The launcher's healProfilesModuleFallback BFS-links the apps/cli dependency closure into the profile's node_modules (the pwsh-local precedent): without dsh-pwsh-sandbox in apps/cli dependencies, a fresh Windows host cannot resolve the inserted row. Assert the closure reaches both inserted packages in the real-layers composition; rewrap the windows-shell layer comment.
Invocation is an ordinary session.prompt again: the pre-step gesture
boundary makes it deterministic host-side for every front end, so the
dedicated RPC (handler, wire schema, error codes, client face, fixtures)
and ui-skill's claim machinery are net deletions. The menu keeps decision
21 exactly — a pick lands literal /name text — plus the user-only marker
from skill.list's modelInvocable flag.
The web app project excludes every e2e file (they are host-plane programs)
and tsconfig.host.json includes them one by one; the new scenario joins
both lists so it keeps typecheck coverage without dragging host sources
into the client project.
The Windows platform layer re-enables the base fs-sandbox row (removing its disable), but still inserted dsh-fs-local: both extend FileSystem and provide ctx.fs, so every shipped win32 profile failed at load. Delete the insert; fs-sandbox stays the single fs provider exactly as on POSIX. Sync the roster specs, the base/reference README pairs, the sandbox core doc (read-only grants no sink on Windows), the windows-shell JSDoc, and re-record the i18n pairings.
The policy scenario now expects the user-only quadrant in the menu with
its marker (riding the description — the hint field is claim-state ghost
text, which the menu never renders), and a new skill-user-invoke scenario
drives /name args through the composer against the real host: the claim
lands skill.invoke, the transcript shows the dedicated card with the
collapsed <skill_content> body, and a paced replay answers the injected
turn deterministically.
A model entry's reasoningEfforts dict declares its selectable thinking
levels — key = offered level, value = the wire spelling dispatch sends;
only off may leave the value empty (supported, send nothing). false
strips reasoning from a catalog model; every level is materialized
explicitly into pi-ai's thinkingLevelMap so nobody has to know pi-ai's
asymmetric absent-key defaulting. compat.thinkingFormat and
compat.supportsReasoningEffort become configurable on the route and per
model (model > route > catalog entry > pi-ai's URL-derived guess),
openai-completions only, so a private gateway speaking the DeepSeek
reasoning dialect no longer depends on its URL being recognizable.
Record-typed drift gates pin both enums to pi-ai's, and an unserviceable
declaration is refused at the write that produced it, naming route,
model, and level.
Source-level environment and credential tests prove the individual loaders, but they do not prove that the published launcher runs them before Loader evaluates a shipped profile.
Start the built dsh binary with the shipped base bundle and a test-only LLM probe. Put the endpoint in $DSH_HOME/.env, put the bearer token only in $DSH_HOME/.credentials.yaml, remove inherited DeepSeek overrides, and assert the mock request received both without leaking the token. This covers launch order, profile composition, the adapter, and the credential seam without a real API.
Code already treats $DSH_HOME/.env as ordinary launch environment and stores managed credentials in .credentials.yaml, but public docs still described the old store, old precedence, removed literal adapter keys, and the deleted TUI. That directed users to the wrong file and overstated the supported configuration surface.
Update the existing English and Chinese owners in place, document inherited > managed > project > user credential resolution, and record the loadLayeredEnv export. Regenerate only pairing records and the source-line catalog; add no new section or site route.
The resolver suite previously exercised only fixture patch lists, leaving
the real shipped windows.cordis.patch.yml and the bundle→windows→user
composition ordering untested on Linux CI. The new cases load a temp
profile whose bundle layers resolve from the real dsh-base/dsh-web-app
packages (app installation anchor), apply the platform layer through the
boot's own composeEntries algorithm with the platform injected, and
assert the win32 danger-full-access roster (eight disables, three
inserts, no warnings on the web profile). A second case pins POSIX
unchanged and the base-only-profile ui-permission no-match warning as
warned-but-harmless, matching the patch header's documented contract.
- apps/cli/reference/README: state the win32 permission/sandbox/approval
degradation so the workspace-write promise no longer misleads Windows users
- bundle README + agent note: give the complete bash-restore recipe (disable
pwsh-local/tool-pwsh and re-enable bash-sandbox/tool-bash), since both
executors register the same bash service and an incomplete recipe fails
loud at load
- windows.cordis.patch.yml: header comment notes the recipe and that the
ui-permission row belongs to dsh-web-app (base-only profiles get a
harmless no-match warning)
- profile-boot.ts: rewrap composeProfile JSDoc
- re-record i18n hashes for the touched bilingual pairs
Master added this scenario while this branch was open, so its golden froze
the composer seat's "Select model" fallback. The scaffold's route-only
adapter (added here for fixture-less scenarios) makes the seat resolve the
model those scenarios actually route to, which is what the other eight
goldens on this branch already show. Only the two seat lines move.
**Reasoning effort leaves the provider cards entirely.** It is a per-MODEL
capability and the models under one provider disagree about which levels
they accept: setting `anthropic` to `max` made six of its eight models
throw UNSUPPORTED_REASONING_EFFORT, and because the catalog build catches
per provider, the whole provider vanished from the picker behind one error
row. A provider-scoped control can only ever be set to a value some of its
models reject.
The composer's model picker already offers each model its own levels, and
a switch there now records provider, model, and effort together as the
next session's default — so the setting has a better home at the right
granularity. The profile field stays in `settings.yaml` for a deployment
that knows its route; only the control is gone, from both cards and both
adapter families. Two `components.spec` cases used the control as the
vehicle for their op assertions and now use `baseURL`, which is what they
were actually testing.
**A rejected Provider ID now reads as a fault.** It shared the neutral
hint paragraph with the field's guidance, so the copy telling the user
what they got wrong looked like advice. Reuses the existing `.error`
style, matching the split the key field already makes.
Huanqi Cao
imccyu
Yichen Jiang
Turtle
Tianyi Cui