[detection-stats] Detection stats for github/gh-aw - 2026-08-30

602 external-detector runs analysed, detection-job error rate 0%, 53 soft failures, threat rate 0% (of runs with a verdict).

## Summary

# gh-aw detection statistics - 2026-08-30 (UTC)

Repository: `github/gh-aw`
Window: `2026-08-30T00:00:00Z` .. `2026-08-30T23:59:59Z`
API requests: 1712, rate-limit pauses: 1
Data complete: yes

## Totals

| Metric | Count |
|---|---|
| Workflow runs in window | 1693 |
| Agentic runs (`*.lock.yml`) | 938 |
| Runs with a `detection` job | 605 |
| ... using the external detector | 602 |
| ... using the built-in detector | 3 |
| ... detector could not be determined | 0 |
| Agentic runs without a `detection` job | 333 |

All rates below are over the **external detector** population (602 runs). Since gh-aw #54111 the external detector is the compile-time default; a run counts as external when its `detection` job showed the `Install threat-detect binary` step, when another run of the same workflow did that day, or when the workflow is `.lock.yml` and no run showed the built-in shape (a completed detection job with steps but no marker). Runs on a workflow that opted out with `features: gh-aw-detection: false` count as built-in.

## Detection job outcomes

| Outcome | Count | Rate |
|---|---|---|
| `success` | 401 | 66.61% |
| `skipped` | 200 | 33.22% |
| `cancelled` | 1 | 0.17% |

**Error rate (failure/timed_out/action_required): 0%**

## Verdict availability

| State | Meaning | Count |
|---|---|---|
| `present` | detection artifact downloaded and parsed | 273 |
| `absent` | detection job ran but published no artifact (soft failure) | 53 |
| `skipped` | detection job was skipped or was still running (nothing to fetch) | 200 |
| `unreadable` | artifact zip could not be unpacked | 76 |

Green detection jobs that published no verdict: **53** (detection steps are `continue-on-error`, so a missing verdict artifact is the only reliable signal for these).

## Detection results

| Result | Count |
|---|---|
| Runs with a parsed verdict | 273 |
| Clean (no threat) | 273 |
| Any threat | 0 |
| `prompt_injection` | 0 |
| `secret_leak` | 0 |
| `malicious_patch` | 0 |

**Threat rate (of runs with a verdict): 0%**

## Reasons reported by gh-aw

From the `[aw] Detection Runs` tracking issue (warning/failure conclusions only), restricted to runs in the external-detector population above.

| Reason | Count |
|---|---|
| `parse_error` | 52 |
| `agent_failure` | 45 |

## By workflow

| Workflow | Runs | Failed | Cancelled | Skipped | No verdict | Threats |
|---|---|---|---|---|---|---|
| PR Sous Chef | 86 | 0 | 0 | 0 | 1 | 0 |
| Daily Trajectory Grader Implementer | 47 | 0 | 1 | 10 | 0 | 0 |
| Issue Monster | 47 | 0 | 0 | 1 | 0 | 0 |
| [aw] Failure Investigator (6h) | 44 | 0 | 0 | 40 | 1 | 0 |
| Q | 41 | 0 | 0 | 41 | 0 | 0 |
| Deployment Incident Monitor | 28 | 0 | 0 | 28 | 0 | 0 |
| Avenger | 24 | 0 | 0 | 21 | 3 | 0 |
| Test Quality Sentinel | 14 | 0 | 0 | 0 | 0 | 0 |
| Design Decision Gate 🏗️ | 13 | 0 | 0 | 0 | 1 | 0 |
| Impeccable Skills Reviewer | 13 | 0 | 0 | 0 | 0 | 0 |
| Matt Pocock Skills Reviewer | 13 | 0 | 0 | 0 | 0 | 0 |
| PR Code Quality Reviewer | 13 | 0 | 0 | 0 | 0 | 0 |
| Ponytail Reviewer | 13 | 0 | 0 | 0 | 0 | 0 |
| Daily Go Test Parallelizer | 12 | 0 | 0 | 1 | 11 | 0 |
| Auto-Triage Issues | 8 | 0 | 0 | 3 | 0 | 0 |
| Squad — ``@copilot`` run pr-finisher skill | 7 | 0 | 0 | 7 | 0 | 0 |
| Contribution Check | 6 | 0 | 0 | 0 | 0 | 0 |
| Code Scanning Fixer | 4 | 0 | 0 | 0 | 1 | 0 |
| Deep Report | 4 | 0 | 0 | 0 | 1 | 0 |
| PR Triage Agent | 4 | 0 | 0 | 0 | 0 | 0 |
| Workflow Generator | 4 | 0 | 0 | 4 | 0 | 0 |
| PR Description Updater | 3 | 0 | 0 | 0 | 0 | 0 |
| Daily Credit Limit Test | 2 | 0 | 0 | 1 | 0 | 0 |
| Agent Job Health Monitor | 1 | 0 | 0 | 0 | 1 | 0 |

_149 further workflows omitted; see `stats.json`._

## Notable runs

See the full `detection-stats-` artifact for the complete notable-runs list (76 total, top entries dominated by `Daily Go Test Parallelizer`, `Avenger`, and `Issue Monster` soft failures / unreadable-artifact cases).

## Change since 2026-08-29

| Metric | 2026-08-29 | 2026-08-30 | Δ | 7-day mean |
|---|---|---|---|---|
| External detector runs | 702 | 602 | -100 | 934.3 |
| Error rate | 0.0% | 0.0% | 0.0pp | 0.0% |
| Soft failures | 28 | 53 | +25 | 48.6 |
| Runs with verdict | 305 | 273 | -32 | 358.9 |
| Any threat | 0 | 0 | 0 | 1.4 |
| Threat rate | 0.0% | 0.0% | 0.0pp | 0.42% |

## Watch list

- Daily Go Test Parallelizer — 0 failed, 11 without a verdict, out of 12 runs
- Avenger — 0 failed, 3 without a verdict, out of 24 runs
- PR Sous Chef — 0 failed, 1 without a verdict, out of 86 runs
- [aw] Failure Investigator (6h) — 0 failed, 1 without a verdict, out of 44 runs
- Design Decision Gate 🏗️ — 0 failed, 1 without a verdict, out of 13 runs

Collected by: https://github.com/github/gh-aw-threat-detection/actions/runs/33355391490
Full data: the detection-stats-33355391490 artifact on that run.

> Generated by [Detection Stats Daily](https://github.com/github/gh-aw-threat-detection/actions/runs/33355391490) · copilot · auto · 30.8 AIC · ⌖ 7.75 AIC · ⊞ 11.3K · [◷](https://github.com/search?q=repo%3Agithub%2Fgh-aw-threat-detection+is%3Aissue+%22gh-aw-workflow-call-id%3A+github%2Fgh-aw-threat-detection%2Fdetection-stats-daily%22&type=issues)

This is an automatically generated detection report and does not identify a requested code or documentation change. Start with the linked detection-stats-33355391490 artifact and the Detection Stats Daily workflow run; a contribution would need an explicit follow-up scope and acceptance criteria before completion can be judged.

Written by the indexing model from the issue text.