apply context filtering. */
+  /**
+   * Resolve listeners for one dispatch and apply context filtering.
+   *
+   * @param type — the dispatch mode, reported on `internal/dispatch`.
+   * @param args — the raw dispatch arguments; consumed up to the event name.
+   * @returns the matching listener callbacks, bound to the dispatch `this`.
+   */
   dispatch(type: string, args: any[]) {
     const thisArg = typeof args[0] === 'object' || typeof args[0] === 'function' ? args.shift() : null
     const name: string = args.shift()
@@ -104,19 +173,33 @@ export class EventsService {
       .map(hook => hook.callback.bind(thisArg))
   }
 
-  /** Run listeners concurrently and wait for all of them. */
+  /**
+   * Run listeners concurrently and wait for all of them.
+   *
+   * @param args — optional `this`, the event name, then listener arguments.
+   * @returns a promise resolving once every listener has settled.
+   */
   async parallel(...args: any[]) {
     const results = await Promise.allSettled(this.dispatch('emit', args).map(async cb => cb(...args)))
     const errors = results.filter((result): result is PromiseRejectedResult => result.status === 'rejected')
     if (errors.length) throw new AggregateError(errors.map(error => error.reason))
   }
 
-  /** Run listeners synchronously without waiting for returned promises. */
+  /**
+   * Run listeners synchronously without waiting for returned promises.
+   *
+   * @param args — optional `this`, the event name, then listener arguments.
+   */
   emit(...args: any[]) {
     this.dispatch('emit', args).map(cb => cb(...args))
   }
 
-  /** Run listeners in order until one returns a bail value. */
+  /**
+   * Run listeners in order, awaiting each, until one returns a bail value.
+   *
+   * @param args — optional `this`, the event name, then listener arguments.
+   * @returns the first bail value (see {@link isBailed}), if any.
+   */
   async serial(...args: any[]) {
     for (const cb of this.dispatch('serial', args)) {
       const result = await cb(...args)
@@ -124,7 +207,12 @@ export class EventsService {
     }
   }
 
-  /** Run listeners synchronously until one returns a bail value. */
+  /**
+   * Run listeners synchronously until one returns a bail value.
+   *
+   * @param args — optional `this`, the event name, then listener arguments.
+   * @returns the first bail value (see {@link isBailed}), if any.
+   */
   bail(...args: any[]) {
     for (const cb of this.dispatch('bail', args)) {
       const result = cb(...args)
@@ -132,7 +220,16 @@ export class EventsService {
     }
   }
 
-  /** Compose listeners around the final `next` callback. */
+  /**
+   * Compose listeners around the final `next` callback.
+   *
+   * The last dispatch argument is treated as the innermost `next`. Listeners
+   * run outermost-first; a listener that does not call `next()` vetoes the
+   * rest of the chain, including the built-in behavior.
+   *
+   * @param args — optional `this`, the event name, listener arguments, then `next`.
+   * @returns the outermost listener's return value.
+   */
   waterfall(...args: any[]) {
     const cbs = this.dispatch('waterfall', args)
     const inner = args.pop()
@@ -144,6 +241,15 @@ export class EventsService {
     return next()
   }
 
+  /**
+   * Store a listener record as an effect on the current fiber.
+   *
+   * @param label — effect label shown in fiber diagnostics.
+   * @param hooks — the listener list for one event.
+   * @param callback — the listener to store.
+   * @param options — placement and filtering options.
+   * @returns a disposer that unregisters the listener.
+   */
   register(label: string, hooks: Hook[], callback: any, options: EventOptions): () => void {
     const method = options.prepend ? 'unshift' : 'push'
     return this.ctx.fiber.effect(() => {
@@ -152,6 +258,13 @@ export class EventsService {
     }, label)
   }
 
+  /**
+   * Remove a stored listener record.
+   *
+   * @param hooks — the listener list for one event.
+   * @param callback — the listener to remove.
+   * @returns `true` if the listener was found and removed.
+   */
   unregister(hooks: Hook[], callback: any) {
     const index = hooks.findIndex(hook => hook.callback === callback)
     if (index >= 0) {
@@ -160,7 +273,17 @@ export class EventsService {
     }
   }
 
-  /** Register an event listener owned by the current fiber. */
+  /**
+   * Register an event listener owned by the current fiber.
+   *
+   * The listener is removed automatically when the fiber unloads. Throws
+   * `CordisError('INACTIVE_EFFECT')` if the fiber is already disposed.
+   *
+   * @param name — the event name to listen for.
+   * @param listener — called with the dispatch arguments.
+   * @param options — listener options; a boolean is shorthand for `prepend`.
+   * @returns a disposer removing the listener; `true` if it was still registered.
+   */
   on(name: string | symbol, listener: (...args: any) => any, options?: boolean | EventOptions) {
     if (typeof options !== 'object') {
       options = { prepend: options }
@@ -177,7 +300,14 @@ export class EventsService {
     return this.register(label, hooks, listener, options)
   }
 
