📊 Copilot Token Usage Report2026-05-20

[github-actions]

Overview

Period: 2026-05-19 to 2026-05-20
Runs analyzed: 50 (48 had token data)
Total tokens: 16,271K across all workflows
Effective tokens (cache-normalized): 164,861K
Total Actions minutes: 246 min

All 50 runs used claude-sonnet-4.6 via the Copilot engine. No cost data was available (estimated_cost = $0 for all runs — api-proxy cost estimation may not be enabled for this engine). Token counts are available and cache efficiency is generally high (80–93%).

Workflow Summary

Workflow	Runs	Total Tokens	Avg Tokens/Run	Avg Turns	Cache Hit%
Build Test Suite	5	3,128K	625K	1.0	93%
API Surface & Export Audit	7	2,821K	403K	1.0	90%
Test Coverage Improver	1	2,682K	2,682K	1.0	90%
Test Coverage Reporter	8	2,216K	277K	1.0	86%
Smoke Services	6	1,557K	259K	1.0	89%
Smoke Copilot BYOK	6	879K	146K	1.0	82%
Smoke Copilot	6	853K	142K	1.2	84%
Dependency Security Monitor	1	571K	571K	1.0	92%
CI/CD Pipelines & Integration Tests Gap Assessment	1	471K	471K	1.0	86%
Daily Claude Token Usage Analyzer	1	462K	462K	1.0	88%
Firewall Issue Dispatcher	2	309K	154K	1.0	78%
Smoke Chroot	4	317K	79K	1.0	69%
Daily Claude Token Optimization Advisor ⚠️	1	—	—	—	N/A
Daily Copilot Token Usage Analyzer ⚠️	1	—	—	—	N/A

⚠️ = no token data (api-proxy may not be active for these runs)

🔍 Optimization Opportunities

1. Test Coverage Improver — 2,682K tokens in a single run

   • Highest single-run token count observed. If this runs regularly, review whether the task scope can be narrowed or decomposed into smaller units.
   • Cache hit 90% is healthy — the high cost is driven by large input context.

2. Build Test Suite — avg 625K tokens/run, 5 runs (3,128K total)

   • Second-highest per-run consumption and largest aggregate consumer. With 93% cache hit, the context is large but stable. Consider whether all 5 runs are necessary or if some are redundant trigger events.

3. Dependency Security Monitor — 571K tokens in a single run

   • 92% cache hit is the best observed. Still a large single-run context; verify the system prompt isn't unnecessarily large.

4. API Surface & Export Audit — avg 403K tokens/run, 7 runs (2,821K total)

   • High run frequency with 7 runs is the main aggregate driver. Review trigger conditions if some runs are redundant.

5. All workflows — avg turns ≈ 1.0

   • Every workflow completes in approximately a single turn, which is efficient. No multi-turn inefficiency detected.

Historical Trend

This is the first Copilot token usage report — no prior baseline exists for comparison.

Per-Workflow Run Details

Build Test Suite

• Runs: 5 | Total tokens: 3,128K (avg 625K) | Turns: 5 | Cache hit: 93%
• §26139876448 — 2026-05-20, 831K tokens, success
• §26134221008 — 2026-05-20, 635K tokens, success
• §26134103328 — 2026-05-20, 680K tokens, success
• §26133977519 — 2026-05-20, 554K tokens, success
• §26131829166 — 2026-05-19, 426K tokens, success

API Surface & Export Audit

• Runs: 7 | Total tokens: 2,821K (avg 403K) | Turns: 7 | Cache hit: 90%
• §26133838253 — 655K, §26131607758 — 577K, §26131589888 — 523K, §26139714873 — 510K, §26141543362 — 197K, §26132441643 — 196K, §26139705829 — 160K

Test Coverage Improver

• Runs: 1 | Total tokens: 2,682K | Turns: 1 | Cache hit: 90%
• §26153002679 — 2026-05-20, success

Test Coverage Reporter

• Runs: 8 | Total tokens: 2,216K (avg 277K) | Turns: 8 | Cache hit: 86%
• §26144085951 282K · §26141543363 246K · §26139714895 281K · §26139705880 253K · §26133838256 285K · §26132441602 251K · §26131607763 284K · §26131589867 331K

Smoke Services

• Runs: 6 | Total tokens: 1,557K (avg 259K) | Turns: 6 | Cache hit: 89%
• §26139876492 221K · §26135286155 254K · §26134188921 254K · §26134103329 221K · §26133977494 285K · §26131829124 318K

Smoke Copilot BYOK

• Runs: 6 | Total tokens: 879K (avg 146K) | Turns: 6 | Cache hit: 82%
• §26139876446 · §26135472656 · §26134188677 · §26134103327 · §26133977517 · §26131829201

Smoke Copilot

• Runs: 6 | Total tokens: 853K (avg 142K) | Turns: 7 (avg 1.2) | Cache hit: 84%
• §26139876443 · §26135473948 · §26134188759 · §26134103321 · §26133977492 · §26131829095

Dependency Security Monitor

• Runs: 1 | Total tokens: 571K | Turns: 1 | Cache hit: 92%
• §26146878471 — 2026-05-20, success

CI/CD Pipelines & Integration Tests Gap Assessment

• Runs: 1 | Total tokens: 471K | Turns: 1 | Cache hit: 86%
• §26144348815 — 2026-05-20, success

Daily Claude Token Usage Analyzer

• Runs: 1 | Total tokens: 462K | Turns: 1 | Cache hit: 88%
• §26153516523 — 2026-05-20, success

Firewall Issue Dispatcher

• Runs: 2 | Total tokens: 309K (avg 154K) | Turns: 2 | Cache hit: 78%
• §26148278942 · §26135423738

Smoke Chroot

• Runs: 4 | Total tokens: 317K (avg 79K) | Turns: 4 | Cache hit: 69%
• §26139876450 · §26134188993 · §26134103326 · §26131829127

Workflows Without Token Data

Runs where token_usage was null or 0 — these may not have the api-proxy active, or the runs had not yet completed token accounting at time of export:

• Daily Claude Token Optimization Advisor — §26153838797 (2026-05-20, conclusion unknown at time of export)
• Daily Copilot Token Usage Analyzer — §26154216292 (this run — in progress when logs were captured)

References:

• §26154216292 — this report run
• §26153002679 — highest single-run token consumer (Test Coverage Improver)
• §26139876448 — largest Build Test Suite run

Generated by Daily Copilot Token Usage Analyzer · ● 4.4M · ