[detection-analysis] Detection Analysis Report — 2026-07-19 (last 24h)

[github-actions[bot]]

Summary

• Window evaluated: last 24 full hours (UTC) — 2026-07-18T23:00:00Z to 2026-07-19T23:00:00Z
• Observed run data span: 2026-07-19T11:33:43.137Z → 2026-07-19T22:57:57.167Z
• Total runs analyzed: 120 across 67 workflows
• Detection-enabled runs: 34 (28%)
• Regular runs: 86 (72%)
• Misconfigured workflows found: 4

Warning

4 workflow(s) show detection-configuration issues in this window. Detection-enabled runs had a 94.1% success rate vs 91.9% for regular runs, and used 63,785 avg tokens vs 112,577 (regular average is skewed upward by a few high-token smoke tests).

Comparison Chart

Stacked bars show success (green) / failure (red) per group; the blue line overlays success rate (right axis); the pink band annotates the misconfigured-workflow count.

Metric	Regular Runs	Detection Runs
Run count	86	34
Success rate	91.9%	94.1%
Failures	7	2
Avg tokens	112,577	63,785

Misconfigured Workflows

Workflow	Misconfiguration type	Runs	Example run	Recommended fix
Daily Regulatory Report Generator	Analysis-type name but detection not enabled	1	§29704388802	Add gh-aw-detection: true to frontmatter
Daily Security Observability Report	Analysis-type name but detection not enabled	1	§29694306894	Add gh-aw-detection: true to frontmatter
GitHub MCP Remote Server Tools Report Generator	Analysis-type name but detection not enabled	1	§29687383219	Add gh-aw-detection: true to frontmatter
Smoke Copilot	Detection disabled (false) on frequently-run workflow (>3 runs in window)	4	§29704421484	Set gh-aw-detection: true in workflow frontmatter

Note on rule 3 (detection step failed): two detection-enabled runs — Daily Issues Report Generator (§29691397935) and Daily Max Ai Credits Test (§29685348389) — ended in failure, but the failure was in the upstream agent/Execute Copilot CLI step and the threat-detection steps were skipped, not failed. They are therefore not counted as detection-step failures. Daily Agentic Workflow AIC Usage Audit is excluded from rule 2 as a documented upstream-mirrored opt-out.

Note on rule 1: only a ~24h window of logs was available, not a full 7 days; Smoke Copilot is flagged on its 4 in-window runs (a smoke test may intentionally opt out of detection — verify before changing).