-  /** Register an event listener that disposes itself after the first call. */
+  /**
+   * Register an event listener that disposes itself after the first call.
+   *
+   * @param name — the event name to listen for.
+   * @param listener — called at most once with the dispatch arguments.
+   * @param options — listener options; a boolean is shorthand for `prepend`.
+   * @returns a disposer removing the listener; `true` if it was still registered.
+   */
   once(name: string, listener: (...args: any) => any, options?: boolean | EventOptions) {
     const dispose = this.on(name, function (...args: any[]) {
       dispose()
@@ -196,12 +326,20 @@ export class EventsService {
  * diagnostics before public events are delivered.
  */
 export interface Events {
+  /** A plugin fiber was created or its uid was cleared on disposal. */
   'internal/plugin'(fiber: Fiber): void
+  /** A fiber changed lifecycle state; receives the fiber and its previous state. */
   'internal/status'(fiber: Fiber, oldValue: FiberState): void
+  /** Interception hook for a service binding (no core producer). */
   'internal/service'(this: Context, name: string, value: any): void
+  /** Waterfall: a fiber config update is being applied; skip `next()` to veto. */
   'internal/update'(this: Fiber, config: any, noSave: boolean, next: () => void): void
+  /** Waterfall: a service is being read through the context proxy. */
   'internal/get'(ctx: Context, name: string, error: Error, next: () => any): any
+  /** Waterfall: a service is being written through the context proxy. */
   'internal/set'(ctx: Context, name: string, value: any, error: Error, next: () => boolean): boolean
+  /** Bail: a listener is being registered; a non-null result replaces registration. */
   'internal/listener'(this: Context, name: string, listener: any, prepend: boolean): void
+  /** An event is being dispatched to listeners (fired for non-internal events only). */
   'internal/dispatch'(mode: DispatchMode, name: string, args: any[], thisArg: any): void
 }
diff --git a/vendor/cordis/src/fiber.ts b/vendor/cordis/src/fiber.ts
index 43a320142f..6beb7bbc24 100644
--- a/vendor/cordis/src/fiber.ts
+++ b/vendor/cordis/src/fiber.ts
@@ -7,6 +7,7 @@ import { StandardSchemaV1 } from '@standard-schema/spec'
 
 declare module './context.ts' {
   export interface Context extends Pick<Fiber, 'effect'> {
+    /** The fiber (plugin runtime instance) that owns this context. */
     fiber: Fiber
   }
 }
@@ -17,6 +18,11 @@ const kValidationError = Symbol.for('ValidationError')
 export class ValidationError extends TypeError {
   name = 'ValidationError'
 
+  /**
+   * Build the aggregated message from schema issues.
+   *
+   * @param issues — the standard-schema issues, one message line each.
+   */
   constructor(issues: readonly StandardSchemaV1.Issue[]) {
     super(`invalid config:\n` + issues.map(issue => {
       if (issue.path) {
@@ -32,7 +38,14 @@ Object.defineProperty(ValidationError.prototype, kValidationError, {
   value: true,
 })
 
-/** Validate and normalize config for a plugin runtime before it starts. */
+/**
+ * Validate and normalize config for a plugin runtime before it starts.
+ *
+ * @param runtime — the plugin runtime whose `Config` schema to apply.
+ * @param config — the raw user config.
+ * @returns the validated config, or `config` unchanged if the runtime has no schema.
+ * @throws {ValidationError} when validation reports issues.
+ */
 export function resolveConfig(runtime: Plugin.Runtime, config: any) {
   if (!runtime.Config) return config
   // TODO: async validation
@@ -51,10 +64,21 @@ interface AsyncDisposable<T extends Awaitable<void> = Awaitable<void>> extends P
   (): T
 }
 
-/** Function returned by an effect to release resources during disposal. */
+/**
+ * Function returned by an effect to release resources during disposal.
+ *
+ * Disposers run in reverse registration order when the owning fiber unloads;
+ * they may be async, in which case unloading awaits them.
+ */
 export type Disposable<T = any> = () => T
 
-/** Effect body result accepted by `ctx.effect()` and plugin startup. */
+/**
+ * Effect body result accepted by `ctx.effect()` and plugin startup.
+ *
+ * Either a single disposer, a promise of one, or a (possibly async) iterable
+ * yielding several — generator effects register each yielded disposer as it
+ * is produced.
+ */
 export type Effect<T = any> =
   | SyncEffect<T>
   | AsyncEffect<T>
@@ -69,7 +93,9 @@ type AsyncEffect<T = any> =
 
 /** Tree node used to expose nested effect labels for diagnostics. */
 export interface EffectMeta {
+  /** Human-readable effect label, e.g. `ctx.on("event")` or `ctx.provide("name")`. */
   label: string
+  /** Metadata of nested effects registered while this effect ran. */
   children: EffectMeta[]
 }
 
