[self-care:open-source-failures] Open-source workflow failure digest for 2026-09-05

Snapshot 2026-09-05T03:27:54.537Z covering the last 168 hours found 958 failed runs across 7 public projects. The newest 100 failed runs were available, so this digest is based on a bounded sample and may undercount older recurring failures.

**Action:** CAO maintainers should fix the target-authority regressions first; accept when new live-mode precompute runs for the affected workflows stop failing with target-authority errors.

Critical findings:
- 3 P0 clusters, 10 P1 clusters, and 24 P2 clusters were identified from the bounded sample.
- The top untracked blocker is a cross-workflow live-mode precompute failure in `githubnext/gh-aw-cao` caused by missing or invalid target-authority declarations, observed in 20 runs across AW Doctor and Dependabot workflows.
- Additional recurring P1 clusters include 5 safe-output processing failures in `Dev Practices / Well-Architected`, 4 GitHub App token generation failures in `AW Optimization / AI Credit Savings`, and 4 activation artifact upload failures across `githubnext/gh-aw-workshop` workflows.
- Existing coverage matching this worker's `[self-care:open-source-failures]` prefix: 0 open issues found.

Prioritized remediation list

1. `githubnext/gh-aw-cao` — missing or invalid live-mode target-authority declarations across AW Doctor and Dependabot workflows (P0, 20 runs).
2. `githubnext/gh-aw-cao` — safe-output processing failures in `.github/workflows/software-development-practices-github-well-architected.lock.yml` (P1, 5 runs).
3. `githubnext/gh-aw-cao` — GitHub App token generation failures in `.github/workflows/optimization-ai-credit-optimizer.lock.yml` (P1, 4 runs).
4. `githubnext/gh-aw-workshop` — activation artifact upload failures across four workflows (P1, 4 runs).

Cluster summary

- P0
- `githubnext/gh-aw-cao` — missing target authority for live mode; 18 runs; workflows: `aw-failures-investigator`, `aw-maintenance-compiler-security`, `aw-maintenance-upgrade`, `dependabot-release-train-updater`; representative run: https://github.com/githubnext/gh-aw-cao/actions/runs/33940267555
- `githubnext/gh-aw-cao` — invalid target authority declaration for aw-doctor; 1 run; representative run: https://github.com/githubnext/gh-aw-cao/actions/runs/33940189340
- `githubnext/gh-aw-cao` — invalid target authority declaration for dependabot; 1 run; representative run: https://github.com/githubnext/gh-aw-cao/actions/runs/33939429479
- P1
- `githubnext/gh-aw-cao` — safe outputs processing failed; 5 runs; representative run: https://github.com/githubnext/gh-aw-cao/actions/runs/33939650717
- `githubnext/gh-aw-cao` — GitHub App token generation failed; 4 runs; representative run: https://github.com/githubnext/gh-aw-cao/actions/runs/33939504516
- `githubnext/gh-aw-workshop` — activation artifact upload failed; 4 runs; representative run: https://github.com/githubnext/gh-aw-workshop/actions/runs/33939477274
- Several additional workflow-local conclusion-only recurrences remain P1 but have insufficient evidence for remediation issue creation in this pass.
- P2
- 24 isolated or weakly evidenced clusters, mostly single-run CLI execution failures or conclusion-only failures.

### Control Plane
- Correlation ID: `33942748065-197`
- Central repository: `githubnext/gh-aw-cao`
- Control plane run: https://github.com/githubnext/gh-aw-cao/actions/runs/33942748065

> Generated by [SelfCare / Open Source Failures](https://github.com/githubnext/gh-aw-cao/actions/runs/33942908741) · pi · gpt54 · 30.7 AIC · ⌖ 8.57 AIC · ⊞ 8.3K · [◷](https://github.com/search?q=repo%3Agithubnext%2Fgh-aw-cao+is%3Aissue+%22gh-aw-workflow-call-id%3A+githubnext%2Fgh-aw-cao%2Fself-care-open-source-failures%22&type=issues)
> - [x] expires on Sep 19, 2026, 3:54 AM UTC

Start with the listed live-mode workflows in githubnext/gh-aw-cao: aw-failures-investigator, aw-maintenance-compiler-security, aw-maintenance-upgrade, and dependabot-release-train-updater. Inspect representative run 33940267555 and compare the target-authority declarations with the aw-doctor and dependabot failures. Done means new live-mode precompute runs for the affected workflows no longer fail with target-authority errors.

Written by the indexing model from the issue text.