Repository navigation
chore(security): regenerate the approved publish-workflow revisions - #5276
Conversation
@coderabbitai full review |
|
|
Important Review skippedThis PR was authored by a bot without an assigned CodeRabbit review seat. Ask an organization administrator to assign a seat to the bot. See the seat-assignment guide. After the seat is assigned, request a review again.
|
CodeRabbit’s authenticated response to this request reports no review for this bot-authored PR (cause class: quota/billing). The no-gate guard independently returned ADMIT for the unchanged head. |
@codex review |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
Codex Review: Didn't find any major issues. Breezy! Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
Evaluated the complete one-file change at a48007a as the release maintainer. It refreshes the observed publication inventory; the approved identity set, executable workflow and runtime matchers do not change. The newly recorded publisher revision was already approved for this consumer. Observed the real generator run 40071335218: it resolved the deployed publications and default-branch pins, verified all five consumer matchers, and created this exact signed head and PR. Current-head CI 40078638228 passed the repository guards, regeneration controls and 53 ablations. Its rendered-source guard verified all nine repositories retain signature verification; the matcher job passed actual package signatures and both identity-negative controls. All 39 reported check runs are success/skipped, all readable required gates passed, and there are zero review threads or body findings. The native code-quality rule remains UNVERIFIED by the helper; GitHub reports CLEAN and remains the authority for that rule. Codex completed a clean review at this exact head (receipt). The programmed-updater classifier does not recognize this publisher-refresh shape, so I used the complete static review and hosted-behaviour path and executed no PR branch locally. No new approval or trust exception is inferred. Ready for the protected merge queue; deployment and post-merge verification remain outstanding. |
Post-merge verification completed for source The protected merge-group CI and deployment succeeded, including the actual publication-signature matcher checks. All ten actual-main validation jobs passed. Its production-convergence job reported |
Why
The cluster only trusts releases signed by publishing-workflow revisions it
has approved. A consumer's deployed release or workflow pin has moved,
so the approved list is out of date.
What
Refresh the approved list and each consumer's matcher from what is deployed
and what each consumer currently pins. Every consumer resolved on both halves,
and each matcher accepts exactly its approved revisions.
No issue: trivial fix.