[detection-analysis] Detection Analysis Report — 2026-07-29

[github-actions[bot]]

Summary

• Window evaluated: last 24 full hours (UTC), approximately 2026-07-28T23:41:00Z to 2026-07-29T23:41:00Z
• Total runs analyzed: 231 (231 of 429 fetched run directories contained complete logs; 198 were empty/incomplete downloads and excluded — see note below)
• Detection-enabled runs: 75 (32.5%)
• Regular runs: 156 (67.5%)
• Misconfigured workflows found: 10

Warning

10 misconfigured workflow(s) detected — see details below.

Comparison Chart

• Regular runs: 156 total, 92.9% success rate, avg 28,724 tokens/run
• Detection runs: 75 total, 90.7% success rate, avg 81,581 tokens/run

Misconfigured Workflows

Workflow	Misconfiguration Type	Run Count	Recommended Fix
PR Description Updater	High-frequency workflow without gh-aw-detection enabled (24h proxy for 7-day rule)	6	Enable gh-aw-detection: true in frontmatter given this workflow's run frequency, to benefit from anomaly/prompt-injection detection.
PR Sous Chef	High-frequency workflow without gh-aw-detection enabled (24h proxy for 7-day rule)	21	Enable gh-aw-detection: true in frontmatter given this workflow's run frequency, to benefit from anomaly/prompt-injection detection.
PR Code Quality Reviewer	High-frequency workflow without gh-aw-detection enabled (24h proxy for 7-day rule)	17	Enable gh-aw-detection: true in frontmatter given this workflow's run frequency, to benefit from anomaly/prompt-injection detection.
Design Decision Gate 🏗️	High-frequency workflow without gh-aw-detection enabled (24h proxy for 7-day rule)	17	Enable gh-aw-detection: true in frontmatter given this workflow's run frequency, to benefit from anomaly/prompt-injection detection.
Impeccable Skills Reviewer	High-frequency workflow without gh-aw-detection enabled (24h proxy for 7-day rule)	18	Enable gh-aw-detection: true in frontmatter given this workflow's run frequency, to benefit from anomaly/prompt-injection detection.
Matt Pocock Skills Reviewer	High-frequency workflow without gh-aw-detection enabled (24h proxy for 7-day rule)	20	Enable gh-aw-detection: true in frontmatter given this workflow's run frequency, to benefit from anomaly/prompt-injection detection.
Issue Monster	High-frequency workflow without gh-aw-detection enabled (24h proxy for 7-day rule)	11	Enable gh-aw-detection: true in frontmatter given this workflow's run frequency, to benefit from anomaly/prompt-injection detection.
Deep Report	Name implies audit/monitoring role but gh-aw-detection is not enabled	1	Add gh-aw-detection: true to frontmatter, matching the monitoring/audit intent implied by the workflow name.
Daily Security Observability Report	Name implies audit/monitoring role but gh-aw-detection is not enabled	1	Add gh-aw-detection: true to frontmatter, matching the monitoring/audit intent implied by the workflow name.
Daily Regulatory Report Generator	Name implies audit/monitoring role but gh-aw-detection is not enabled	1	Add gh-aw-detection: true to frontmatter, matching the monitoring/audit intent implied by the workflow name.

_{Note: rule "gh-aw-detection: false with >3 runs" is specified against a 7-day window, but only the 24h log window was fetched for this report; the 24h run count was used as a conservative proxy. Daily Agentic Workflow AIC Usage Audit was excluded from the name-based check per its documented upstream opt-out.}

