You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Over the reporting period (2026-02-23 → 2026-02-24), Copilot-powered agentic workflows consumed 123,484,939 tokens across 141 runs covering 23 unique workflows, with a combined run time of 11.8 hours. The overall success rate was 61% (86/141 runs) — significantly dragged down by the Issue Monster workflow, which failed all 50 of its runs (0% success, 0 tokens). Excluding Issue Monster, the success rate rises to 94% across 91 runs.
Cost data is not available for this reporting period (the estimated_cost field returned 0 across all runs), so analysis focuses on token consumption and operational efficiency metrics.
Key Highlights
Metric
Value
🔢 Total Tokens
123.5M
🔄 Total Runs
141 (23 workflows)
✅ Success Rate
61% overall / 94% excl. Issue Monster
⏱️ Total Duration
11.8 hours
🏆 Top Consumer
CI Failure Doctor — 38.8M tokens (31.4%)
📈 Highest Avg/Run
Documentation Noob Tester — 6.8M avg tokens/run
🚨 Critical Issue
Issue Monster — 50 consecutive failures
📊 Visual Analysis
Top Workflows by Token Consumption
CI Failure Doctor dominates at 31.4% of all tokens, followed by Chroma Issue Indexer (12.5%) and Documentation Noob Tester (11.0%). The top 5 workflows alone account for 73.5% of total token consumption.
Issue Monster accounts for 35% of all runs but 0% of tokens (all failures). The token distribution is highly concentrated among a few high-value workflows.
Issue Monster triggered 50 runs (35% of all runs) with a 0% success rate and 0 tokens consumed. This indicates a systematic configuration or authentication failure — the workflow is failing before the agent even starts. This is the single biggest driver of the 61% overall success rate.
Recommendation: Investigate and fix the root cause immediately. Check authentication, trigger conditions, and workflow configuration. 50 automated failures in a reporting period represent significant CI noise.
⚠️ High Attention: Documentation Noob Tester
With 6.8M avg tokens per run (the highest of any workflow) and only 50% success rate, this workflow has the worst efficiency profile of any meaningful workflow. Each failed run wastes ~6.8M tokens.
Recommendation: Debug the 1 failed run to understand why 50% fail. Consider adding an early exit condition or reducing the scope of each run to lower the token ceiling.
⚠️ Watch: Chroma Issue Indexer
75% success rate across 8 runs (2 failures). At 2.5M avg tokens/run, each failure is expensive.
Recommendation: Add better error handling and retry logic. Investigate the 2 failures to determine if they're systematic.
✅ High Efficiency Highlights
CI Failure Doctor (100% success, 16 runs) — Most active productive workflow, delivering consistent results at 2.4M tokens/run
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
Executive Summary
Over the reporting period (2026-02-23 → 2026-02-24), Copilot-powered agentic workflows consumed 123,484,939 tokens across 141 runs covering 23 unique workflows, with a combined run time of 11.8 hours. The overall success rate was 61% (86/141 runs) — significantly dragged down by the Issue Monster workflow, which failed all 50 of its runs (0% success, 0 tokens). Excluding Issue Monster, the success rate rises to 94% across 91 runs.
Cost data is not available for this reporting period (the
estimated_costfield returned 0 across all runs), so analysis focuses on token consumption and operational efficiency metrics.Key Highlights
📊 Visual Analysis
Top Workflows by Token Consumption
CI Failure Doctor dominates at 31.4% of all tokens, followed by Chroma Issue Indexer (12.5%) and Documentation Noob Tester (11.0%). The top 5 workflows alone account for 73.5% of total token consumption.
Efficiency: Success Rate vs Token Consumption
Most high-token workflows maintain excellent success rates (≥80%, shown in green). Notable exceptions: Documentation Noob Tester (50% success, 6.8M avg tokens/run) and Daily Malicious Code Scan Agent (50% success, 863K avg tokens/run).
Token Share & Run Activity Distribution
Issue Monster accounts for 35% of all runs but 0% of tokens (all failures). The token distribution is highly concentrated among a few high-value workflows.
🏆 Top 10 Most Expensive Workflows
All 23 Workflows — Complete Statistics
💡 Insights & Recommendations
🚨 Critical: Issue Monster — 50 Consecutive Failures
Issue Monster triggered 50 runs (35% of all runs) with a 0% success rate and 0 tokens consumed. This indicates a systematic configuration or authentication failure — the workflow is failing before the agent even starts. This is the single biggest driver of the 61% overall success rate.
With 6.8M avg tokens per run (the highest of any workflow) and only 50% success rate, this workflow has the worst efficiency profile of any meaningful workflow. Each failed run wastes ~6.8M tokens.
75% success rate across 8 runs (2 failures). At 2.5M avg tokens/run, each failure is expensive.
✅ High Efficiency Highlights
Optimization Opportunities
Fix Issue Monster immediately
Documentation Noob Tester — reduce scope or add early exit
Chroma Issue Indexer — improve reliability
Daily Malicious Code Scan Agent — 50% success
Token concentration risk: Top 5 workflows = 73.5% of tokens
Methodology & Data Quality Notes
Methodology
/tmp/gh-aw/copilot-logs.json)Data Quality Notes
estimated_costreturned 0 for all runs; cost analysis not possible this periodtotal_turnsreturned 0 in summary; per-run turn analysis not availableReferences:
Beta Was this translation helpful? Give feedback.
All reactions