Repository navigation
Conversation
…l lint, fix eval-test docs - Rejected /oc-review comments now show one line in the summary and the job log with the reason; verdict and risk do not change. - Graph-off evidence ignores conditional or negated text before the phrase. - eval.yml lint rejects workflow-level and job-level defaults and step shell. - Docs and comments match live eval-test (prevent self-review off) and the approval-free eval run on main. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Audit mode and all allowlists stay. Upstream issue 708 is fixed in v2.22.4; block mode waits for the issue 24 soak. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
🔴 Hold deployment: Claude of Duty identified 1 item requiring attentionMission: 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.
Run 42313321619, attempt 1 for head Claude of Duty never approves a PR. A human reviewer gives the final approval. 🔑 Waiver: how a maintainer clears thisThis 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 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 usageReviewed snapshot
Reasons the gate failed
Checks (code, no model)
Signals (context only, never block)
Dependency changes None. Finding details
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
Token usage No token usage was recorded for any pass. |
Summary
/oc-reviewoverride comments with the reason (verdict and risk unchanged). Before, rejected overrides were silent.defaults.run.shell.eval-testenvironment has prevent self-review off (sole reviewer is the run actor); staleevals.pydocstring andrelease.pycomment fixed.1d88384) in all 7 steps; egress stays in audit mode, allowlists unchanged. v2.22.4 carries the upstream fix for Allowed domains blocked on westus2 runners: agent records one A record per domain, drops the others step-security/harden-runner#708. The owner approved pinning before the 2-day release age (recorded as K-16). Issue Deferred: re-test harden-runner v2.22.3 in block mode (phase 5) #24 stays open for the soak and the block-mode switch.Test plan
cd review && uv run pytest tests/unit: 3033 passed (3016 before)make release) to move the review rulesets to the new SHA🤖 Generated with Claude Code