[ThomasK33]

Summary

Fixes #7416. The test "a slow sign-out during section open does not latch journal health" failed in merge-group run 40094239818 (Test / Unit (2/6)): the pending sign-out record was still there after the test awaited the sign-out pass. The cause is test timing, not product behavior. This PR changes only the test file.

Background

• The test gave every sign-out pass a 10 ms budget (reconciliationTimeoutMs: 10). The pass deadline is the lock time plus that budget (installService.ts completePendingSignOuts).
• After release(), the fake logout runs the real oauth.logout, which writes the token store to disk. Locally that write ends 4-5 ms after the pass takes the lock. On CI it took more than 10 ms.
• A logout that outlives the deadline counts as failed, and its URL stays in the record for a later pass. That is the documented fail-closed design (signOutFailure doc comment): the record keeps discovery closed, and the next pass retries.
• The test then awaited pendingSignOutPass, which is one slot. It saw the timed-out pass settle, and no later pass cleared the record.

Implementation

1. The passes get a 2 s budget instead of 10 ms. A test comment explains why (pass deadline vs. disk write under CI load).
2. The test awaits settledSignOutPasses() before slow = true, so a fast-fail pass that is still running cannot meet the slow logout.
3. After release(), the test awaits settledSignOutPasses() (the same access pattern as the other tests in this describe), not the single slot.

The guard stays: 2 s is below the test timeout, so if recover() awaited the pass, the section-open attempt would time out and store false.

Validation

All three proofs ran on Bun 1.4.3, 3 of 3 runs each. Proofs (a) and (b) used a scratch edit that is not in this PR.

Proof	Scratch edit	Result
(a) Red	main's test + await new Promise((r) => setTimeout(r, 60)) before release()	fails at the CI assertion: Received: [ "https://mcp.example.com/mcp" ]
(b) Green	this PR's test + the same forced 60 ms wait	passes
(c) Guard	this PR's test + await this.settledSignOutPasses() after startPendingSignOutPass in recover()	fails: expect(await internals.reconciliationState).toBe(true) gets Received: false after ~2.7 s

The whole file passes (242 of 242) on Bun 1.4.3. make static-check passes.

Risks

Test-only. Under the regression in proof (c), the test now takes about 2.7 s to fail instead of failing at once.

---

Generated with xum • Model: anthropic:claude-opus-5-5 • Thinking: high • Cost: $201.75

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-24T15:38:40.556081Z	87b1165	PR opened
🔒 Security Review	✅ Completed 2026-10-24T15:38:33.843689Z	87b1165	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[ThomasK33]

Perf gate: CLEARED for head 87b1165 (perf owner). One test file only (src/node/services/agentPlugins/installService.test.ts, +11/-7). No product code, no renderer or startup path. The PR base is 8a723f7, a later main than the requested 417fa9b. That is fine.

[github-actions]

🟢 e2e stories (Haiku): 14 passed

27 agent steps · 7 replayed from cache · 31 model calls · 218.7k tokens (70% cached)
Spend: $0.10 of the $1.00 cap (proxy ledger)

All 14 tests in 9 files

	Test	Agent	Time
🟢	chatRestart.e2e.ts · 1 passed	2 steps · 4 calls	23.9s
🟢	a chat survives a restart of the server	2 steps · 4 calls	23.9s
🟢	compaction.e2e.ts · 1 passed	3 steps · 7 calls	44.7s
🟢	/compact replaces the old chat with a summary	3 steps · 7 calls	44.7s
🟢	pluginTrustRestart.e2e.ts · 1 passed	4 steps · 2 calls	9.8s
🟢	a trusted plugin hook stays trusted after the server restarts	4 steps · 2 calls	9.8s
🟢	settingsRoundTrip.e2e.ts · 1 passed	3 steps · 5 calls	29.5s
🟢	a chat survives trips through Settings	3 steps · 5 calls	29.5s
🟢	sidebarPerWorkspace.e2e.ts · 1 passed	4 steps · 1 call	4.5s
🟢	a new workspace starts with its right sidebar collapsed while another keeps it open	4 steps · 1 call	4.5s
🟢	smoke.e2e.ts · 1 passed	2 steps · 1 call	3.9s
🟢	the agent opens the playground workspace	2 steps · 1 call	3.9s
🟢	streamErrors.e2e.ts · 1 passed	1 step · 1 call	7.3s
🟢	stream errors show in the chat, and the chat recovers	1 step · 1 call	7.3s
🟢	todoNudge.e2e.ts · 1 passed	2 steps · 4 calls	24.6s
🟢	the agent finishes open TODO items after the turn-end nudge	2 steps · 4 calls	24.6s
🟢	touchAlignment.e2e.ts · 6 passed	6 steps · 6 calls	26.6s
🟢	the Settings close button is level with the title on a touch screen 390px wide	1 step · 1 call	3.5s
🟢	the Settings close button is level with the title on a touch screen 768px wide	1 step · 1 call	4.0s
🟢	Backup settings checkboxes are small and level with their labels on a touch phone	1 step · 1 call	7.6s
🟢	the heartbeat checkbox is small and level with its label on a touch phone	1 step · 1 call	5.0s
🟢	a reply's model and time stay on one line on a touch phone	1 step · 1 call	2.9s
🟢	the workspace actions menu stays on the title row on a touch phone	1 step · 1 call	3.7s

_{e2e 0.20.0 · 2m 58s · web · run artifacts}

🟢 e2e Electron stories (Haiku): 3 passed

3 agent steps · 3 model calls · 8.0k tokens (21% cached)
Spend: $0.02 of the $1.00 cap (proxy ledger)

All 3 tests in 3 files

	Test	Agent	Time
🟢	electron/deepLink.e2e.ts · 1 passed	1 step · 1 call	5.2s
🟢	a deep link of a second launch starts a chat in the running app	1 step · 1 call	5.2s
🟢	electron/menuBar.e2e.ts · 1 passed	1 step · 1 call	5.9s
🟢	the workspace menu bar is the desktop title bar	1 step · 1 call	5.9s
🟢	electron/smoke.e2e.ts · 1 passed	1 step · 1 call	4.8s
🟢	Xum Electron opens the playground workspace	1 step · 1 call	4.8s

_{e2e 0.20.0 · 17.9s · electron · run artifacts}

🟢 e2e agent repros (Haiku): no tests selected

Spend: $0.00 of the $1.00 cap (proxy ledger)

_{e2e 0.20.0 · 5.3s · 2 targets (web, phone) · run artifacts}