feat(subprocess): migrate lsp-local, subagent-acp, and the env scrubs onto the seam

Review direction (tianyicui, PR #660): in a stacked PR, change all other
process-running places to use the new service.

- lsp-local: LspConnection spawns through ctx.subprocess (piped protocol
  streams + a no-spill collected stderr tail); its private process-tree
  helpers (POSIX group signalling, Windows taskkill, liveness polling) are
  deleted in favor of the seam's handle verbs, and its buildChildEnv now
  rides scrubbedParentEnv (LSP children also stop inheriting stale DSH_*).
  The plugin injects 'subprocess'; compositions/tests mount
  dsh-subprocess-local.
- subagent-acp: the ACP child spawns through the seam (piped ndjson streams,
  inherited stderr); spawn failure surfaces through done-rejection into the
  same startup race; disposal is handle.dispose with the plugin's configured
  graces. dsh-subagent-subprocess is DELETED — its dispose ladder and scrub
  are the seam's, and the isolated-config-dir helper had no consumer.
- mcp-client, pty-local, sdk-helper: adopt scrubbedParentEnv as the one
  scrub definition (their spawns stay put by ownership: the MCP SDK and
  node-pty own those calls; the SDK wizard runs outside any composition).
- Coverage: per-file 100% over every touched src file, with each v8 ignore
  carrying a platform or contract reason; new suites cover stdio
  dispositions, the dispose ladder tiers, injected-win32 tree semantics,
  waitForExit, settled-kill/terminate no-ops, and spawn-failure disposal.
- Docs: consumer-migration Agent Note (en; zh follows in this PR), seam note
  updated in place, subprocess.md rewritten for the reshaped vocabulary
  (type-equiv re-registered), READMEs and SERVICE_ROLES updated, taskkill
  added to knip ignoreBinaries.
This commit is contained in:
Tianyi Cui
2026-07-26 15:27:59 +08:00
parent fd7e6d4d11
commit 3672cd25b4
57 changed files with 750 additions and 1225 deletions

View File

@@ -4,13 +4,14 @@ The subprocess seam (`ctx.subprocess`). The abstract `SubprocessService` exposes
## Contract
- `spawn(spec)` returns immediately with a live handle; `done` resolves at process close and rejects only for spawn-level failures.
- The spec is fully explicit — argv, cwd, per-stream byte caps, spill cap, grace — because deployment-varying defaults belong to the calling seam's config, not to a hidden subprocess-service default (the `dsh-bash` request/spec split is the owning template). `argv` is never shell-interpreted here; a consumer that wants a shell passes `['bash', '-c', command]` itself.
- Output readers take whole-stream byte offsets and never consume: independent readers cannot steal one another's deltas. A read whose offset slid out of the in-memory tail is `lossy` and points at the full-stream spill file when one exists.
- `kill()` and the spec's abort signal escalate SIGTERM→grace→SIGKILL across the whole detached group; the service reacts to the abort but never classifies why (callers own deadlines and cause classification).
- Disposal kills all still-running managed processes and awaits their exit.
- `spawn(spec)` returns immediately with a live handle; `done` resolves at process close with exit facts (`SubprocessOutcome` carries no output and no cause classification) and rejects only for spawn-level failures.
- The spec is fully explicit — argv, cwd, per-stream stdio dispositions, grace — because deployment-varying defaults belong to the calling seam's config, not to a hidden subprocess-service default (the `dsh-bash` request/spec split is the owning template). `argv` is never shell-interpreted; a consumer that wants a shell passes `['bash', '-c', command]` itself.
- Stdio is Node-shaped per stream: `'pipe'` hands the caller the raw stream for its own protocol framing (LSP JSON-RPC, ACP ndjson), `'inherit'` passes the parent descriptor through for diagnostics, and collect mode (`{ maxBytes, spill? }`) buffers a bounded tail with an optional full-stream spill file. Collect readers take whole-stream byte offsets and never consume, so independent readers cannot steal one another's deltas; a read whose offset slid out of the in-memory tail is `lossy` and points at the spill file when one exists. Collected output stays readable after settlement.
- Termination is tree-scoped on every platform (POSIX detached groups with direct-child fallback; Windows `taskkill /T`): `kill(signal)` sends one signal Node-style and is a no-op after settlement, `terminate()` (and the spec's abort signal) escalates SIGTERM→grace→SIGKILL, `waitForExit()` observes the whole tree, and `dispose(graces)` runs the cooperative stdin-EOF→SIGTERM→SIGKILL ladder out-of-process children need — the manager reacts but never classifies why (callers own deadlines and cause classification).
- `scrubbedParentEnv()` / `SENSITIVE_ENV_PATTERN` are the one shared scrub definition: ambient credential-shaped and `DSH_*` names are dropped, explicit `env` merges after the scrub (a deliberately forwarded key survives), and `dshEnv` carries current harness facts on its own validated channel. Spawners that cannot route through the service (node-pty backends, SDK-managed transports) import the function.
- Disposal of the service terminates all still-running managed processes and awaits their exit.
See the [process data-structure catalog](../../../docs/core-data-structures/subprocess.md) and the [seam Agent Note](../../../.agents/notes/implemented/architecture/2026-07-26-subprocess-seam.md).
See the [subprocess data-structure catalog](../../../docs/core-data-structures/subprocess.md) and the [seam Agent Note](../../../.agents/notes/implemented/architecture/2026-07-26-subprocess-seam.md).
## Model Experience
@@ -22,5 +23,5 @@ No direct invalidation; the named consumers own any request-prefix changes.
## Known Limitations and Deferred Work
- **One consumer family so far** — the seam's shape is proven against the bash executors only; the other in-repo spawn sites (LSP servers, PTY backends, subagent transports) keep their own bespoke process handling until their stream/lifecycle needs are re-examined against this contract.
- **POSIX group semantics are assumed** — the handle vocabulary (`pid` as group leader, group kills, SIGTERM/SIGKILL escalation) has no Windows story.
- **node-pty and SDK-managed spawns share only the scrub** — the PTY backend's terminal fork and the MCP SDK's own stdio transport cannot route their spawns through this seam (the library owns the fork/spawn call); they import `scrubbedParentEnv` so the environment policy stays single-sourced.
- **The dispose ladder assumes stdin-EOF cooperation** — a child that quiesces on a different signal (SIGHUP conventions, control sockets) needs its own tier-1 before the generic ladder fits.