Context
Identified by DeepReport intelligence briefing 2026-05-12 (workflow run §25744758445).
The Agent Performance Report (discussion #31690) reports Daily Fact About gh-aw has experienced 15+ consecutive failures, with 3/3 runs failing in the latest window. Issues #31432 and #31524 are still open and PR #31411 was merged but failures continue. The workflow has no circuit-breaker — it keeps retrying endlessly with zero useful outputs.
Expected Impact
- Stop daily noise from failure issues being filed for the same root cause
- Free up Copilot runner minutes (estimated 3 wasted runs/day × ~minutes per run)
- Force human attention on the underlying parse failure rather than auto-retrying
Suggested Approach
- Add a
concurrency.cancel-in-progress: false and a if gate that skips runs when the last N runs were all failures
- Or add a small step early in the workflow that checks recent run history via
gh run list and exits cleanly if N consecutive failures detected
- Document the pattern so other long-running daily workflows can adopt it
Suggested Agent
Any coding agent — small focused workflow change in .github/workflows/daily-fact-about-gh-aw.md.
Estimated Effort
~1 hour of agent time.
Source: DeepReport 2026-05-12 #aw_dr1
Generated by DeepReport - Intelligence Gathering Agent · ● 9.5M · ◷
Context
Identified by DeepReport intelligence briefing 2026-05-12 (workflow run §25744758445).
The Agent Performance Report (discussion #31690) reports Daily Fact About gh-aw has experienced 15+ consecutive failures, with 3/3 runs failing in the latest window. Issues #31432 and #31524 are still open and PR #31411 was merged but failures continue. The workflow has no circuit-breaker — it keeps retrying endlessly with zero useful outputs.
Expected Impact
Suggested Approach
concurrency.cancel-in-progress: falseand aifgate that skips runs when the last N runs were all failuresgh run listand exits cleanly if N consecutive failures detectedSuggested Agent
Any coding agent — small focused workflow change in
.github/workflows/daily-fact-about-gh-aw.md.Estimated Effort
~1 hour of agent time.
Source: DeepReport 2026-05-12 #aw_dr1