Raise pr-review job timeout to 30 minutes - #108
Merged
Conversation
The opus-5 reviewer runs longer per review than opus-4-8, and the pr-review (15m) and general-pr-review (20m) caps were cancelling reviews mid-run (they surface as a 'did not finish in time' non-verdict). Raise both to 30 minutes.
Moves the pin from v1.0.130 (109 commits behind) to v1.0.185. Net-hardening: stronger untrusted-content sanitization and token redaction, plus the #1350 fix for the silent Bash-permission-widening parser bug that v1.0.130 lacked. No new default token perms or tool exposure for our config. Replaces the bare SHA with a SHA + version comment (still pinned, not @v1).
jerenkrantz
approved these changes
Aug 7, 2026
jugonzalez12
approved these changes
Aug 7, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
The reviewer model is now
claude-opus-5, which runs longer per review than opus-4-8 (and the automatic cybersecurity→opus-4.8 safety fallback can stack a second run into the same job). On the ductone side the same model bump drove judge PR-review timeouts from 0/809 runs to 20/191 (10.5%) against a 15-minute cap — cancelled mid-review, surfacing as a "did not finish in time … not a verdict" notice rather than a real verdict. This raises the caps here before the same regression shows up.What this changes
Raises
timeout-minutesto 30 on both PR-review reusable-workflow jobs:pr-review.yaml(15→30) andgeneral-pr-review.yaml(20→30). Trade-off: a longer-running or wedged review now uses up to 30 runner-minutes before being cut off.