View All Run Metrics

• Total distinct workflows observed: 99

Workflow	Runs	Detection	Success Rate	Avg Tokens	Engine
PR Sous Chef	21	—	71.4%	47,425	pi
Matt Pocock Skills Reviewer	20	—	100.0%	11,902	copilot
Impeccable Skills Reviewer	18	—	100.0%	5,858	copilot
PR Code Quality Reviewer	17	—	76.5%	27,280	copilot
Test Quality Sentinel	17	✅	100.0%	11,557	copilot
Design Decision Gate 🏗️	17	—	100.0%	11,001	claude
Issue Monster	11	—	100.0%	18,443	pi
PR Description Updater	6	—	100.0%	16,066	copilot
Auto-Triage Issues	3	✅	100.0%	24,790	pi
Contribution Check	3	✅	100.0%	23,952	copilot
Avenger	3	✅	0.0%	0	claude
Release	2	—	100.0%	11,726	copilot
Smoke Copilot	2	—	100.0%	395,093	copilot
PR Triage Agent	2	—	100.0%	18,832	copilot
Discussion Task Miner - Code Quality Improvement Agent	2	—	100.0%	21,222	copilot
[aw] Failure Investigator (6h)	2	✅	100.0%	50,228	claude
Code Scanning Fixer	2	✅	100.0%	10,348	copilot
Daily Ambient Context Optimizer	2	✅	100.0%	34,147	copilot
Daily MCP Tool Concurrency Analysis	1	✅	100.0%	15,219	copilot
Dependabot Dependency Checker	1	—	100.0%	12,798	copilot
Package Specification Extractor	1	—	100.0%	26,060	copilot
Daily Spending Forecast	1	—	100.0%	14,804	copilot
Copilot Agent Prompt Clustering Analysis	1	✅	100.0%	38,878	claude
Glossary Maintainer	1	—	100.0%	36,723	copilot
Copilot PR Conversation NLP Analysis	1	✅	100.0%	22,157	copilot
Daily Documentation Updater	1	✅	100.0%	85,615	pi
Agent Container Smoke Test	1	—	100.0%	4,889	copilot
Smoke Gemini	1	—	100.0%	310,413	gemini
Smoke Copilot Auto	1	—	100.0%	3,227	copilot
Smoke Pi	1	—	100.0%	10,911	pi
GitHub API Consumption Report Agent	1	✅	100.0%	40,540	claude
Draft PR Cleanup	1	—	100.0%	7,004	copilot
Smoke Copilot - AOAI (apikey)	1	—	100.0%	185,021	copilot
Smoke Copilot - AOAI (Entra)	1	—	100.0%	130,104	copilot
Smoke Copilot MAI	1	—	100.0%	84,461	copilot
Smoke Copilot Small	1	—	100.0%	3,273	copilot
Smoke OTEL	1	—	100.0%	22,945	copilot
Daily Choice Type Test	1	✅	100.0%	4,001	claude
Daily Max Ai Credits Test	1	✅	0.0%	3,780	copilot
Typist - Go Type Analysis	1	✅	100.0%	111,345	claude
Package Specification Enforcer	1	—	100.0%	30,445	pi
Constraint Solving — Problem of the Day	1	✅	100.0%	8,026	copilot
Blog Auditor	1	✅	100.0%	19,420	claude
Team Status	1	—	100.0%	5,628	copilot
Daily Agentic Workflow AIC Usage Audit	1	—	100.0%	9,480	copilot
Daily Go Function Namer	1	✅	100.0%	37,041	pi
Daily AIC Consumption Report (Sentry + Grafana OTel)	1	✅	100.0%	77,981	claude
GitHub MCP Structural Analysis	1	✅	100.0%	31,307	claude
Claude Code User Documentation Review	1	✅	100.0%	76,213	claude
Agent Performance Analyzer - Meta-Orchestrator	1	✅	100.0%	39,500	copilot
Workflow Normalizer	1	—	100.0%	6,672	copilot
Repository Quality Improvement Agent	1	—	100.0%	30,481	copilot
Daily Credit Limit Test	1	✅	100.0%	3,626	copilot
CI Optimization Coach	1	✅	0.0%	0	copilot
CLI Consistency Checker	1	✅	100.0%	43,231	copilot
Daily Malicious Code Scan Agent	1	✅	100.0%	14,231	copilot
Daily Fact	1	✅	100.0%	302,897	codex
Dead Code Removal Agent	1	—	100.0%	44,950	copilot
Architecture Guardian	1	✅	100.0%	5,406	copilot
Daily Agent of the Day Blog Writer	1	✅	0.0%	30,071	copilot
Daily Issues Report Generator	1	✅	100.0%	17,110	copilot
Deep Report	1	—	100.0%	36,704	claude
Issue Triage Agent	1	—	100.0%	5,883	copilot
Delight	1	—	0.0%	18,085	copilot
Daily CLI Performance Agent	1	✅	100.0%	29,205	copilot
Agentic Workflow AIC Usage Optimizer	1	—	100.0%	23,755	copilot
Breaking Change Checker	1	✅	100.0%	9,118	copilot
Daily Copilot PR Merged Report	1	✅	100.0%	13,872	copilot
The Daily Repository Chronicle	1	✅	100.0%	35,393	copilot
Daily Formal Spec Verifier	1	✅	100.0%	19,443	copilot
UK AI Operational Resilience	1	—	100.0%	16,622	copilot
Agent Persona Explorer	1	✅	100.0%	61,332	pi
Daily SPDD Spec Planner	1	—	100.0%	17,656	copilot
Daily Security Observability Report	1	—	100.0%	15,482	copilot
Slide Deck Maintainer	1	—	100.0%	16,292	copilot
GEO Optimizer Daily Audit	1	✅	100.0%	11,450	copilot
Linter Miner	1	—	100.0%	26,933	copilot
Daily Secrets Analysis Agent	1	—	100.0%	11,626	copilot
Daily Testify Uber Super Expert	1	✅	100.0%	10,045	copilot
Copilot Agent PR Analysis	1	✅	100.0%	23,010	claude
Daily Cache Strategy Analyzer	1	✅	100.0%	3,236,670	codex
Daily Code Metrics and Trend Tracking Agent	1	✅	100.0%	76,072	claude
Daily Safe Output Integrator	1	—	100.0%	9,317	copilot
Smoke Copilot Sub Agents	1	—	100.0%	18,509	copilot
Daily Graft Intelligence	1	✅	100.0%	8,154	copilot
Daily Team Evolution Insights	1	—	100.0%	23,481	claude
Lockfile Statistics Analysis Agent	1	—	100.0%	13,758	claude
Copilot PR Prompt Pattern Analysis	1	✅	100.0%	15,609	copilot
Daily Project Performance Summary Generator (Using MCP Scripts)	1	✅	100.0%	19,637	copilot
Daily Caveman Optimizer	1	✅	100.0%	8,733	claude
Daily Safe Output Tool Optimizer	1	—	100.0%	16,040	claude
Agentic Workflow Audit Agent	1	✅	100.0%	32,928	claude
Daily Regulatory Report Generator	1	—	100.0%	22,313	copilot
Plan Command	1	—	100.0%	10,710	copilot
Duplicate Code Detector	1	✅	100.0%	840,552	codex
ESLint Monster	1	—	100.0%	14,255	pi
Daily AWF Spec Compiler Surfacing Review	1	✅	100.0%	80,562	pi
Daily BYOK Ollama Test	1	✅	0.0%	0	copilot
Daily Reliability Review	1	✅	100.0%	27,077	claude

