[detection-analysis] Detection Analysis Report — 2026-07-04 (last 24h) #43390
Closed
Replies: 1 comment
|
This discussion has been marked as outdated by Detection Analysis Report. A newer discussion is available at Discussion #43550. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Detection Analysis Report — Last 24h
Summary
2026-07-04T08:14:25Z→2026-07-04T12:48:38Zfeatures.gh-aw-detection: true): 21 (21%)false/absent): 79 (79%)Warning
2 workflows are flagged as misconfigured this window. See the Misconfigured Workflows section for details and recommended fixes.
1 Token note:
TokenUsage/EffectiveTokenswere0across all 100 runs in this window'srun_summary.json(token accounting not populated), so AIC (AI-cost units) fromaw_info/summary is used as the usage proxy. Treat absolute values as directional.Comparison Chart
Detection-enabled and regular runs performed almost identically on success rate (81.0% vs 82.3%) — detection is not degrading run reliability. Detection adoption sits at 21% of all runs, spread across copilot (13), claude (3), pi (3), and codex (2) engines.
Misconfigured Workflows
gh-aw-detection: true(name contains "Analysis")gh-aw-detection: true— analysis/monitoring workflows should participate in detection. Example run §28703236869.gh-aw-detection: falseon an active workflow (7 runs in window)Not flagged as misconfig: The 4 detection-enabled failures (Daily Max Ai Credits Test, Changeset Generator, daily-experiment-report, Daily Rendering Scripts Verifier) failed on ordinary workflow errors — no detection-job-specific error patterns were found, so they are not counted as detection misconfigurations. No workflows showed inconsistent (mixed) detection state within the window.
View All Run Metrics (per-workflow breakdown)
Legend: Det = ✅ detection-enabled this window, — = regular. AIC = AI-cost proxy.
View Historical Trend (last 30 days)
Run counts (both regular and detection) and detection success rate are consistent metrics across the 8-day history and are shown above. Historical
avg_tokensvalues mix raw tokens (earlier days) with the AIC proxy (recent days), so the token axis is intentionally omitted from the trend to avoid a misleading unit mismatch. Detection success rate has stayed in the 80–95% band; misconfigured count has trended down (7 → 5 → 5 → 2).Recommendations
gh-aw-detection: true; as an analysis/clustering workflow it should be in scope for detection.TokenUsageis0for every run in this window. If token-based reporting matters, verify the token-accounting step is emitting intorun_summary.json/aw_info.json; until then reports fall back to the AIC proxy.References:
Warning
Firewall blocked 1 domain
The following domain was blocked by the firewall during workflow execution:
awmgmcpgSee Network Configuration for more information.
All reactions