chore: collect maintenance notes and stabilize checks

[steipete]

Collect the batch-six notes in the existing 0.59.1 Unreleased section, ordered by user impact, and fix a process-check setup flake observed repeatedly on a loaded Mac.

The Claude usage-retry fixtures now set their authenticated-CLI prerequisite through the existing task-scoped test hook, removing an unrelated shell startup and its potential 30-second timeout. Retry counts, timeouts, and dedicated authentication-probe tests are unchanged.

The sentinel's interpreter/file startup now gets up to ten seconds before cleanup timing begins. The actual two-second cleanup timeout, bounded teardown, child-ownership checks, and unrelated-sentinel assertions are unchanged. The 101-test process-cleanup suite passes with real subprocesses, and independent P0–P2 review is clean.

This is the only batch PR that changes CHANGELOG.md. Merge after #3539 (Linux Cursor), #3566 (Codex catch-up timing), #3547 (opt-in host hiding), #3525 (Windows companion link), and #3544 (AI Monitor link). Existing notes and contributor credit are retained; released sections are unchanged.

No newly stale small dependency was found beyond #3564's refresh. Previously held compiler/major-version migrations remain separate. The queue still needs another batch, so this does not authorize a tag or release. Exact-head CI passed: https://github.com/steipete/CodexBar/actions/runs/34659109403. The exact notes head also passed 27 focused Swift tests across retry, auth-status, and background-availability suites.

[clawsweeper]

🦞👀
ClawSweeper picked this up.

Pull request received. I will update this pull request when review starts.

ClawSweeper review complete

ClawSweeper finished reviewing this revision. The review result is being finalized.

View the workflow run.

[clawsweeper]

Codex review: blocked before merge. Reviewed September 11, 2026, 8:33 PM ET / September 12, 2026, 00:33 UTC (Revision 4).

ClawSweeper review

What this changes

Updates unreleased maintenance notes, isolates Claude retry fixtures from authentication subprocess startup, and gives the process-cleanup sentinel more startup time.

Merge readiness

⛔ Blocked before merge - 2 items remain

The patch is focused and has no actionable correctness findings. It remains useful beyond current main, and the owner's explicit landing order still applies: all five prerequisite PRs remain open.

Priority: P3
Reviewed head: 747020c95cb21411110f126d4447d591758fb63f

Review scores

Measure	Result	What it means
Overall readiness	🐚 platinum hermit (4/6)	A focused maintenance patch with preserved assertions and reported passing validation; landing order remains an operational prerequisite.
Proof confidence	🌊 off-meta tidepool	Not applicable: The owner-authored maintenance PR is exempt from contributor runtime proof. Reported subprocess and Swift runs target the changed fixtures; the inspected desk-display image belongs to a related integration and does not prove these changes.
Patch quality	🐚 platinum hermit (4/6)	No actionable review findings were identified.

Verification

Check	Result	Evidence
Real behavior	Not applicable	Not applicable: The owner-authored maintenance PR is exempt from contributor runtime proof. Reported subprocess and Swift runs target the changed fixtures; the inspected desk-display image belongs to a related integration and does not prove these changes.
Evidence reviewed	6 items	Complete policy and introduced diff inspected: Read the full root AGENTS.md; no nested AGENTS.md or maintainer-notes directory was present. The complete pinned diff changes only the changelog and two test files; released changelog sections and production code are untouched. Authentication prerequisite remains isolated: The replacement uses the existing TaskLocal result override, which returns before subprocess execution. Retry counts, timeout assertions, and separate auth-status subprocess tests remain intact. Cleanup timing remains bounded: Only sentinel setup changes from the helper's three-second default to ten seconds. Timing starts afterward; command timeout, child termination, unrelated-sentinel survival, and nine-second bounded cleanup assertions remain unchanged.
Findings	None	None.
Security	None	None.

How this fits together

CodexBar's checks exercise provider retry behavior and subprocess cleanup before changes ship. This patch adjusts test setup and collects release notes without changing application behavior.

flowchart TD
 A[Claude retry fixtures] --> B[Task-scoped authentication prerequisite]
 B --> C[Existing retry assertions]
 D[Process sentinel startup] --> E[Existing timed cleanup assertions]
 F[Related code and documentation PRs] --> G[Unreleased maintenance notes]

Loading

Before merge

• Resolve merge risk (P1) - Landing before the five prerequisites would advertise behavior and integration links that have not landed.
• Complete next step (P2) - Land only after fix(linux): restore Cursor app authentication #3539, fix(codex): exclude queue waiting from catch-up delays #3566, feat(sessions): make failed-host hiding opt-in #3547, docs: add CodexBar for Windows #3525 and README: add AI Monitor desk display to integrations #3544 have merged.

Agent review details

Security

None.

Review metrics

Metric	Value	Why it matters
Landing prerequisites	5 open, 0 merged	The owner requires these code and documentation changes to precede their centralized notes.

Merge-risk options

Maintainer options:

1. Decide the mitigation before merge
   Keep the focused fixture changes and publish the consolidated notes after their corresponding code and documentation land.
2. Pause or close
   Do not merge this PR until maintainers decide whether the risk is worth taking.

Technical review

Best possible solution:

Keep the focused fixture changes and publish the consolidated notes after their corresponding code and documentation land.

Do we have a high-confidence way to reproduce the issue?