View Historical Trend

27 days of history in cache-memory (trending/detection-metrics/history.jsonl). Regular-run volume has grown sharply since mid-July while detection-enabled adoption has grown more slowly — detection coverage is currently ~32% of runs, down from the recent peak proportionally as regular-workflow volume outpaced it.

Recommendations

• Enable gh-aw-detection: true on the 7 high-frequency PR/issue-triggered workflows flagged above (PR Sous Chef, PR Code Quality Reviewer, Design Decision Gate, Impeccable Skills Reviewer, Matt Pocock Skills Reviewer, Issue Monster, PR Description Updater) — these process untrusted PR/issue content most frequently and benefit most from detection coverage.
• Review the 3 report/monitoring-named workflows without detection (Deep Report, Daily Security Observability Report, Daily Regulatory Report Generator) and add gh-aw-detection: true unless there is a documented opt-out like the AIC Usage Audit workflow.
• Detection-enabled runs average ~2.8x the tokens of regular runs (81,581 vs 28,724) — worth monitoring cost impact as adoption grows, though this window has a small/skewed sample (75 runs) dominated by a few heavy workflows.
• Re-run this analysis with a --start-date -7d log fetch when feasible to evaluate the frequency-based rule against its intended 7-day window rather than the 24h proxy used here.

Generated by 🔎 Detection Analysis Report · age00 · 142.6 AIC · ⌖ 21.9 AIC · ⊞ 7.2K · ◷

• expires on Aug 1, 2026, 3:45 PM UTC-08:00

[github-actions[bot]]

This discussion was automatically closed because it expired on 2026-08-01T23:45:07.538Z.

Closed by Workflow