[proerror77]

Change

CI policy contracts ran test-release-once.sh twice on the same checkout: through the selector suite and through the release artifact suite. Remove the selector's duplicate invocation. The release artifact suite retains all release admission and source identity cases. Workflow selection, required gates, production scripts, and test assertions are unchanged. Standalone selector tests now cover scheduling without also running the release admission suite.

The previous CI history report ranked only four installation candidates. This is the next independently actionable improvement after reassessment, not a previously measured fifth-place ranking. Large research builds, nextest shards and collector parallelization remain owned by other PRs.

Issue relationship

None

Validation

• Base: a84531f51092f9dc52bbee2efac5edcbce8c2778. Full selector suite passed before and after: 87.417 s → 60.259 s locally, a single sequential pair (27.158 s saved; timing varies by host/load).
• Unchanged release artifact suite passed; standalone release admission suite passed in 18.708 s locally.
• Injected exit 73 separately into retained release admission and source identity tests. Both failures propagated through the actual CI policy step extracted from YAML. Unrelated test scripts were stubbed for these failure-propagation checks.
• git diff --check passed. Diff review confirms only the redundant invocation and its explanatory comment changed.
• Previous PR CI and main CI each logged two release admission completions and two source identity completions. Each repeated admission segment spans at least 2.73–3.17 s in hosted logs; local timings must not be extrapolated to hosted runners.
• New PR CI: exactly one release admission completion and one source identity completion; retained artifact and selector suites passed. All three required gates passed for 3003131fbba2d03bddc42100e7f858b6a3da959c.
• The complete hosted policy step took 179 s, versus 170 s and 164 s in the two previous runs. This does not establish a reduction in total hosted job latency. The verified saving is one redundant test execution (a roughly 3 s visible segment on prior hosted runners), plus the local paired selector observation above. Host variation and other tests dominate the end-to-end comparison.

Runtime impact and rollback

CI test orchestration only. Revert this commit to restore the duplicate execution. Existing automatic publication selection is unchanged. No production transition is part of this change.

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

• Configuration used: defaults
• Review profile: CHILL
• Plan: Advanced
• Run ID: 0c1811d3-0d67-472a-b9fc-8d098ddde45d

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[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 2026-10-16T07:29:23.663514Z	3003131	PR opened
🔒 Security Review	✅ Completed 2026-10-16T07:29:10.185520Z	3003131	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.