Not applicable to a product bug: this patch stabilizes test prerequisites. The owner reports successful subprocess and focused Swift runs; this read-only review did not execute them.

Is this the best way to solve the issue?

Yes. Reusing the task-scoped authentication hook and extending setup time preserve the assertions that actually test retry and cleanup behavior.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning medium; reviewed against 830cdd86c4b6.

Labels

Label justifications:

• P3: This is bounded test stabilization and unreleased-note maintenance with no application behavior change.
• rating: 🐚 platinum hermit: Overall readiness is 🐚 platinum hermit; proof is 🌊 off-meta tidepool and patch quality is 🐚 platinum hermit.
• status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Not applicable: The owner-authored maintenance PR is exempt from contributor runtime proof. Reported subprocess and Swift runs target the changed fixtures; the inspected desk-display image belongs to a related integration and does not prove these changes.

Evidence

What I checked:

• Complete policy and introduced diff inspected: Read the full root AGENTS.md; no nested AGENTS.md or maintainer-notes directory was present. The complete pinned diff changes only the changelog and two test files; released changelog sections and production code are untouched. (AGENTS.md:1, 747020c95cb2)
• Authentication prerequisite remains isolated: The replacement uses the existing TaskLocal result override, which returns before subprocess execution. Retry counts, timeout assertions, and separate auth-status subprocess tests remain intact. (Sources/CodexBarCore/Providers/Claude/ClaudeCLIAuthStatusProbe.swift:54, 747020c95cb2)
• Cleanup timing remains bounded: Only sentinel setup changes from the helper's three-second default to ten seconds. Timing starts afterward; command timeout, child termination, unrelated-sentinel survival, and nine-second bounded cleanup assertions remain unchanged. (Scripts/test_swift_test_process_cleanup.py:253, 747020c95cb2)
• Owner-confirmed sequencing and validation: The owner explicitly requires landing last after fix(linux): restore Cursor app authentication #3539, fix(codex): exclude queue waiting from catch-up delays #3566, feat(sessions): make failed-host hiding opt-in #3547, docs: add CodexBar for Windows #3525 and README: add AI Monitor desk display to integrations #3544. Separate REST reads confirmed all five remain open and unmerged. The supplied comment reports 101 subprocess tests, 27 focused Swift tests, make check, and successful CI for this head: chore: collect maintenance notes and stabilize checks #3567 (comment). (CHANGELOG.md:10, 747020c95cb2)
• Main and release comparison: The pinned main diff still contains both fixture improvements and the additional notes. The supplied latest release, v0.59.0, also uses the earlier sentinel startup wait; this maintenance change is not already shipped. (Scripts/test_swift_test_process_cleanup.py:252, dca9c5f4b7a2)
• Area history and review limits: Main history connects Peter Steinberger to earlier fixture-startup and Claude fallback work. Detailed historical blame/patch inspection encountered unavailable objects, so no source-line introduction attribution is claimed. Builds and tests were not rerun under the read-only review contract; git diff --check passed and the checkout remained clean. (Scripts/test_swift_test_process_cleanup.py, 1680b4ed5bca)

Likely related people:

• steipete: Suggested for follow-up; no historical authorship or introduction is verified. (role: unverified routing candidate; confidence: low)

Rating scale

Score	Internal tier	Crab rank	Meaning
6/6	S	🦀 challenger crab	Exceptional readiness
5/6	A	🦞 diamond lobster	Very strong readiness
4/6	B	🐚 platinum hermit	Good normal PR; ordinary maintainer review
3/6	C	🦐 gold shrimp	Useful, but confidence is limited
2/6	D	🦪 silver shellfish	Proof or implementation needs work
1/6	F	🧂 unranked krab	Not merge-ready
N/A	NA	🌊 off-meta tidepool	Rating does not apply

Overall follows the weaker of proof and patch quality.
Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

Workflow

• ClawSweeper keeps one durable marker-backed review comment per issue or PR.
• Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
• A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
• PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
• Maintainers can also comment @clawsweeper review to request a fresh review only.
• Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
• Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
• Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.

History

Review history (3 earlier review cycles)

• reviewed 2026-09-11T23:17:10.563Z sha 4e88b77 :: blocked before merge. :: none
• reviewed 2026-09-11T23:33:51.687Z sha aeba0e6 :: blocked before merge. :: none
• reviewed 2026-09-11T23:46:54.076Z sha 747020c :: blocked before merge. :: none

[steipete]

Verified head 747020c95cb21411110f126d4447d591758fb63f: exact-head CI passed at https://github.com/steipete/CodexBar/actions/runs/34659109403.

The real-subprocess cleanup suite ran 101 tests successfully (one existing platform skip). Its measured cleanup deadlines, ownership checks, and unrelated-sentinel checks remain intact. On this exact notes head, 27 Swift tests across Claude retry, auth-status, and background-availability suites passed; the retry fixtures no longer start an unrelated authentication shell. Make check and final independent P0–P2 branch review pass.

Released changelog sections were compared byte-for-byte and are unchanged. The Unreleased section centralizes the batch's behavior notes and contributor credit. No additional safe dependency bump was needed after #3564.

Ready to squash last, after #3539, #3566, #3547, #3525 and #3544. The sequencing warning is intentional: do not put their notes on main before their code/docs. No release or tag is authorized by this PR.