test(codex-cloud): budget the shell-policy case for Windows Git Bash spawns

[BigSimmo]

Summary

• Give codex-cloud-setup › "writes managed shell policy behaviorally and preserves unrelated Codex config" an explicit 120_000 ms budget, matching the sibling case directly below it, and record why in a comment.

This closes out the first of the five failures reported from a full npm run test on a Windows workstation. It is not the state failure it looked like.

What the Windows evidence showed

Re-running the two deterministic cases on that machine in isolation, both passed — but the timings are the finding:

✓ tests/hosted-migration-role-guard.test.ts (5 tests) 1557ms
✓ accepts the current repository state 1550ms
✓ tests/codex-cloud-setup.test.ts (34 tests) 149859ms
✓ writes managed shell policy behaviorally and preserves unrelated Codex config 24958ms
✓ pins connected retrieval mode and rejects unsafe shell-policy configs 112948ms

The failing case spawns bash scripts/setup-codex-cloud.sh twice. That is cheap on Linux and expensive on Windows, where every spawn goes through Git Bash. 24.96 s against the 30 s testTimeout in vitest.config.mts is 83% of the budget with nothing else running. A full npm run test puts four workers (maxWorkers: 4) on the machine at once, and it tips over — which is exactly why it failed there in the full suite, passed there in isolation, and passes everywhere on Linux.

The sibling case immediately below already carries }, 120_000) for the same reason, and measures 112.9 s on that machine. This one was simply missed when that budget was added.

Correcting the prior analysis

PR #2252 recorded these two cases as deterministic state assertions, and reasoned that a Windows-only failure in either would therefore be a real Windows-only defect rather than load noise. That reasoning was wrong for this case: the assertions are deterministic, but the runtime is not, because of the subprocess cost. The conclusion it led to — "prioritise item 2, it suggests real repo drift" — was also wrong. validateRepository() is happy on both platforms.

hosted-migration-role-guard › "accepts the current repository state" also passes now at 1.55 s, nowhere near its budget. Its original failure is therefore still unexplained: 1.55 s would have to degrade roughly twentyfold under load to time out, which is a stretch, and nothing in this PR addresses it. It may have been a transient condition on that machine — a stray untracked file under a guarded path would do it, since repositoryEntries() grades untracked files — that has since cleared. Left open rather than assumed fixed.

Verification

• npx vitest run tests/codex-cloud-setup.test.ts — Tests 34 passed (34)
• npm run lint — passed (4371 input files)
• npm run typecheck — passed (4371 input files)
• npm run format — clean; tests/codex-cloud-setup.test.ts (unchanged)

Verification not run: npm run test (full suite) — the full offline suite was run green on this content's parent in PR #2252 (696 passed (696) files, 7770 passed | 4 skipped); this diff changes one test's time budget and no assertion or source behaviour, so it cannot alter any other file's result.

UI verification not run: no production UI, routing, styling, or browser behaviour in the diff.

Risk and rollout

• Risk: Very low. One argument on one it() call plus a comment. No assertion changes, no source changes, no behaviour change — the test does the same work and checks the same things; only the time it is allowed to take moves. The risk of raising a timeout is that a genuine hang takes longer to surface; bounded here by the same 120 s the neighbouring case already uses.
• Rollback: git revert the single commit.
• Provider or production effects: None.
• RAG impact: none

Notes

Four of the five originally reported failures remain open:

• hosted-migration-role-guard › "accepts the current repository state" — passes now at 1.55 s; original cause unexplained, see above.
• http-readiness › "honours the overall deadline when every response remains open" — untested on Windows so far; a deadline shape, so the same load sensitivity is plausible.
• The two document-viewer-page-virtualization cases — PR test(document-viewer): name the cause when render-ahead is never scheduled, and record the five unreproducible Windows failures #2252 made these fail legibly if they recur, which is what a re-run on that machine would now reveal.

Worth running on the Windows workstation when convenient, since the timings are the diagnostic:

npx vitest run tests/http-readiness.test.ts tests/document-viewer-page-virtualization.dom.test.tsx

---

Generated by Claude Code

[supabase]

This pull request has been ignored for the connected project sjrfecxgysukkwxsowpy because there are no changes detected in supabase directory. You can change this behaviour in Project Integrations Settings ↗︎.

---

Preview Branches by Supabase.
Learn more about Supabase Branching ↗︎.

[coderabbitai]

Warning

Review limit reached

You’ve reached a temporary PR review limit under our Fair Usage Limits Policy.

Your current included review allowance is based on your included PR review attempts over the past 7 days.

Next review available in:27 minutes

Limit details: You’ve used the included review currently available. Your 88 included PR review attempts over the past 7 days set your current allowance at 1 review per hour.

Your organization has reached its usage spending cap. Adjust your spending cap in the billing tab.

How can I continue?

Wait for the limit to reset, then comment @coderabbitai review or push new commits to the PR.

An organization admin can change what happens after included review limits in Billing.

How do review limits work?

CodeRabbit enforces per-developer PR review limits within each organization.

For paid Pro and Pro+ reviews, CodeRabbit uses a developer's included PR review attempts over the past 7 days to set the current hourly allowance. At typical activity levels, the full plan allowance applies. Higher sustained activity can lower the allowance until earlier attempts leave the 7-day window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: aff41ab7-bbf5-453a-a147-b84cf9e71b86

📥 Commits

Reviewing files that changed from the base of the PR and between 20e9aaf and c962da7.

📒 Files selected for processing (1)

• tests/codex-cloud-setup.test.ts

---

_{Comment @coderabbitai help to get the list of available commands.}