📊 Daily Copilot Token Consumption Report - February 12, 2026

[github-actions[bot]]

Executive Summary

Over the last 30 days, Copilot-powered agentic workflows consumed 215,998,387 tokens at an estimated cost of $21.60, across 333 workflow runs covering 49 unique workflows.

Key Highlights:

• Highest consuming workflow: Test Workflow (55.6M tokens, $5.56)
• Most active workflow: Test Dispatcher Workflow (132 runs)
• Average cost per run: $0.065
• Average tokens per run: 648,644
• Trend: First day of historical tracking - baseline established for future trend analysis

---

📈 Token Usage Trends

Overall Trends

The chart above shows daily token consumption. This is the first day of tracking, establishing a baseline of 216M tokens across 333 workflow runs. Future reports will show trends and patterns as we accumulate more data.

Cost Trends

Daily cost trends show an initial spend of $21.60. As we collect more data, this chart will reveal spending patterns and help identify optimization opportunities.

---

🏆 Top Workflows by Token Consumption

Top 10 Most Expensive Workflows

Rank	Workflow	Total Tokens	Total Cost	Runs	Avg Tokens/Run	Avg Cost/Run	Avg Duration
1	Test Workflow	55,562,640	$5.56	74	750,846	$0.0751	4.5m
2	CI Failure Doctor	53,017,438	$5.30	18	2,945,413	$0.2945	10.9m
3	Chroma Issue Indexer	19,005,364	$1.90	9	2,111,707	$0.2112	7.9m
4	Test Dispatcher Workflow	8,765,842	$0.88	132	66,408	$0.0066	4.1m
5	Auto-Triage Issues	6,901,155	$0.69	7	985,879	$0.0986	7.5m
6	PR Triage Agent	5,540,825	$0.55	6	923,471	$0.0923	6.1m
7	Example: Custom Error Patterns	4,723,648	$0.47	1	4,723,648	$0.4724	13.5m
8	Documentation Noob Tester	4,687,972	$0.47	2	2,343,986	$0.2344	10.3m
9	GPL Dependency Cleaner (gpclean)	4,206,600	$0.42	5	841,320	$0.0841	5.8m
10	The Great Escapi	3,806,947	$0.38	2	1,903,474	$0.1903	11.6m

Per-Workflow Detailed Statistics (All 49 Workflows)