@@ -109,7 +135,14 @@ function emitPluginDisposed(context: Context, fiber: Fiber) {
   }
 }
 
-/** Lifecycle state for one plugin fiber. */
+/**
+ * Lifecycle state for one plugin fiber.
+ *
+ * `PENDING` — waiting for required services; `LOADING` — the plugin callback
+ * is running; `ACTIVE` — loaded and providing; `FAILED` — the callback or its
+ * config threw; `UNLOADING` — disposers are running; `DISPOSED` — the fiber
+ * was removed and cannot restart.
+ */
 export const enum FiberState {
   PENDING,
   LOADING,
@@ -121,6 +154,10 @@ export const enum FiberState {
 
 /** Framework error with a stable machine-readable code. */
 export class CordisError extends Error {
+  /**
+   * @param code — the stable error code; also the default message.
+   * @param message — optional human-readable override.
+   */
   constructor(public code: CordisError.Code, message?: string) {
     super(message ?? CordisError.Code[code])
   }
@@ -144,12 +181,19 @@ const INACTIVE = '__INACTIVE__'
  * cleanup for the plugin context returned by `ctx.plugin()`.
  */
 export class Fiber {
+  /** Unique id within the registry; 0 for the root fiber, `null` once disposed. */
   public uid: number | null
+  /** The context this fiber's plugin runs in (extends the parent context). */
   public readonly ctx: Context
+  /** The validated plugin config (updated by `update()`). */
   public config: any
+  /** Current lifecycle state; transitions emit `internal/status`. */
   public state = FiberState.PENDING
+  /** Dispose this fiber: unload the plugin, then settle once cleanup finished. */
   public readonly dispose: () => Promise<void>
+  /** Snapshot of required service implementations while loaded; `undefined` otherwise. */
   public store: Dict<Impl> | undefined
+  /** The in-flight load/unload transition, if one is currently running. */
   public inertia: Promise<void> | undefined
 
   public readonly _hooks: Dict<DisposableList<Function>> = Object.create(null)
@@ -162,6 +206,16 @@ export class Fiber {
   private _runner: EffectRunner<string>
   private _store: Dict<Impl> = Object.create(null)
 
+  /**
+   * Create a fiber. Plugin authors normally obtain fibers from `ctx.plugin()`
+   * rather than constructing them directly.
+   *
+   * @param parent — the context the plugin was loaded from.
+   * @param config — raw config, validated against the runtime's schema.
+   * @param inject — resolved dependency map (service name → intercept config).
+   * @param runtime — the shared plugin runtime, or `null` for the root fiber.
+   * @param getOuterStack — captures the caller stack for effect diagnostics.
+   */
   constructor(
     public parent: Context,
     config: any,
@@ -282,6 +336,7 @@ export class Fiber {
     }
   }
 
+  /** The plugin's display name, inherited from the nearest named ancestor, else `'root'`. */
   get name() {
     let fiber: Fiber = this
     do {
@@ -291,7 +346,12 @@ export class Fiber {
     return 'root'
   }
 
-  /** Throw if the fiber has already been disposed. */
+  /**
+   * Throw if the fiber has already been disposed.
+   *
+   * @returns nothing when the fiber is still active.
+   * @throws {CordisError} `INACTIVE_EFFECT` when the fiber's uid has been cleared.
+   */
   assertActive() {
     if (this.uid !== null) return
     throw new CordisError('INACTIVE_EFFECT')
@@ -343,8 +403,21 @@ export class Fiber {
     }, runner.getOuterStack)
   }
 
-  /** Register a cleanup-aware effect on this fiber. */
+  /**
+   * Register a cleanup-aware effect on this fiber.
+   *
+   * `execute` runs immediately; the disposers it produces are collected and
+   * run (in reverse order) either when the returned disposer is called or
+   * when the fiber unloads, whichever comes first. Calling the disposer twice
+   * is a no-op. Throws `CordisError('INACTIVE_EFFECT')` if the fiber is
+   * already disposed, and `TypeError` if `execute` returns an invalid shape.
+   *
+   * @param execute — the effect body; see {@link Effect} for accepted shapes.
+   * @param label — effect label shown in `getEffects()` diagnostics.
+   * @returns a disposer that tears the effect down and settles once done.
+   */
   effect(execute: () => SyncEffect, label?: string): Disposable<Promise<void>>
+  /** Same as above for async effects; the disposer is also awaitable. */
   effect(execute: () => Effect, label?: string): AsyncDisposable<Promise<void>>
   effect(execute: () => Effect, label = 'anonymous'): any {
     this.assertActive()
@@ -491,7 +564,11 @@ export class Fiber {
     return wrapper
   }
 
-  /** Return metadata for currently registered effects. */
+  /**
+   * Return metadata for currently registered effects.
+   *
+   * @returns one {@link EffectMeta} tree per labeled live effect.
+   */
   getEffects() {
     return [...this._disposables]
       .map<EffectMeta>(dispose => dispose[symbols.effect])
@@ -615,7 +692,12 @@ export class Fiber {
     })
   }
 
