📊 Copilot Token Usage Report2026-04-18

[github-actions]

Overview

Period: 2026-04-17 to 2026-04-18
Runs analyzed: 50 (47 had token data)
Total tokens: 11,495,982 (~11.5M) across all workflows
Effective tokens: ~12.8M (includes cache write weighting)
Estimated total cost: $0.00 (api-proxy cost tracking not yet emitting USD values)
Total Actions minutes: 239 min
All models: claude-sonnet-4.6 (via Copilot API proxy)

Workflow Summary

Workflow	Runs	Total Tokens	Avg Tokens/Run	Total Turns	Avg Turns	Errors
Build Test Suite	8	4,555K	569K	8	1.0	0
Smoke Services	9	2,086K	232K	8	0.9	2
Smoke Copilot BYOK	10	1,223K	122K	10	1.0	0
Smoke Copilot	9	995K	111K	8	0.9	1
Firewall Issue Dispatcher	3	581K	194K	3	1.0	0
Daily Claude Token Usage Analyzer	1	553K	553K	1	1.0	0
Smoke Chroot	5	425K	85K	5	1.0	0
Issue Duplication Detector	2	418K	209K	2	1.0	0
Pelis Agent Factory Advisor	1	339K	339K	1	1.0	0
Dependency Security Monitor	1	319K	319K	1	1.0	0
Daily Copilot Token Usage Analyzer	1	0K	0K	0	0.0	0

🔍 Optimization Opportunities

1. Build Test Suite — 569K tokens/run (highest consumer, 40% of total)

   • Each run uses 1 turn but consumes ~569K tokens — likely a large system prompt or extensive test output fed back to the model
   • Consider reducing context passed to the agent; pre-filter test results before sending to LLM

2. Daily Claude Token Usage Analyzer — 553K tokens/run (single-run, one-shot workflow)

   • Large token footprint for a reporting workflow; review whether all log data needs to be in-context
   • Consider summarizing logs before feeding to the model

3. Pelis Agent Factory Advisor — 339K tokens/run

   • Single run; monitor to see if this is a consistently heavy workflow

4. Dependency Security Monitor — 319K tokens/run

   • Single run; may be ingesting large dependency manifests; consider pre-filtering to changed or flagged deps only

5. Firewall Issue Dispatcher — 194K avg tokens/run, high variance (104K–303K)

   • Variance suggests input size varies significantly; the 303K run is 3× the smallest; investigate what drives large inputs

6. Smoke Services — 2/9 runs errored (22% error rate, approaching 30% threshold)

   • One failure had 0 token data (early exit); the 471K-token failure used 2× normal tokens, suggesting retry/recovery loops

Per-Workflow Details

Build Test Suite

• Runs: 8
• Total tokens: 4,555K (avg 569K/run)
• Effective tokens: 5,100K
• Turns: 8 total (avg 1.0/run)
• Model: claude-sonnet-4.6
• Error rate: 0/8 runs
  • ✅ 24593695599 — 522K tokens, 6.1m, 2026-04-18
  • ✅ 24593538044 — 676K tokens, 6.1m, 2026-04-18
  • ✅ 24593463065 — 1,064K tokens, 6.4m, 2026-04-18
  • ✅ 24591315629 — 430K tokens, 5.5m, 2026-04-17
  • ✅ 24590789048 — 113K tokens, 3.5m, 2026-04-17
  • ✅ 24586966574 — 113K tokens, 3.8m, 2026-04-17
  • ✅ 24583912074 — 505K tokens, 6.7m, 2026-04-17
  • ✅ 24577008466 — 1,127K tokens, 6.6m, 2026-04-17

Smoke Services

• Runs: 9
• Total tokens: 2,086K (avg 232K/run)
• Effective tokens: 2,298K
• Turns: 8 total (avg 0.9/run)
• Model: claude-sonnet-4.6
• Error rate: 2/9 runs
  • ✅ 24593695592 — 269K tokens, 4.5m, 2026-04-18
  • ✅ 24593538037 — 239K tokens, 4.7m, 2026-04-18
  • ✅ 24593463083 — 239K tokens, 4.5m, 2026-04-18
  • ❌ 24593185767 — 471K tokens, 5.3m, 2026-04-18
  • ✅ 24591315630 — 178K tokens, 3.8m, 2026-04-17
  • ✅ 24590789078 — 208K tokens, 3.9m, 2026-04-17
  • ✅ 24586966613 — 240K tokens, 4.2m, 2026-04-17
  • ✅ 24583912053 — 239K tokens, 4.9m, 2026-04-17
  • ❌ 24577008397 — 0K tokens, 1.5m, 2026-04-17