Workflow	Total Tokens	Total Cost	Runs	Avg Tokens	Avg Cost	Avg Duration
Test Workflow	55,562,640	$5.56	74	750,846	$0.0751	4.5m
CI Failure Doctor	53,017,438	$5.30	18	2,945,413	$0.2945	10.9m
Chroma Issue Indexer	19,005,364	$1.90	9	2,111,707	$0.2112	7.9m
Test Dispatcher Workflow	8,765,842	$0.88	132	66,408	$0.0066	4.1m
Auto-Triage Issues	6,901,155	$0.69	7	985,879	$0.0986	7.5m
PR Triage Agent	5,540,825	$0.55	6	923,471	$0.0923	6.1m
Example: Custom Error Patterns	4,723,648	$0.47	1	4,723,648	$0.4724	13.5m
Documentation Noob Tester	4,687,972	$0.47	2	2,343,986	$0.2344	10.3m
GPL Dependency Cleaner (gpclean)	4,206,600	$0.42	5	841,320	$0.0841	5.8m
The Great Escapi	3,806,947	$0.38	2	1,903,474	$0.1903	11.6m
Auto-Assign Issue	3,071,085	$0.31	2	1,535,542	$0.1536	5.3m
jsweep - JavaScript Unbloater	2,985,186	$0.30	2	1,492,593	$0.1493	8.0m
Daily Regulatory Report Generator	2,789,986	$0.28	2	1,394,993	$0.1395	7.8m
Daily Syntax Error Quality Check	2,786,936	$0.28	2	1,393,468	$0.1393	6.0m
Functional Pragmatist	2,425,220	$0.24	1	2,425,220	$0.2425	8.1m
Daily News	2,344,967	$0.23	2	1,172,484	$0.1172	9.0m
CI Cleaner	2,264,227	$0.23	1	2,264,227	$0.2264	14.6m
Smoke Project	1,937,534	$0.19	13	149,041	$0.0149	3.3m
Daily MCP Tool Concurrency Analysis	1,913,020	$0.19	2	956,510	$0.0957	5.9m
Agent Container Smoke Test	1,861,987	$0.19	16	116,374	$0.0116	3.2m
Glossary Maintainer	1,680,189	$0.17	2	840,094	$0.0840	4.8m
Copilot PR Conversation NLP Analysis	1,612,997	$0.16	2	806,498	$0.0806	16.8m
Code Simplifier	1,543,133	$0.15	1	1,543,133	$0.1543	6.7m
Issue Triage Agent	1,356,649	$0.14	1	1,356,649	$0.1357	6.4m
Daily CLI Performance Agent	1,306,967	$0.13	1	1,306,967	$0.1307	7.1m
Workflow Skill Extractor	1,223,235	$0.12	1	1,223,235	$0.1223	7.6m
Terminal Stylist	1,165,444	$0.12	2	582,722	$0.0583	5.3m
Sub-Issue Closer	1,163,551	$0.12	1	1,163,551	$0.1164	5.7m
Layout Specification Maintainer	1,094,359	$0.11	1	1,094,359	$0.1094	9.5m
Workflow Health Manager - Meta-Orchestrator	1,088,213	$0.11	1	1,088,213	$0.1088	7.4m
Daily Copilot Token Consumption Report	942,249	$0.09	1	942,249	$0.0942	25.3m
Daily Semgrep Scan	938,829	$0.09	1	938,829	$0.0939	4.9m
Breaking Change Checker	893,506	$0.09	1	893,506	$0.0894	5.8m
Copilot CLI Deep Research Agent	841,801	$0.08	1	841,801	$0.0842	7.6m
Daily Workflow Updater	810,427	$0.08	1	810,427	$0.0810	4.0m
Plan Command	780,511	$0.08	2	390,256	$0.0390	4.2m
GitHub Remote MCP Authentication Test	729,450	$0.07	2	364,725	$0.0365	3.8m
Slide Deck Maintainer	684,813	$0.07	1	684,813	$0.0685	5.5m
Daily Compiler Quality Check	679,213	$0.07	1	679,213	$0.0679	6.8m
Daily Malicious Code Scan Agent	647,814	$0.06	1	647,814	$0.0648	4.2m
Daily Copilot PR Merged Report	627,403	$0.06	1	627,403	$0.0627	5.2m
The Daily Repository Chronicle	613,948	$0.06	1	613,948	$0.0614	6.7m
Dependabot Dependency Checker	590,698	$0.06	1	590,698	$0.0591	3.8m
Copilot PR Prompt Pattern Analysis	510,677	$0.05	1	510,677	$0.0511	6.9m
Delight	449,753	$0.04	1	449,753	$0.0450	4.9m
Repository Quality Improvement Agent	421,498	$0.04	1	421,498	$0.0421	7.8m
Daily Secrets Analysis Agent	371,898	$0.04	1	371,898	$0.0372	8.9m
CLI Consistency Checker	327,327	$0.03	1	327,327	$0.0327	3.4m
Draft PR Cleanup	303,256	$0.03	1	303,256	$0.0303	3.2m

---

💡 Insights & Recommendations

High-Cost Workflows

The following workflows account for the majority of token consumption:

1. Test Workflow - $5.56 (25.7% of total)

   • Observation: Most frequently run workflow (74 runs) with moderate per-run cost ($0.075). Primarily used for testing and development.
   • Recommendation: This is a test workflow with reasonable efficiency (750K tokens/run). Consider monitoring for unnecessary repeated runs.

2. CI Failure Doctor - $5.30 (24.5% of total)

   • Observation: High per-run cost ($0.29) with moderate frequency (18 runs). This workflow diagnoses CI failures, which typically requires extensive context analysis.
   • Recommendation: This is working as intended - CI failure diagnosis requires deep analysis. Consider batching CI failure investigations to reduce duplicate analysis.

3. Chroma Issue Indexer - $1.90 (8.8% of total)

   • Observation: High per-run cost ($0.21) with 9 runs. Indexes issues into a searchable knowledge base.
   • Recommendation: Consider incremental indexing rather than full re-indexing to reduce token consumption.

Optimization Opportunities

Efficiency Analysis by Workflow Type

High-Frequency, Low-Cost Workflows (Good Efficiency):

• Test Dispatcher Workflow: 132 runs at $0.0066/run - Very efficient automated triage
• Smoke Project: 13 runs at $0.0149/run - Efficient testing workflow
• Agent Container Smoke Test: 16 runs at $0.0116/run - Good efficiency for testing

Low-Frequency, High-Cost Workflows (Expected):

• Example: Custom Error Patterns: $0.47/run - One-time analysis, reasonable cost
• Documentation Noob Tester: $0.23/run - Documentation testing requires context
• Daily Copilot Token Consumption Report: $0.09/run - This report! Meta-analysis is moderately expensive