-  /** Wait for current lifecycle work and rethrow startup errors. */
+  /**
+   * Wait for current lifecycle work and rethrow startup errors.
+   *
+   * @returns this fiber, once it has settled into a stable state.
+   * @throws the config-validation or plugin-startup error, if any.
+   */
   async await() {
     while (this.inertia) {
       await this.inertia
@@ -624,7 +706,12 @@ export class Fiber {
     return this
   }
 
-  /** Dispose and immediately reload this plugin with its current config. */
+  /**
+   * Dispose and immediately reload this plugin with its current config.
+   *
+   * @returns a promise resolving once the reload settled.
+   * @throws {CordisError} `INACTIVE_EFFECT` when the fiber is already disposed.
+   */
   async restart() {
     this.assertActive()
     this._setEpoch(INACTIVE)
@@ -632,7 +719,17 @@ export class Fiber {
     await this.await()
   }
 
-  /** Validate and apply new config, then restart the plugin. */
+  /**
+   * Validate and apply new config, then restart the plugin.
+   *
+   * Runs the `internal/update` waterfall first, so update hooks (and HMR)
+   * can veto or replace the restart.
+   *
+   * @param config — the new raw config; validated before anything restarts.
+   * @param noSave — hint for persistence hooks not to write the change back.
+   * @returns nothing; the restart runs behind the `internal/update` waterfall.
+   * @throws {ValidationError} when the new config fails validation.
+   */
   update(config: any, noSave = false) {
     this.assertActive()
     config = resolveConfig(this.runtime!, config)
diff --git a/vendor/cordis/src/logger.ts b/vendor/cordis/src/logger.ts
index a1e97c165a..3c5ad10525 100644
--- a/vendor/cordis/src/logger.ts
+++ b/vendor/cordis/src/logger.ts
@@ -62,8 +62,11 @@ export const defaultFormatters: Record<string, Formatter> = {
 
 /** Options used when creating a named logger facade. */
 export interface LoggerOptions {
+  /** The logger name shown with each message. */
   name: string
+  /** Message fields merged into every record from this logger. */
   meta?: Partial<Message>
+  /** Default maximum level exported when an exporter has no own threshold. */
   level?: number
 }
 
@@ -220,7 +223,12 @@ export class LoggerService {
     return self
   }
 
-  /** Register an exporter and dispose it with the current fiber. */
+  /**
+   * Register an exporter and dispose it with the current fiber.
+   *
+   * @param exporter — the sink that receives structured log messages.
+   * @returns a disposer that removes the exporter.
+   */
   exporter(exporter: Exporter) {
     return this.ctx.effect(() => {
       this.exporters.set(++this._snExporter, exporter)
diff --git a/vendor/cordis/src/reflect.ts b/vendor/cordis/src/reflect.ts
index c084067af1..e91cb63fea 100644
--- a/vendor/cordis/src/reflect.ts
+++ b/vendor/cordis/src/reflect.ts
@@ -5,14 +5,66 @@ import { Fiber, FiberState } from './fiber.ts'
 
 declare module './context.ts' {
   interface Context {
+    /**
+     * Read a service from the store without the inject requirement.
+     *
+     * @param name — the service name.
+     * @param strict — when `true` (default), only return implementations
+     * whose providing fiber is currently active.
+     * @returns the service value, or `undefined` when not (yet) provided.
+     */
     get<K extends string & keyof this>(name: K, strict?: boolean): undefined | this[K]
+    /** Same as above for service names outside the typed `Context` surface. */
     get(name: string, strict?: boolean): any
+    /**
+     * Overwrite a provided service's value.
+     *
+     * Only the fiber that provided the service may set it; setting an
+     * unprovided name throws.
+     *
+     * @param name — the service name.
+     * @param value — the new service value.
+     */
     set<K extends string & keyof this>(name: K, value: undefined | this[K]): void
+    /** Same as above for service names outside the typed `Context` surface. */
     set(name: string, value: any): void
+    /**
+     * Register a service implementation owned by the current fiber.
+     *
+     * The service becomes visible to dependents in the same isolation scope
+     * once the fiber is active; it is unregistered (waking dependents) when
+     * the returned disposer runs or the fiber unloads. Throws if the name is
+     * already provided in this scope or declared as an accessor.
+     *
+     * @param name — the service name.
+     * @param value — the service value.
+     * @returns a disposer that unregisters the service.
+     */
     provide<K extends string & keyof this>(name: K, value: undefined | this[K]): () => void
+    /** Same as above for service names outside the typed `Context` surface. */
     provide(name: string, value?: any): () => void
+    /**
+     * Define a computed context property backed by get/set hooks.
+     *
+     * The accessor is removed when the current fiber unloads. Throws if the
+     * name is already declared.
+     *
+     * @param name — the context property name.
+     * @param options — the `get` hook and optional `set` hook.
+     */
     accessor(name: string, options: Omit<Property.Accessor, 'type'>): void
+    /**
+     * Expose selected members of a service directly on `ctx`.
+     *
+     * Each mixed-in key becomes an accessor that forwards to the service
+     * (binding methods to it), so e.g. `ctx.on` forwards to `ctx.events.on`.
+     * Mixins are removed when the current fiber unloads.
+     *
+     * @param name — the context property holding the source service.
+     * @param mixins — keys to forward, or a source-key → ctx-key map.
+     */
     mixin<K extends string & keyof this>(name: K, mixins: (keyof this & keyof this[K])[] | Dict<string>): void
+    /** Same as above with a source object instead of a context property name. */
     mixin<T extends {}>(source: T, mixins: (keyof this & keyof T)[] | Dict<string>): void
   }
 }
@@ -44,22 +96,30 @@ export type Property = Property.Service | Property.Accessor
 export namespace Property {
   /** Service property backed by a provided implementation. */
   export interface Service {
+    /** Discriminator. */
     type: 'service'
   }
 
   /** Computed context property backed by custom get/set hooks. */
   export interface Accessor {
+    /** Discriminator. */
     type: 'accessor'
+    /** Compute the property value; `error` carries the caller stack for diagnostics. */
     get: (this: Context, receiver: any, error: Error) => any
+    /** Optional setter; return `false` to reject the write. */
     set?: (this: Context, value: any, receiver: any, error: Error) => boolean
   }
 }
 
 /** Concrete service implementation record stored in the root reflect service. */
 export interface Impl {
+  /** The service name. */
   name: string
+  /** The fiber that provided the service (owns its lifetime). */
   fiber: Fiber
+  /** The current service value. */
   value?: any
+  /** Optional availability predicate consulted before dependents may load. */
   check?: () => boolean
 }
 
@@ -70,6 +130,7 @@ export interface Impl {
  * the mixins that expose core service methods directly on `ctx`.
  */
 export class ReflectService {
+  /** Proxy traps implementing service resolution for every context object. */
   static handler: ProxyHandler<Context> = {
     get: (target, prop, ctx: Context) => {
       if (isSpecialProperty(prop)) {
@@ -143,7 +204,9 @@ export class ReflectService {
     },
   }
 
+  /** Service implementations, keyed by isolation label. */
   public store: Dict<Impl, symbol> = Object.create(null)
+  /** Declared context properties (services and accessors), by name. */
   public props: Dict<Property> = Object.create(null)
 
   constructor(public ctx: Context) {
@@ -158,6 +221,14 @@ export class ReflectService {
     this.mixin('events', ['on', 'once', 'parallel', 'emit', 'serial', 'bail', 'waterfall'])
   }
 
+  /**
+   * Read a service from the store without the inject requirement.
+   *
+   * @param name — the service name.
+   * @param strict — when `true`, only return implementations whose providing
+   * fiber is currently active.
+   * @returns the service value, or `undefined` when not (yet) provided.
+   */
   get(name: string, strict = true) {
     return getTraceable(this.ctx, this._getImpl(name, strict)?.value)
   }
@@ -170,6 +241,15 @@ export class ReflectService {
     return impl
   }
 
+  /**
+   * Overwrite a provided service's value.
+   *
+   * @param name — the service name.
+   * @param value — the new service value.
+   * @param error — carrier for the caller stack in diagnostics.
+   * @returns `true` on success.
+   * @throws when `name` was never provided, or was provided by another fiber.
+   */
   set(name: string, value: any, error?: Error) {
     const key = this.ctx[symbols.isolate][name]
     const impl = this.store[key]
@@ -183,6 +263,16 @@ export class ReflectService {
     return true
   }
 
+  /**
+   * Register a service implementation owned by the current fiber.
+   *
+   * See the `ctx.provide()` overload above for the full contract.
+   *
+   * @param name — the service name.
+   * @param value — the service value.
+   * @param check — optional availability predicate for dependents.
+   * @returns a disposer that unregisters the service.
+   */
   provide(name: string, value?: any, check?: () => boolean) {
     return this.ctx.fiber.effect(() => {
       if (!this.props[name]) {
@@ -213,6 +303,13 @@ export class ReflectService {
     }, `ctx.provide(${JSON.stringify(name)})`)
   }
 
+  /**
+   * Re-evaluate every fiber that requires one of the given services.
+   *
+   * @param names — the service names that changed.
+   * @param filter — restricts notification to matching isolation scopes.
+   * @returns the fibers whose dependency state was refreshed.
+   */
   notify(names: string[], filter = (ctx: Context, name: string) => ctx[symbols.isolate][name] === this.ctx[symbols.isolate][name]) {
     const fibers: Fiber[] = []
     for (const runtime of this.ctx.registry.values()) {
@@ -237,6 +334,13 @@ export class ReflectService {
     return fibers
   }
 