Smoke Copilot BYOK

• Runs: 10
• Total tokens: 1,223K (avg 122K/run)
• Effective tokens: 1,346K
• Turns: 10 total (avg 1.0/run)
• Model: claude-sonnet-4.6
• Error rate: 0/10 runs
  • ✅ 24593695587 — 101K tokens, 3.1m, 2026-04-18
  • ✅ 24593538055 — 127K tokens, 3.5m, 2026-04-18
  • ✅ 24593463087 — 127K tokens, 3.2m, 2026-04-18
  • ✅ 24593219339 — 130K tokens, 3.3m, 2026-04-18
  • ✅ 24591315666 — 152K tokens, 3.5m, 2026-04-17
  • ✅ 24590789050 — 101K tokens, 3.1m, 2026-04-17
  • ✅ 24586966571 — 101K tokens, 3.4m, 2026-04-17
  • ✅ 24583912059 — 127K tokens, 3.5m, 2026-04-17
  • ✅ 24577008408 — 126K tokens, 3.9m, 2026-04-17
  • ✅ 24572922797 — 127K tokens, 3.4m, 2026-04-17

Smoke Copilot

• Runs: 9
• Total tokens: 995K (avg 111K/run)
• Effective tokens: 1,095K
• Turns: 8 total (avg 0.9/run)
• Model: claude-sonnet-4.6
• Error rate: 1/9 runs
  • ✅ 24593695588 — 100K tokens, 3.2m, 2026-04-18
  • ✅ 24593538060 — 101K tokens, 3.5m, 2026-04-18
  • ✅ 24593463077 — 127K tokens, 3.3m, 2026-04-18
  • ✅ 24593072788 — 181K tokens, 3.2m, 2026-04-18
  • ✅ 24591315631 — 126K tokens, 3.5m, 2026-04-17
  • ✅ 24590789061 — 100K tokens, 3.3m, 2026-04-17
  • ✅ 24586966572 — 100K tokens, 3.4m, 2026-04-17
  • ✅ 24583912103 — 156K tokens, 3.4m, 2026-04-17
  • ❌ 24577008415 — 0K tokens, 1.4m, 2026-04-17

Firewall Issue Dispatcher

• Runs: 3
• Total tokens: 581K (avg 194K/run)
• Effective tokens: 644K
• Turns: 3 total (avg 1.0/run)
• Model: claude-sonnet-4.6
• Error rate: 0/3 runs
  • ✅ 24599044648 — 104K tokens, 2.9m, 2026-04-18
  • ✅ 24593111713 — 173K tokens, 3.8m, 2026-04-18
  • ✅ 24580718318 — 303K tokens, 3.8m, 2026-04-17

Daily Claude Token Usage Analyzer

• Runs: 1
• Total tokens: 553K (avg 553K/run)
• Effective tokens: 638K
• Turns: 1 total (avg 1.0/run)
• Model: claude-sonnet-4.6
• Error rate: 0/1 runs
  • ✅ 24584312276 — 552K tokens, 18.6m, 2026-04-17

Smoke Chroot

• Runs: 5
• Total tokens: 425K (avg 85K/run)
• Effective tokens: 460K
• Turns: 5 total (avg 1.0/run)
• Model: claude-sonnet-4.6
• Error rate: 0/5 runs
  • ✅ 24593538063 — 84K tokens, 4.5m, 2026-04-18
  • ✅ 24591315732 — 84K tokens, 4.4m, 2026-04-17
  • ✅ 24586966576 — 84K tokens, 4.4m, 2026-04-17
  • ✅ 24583912062 — 84K tokens, 4.8m, 2026-04-17
  • ✅ 24577008418 — 85K tokens, 4.6m, 2026-04-17

Issue Duplication Detector

• Runs: 2
• Total tokens: 418K (avg 209K/run)
• Effective tokens: 462K
• Turns: 2 total (avg 1.0/run)
• Model: claude-sonnet-4.6
• Error rate: 0/2 runs
  • ✅ 24593182740 — 201K tokens, 3.8m, 2026-04-18
  • ✅ 24575076475 — 216K tokens, 3.9m, 2026-04-17

Pelis Agent Factory Advisor

• Runs: 1
• Total tokens: 339K (avg 339K/run)
• Effective tokens: 383K
• Turns: 1 total (avg 1.0/run)
• Model: claude-sonnet-4.6
• Error rate: 0/1 runs
  • ✅ 24587829426 — 338K tokens, 5.2m, 2026-04-17

Dependency Security Monitor

• Runs: 1
• Total tokens: 319K (avg 319K/run)
• Effective tokens: 356K
• Turns: 1 total (avg 1.0/run)
• Model: claude-sonnet-4.6
• Error rate: 0/1 runs
  • ✅ 24591181416 — 319K tokens, 4.1m, 2026-04-17

Workflows Without Token Data

Runs where token_usage was null or 0 — these may have exited early before any LLM call, or have not completed yet:

• Daily Copilot Token Usage Analyzer — 24602654260 — in_progress (this run)
• Smoke Services — 24577008397 — failed early exit (1.5m)
• Smoke Copilot — 24577008415 — failed early exit (1.4m)

Historical Trend

This is the first Copilot token usage report. Future reports will compare against this baseline.

References:

• §24602654260 — this analyzer run

Generated by Daily Copilot Token Usage Analyzer · ● 386.5K · ◷