[irvingpop]

Summary

• Summary comment: the TIP box now names the signing step when unsigned commits exist; the unsigned-commit row uses correct singular/plural forms; the "Dependency graph disabled" setting label needs one narrow affirmative phrase (negated text no longer matches); collect.py catches http.client.IncompleteRead and still fails closed.
• Eval environment split: a no-permission select-environment job picks eval only for exactly refs/heads/main; every other ref (including MAIN, tags, PR refs) gets eval-test. Federation values are per environment; the eval-test rule id is unset, so branch evals fail closed until the owner creates the rule. Boundary lint updated.
• Phase 6 scope (owner decisions 2026-11-05): phase 4 exit evidence = 5 real PRs, 2 weekly reports, 7 clean days; START_HERE is not a target (to be archived, Move START_HERE content into the org .github repo and archive START_HERE #27); the pilot bypass stays (gate switch is follow-up work); security configs recorded as live, S1 skipped; final acceptance checks rewritten for this shape.
• Pins: harden-runner v2.22.2 -> v2.22.3 (audit mode kept; Deferred: re-test harden-runner v2.22.3 in block mode (phase 5) #24 stays open, the westus2 fix is not released yet); setup-uv v10.3.2 -> v10.3.3 with releases.astral.sh added to the setup-uv job allowlists.

Review

Adversarial review (Codex) found 3 Important + 1 Minor; all fixed in a8a8de4 and closed by a scoped re-review. Unit tests: 2928 -> 3016 passed. All commits signed.

After merge (owner)

1. Create environment eval-test (required reviewer, prevent self-review).
2. Create the Anthropic Console federation rule for the eval-test subject.
3. Small PR that sets the eval-test rule id.
4. Release (comment, collect and pin changes reach the pilot repos only after a release).

🤖 Generated with Claude Code

[github-actions]

🔴 Hold deployment: Claude of Duty identified 1 item requiring attention

Mission: Review the code. Protect the branch. Deploy the future.

Caution

Out of range. This change is too large for the model review. File checks ran, but model checks did not. A maintainer can waive this.
Next step for the author: split the change, or ask a maintainer for a waiver (under "Waiver" below).

Severity	Effect	Where	Finding	Found by
⚠️ Not run	Blocks, waivable	Model passes	Too large for automatic review. File checks ran, model checks did not.	Workflow (code, no model)

File checks	Review	Security	Supply chain	Dependency review	Coverage
✅ ok, 7 signals	⚠️ skipped, too large	⚠️ skipped, too large	⚠️ skipped, too large	✅ ok	✅ complete

_{Run 42143423528, attempt 2 for head a8a8de4 into main. The required workflow in the merge box is the result that counts.}

Claude of Duty never approves a PR. A human reviewer gives the final approval.

🔑 Waiver: how a maintainer clears this

This PR fails for waivable reasons only. A person with write or admin access, who is not the PR author, posts this exact line as a new comment and replaces <reason> with the reason. Then they re-run the newest run of the required workflow through the GitHub REST API (POST .../actions/runs/{run_id}/rerun). Do not edit the comment afterwards. Do not close and reopen the PR to apply the waiver.

/oc-review override a8a8de4c8d23f8e27cbacc0e5652f7bc41a2419f 364058df01f84ffe70d1c9ea6cd60a7930fa81b7 <reason>

A waiver covers the code at this exact head and merge base. Any new push needs a new waiver.

📋 Full review record: reasons, checks, signals, finding details, model summaries, token usage

Reviewed snapshot

Head SHA	Merge base SHA	Base branch	Run
a8a8de4c8d23f8e27cbacc0e5652f7bc41a2419f	364058df01f84ffe70d1c9ea6cd60a7930fa81b7	main	42143423528, attempt 2

Reasons the gate failed

Reason	Waivable?	Detail
Change too large for the model review	✅ yes	too large for automatic review

Checks (code, no model)

Check	Result	Detail
Hard blocks	✅ none
Coverage	✅ complete	files in comparison 40, files collected 40, trees complete
Dependency changes	✅ none
Dependency review job	✅ success

Signals (context only, never block)

• ⚠️ Sensitive path changed: change to workflow, in .github/workflows/claude-review.yml
• ⚠️ Sensitive path changed: change to workflow, in .github/workflows/eval.yml
• ⚠️ Sensitive path changed: change to workflow, in .github/workflows/test.yml
• ℹ️ Very long line: lines longer than 500 characters, in docs/build-plan.md, lines 61 110 246 247
• ℹ️ Text aimed at an AI reviewer: pattern reviewer-address, in docs/claude-review-design.md, lines 706
• ℹ️ Very long line: lines longer than 500 characters, in docs/claude-review-design.md, lines 43 199 211 611 614 615 693 694 711 762 and more
• ℹ️ Credential-like string: credential-assignment (value redacted), in review/tests/unit/test_workflows.py, lines 942

Dependency changes

None.

Finding details

Model output from untrusted input. Check it before you act on it. Verified means only that the quoted text is at that place. It does not mean the claim is true.

Blocking is decided by code from category and severity, never by the model. A finding in a blocking category blocks only at high or critical severity.

No findings from the passes that ran. The Review, Security, Supply chain pass did not run, so this is not a clean result.

Model summaries

Model output from untrusted input. Check it before you act on it. Verified means only that the quoted text is at that place. It does not mean the claim is true.

Pass	Status	Model	Summary
Review	❌ skipped-size		No summary. Reason: bundle status skipped-size
Security	❌ skipped-size		No summary. Reason: bundle status skipped-size
Supply chain	❌ skipped-size		No summary. Reason: bundle status skipped-size

Token usage

No token usage was recorded for any pass.

[irvingpop]

/oc-review override a8a8de4 364058d because