+  /**
+   * Define a computed context property backed by get/set hooks.
+   *
+   * @param name — the context property name.
+   * @param options — the `get` hook and optional `set` hook.
+   * @returns a disposer that removes the accessor.
+   */
   accessor(name: string, options: Omit<Property.Accessor, 'type'>) {
     return this.ctx.fiber.effect(() => {
       if (name in this.props) {
@@ -247,6 +351,15 @@ export class ReflectService {
     }, `ctx.accessor(${JSON.stringify(name)})`)
   }
 
+  /**
+   * Expose selected members of a service directly on `ctx`.
+   *
+   * See the `ctx.mixin()` overload above for the full contract.
+   *
+   * @param source — a context property name or a source object.
+   * @param mixins — keys to forward, or a source-key → ctx-key map.
+   * @returns a disposer that removes all created accessors.
+   */
   mixin(source: any, mixins: string[] | Dict<string>) {
     const self = this
     return this.ctx.fiber.effect(function* () {
@@ -275,10 +388,22 @@ export class ReflectService {
     }, `ctx.mixin(${JSON.stringify(source)})`)
   }
 
+  /**
+   * Attach this context's tracing wrapper to a value.
+   *
+   * @param value — the value to wrap.
+   * @returns the traceable wrapper (or the value itself when not applicable).
+   */
   trace<T>(value: T) {
     return getTraceable(this.ctx, value)
   }
 
+  /**
+   * Wrap a callback so calls trace `this` and arguments to this context.
+   *
+   * @param callback — the function to wrap.
+   * @returns a proxy delegating to `callback` with traced values.
+   */
   bind<T extends Function>(callback: T) {
     return new Proxy(callback, {
       apply: (target, thisArg, args) => {
diff --git a/vendor/cordis/src/registry.ts b/vendor/cordis/src/registry.ts
index 9dfa10a06b..05fbadcfad 100644
--- a/vendor/cordis/src/registry.ts
+++ b/vendor/cordis/src/registry.ts
@@ -28,6 +28,11 @@ export type InjectKey = keyof {
  * On classes it contributes to the plugin's static `inject` map. On methods it
  * delays the method call until the declared services are available.
  */
+/**
+ * @param name — the required service name.
+ * @param config — optional intercept config applied for that service.
+ * @returns the class or method decorator.
+ */
 export function Inject<K extends InjectKey>(name: K, config?: Context[K] extends { [symbols.config]: infer T } ? T : never) {
   return function (value: any, decorator: ClassDecoratorContext<any> | ClassMethodDecoratorContext<any>) {
     if (decorator.kind === 'class') {
@@ -55,7 +60,13 @@ export function Inject<K extends InjectKey>(name: K, config?: Context[K] extends
 
 /** Utilities for normalizing plugin dependency declarations. */
 export namespace Inject {
-  /** Convert array/object/class-inherited inject metadata into a plain map. */
+  /**
+   * Convert array/object/class-inherited inject metadata into a plain map.
+   *
+   * @param inject — the declaration to normalize; `null`/`undefined` add nothing.
+   * @param result — the map to fill (service name → intercept config or `null`).
+   * @returns `result`.
+   */
   export function resolve(inject: Inject | null | undefined, result: Dict = Object.create(null)) {
     if (!inject) return result
     if (Array.isArray(inject)) {
@@ -86,10 +97,15 @@ export type Plugin<T = any> =
 export namespace Plugin {
   /** Shared metadata understood by the plugin registry and related tooling. */
   export interface Base<T = any> {
+    /** Display name used for fiber diagnostics and logger names. */
     name?: string
+    /** Standard-schema validator applied to config before the plugin starts. */
     Config?: StandardSchemaV1<any, T>
+    /** Services the plugin requires; it only loads while all are available. */
     inject?: Inject
+    /** Service name(s) the plugin provides (read by `Service` and by loaders). */
     provide?: string | string[]
+    /** Service names whose intercept config the plugin declares it consumes. */
     intercept?: Dict<boolean>
   }
 
@@ -117,9 +133,13 @@ export namespace Plugin {
 
   /** Mutable registry record shared by all fibers of one plugin callback. */
   export interface Runtime {
+    /** Display name copied from the first registered plugin shape. */
     name?: string
+    /** Every live fiber of this plugin (one per `ctx.plugin()` call). */
     fibers: DisposableList<Fiber>
+    /** The executable entrypoint all fibers share (registry identity key). */
     callback: globalThis.Function
+    /** Standard-schema validator applied to each fiber's config. */
     Config?: StandardSchemaV1
   }
 }
@@ -142,7 +162,25 @@ type GetPluginConfig<P> =
 
