📊 Copilot Token Usage Report2026-05-27

[github-actions]

Overview

Period: 2026-05-27T03:48Z to 2026-05-27T09:37Z (~5.8 hours)
Runs analyzed: 50 (46 had token data)
Total tokens: 18,150,356
Total effective tokens: 181,848,374
Total turns: 473
Total Actions minutes: 227 min

i️ estimated_cost is $0.00 for all runs — the api-proxy sidecar cost tracking is not yet emitting cost data. Effective tokens (a 10× multiplier proxy metric) are used as the cost-effort proxy in this report.

Workflow Summary

Workflow	Runs	Total Tokens	Avg Tokens	Eff Tokens	Avg Eff Tokens	Turns	Act Min
Build Test Suite	6	3,969K	662K	39,917K	6,653K	120	33
Test Coverage Improver	1	3,931K	3,931K	39,515K	39,515K	64	14
API Surface & Export Audit	4	3,022K	755K	30,374K	7,594K	75	26
Smoke Services	6	1,674K	279K	16,631K	2,772K	53	31
Smoke Copilot	8	1,211K	151K	11,956K	1,494K	39	25
Smoke Copilot BYOK	7	959K	137K	9,430K	1,347K	31	22
Test Coverage Reporter	5	728K	146K	7,402K	1,480K	20	19
CI/CD Gap Assessment	1	630K	630K	6,414K	6,414K	12	6
Smoke OTel Tracing	2	629K	315K	6,278K	3,139K	17	7
Daily Claude Token Analyzer	1	573K	573K	5,805K	5,805K	14	11
Smoke Chroot	5	394K	79K	3,872K	774K	15	25
Firewall Issue Dispatcher	1	241K	241K	2,400K	2,400K	8	3
Dependency Security Monitor	1	186K	186K	1,855K	1,855K	5	3
Daily Claude Token Optimization Advisor	1	0	—	0	—	0	1
Daily Copilot Token Usage Analyzer	1	0	—	0	—	0	1

🔍 Optimization Opportunities

1. Test Coverage Improver — 3,931K tokens, 64 turns in a single run

   • 64 turns is extremely high; consider breaking this into smaller scoped tasks or adding early-exit conditions
   • Review whether the workflow is doing too much in one pass (e.g., split by module)

2. Build Test Suite — avg 662K tokens, 20 turns/run across 6 runs

   • 20 turns/run consistently suggests iterative back-and-forth; consider pre-computing deterministic steps outside the agent turn loop
   • 6 runs × 662K = 3.97M tokens — highest aggregate consumer

3. API Surface & Export Audit — avg 755K tokens, 18.8 turns/run

   • High token count per run; consider reducing MCP tool surface or narrowing the prompt scope per run
   • 4 runs in the period suggests it may be re-triggered frequently

4. Test Coverage Reporter — 40% error rate (2/5 runs with errors)

   • Reliability issue; investigate root cause of failures before optimizing token usage
   • 2 runs had missing token data, which may be related to failures

5. Smoke OTel Tracing — avg 315K tokens/run

   • Higher than other smoke tests (Chroot: 79K, Copilot: 151K); review if this test requires additional context or tool calls

6. Daily Claude Token Usage Analyzer (run by Copilot) — 573K tokens, 14 turns

   • High for a reporting workflow; consider reducing context size passed to the agent

Per-Workflow Details

Build Test Suite

• Runs: 6 — #1966, #1965, #1964, #1963, #1962, #1961
• Total tokens: 3,969,352 (avg 661,558/run)
• Effective tokens: 39,916,587 (avg 6,652,764/run)
• Turns: 120 total (avg 20.0/run)
• Action minutes: 33
• Error rate: 0/6 runs ✅

Test Coverage Improver

• Runs: 1 — #32
• Total tokens: 3,931,484 (avg 3,931,484/run)
• Effective tokens: 39,515,406 (avg 39,515,406/run)
• Turns: 64 total (avg 64.0/run)
• Action minutes: 14
• Error rate: 0/1 runs ✅

API Surface & Export Audit

• Runs: 4 — #309, #307, #305, #304
• Total tokens: 3,021,651 (avg 755,412/run)
• Effective tokens: 30,374,230 (avg 7,593,557/run)
• Turns: 75 total (avg 18.8/run)
• Action minutes: 26
• Error rate: 0/4 runs ✅