View All Run Metrics

Workflow	Detection	Runs	Success	Avg tokens
PR Sous Chef	absent	24	23/24 (96%)	45,300
Tidy	absent	7	6/7 (86%)	5,648
Smoke Copilot	off (false)	4	2/4 (50%)	832,238
Contribution Check	on	3	3/3 (100%)	22,299
Issue Monster	absent	3	3/3 (100%)	11,566
PR Description Updater	absent	3	3/3 (100%)	7,682
Auto-Triage Issues	on	2	2/2 (100%)	12,988
Code Scanning Fixer	on	2	2/2 (100%)	5,332
Design Decision Gate 🏗️	absent	2	2/2 (100%)	10,999
Discussion Task Miner - Code Quality Improvement Agent	absent	2	2/2 (100%)	12,138
Impeccable Skills Reviewer	absent	2	2/2 (100%)	5,444
Matt Pocock Skills Reviewer	absent	2	2/2 (100%)	11,480
PR Code Quality Reviewer	absent	2	2/2 (100%)	18,081
PR Triage Agent	absent	2	2/2 (100%)	14,694
Smoke Copilot - AOAI (Entra)	off (false)	2	2/2 (100%)	704,456
Smoke Copilot - AOAI (apikey)	off (false)	2	2/2 (100%)	1,053,526
Smoke Copilot MAI	off (false)	2	2/2 (100%)	102,736
Smoke Copilot Small	off (false)	2	2/2 (100%)	3,276
Smoke OTEL	off (false)	2	2/2 (100%)	17,563
Test Quality Sentinel	on	2	2/2 (100%)	10,548
[aw] Failure Investigator (6h)	on	2	2/2 (100%)	31,289
Agent Container Smoke Test	off (false)	1	1/1 (100%)	4,117
Agent Performance Analyzer - Meta-Orchestrator	on	1	1/1 (100%)	19,736
Agent Persona Explorer	on	1	1/1 (100%)	74,456
Agentic Workflow Audit Agent	on	1	1/1 (100%)	38,147
Changeset Generator	on	1	1/1 (100%)	44,664
Claude Code User Documentation Review	on	1	1/1 (100%)	48,152
Constraint Solving — Problem of the Day	on	1	1/1 (100%)	7,551
Copilot Agent PR Analysis	on	1	1/1 (100%)	19,956
Copilot PR Prompt Pattern Analysis	on	1	1/1 (100%)	11,901
Daily AWF Spec Compiler Surfacing Review	on	1	1/1 (100%)	120,361
Daily Ambient Context Optimizer	on	1	1/1 (100%)	27,759
Daily CLI Performance Agent	on	1	1/1 (100%)	16,900
Daily Cache Strategy Analyzer	on	1	1/1 (100%)	766,206
Daily Caveman Optimizer	on	1	1/1 (100%)	15,658
Daily Code Metrics and Trend Tracking Agent	on	1	1/1 (100%)	27,087
Daily Formal Spec Verifier	on	1	1/1 (100%)	53,468
Daily Go Function Namer	on	1	1/1 (100%)	24,499
Daily Issues Report Generator	on	1	0/1 (0%)	0
Daily Malicious Code Scan Agent	on	1	1/1 (100%)	8,853
Daily Max Ai Credits Test	on	1	0/1 (0%)	3,014
Daily Project Performance Summary Generator (Using MCP Scripts)	on	1	1/1 (100%)	19,047
Daily Regulatory Report Generator	absent	1	1/1 (100%)	12,476
Daily SPDD Spec Planner	absent	1	1/1 (100%)	8,221
Daily Safe Output Integrator	absent	1	1/1 (100%)	8,428
Daily Secrets Analysis Agent	absent	1	1/1 (100%)	11,999
Daily Security Observability Report	absent	1	1/1 (100%)	15,161
Daily Team Evolution Insights	absent	1	1/1 (100%)	14,308
Daily Testify Uber Super Expert	on	1	1/1 (100%)	10,205
Dead Code Removal Agent	absent	1	1/1 (100%)	9,224
Delight	absent	1	1/1 (100%)	15,198
Duplicate Code Detector	on	1	1/1 (100%)	548,023
ESLint Monster	absent	1	1/1 (100%)	15,417
GEO Optimizer Daily Audit	on	1	1/1 (100%)	12,049
GitHub MCP Remote Server Tools Report Generator	absent	1	0/1 (0%)	74,737
Linter Miner	absent	1	1/1 (100%)	12,601
Lockfile Statistics Analysis Agent	absent	1	1/1 (100%)	16,410
Package Specification Enforcer	absent	1	1/1 (100%)	68,880
Package Specification Librarian	absent	1	1/1 (100%)	19,165
Smoke Antigravity	off (false)	1	1/1 (100%)	0
Smoke Claude	off (false)	1	0/1 (0%)	0
Smoke Codex	off (false)	1	1/1 (100%)	566,995
Smoke Copilot Sub Agents	off (false)	1	0/1 (0%)	19,114
Smoke Gemini	absent	1	1/1 (100%)	212,883
Smoke Pi	off (false)	1	1/1 (100%)	12,478
Team Status	absent	1	1/1 (100%)	90,457
Workflow Normalizer	absent	1	1/1 (100%)	11,000

View Historical Trend

30-day trend of regular vs detection run volume (left axis) with detection success rate shaded (right axis). Today's metrics were appended to the cache-memory trending store (22 data points).

Recommendations

1. Enable detection on analysis/report workflows — Daily Regulatory Report Generator, Daily Security Observability Report, and GitHub MCP Remote Server Tools Report Generator process external/generated content and match analysis-type naming but run without gh-aw-detection: true. Add it to their frontmatter unless there is a documented reason to opt out. Note GitHub MCP Remote Server Tools Report Generator also failed this run (§29687383219) — worth a separate look.
2. Review Smoke Copilot — runs frequently with gh-aw-detection: false and had 2/4 failures this window. If it is a pure smoke test, document the opt-out (as done for the AIC Usage Audit); otherwise enable detection.
3. Investigate the two failed detection runs — Daily Issues Report Generator and Daily Max Ai Credits Test failed at the agent/Copilot-CLI step (detection steps were skipped). Root-cause the agent-step failure; these are not detection misconfigurations.
4. No alternating-config (rule 4) issues were detected in this window — detection settings were consistent per workflow.

Warning

Firewall blocked 1 domain

The following domain was blocked by the firewall during workflow execution:

• awmgmcpg

To allow these domains, add them to the network.allowed list in your workflow frontmatter:

network:
  allowed:
    - defaults
    - "awmgmcpg"

See Network Configuration for more information.

Generated by 🔎 Detection Analysis Report · 211.7 AIC · ⌖ 43.6 AIC · ⊞ 6.9K · ◷

• expires on Jul 22, 2026, 3:45 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 #46933.