[agentic-token-audit] Daily AIC Usage Audit — 2026-08-24

Daily Agentic Workflow AIC Usage Audit · issue · closed

Filter2mode:review mode:live
All recorded Export JSON
github-actions[bot]

published Aug 24, 2026, 11:55 AM · updated Aug 25, 2026, 11:52 AM

📊 Executive Summary

  • Period: last 24 hours (2026-08-24 collection window)
  • Total runs: 100
  • Total AIC: 4827.85
  • Total Actions minutes: 1065.0 min
  • Active workflows: 68

🏆 Top 5 Workflows by AIC Usage

Workflow Runs Total AIC Avg AIC
Instructions Janitor 1 480.40 480.40
Daily VulnHunter Scan 1 454.14 454.14
Daily Rendering Scripts Verifier 1 354.32 354.32
Copilot Session Insights 1 255.21 255.21
Typist - Go Type Analysis 1 229.46 229.46

📈 Trends

AIC by Workflow

Historical AIC Trend

Total AIC dropped from 7887.30 (2026-07-08, previous snapshot) to 4827.85 today, a decrease of ~38.8%. Note there is a gap in daily snapshots between 2026-07-08 and today, so this comparison spans more than one day.

Full Per-Workflow Breakdown
Workflow Runs Total AIC Avg AIC Errors Warnings
Instructions Janitor 1 480.40 480.40 0 0
Daily VulnHunter Scan 1 454.14 454.14 0 0
Daily Rendering Scripts Verifier 1 354.32 354.32 0 0
Copilot Session Insights 1 255.21 255.21 0 0
Typist - Go Type Analysis 1 229.46 229.46 0 0
GitHub API Consumption Report Agent 1 192.92 192.92 0 0
ESLint Miner 1 184.97 184.97 0 0
Package Specification Extractor 1 171.06 171.06 0 0
Weekly Workflow Analysis 1 159.31 159.31 0 0
PR Sous Chef 15 159.10 10.61 0 0
Go Fan 1 150.82 150.82 0 0
[aw] Failure Investigator (6h) 1 145.28 145.28 0 0
Daily Harness Experiment Proposer 1 138.73 138.73 1 0
Deep Report 1 136.50 136.50 0 0
Code Scanning Fixer 1 127.23 127.23 1 0
Glossary Maintainer 1 110.90 110.90 0 0
Archivx — Animated Workflow Visualizer 1 109.78 109.78 0 0
Daily AgentRx Trace Optimizer 1 106.35 106.35 0 0
Organization Health Report 1 83.35 83.35 0 0
Matt Pocock Skills Reviewer 3 80.92 26.97 0 0
Tidy 1 79.43 79.43 0 0
arXiv Paper Researcher: GitHub Agentic Workflows 1 73.24 73.24 0 0
Avenger 1 63.79 63.79 0 0
Terminal Stylist 1 57.33 57.33 1 0
Contribution Check 1 55.40 55.40 0 0
Daily Storify 1 53.64 53.64 0 0
PR Description Updater 1 53.64 53.64 0 0
Impeccable Skills Reviewer 2 46.63 23.32 0 0
Daily PR Code Quality Review 1 41.04 41.04 0 0
Daily AW Cross-Repo Compile Check 1 39.89 39.89 0 0
Daily News 1 39.19 39.19 0 0
PR Triage Agent 1 38.46 38.46 0 0
Layout Specification Maintainer 1 36.85 36.85 1 0
PR Code Quality Reviewer 3 35.86 11.95 0 0
Daily MCP Tool Concurrency Analysis 1 31.46 31.46 0 0
Design Decision Gate 🏗️ 2 31.36 15.68 0 0
Team Status 1 26.89 26.89 0 0
Sub-Issue Closer 1 22.71 22.71 0 0
Daily Safe Outputs Conformance Checker 1 19.90 19.90 0 0
Daily Elixir Credo Snippet Audit 1 18.95 18.95 0 0
Daily Choice Type Test 1 17.24 17.24 0 0
Package Specification Enforcer 1 15.94 15.94 0 0
Weekly Network Domain Set Audit 1 15.64 15.64 0 0
Daily Workflow Updater 1 14.77 14.77 0 0
Draft PR Cleanup 1 14.60 14.60 0 0
Test Quality Sentinel 2 12.28 6.14 0 0
Daily Assign Issue To User 1 10.56 10.56 0 0
Multi-Device Docs Tester 1 10.40 10.40 0 0
Ponytail Reviewer 2 7.50 3.75 0 0
Issue Monster 4 6.40 1.60 1 0
Constraint Solving — Problem of the Day 1 5.47 5.47 0 0
Auto-Triage Issues 1 0.68 0.68 1 0
Daily Max Ai Credits Test 1 0.00 0.00 1 0
Smoke Claude on Copilot 1 0.00 0.00 1 0
Daily Documentation Updater 1 0.00 0.00 0 0
Copilot PR Conversation NLP Analysis 1 0.00 0.00 1 0
Daily Go Test Parallelizer 3 0.00 0.00 3 0
Dev 1 0.00 0.00 1 0
Copilot Agent Prompt Clustering Analysis 1 0.00 0.00 1 0
Dependabot Dependency Checker 1 0.00 0.00 1 0
1. List all Go packages with their doc comments 1 0.00 0.00 1 0
Daily Spending Forecast 1 0.00 0.00 1 0
daily-experiment-report 1 0.00 0.00 1 0
AI Moderator 6 0.00 0.00 6 0
Daily Syntax Error Quality Check 1 0.00 0.00 1 0
Daily Evals Feature Report 1 0.00 0.00 1 0
Schema Feature Coverage Checker 1 0.00 0.00 1 0
Daily Container Image Security Scan 1 0.00 0.00 1 0

💡 Observations

  • No single workflow exceeded 30% of total AIC (largest was Instructions Janitor at ~10.0%), so there are no "heavy hitter" concentration risks today.
  • All 100 completed runs succeeded with only isolated single error entries recorded against otherwise-successful runs (e.g. Daily Harness Experiment Proposer, Code Scanning Fixer, Terminal Stylist, Layout Specification Maintainer, Issue Monster, Auto-Triage Issues, Daily Max Ai Credits Test, Smoke Claude on Copilot, Copilot PR Conversation NLP Analysis, Daily Go Test Parallelizer) — none had elevated error rates relative to run counts.
  • 27 workflows exceeded an average of 10.00 AIC per run; the highest averages were Instructions Janitor (480.40), Daily VulnHunter Scan (454.14), Daily Rendering Scripts Verifier (354.32), Copilot Session Insights (255.21), and Typist - Go Type Analysis (229.46) — each ran only once in this window, so per-run costs equal totals.

References:

Data snapshot: memory/token-audit/2026-08-24.json

Generated by Daily Agentic Workflow AIC Usage Audit · copilot · auto · 53.8 AIC · ⌖ 5.06 AIC · ⊞ 13.4K ·

  • expires on Aug 27, 2026, 3:55 AM UTC-08:00