[detection-analysis] Detection Analysis Report — 2026-10-22

[github-actions[bot]]

Summary

• Window requested: last 24 full hours (2026-10-21T23:19Z → 2026-10-22T23:19Z UTC)
• Actual data coverage: 2026-10-22T02:18Z → 2026-10-22T23:19Z (~21 of 24 hours) — the remaining ~3 hours near the start of the window did not have per-run gh-aw-detection feature detail downloaded, so that slice is not included below
• Total runs analyzed: 421
• Detection-enabled runs: 415 (98.6%)
• Regular runs (gh-aw-detection: false): 6 (1.4%)
• Misconfigured workflows found: 0

Note

No misconfigured workflows were detected in the analyzed window. All 6 "regular" runs belong to experimental/smoke-test workflows (Smoke Pi Routed, Engine Conformance Agy, Engine Conformance Aider, Smoke Copilot SDK Routed, Daily GitHub Docs SEO Optimizer) with ≤3 runs each — none crossed the >3-runs-in-7-days threshold for rule 1, no detection job showed a failing conclusion (rule 2, checked across 413 runs where a detection job actually executed), and no workflow alternated between enabled/disabled within the window (rule 3).

Comparison Chart

Metric	Regular (n=6)	Detection (n=415)
Success rate	66.7%	85.8%
Failures	2	59
Avg. tokens/run	25,948	195,989

Caveat: the regular-run sample (n=6) is too small for a statistically meaningful success-rate or token comparison — it's dominated by a couple of intentionally-failing engine-conformance smoke tests (Engine Conformance Agy, Engine Conformance Aider), and the large token gap mostly reflects a few high-volume detection-enabled production workflows (e.g. Metrics Collector, Daily Go Test Parallelizer, Avenger) rather than detection overhead itself.

Misconfigured Workflows

No misconfigured workflows detected in this window.

View All Run Metrics (top 25 of 167 workflows by run count)

Workflow	Total Runs	Detection	Regular	Success %	Avg Tokens	Engines
PR Sous Chef	74	74	0	100.0	195	pi
Issue Monster	41	41	0	97.6	88,234	pi
Matt Pocock Skills Reviewer	21	21	0	81.0	35,255	copilot
Ponytail Reviewer	20	20	0	80.0	113,587	codex
PR Code Quality Reviewer	20	20	0	75.0	25,303	copilot
Design Decision Gate 🏗️	18	18	0	77.8	7,158	pi
Test Quality Sentinel	18	18	0	77.8	11,182	copilot
Avenger	17	17	0	100.0	1,014,353	copilot
Daily Go Test Parallelizer	11	11	0	100.0	1,359,373	codex
[aw] Failure Investigator (6h)	7	7	0	85.7	378	claude
Auto-Triage Issues	6	6	0	100.0	128,469	codex
Contribution Check	5	5	0	80.0	18,816	copilot
PR Triage Agent	3	3	0	100.0	45,246	copilot
Code Scanning Fixer	3	3	0	100.0	10,958	copilot
ESLint Monster	2	2	0	0.0	113,305	codex
Smoke Claude Dynamic Workflow	2	2	0	100.0	1,681	claude
AI Moderator	2	2	0	100.0	850	pi
Smoke Pi Routed	2	0	2	100.0	61,870	pi
Daily Regression Audit	1	1	0	100.0	5,937	copilot
Semantic Function Refactoring	1	1	0	100.0	10,538	claude
Daily Community Attribution Updater	1	1	0	0.0	12,156	copilot
Daily Compiler Threat Spec Optimizer	1	1	0	0.0	8,896	copilot
Daily action/setup/* Security Audit	1	1	0	100.0	13,575	copilot
Metrics Collector	1	1	0	100.0	10,628,210	codex
LintMonster	1	1	0	100.0	1,215,622	codex

...and 142 additional workflows with 1 run each in this window (omitted for brevity).

View Historical Trend

23 days of history now exist in cache-memory; the chart above plots the available 30-day window (12 days with actual data, gaps shown for missing days).

Recommendations

1. No action required on misconfiguration — the gh-aw-detection feature is enabled by default across the fleet (98.6% of runs) and no workflow shows a failing detection job or inconsistent opt-out behavior.
2. Coverage gap: log collection for this run hit the count/storage budget before reaching the full 24h window — the oldest ~3 hours (2026-10-21T23:19Z–2026-10-22T02:18Z) lack per-run features.gh-aw-detection detail. Consider raising max_storage/count for this report's log pre-fetch, or splitting it across two smaller windows, if full 24h coverage is required going forward.
3. Small regular-run sample: if gh-aw-detection: false is intentional for the 5 smoke/conformance workflows currently using it, consider documenting that opt-out directly in each workflow's frontmatter comment so future audits don't need to re-derive the rationale.

References:

• §39851095785 — detection job ran cleanly (Test Quality Sentinel)

Generated by 🔎 Detection Analysis Report · claude · agent · ~162 AIC · ⌖ ~1 AIC · ⊞ 6.7K · ◷

• expires on Oct 25, 2026, 3:34 PM UTC-08:00

[github-actions[bot]]

This discussion has been marked as outdated by Detection Analysis Report.

A newer discussion is available at Discussion #71510.