 declare module './context.ts' {
   export interface Context {
+    /**
+     * Run a callback once the requested services are available.
+     *
+     * Shorthand for `ctx.plugin({ inject, apply: callback })`: the callback
+     * is unloaded and re-run whenever a required service changes.
+     *
+     * @param deps — required services, as an array or a name → config map.
+     * @param callback — plugin body called with `(ctx, config)`.
+     * @returns the fiber; awaiting it settles once loading finished.
+     */
     inject(deps: Inject, callback: Plugin.Function<void>): Fiber & PromiseLike<Fiber>
+    /**
+     * Load a plugin in the current context.
+     *
+     * @param plugin — a function, class, or `{ apply }` object plugin.
+     * @param args — the plugin config, validated against its `Config` schema.
+     * @returns the fiber; awaiting it settles once loading finished
+     * (rejecting on config or startup errors).
+     */
     plugin<P extends Plugin>(plugin: P, ...args: Spread<GetPluginConfig<P>>): Fiber & PromiseLike<Fiber>
   }
 }
@@ -164,15 +202,22 @@ export class RegistryService {
     })
   }
 
+  /** Allocate the next fiber uid (increments on every read). */
   get counter() {
     return ++this._counter
   }
 
+  /** Number of registered plugin runtimes. */
   get size() {
     return this._internal.size
   }
 
-  /** Resolve a supported plugin shape to its executable callback. */
+  /**
+   * Resolve a supported plugin shape to its executable callback.
+   *
+   * @param plugin — a function, class, or `{ apply }` object plugin.
+   * @returns the callback identifying the plugin, or `undefined` if invalid.
+   */
   resolve(plugin: Plugin): Function | undefined {
     // plugin.apply may throw
     try {
@@ -181,17 +226,34 @@ export class RegistryService {
     } catch {}
   }
 
+  /**
+   * Look up the runtime record for a plugin.
+   *
+   * @param plugin — any supported plugin shape.
+   * @returns the runtime, or `undefined` when the plugin is not registered.
+   */
   get(plugin: Plugin) {
     const key = this.resolve(plugin)
     return key && this._internal.get(key)
   }
 
+  /**
+   * Check whether a plugin has a registered runtime.
+   *
+   * @param plugin — any supported plugin shape.
+   * @returns `true` when at least one fiber of the plugin exists.
+   */
   has(plugin: Plugin) {
     const key = this.resolve(plugin)
     return !!key && this._internal.has(key)
   }
 
-  /** Dispose every running fiber for a plugin and remove its runtime record. */
+  /**
+   * Dispose every running fiber for a plugin and remove its runtime record.
+   *
+   * @param plugin — any supported plugin shape.
+   * @returns the removed runtime, or `undefined` when none was registered.
+   */
   delete(plugin: Plugin) {
     const key = this.resolve(plugin)
     const runtime = key && this._internal.get(key)
@@ -203,28 +265,53 @@ export class RegistryService {
     return runtime
   }
 
+  /** Iterate the registered plugin callbacks. */
   keys() {
     return this._internal.keys()
   }
 
+  /** Iterate the registered plugin runtimes. */
   values() {
     return this._internal.values()
   }
 
+  /** Iterate `[callback, runtime]` pairs. */
   entries() {
     return this._internal.entries()
   }
 
+  /**
+   * Visit every registered runtime.
+   *
+   * @param callback — receives each runtime and its identifying callback.
+   */
   forEach(callback: (value: Plugin.Runtime, key: Function) => void) {
     return this._internal.forEach(callback)
   }
 
-  /** Start a callback once the requested dependencies are available. */
+  /**
+   * Start a callback once the requested dependencies are available.
+   *
+   * @param inject — required services, as an array or a name → config map.
+   * @param callback — plugin body called with `(ctx, config)`.
+   * @returns the fiber; awaiting it settles once loading finished.
+   */
   inject(inject: Inject, callback: Plugin.Function<void>) {
     return this.plugin({ inject, apply: callback, name: callback.name })
   }
 
-  /** Start a plugin in the current context and return its fiber. */
+  /**
+   * Start a plugin in the current context and return its fiber.
+   *
+   * Creates (or reuses) the plugin's runtime record, then starts a new fiber
+   * under the current context. Throws if `plugin` is not a supported shape or
+   * if the current fiber is already disposed.
+   *
+   * @param plugin — a function, class, or `{ apply }` object plugin.
+   * @param config — the plugin config, validated against its `Config` schema.
+   * @param getOuterStack — captures the caller stack for effect diagnostics.
+   * @returns the fiber; awaiting it settles once loading finished.
+   */
   plugin(plugin: Plugin, config?: any, getOuterStack = buildOuterStack()) {
     // check if it's a valid plugin
     const callback = this.resolve(plugin)
diff --git a/vendor/cordis/src/service.ts b/vendor/cordis/src/service.ts
index 30895247c1..dc6622b68f 100644
--- a/vendor/cordis/src/service.ts
+++ b/vendor/cordis/src/service.ts
@@ -9,19 +9,36 @@ import { createCallable, joinPrototype, symbols, Tracker } from './utils.ts'
  * registered immediately and is automatically removed with the owning fiber.
  */
 export abstract class Service<out T = never> {
+  /** Symbol key of an instance method run after construction (class plugins). */
   static readonly init: unique symbol = symbols.init
+  /** Symbol key of the availability predicate passed to `ctx.provide()`. */
   static readonly check: unique symbol = symbols.check
+  /** Symbol key of the phantom intercept-config type parameter. */
   static readonly config: unique symbol = symbols.config
+  /** Symbol key of the call body making a service callable (e.g. `ctx.logger()`). */
   static readonly invoke: unique symbol = symbols.invoke
+  /** Symbol key of the helper deriving an extended service instance. */
   static readonly extend: unique symbol = symbols.extend
+  /** Symbol key of the tracker metadata used for context tracing. */
   static readonly tracker: unique symbol = symbols.tracker
+  /** Symbol key of the intercept-config resolution helper below. */
   static readonly resolveConfig: unique symbol = symbols.resolveConfig
 