Smoke Services

• Runs: 6 — #1744, #1743, #1742, #1741, #1740, #1739
• Total tokens: 1,674,394 (avg 279,065/run)
• Effective tokens: 16,630,913 (avg 2,771,818/run)
• Turns: 53 total (avg 8.8/run)
• Action minutes: 31
• Error rate: 0/6 runs ✅

Smoke Copilot

• Runs: 8 — #3853, #3852, #3851, #3850, #3849, #3848, #3847, #3846
• Total tokens: 1,211,345 (avg 151,418/run)
• Effective tokens: 11,955,557 (avg 1,494,444/run)
• Turns: 39 total (avg 4.9/run)
• Action minutes: 25
• Error rate: 0/8 runs ✅

Smoke Copilot BYOK

• Runs: 7 — #1293, #1292, #1291, #1290, #1289, #1288, #1287
• Total tokens: 959,368 (avg 137,052/run)
• Effective tokens: 9,429,615 (avg 1,347,087/run)
• Turns: 31 total (avg 4.4/run)
• Action minutes: 22
• Error rate: 0/7 runs ✅

Test Coverage Reporter

• Runs: 5 — #334, #333, #330, #329, #328
• Total tokens: 727,875 (avg 145,575/run)
• Effective tokens: 7,402,087 (avg 1,480,417/run)
• Turns: 20 total (avg 4.0/run)
• Action minutes: 19
• Error rate: 2/5 runs ⚠️ (2 runs also missing token data)

CI/CD Pipelines and Integration Tests Gap Assessment

• Runs: 1 — #159
• Total tokens: 630,374 (avg 630,374/run)
• Effective tokens: 6,414,271 (avg 6,414,271/run)
• Turns: 12 total (avg 12.0/run)
• Action minutes: 6
• Error rate: 0/1 runs ✅

Smoke OTel Tracing

• Runs: 2 — #86, #85
• Total tokens: 629,495 (avg 314,747/run)
• Effective tokens: 6,278,157 (avg 3,139,078/run)
• Turns: 17 total (avg 8.5/run)
• Action minutes: 7
• Error rate: 0/2 runs ✅

Daily Claude Token Usage Analyzer

• Runs: 1 — #58
• Total tokens: 573,243 (avg 573,243/run)
• Effective tokens: 5,805,016 (avg 5,805,016/run)
• Turns: 14 total (avg 14.0/run)
• Action minutes: 11
• Error rate: 0/1 runs ✅

Smoke Chroot

• Runs: 5 — #1666, #1665, #1664, #1663, #1662
• Total tokens: 394,351 (avg 78,870/run)
• Effective tokens: 3,871,839 (avg 774,367/run)
• Turns: 15 total (avg 3.0/run)
• Action minutes: 25
• Error rate: 0/5 runs ✅

Firewall Issue Dispatcher

• Runs: 1 — #248
• Total tokens: 241,490 (avg 241,490/run)
• Effective tokens: 2,400,062 (avg 2,400,062/run)
• Turns: 8 total (avg 8.0/run)
• Action minutes: 3
• Error rate: 0/1 runs ✅

Dependency Security Monitor

• Runs: 1 — #129
• Total tokens: 185,934 (avg 185,934/run)
• Effective tokens: 1,854,634 (avg 1,854,634/run)
• Turns: 5 total (avg 5.0/run)
• Action minutes: 3
• Error rate: 0/1 runs ✅

Workflows Without Token Data

Runs where token_usage was null or 0 — these may not have the api-proxy enabled or are still in-flight:

• Daily Claude Token Optimization Advisor — 1/1 runs missing token data (#62)
• Daily Copilot Token Usage Analyzer — 1/1 runs missing token data (#59) (this run)
• Test Coverage Reporter — 2/5 runs missing token data

Historical Trend

This is the first Copilot token usage report — no previous data to compare against.

References:

• §26503330304 — this analyzer run

Generated by Daily Copilot Token Usage Analyzer · sonnet46 1.2M · ◷

[github-actions]

This issue is being closed as outdated. A newer issue has been created: #3947

View newer issue

---

This action was performed automatically by the Daily Copilot Token Usage Analyzer workflow.