Repository navigation
Conversation
…oken (INF-5559) The Review step was fully advisory (continue-on-error, no step id), so a broken whip harness (failed install, 404ing release asset, renamed binary) left the check green with no comments. Gate the check instead: the pinned whip install is no longer continue-on-error; the Review step gets id loupe-run and stays advisory, and a new Harness health gate step fails the check when the action's status output is failed. quota/rate-limit remain advisory.
The INF-5559 fix in the loupe dogfooding workflow (PR context-labs#59) only guards loupe's own repo. The same silent-green gap exists for every consumer that follows the recommended advisory posture: `continue-on-error: true` turns a broken whip harness (missing/404ing/renamed binary, bad install) into a green check with no comments, so a PR looks reviewed when nothing ran. A step cannot fail its own job from inside an action, so the general fix is a documented gate step in the consuming workflow. Bake the pattern into the 'Recommended posture' section of docs/github-action.md: keep the Review step advisory with an id, then a Harness health gate step fails the job only `when steps.<id>.outputs.status == 'failed'` (harness/setup breakage); quota and rate-limit stay advisory as provider-side issues. Cross-reference it from the existing 'Branching on failure' section.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes INF-5559.
Why
The loupe self-review workflow was fully advisory: the Review step was
continue-on-error: truewith no step id, so the action'sstatusoutput was never read. When the whip harness broke (install failure, 404ing release asset, renamed binary — see #52-#57), the check stayed green and no review comments appeared.What changes
continue-on-error— a failed deterministic install fails the check directly.id: loupe-run(sostatusis readable) and stayscontinue-on-error, so review findings remain advisory comments.steps.loupe-run.outputs.status == 'failed'(harness/setup breakage).quotaandrate-limitremain advisory — provider-side issues, not a broken harness.This distinguishes "we could not review because the harness is broken" (red check, must fix) from "review completed" (green, advisory).
Note
Low Risk
CI and documentation only; no runtime product or auth/data-path changes.
Overview
Advisory loupe reviews no longer show green when the harness never ran. The self-review workflow treats a broken whip install as a hard failure and adds a follow-up gate when the Review step reports harness/setup failure via
status.The Install whip step drops
continue-on-errorso a pinned deterministic download failure fails the check immediately. The Review step keepscontinue-on-error(findings stay comments-only) but gainsid: loupe-runso workflows can readstatus. A new Harness health gate runs onalways()and fails only whenstatus == 'failed', leavingquotaandrate-limitadvisory.docs/github-action.md documents the same pattern: advisory reviews plus a consuming-workflow gate, since the action cannot fail its own job from inside the composite step.
Reviewed by Cursor Bugbot for commit bc74bea. Bugbot is set up for automated code reviews on this repo. Configure here.