   declare [symbols.config]: T
 
+  /** The service name this instance is registered under. */
   public name!: string
 
-  /** Register this instance as `name` in the current context. */
+  /**
+   * Register this instance as `name` in the current context.
+   *
+   * Calls `ctx.reflect.provide(name, this, this[Service.check])`, so the
+   * service is unregistered automatically when the owning fiber unloads.
+   * Services with a `[Service.invoke]` body return a callable instance.
+   *
+   * @param ctx — the context to register in (stored as `this.ctx`).
+   * @param name — the service name; defaults to the static `provide` field.
+   */
   constructor(protected ctx: Context, name: string) {
     name ??= this.constructor['provide'] as string
 
@@ -55,7 +72,17 @@ export abstract class Service<out T = never> {
     return Object.assign(self, props)
   }
 
-  /** Merge intercept config from ancestors with optional base and head values. */
+  /**
+   * Merge intercept config from ancestors with optional base and head values.
+   *
+   * Entries added closer to the root apply first; `base` is prepended and
+   * `head` appended. Uses `Config.merge` when the service declares one,
+   * otherwise a shallow `Object.assign`.
+   *
+   * @param base — lowest-precedence config merged before all intercepts.
+   * @param head — highest-precedence config merged after all intercepts.
+   * @returns the merged config.
+   */
   [symbols.resolveConfig](base?: T, head?: T): T {
     let intercept = this.ctx[Context.intercept]
     const configs: any[] = []
diff --git a/vitest.snapshot.config.ts b/vitest.snapshot.config.ts
index 903cd3e129..9753176f9d 100644
--- a/vitest.snapshot.config.ts
+++ b/vitest.snapshot.config.ts
@@ -1,6 +1,25 @@
+import { availableParallelism } from 'node:os'
 import tsconfigPaths from 'vite-tsconfig-paths'
 import { defineConfig } from 'vitest/config'
 
+const DEFAULT_SNAPSHOT_MAX_CONCURRENCY = 5
+
+function positiveIntFromEnv(name: string, fallback: number): number {
+  const raw = process.env[name]
+  if (raw === undefined || raw === '') return fallback
+
+  const value = Number(raw)
+  if (!Number.isInteger(value) || value < 1) {
+    throw new Error(`${name} must be a positive integer, got ${JSON.stringify(raw)}`)
+  }
+  return value
+}
+
+const snapshotMaxConcurrency = positiveIntFromEnv(
+  'DSH_SNAPSHOT_MAX_CONCURRENCY',
+  Math.min(DEFAULT_SNAPSHOT_MAX_CONCURRENCY, availableParallelism()),
+)
+
 // Replay is the keyless default: boot the real ACP subprocess from recorded model scripts and diff
 // normalized transcript plus persisted-log goldens. `record` calls the real API and updates fixtures
 // and goldens; `refresh` replays committed scripts and updates only current goldens. Replay/refresh
@@ -20,11 +39,13 @@ export default defineConfig({
   // through the root tsconfig paths map; the native option cannot do this.
   plugins: [tsconfigPaths({ projects: ['./tsconfig.json'] })],
   test: {
-    include: ['examples/*/tests/**/*.snapshot.ts'],
-    // Each test boots a subprocess; give it room, and run files one at a time
-    // (a record run hits the live API, and replay subprocess boot is heavy).
+    include: ['examples/*/tests/**/*.snapshot.ts', 'packages/sdk/*/tests/**/*.snapshot.ts'],
+    // Each test boots a subprocess; give it room and keep the worker file singular. Replay tests
+    // opt into bounded in-file concurrency, while record/refresh stay serial because they write
+    // fixtures. The environment knob restores serial replay with value 1 on constrained machines.
     testTimeout: 120_000,
     hookTimeout: 30_000,
     fileParallelism: false,
+    maxConcurrency: snapshotMaxConcurrency,
   },
 })