Tianyi Cui
e2bde2902c
Implements docs/rfc/.../2026-06-20-extract-example-app-packages.md. Each
example was thick — a hand-rolled start.ts, an infra preamble, nested
base.yml/base-core.yml/acp-tail.yml includes, and a coupled front-door
cluster enforced only by prose. This moves the composition into packages so
each example is a thin leaf cordis.yml: pick the swappable backends, load one
app package.
New packages:
- @deepseek-ai/dsh-agent-core (packages/core/agent-core): one bundle plugin
that loads the providerless/executor-less/UI-less spine (timer + llm +
sessions + system-prompt + tools + agents + invariants + tool-bash +
agent-loop) via ctx.plugin(...) inside apply(), and forwards agent-loop's
`agents` list as its own Config (export const Config = AgentLoop.Config,
default []).
- @deepseek-ai/dsh-stdio-agent (packages/ui/stdio-agent): terminal chat APP —
agent-core + console logger + readline UI + a pre-created `main` agent, with
a bin. The demo:echo/coding front door.
- @deepseek-ai/dsh-acp-agent (packages/ui/acp-agent): ACP server APP —
agent-core + JSONL persistence + the acp bridge, NO stdout logger, with a
bin. The stdout-purity footgun is structurally unreachable from the leaf.
Amendment to the RFC: hmr stays a LEAF cordis.yml entry, not baked into
dsh-stdio-agent. hmr is a Loader-only dev plugin (throws without
--expose-internals; the in-process test tier can't even import its decorator
form), so a package statically importing it 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. With hmr out, all three new
packages carry in-process unit specs at 100%.
Boot glue (Loader tail, .env load, snapshot-mode selection, stdin-dispose
lifecycle) moves into each app's bin; start.ts and base.yml/base-core.yml/
acp-tail.yml are deleted. Each app package gets a keyless real-load-path test
that boots through its bin + the cordis Loader (guarding the unwrapExports
export-shape bug class, postmortem 0001). ACP snapshot replay stays green
against the existing committed goldens (pure boot restructuring). RFC moved
proposed->implemented with the amendment recorded; package/example/architecture
docs and the module graph updated.
coding-agent
The first REAL agent wiring: DeepSeek V4 + the bash tool suite + stdio chat
- JSONL persistence, loaded from
cordis.yml. Where echo-agent proves the skeleton with mocks, this example is a usable coding assistant.
Run it
# repo root .env (gitignored) or exported env:
# DEEPSEEK_API_KEY=sk-…
# DEEPSEEK_BASE_URL=https://… # optional; defaults to the public API
pnpm run demo:coding
Type a coding task. The agent's only tools are bash (+ bash_output / bash_kill for background tasks): file reads, writes, searches, and test runs all happen through shell commands, each in a fresh bash -c (the system prompt tells the model to pass workdir instead of cd). Reasoning streams dimmed; tool calls/results render inline.
> fix the failing test in /path/to/project
[main turn 1] (reasoning…)
[tool call] bash({"command": "node --test", "workdir": "/path/to/project"})
[tool result] … [exit code: 1]
…
Resuming a prior session
Each run starts a fresh session by default (its event log lands under ./.sessions/). To continue a previous conversation, set RESUME_SESSION_ID to that session's id — the main agent then rehydrates the persisted log instead of starting fresh, so the model sees the earlier turns as history:
RESUME_SESSION_ID=<prior-session-id> pnpm run demo:coding
The id is wired through cordis.yml (resumeSessionId: !!js process.env.RESUME_SESSION_ID); unset, the agent starts a new session. A missing/unreadable id is non-fatal — it logs a warning and starts no main agent.
What each plugin demonstrates
| Entry | Demonstrates |
|---|---|
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) + tool-bash |
the executor seam + tool schemas as separate plugins |
agent-loop |
agent created from config with a coding system prompt |
session-persistence (dsh-session-persistence-jsonl) |
durable JSONL persistence (root: ./.sessions): append-only event log per session, crash-safe atomic writes — the shared backend, no per-example file |
src/stdio-chat.ts |
UI as a plugin; copied from echo-agent with reasoning-dimming and an exit-on-idle close handler for piped stdin. Example-local on purpose — extract a shared UI package when a third example needs it |
End-to-end tests (pnpm run test:e2e, key-gated)
tests/full-loop.e2e.ts— the canary: real model runsecho e2e-okthrough the real bash tool; assertstool/call/tool/resultsession events and the final answer.tests/coding-task.e2e.ts— the swebench-style smoke: a temp dir holdsadd.js(witha - bwherea + bbelongs) and a failingadd.test.js; the agent must fix the bug and verify. The test re-runsnode add.test.jsITSELF and inspects the files — agent claims are not trusted.tests/resume.e2e.ts— durable continuity across processes: run 1 tells the real model a secret code and persists the turn to a temp JSONL root, then the whole context is disposed; run 2 is a fresh context over the same root that RESUMES the session id and asks the model to recall the code. The recall can only come from the rehydrated log.
Both self-skip without DEEPSEEK_API_KEY.