Potential Optimization Areas

1. Batch Processing for Repeated Analyses

   • Affected Workflows: CI Failure Doctor, Chroma Issue Indexer
   • Potential Savings: ~$2-3 per month
   • Action: Implement caching for repeated context (e.g., CI logs, issue content) to avoid re-analyzing same data

2. Incremental Updates vs. Full Scans

   • Affected Workflows: Issue indexing, daily reports
   • Potential Savings: ~$1-2 per month
   • Action: Store previous state and only process deltas where possible

3. Test Workflow Consolidation

   • Affected Workflows: Test Workflow, Test Dispatcher Workflow
   • Potential Savings: Minimal, but reduces noise
   • Action: Review if all test runs are necessary or if some can be consolidated

Efficiency Trends

Token Efficiency Metrics

• Most efficient workflows (tokens per run):

  • Test Dispatcher Workflow: 66,408 tokens/run
  • Agent Container Smoke Test: 116,374 tokens/run
  • Smoke Project: 149,041 tokens/run

• Least efficient workflows (tokens per run):

  • Example: Custom Error Patterns: 4,723,648 tokens/run (one-time deep analysis)
  • CI Failure Doctor: 2,945,413 tokens/run (requires extensive context)
  • Documentation Noob Tester: 2,343,986 tokens/run (comprehensive testing)

Run Pattern Analysis

• Peak activity: Test-related workflows dominate run frequency
• Triage workflows: Very efficient with low per-run costs
• Analysis workflows: Higher costs but provide significant value (CI diagnosis, documentation testing)

Duration vs. Cost Correlation

Workflows with longer duration tend to have higher costs, but efficiency varies:

• Daily Copilot Token Consumption Report: 25.3m duration, $0.09 (this report - data processing takes time)
• Copilot PR Conversation NLP Analysis: 16.8m duration, $0.08 (efficient for the analysis depth)
• CI Cleaner: 14.6m duration, $0.23 (moderate efficiency)

Historical Comparison

Baseline Data

This is the first day of historical tracking, establishing our baseline:

Metric	Today (Feb 12)	Notes
Total Tokens	215,998,387	Baseline established
Total Cost	$21.60	First measurement
Total Runs	333	Across 49 workflows
Avg Cost/Run	$0.065	Baseline efficiency
Workflows Active	49	Repository has diverse automation

What to Watch

As we accumulate more data, we'll track:

• Week-over-week changes: Are we becoming more or less efficient?
• Seasonal patterns: Do certain workflows run more frequently on specific days?
• Cost per workflow trends: Are individual workflows becoming more efficient over time?
• New workflow impact: How do new workflows affect overall consumption?

Next update will include:

• 7-day trend comparison
• Day-of-week patterns
• Workflow efficiency changes

Methodology & Data Quality Notes

Methodology

• Data Source: GitHub Actions workflow run artifacts from last 30 days
• Engine Filter: Copilot engine only (engine: copilot workflows)
• Memory Storage: /tmp/gh-aw/repo-memory/default/history.jsonl
• Analysis Date: 2026-02-12T11:30:38Z
• Historical Data: Day 1 of tracking - baseline established
• Cost Model: Based on Copilot token pricing (~$0.10 per 1M tokens)

Data Quality Notes

• All 333 workflow runs included valid token usage data
• Cost estimates are approximate based on standard Copilot pricing
• Duration data parsed from GitHub Actions run metadata
• First day of tracking means no historical trends yet available
• Future reports will include week-over-week and month-over-month comparisons

Pricing Assumptions

• Copilot token pricing: $0.10 per 1M tokens (approximate)
• Costs are estimates and may vary based on actual GitHub pricing
• Token counts are measured at the workflow level and include all agent interactions

Report Frequency

• Daily reports: Monday-Friday at 11 AM UTC
• Historical retention: 90 days of detailed data
• Trend analysis: Rolling 7-day and 30-day windows

---

🎯 Key Takeaways

1. Baseline Established: First day of tracking with 216M tokens across 333 runs
2. Top Consumers Identified: Test Workflow and CI Failure Doctor account for 50% of costs
3. Efficiency Varies: Dispatcher workflows are highly efficient, while analysis workflows require more tokens
4. Reasonable Costs: Average cost per run is $0.065, indicating efficient automation
5. Future Optimization: As trends emerge, we'll identify specific opportunities to reduce consumption

📅 Next Steps

• Tomorrow's report will include first day-over-day comparison
• Weekly report will show patterns and trends once we have 7 days of data
• Monthly report will provide comprehensive efficiency analysis

---

References:

• §21944377010 - Current workflow run

AI generated by Daily Copilot Token Consumption Report

• expires on Feb 15, 2026, 11:34 AM UTC