[7ttp]

TL;DR

brings apps/cli/tests and apps/cli/vitest.config.ts under the effect lint

whats introduced?

• apps/cli/tests/** and apps/cli/vitest.config.ts join the allow list, replacing the seven per-file tests/helpers entries
• the test helpers, smoke tests and vitest config run on FileSystem, Path, Config and ChildProcess, and promise helpers like runSupabase keep their call shape and reject with the same values
• spawnSupabase and makeTempHome become scoped effects, and seedLegacyStateDecoys and setLinkedRefFile become effects, so their e2e and integration call sites yield them
• a timed out until run now kills its child and removes its temp home instead of leaving both behind

ref

• closes: CLI-2852

[github-actions]

Superseded by a newer AI review

🤖 AI Review

Reviewed both findings against the checked-out code and trusted conventions. The subprocess cleanup finding remains uncertain because the dependency implementation needed to verify teardown behavior is unavailable. The duplication nit is refuted under the documented testing conventions. No runtime checks were possible without installed dependencies.

Findings

Severity	Location	Category	Sources	Claim
🟠 MAJOR	apps/cli/tests/helpers/cli.ts:678	resource-lifecycle	codex	Interruption during subprocess acquisition can bypass the harness's effective cleanup, leaving a long-running CLI alive or blocking teardown.

Refuted findings (kept for transparency, not posted as review comments)

• apps/cli/tests/smoke-test-windows.ts:26 (duplication): The withFinally cleanup algorithm is repeated in live.ts, smoke-test-macos.ts, and smoke-test-windows.ts, creating a potential maintenance drift concern.
  Refuted: The repetition exists, although the implementations are not literally word for word identical. trusted/CLAUDE.md:159 explicitly accepts some duplication instead of unnecessary test abstractions, and trusted/apps/cli/CLAUDE.md:34–38 cautions against automatically extracting repeated helpers. These small local helpers have equivalent behavior; the finding identifies no existing divergence or correctness defect that overrides those conventions.

Stats

Claude findings: 1 · Codex findings: 1 · Confirmed: 0 · Refuted: 1 · Uncertain: 1

---

Models: claude-opus-5-5 + gpt-6.1-sol · Trigger: auto · Workflow run

This review runs once per PR. A maintainer can request another with a /ai-review comment.

[7ttp]

/ai-review

[github-actions]

🤖 AI Review

Verified all three findings from both independent reviews. Confirmed the minor stdin error-reporting race, refuted the storage compatibility concern, and left the cleanup finding uncertain because Effect's spawner implementation is unavailable in this checkout.

Findings

Severity	Location	Category	Sources	Claim
🟡 MINOR	apps/cli/src/commands/experimental/stack/start/start.e2e.test.ts:344	resource-cleanup	claude	Scope closure may not await the log follower's exit, despite the new comment, allowing subsequent stack cleanup to race with the exiting follower.
🟡 MINOR	apps/cli/tests/helpers/cli.ts:783	error-handling	codex	The exit path can silently drop stdin write failures because it never waits for the forked writer.

Refuted findings (kept for transparency, not posted as review comments)

• apps/cli/tests/helpers/storage.ts:318 (test-reliability): Moving fixture writes from setupStorage invocation to first layer acquisition introduces an undocumented compatibility regression that can overwrite intervening test edits.
  Refuted: The timing change is real, but storage.ts:316-317 explicitly documents initialization 'on first use'. The callers do not establish the claimed regression: for example, cp.integration.test.ts:58-77 builds the layer before executing file preparation, and ls.integration.test.ts:358-375 writes ancestor configuration before setupStorage without supplying conflicting fixture contents. No caller depending on eager fixture writes was found. trusted/CLAUDE.md:144-152 permits reshaping internal APIs based on current consumer requirements.

Stats

Claude findings: 2 · Codex findings: 1 · Confirmed: 1 · Refuted: 1 · Uncertain: 1

---

Models: claude-opus-5-5 + gpt-6.1-sol · Trigger: manual · Workflow run

This review runs once per PR. A maintainer can request another with a /ai-review comment.

[github-actions]

Flaky test check: ✅ no flaky tests

Commit 476e3b0e2c77 · focused ×10 (5 executions each), 13963 tests · integration ×10, 5621 tests · unit ×10, 8342 tests · workflow run

_{Each execution is a separate Vitest process; Where lists failing runs, as run.iteration for suites executed more than once per run. The full data is in the flaky-check-report artifact.}