[durandom]

This PR updates the fullsend reusable workflow pin from v0.38.0 to v0.43.0.

[fullsend-ai-review]

🤖 Finished Review · ✅ Success · Started 10:42 AM UTC · Completed 10:46 AM UTC

Commit: 546195c · View workflow run →

Runtime: claude · Model: claude-opus-4-6 · Cost: $0.59

[fullsend-ai-review]

Review

Reason: stale-head

The review agent reviewed commit 546195c777d0ca853d78c0fa62efa29262130b11 but the PR HEAD is now 22bfb7df63634bdbe91f0befc1b1297ea14c9ba8. This review was discarded to avoid approving unreviewed code.

[fullsend-ai-review]

/fs-review

[fullsend-ai-review]

🤖 Finished Review · ✅ Success · Started 10:47 AM UTC · Completed 10:52 AM UTC

Commit: 22bfb7d · View workflow run →

Runtime: claude · Model: claude-opus-4-6 · Cost: $0.73

[fullsend-ai-retro]

🤖 Finished Retro · ✅ Success · Started 10:48 AM UTC · Completed 10:56 AM UTC

Commit: 22bfb7d · View workflow run →

Runtime: claude · Model: opus → claude-opus-4-6 · Effort: high · Cost: $1.97

[fullsend-ai-review]

Review skipped — this PR is already merged.

The /fs-review command only reviews open pull requests.

_{Posted by fullsend post-review check}

[fullsend-ai-retro]

Retro: PR #93 — fullsend version bump v0.38.0 → v0.43.0

Timeline

1. 10:40:56Z — PR created by durandom. Two files changed (+57/−5): version bump in fullsend.yaml (disabling auto-review triggers) and new prioritize.yml workflow.
2. 10:40:59Z — First review auto-dispatched on commit 546195c (old workflow config still active).
3. 10:42–10:46Z — Review agent completed. Cost: $0.59 (claude-opus-4-6).
4. ~10:46Z — Author pushed second commit (22bfb7d), advancing HEAD.
5. 10:46:31Z — First review discarded (stale-head: HEAD moved during review).
6. 10:46:32Z — Platform auto-posted /fs-review to re-dispatch.
7. 10:47:41Z — Second review agent started on 22bfb7d.
8. 10:47:48Z — PR merged — 7 seconds after the second review started. No review result was ever posted.

Observations

Wasted work: The first review ran to completion ($0.59) and was entirely discarded. The second review was dispatched but the PR merged before it could produce results. Five additional issue_comment workflow runs were triggered by bot status comments — all skipped. Total estimated waste: ~$0.60–1.00+.

No review coverage on merged commit: The PR merged without any review — agent or human. For a well-understood config/version-bump PR, this is low risk, but the workflow changes (disabling auto-review triggers, adding a new workflow file) received no validation.

Self-healing change: Ironically, this PR fixes the very problem it experienced. After merge, future PRs on this repo will no longer auto-trigger reviews (the opened/synchronize/ready_for_review event types were removed from pull_request_target). The new /fs- prefix filter on issue_comment will also prevent bot status comments from triggering spurious workflow runs. This is an intentional repo-owner customization, not a bug.

Existing issue coverage

All findings from this retro are already well-covered by existing open issues:

• Stale-head race condition → fullsend-ai/fullsend#1331 (umbrella) plus #7236, #7214, Update adoption-insights workspace to commit d2b42a2 for backstage 1.49.3 on branch main redhat-developer/rhdh-plugin-export-overlays#2399, fix(azure-devops): add embed-package for azure-annotator-processor redhat-developer/rhdh-plugin-export-overlays#1043. This PR provides another data point: $0.59 wasted on a completed review that was discarded.
• Cancel in-flight review when PR merges → fullsend-ai/fullsend#2388. The second review continued running after the PR merged.
• Skip review for fullsend version bump PRs → fullsend-ai/fullsend#7357. This PR is a pure config change where review adds minimal value.
• Require at least one review before merge → rhdh-parasol/rhdh-plugin-export-overlays#17. Branch protection requiring a review would prevent the merge-before-review-completes scenario.

No new proposals are warranted — filing them would create duplicates of well-tracked issues.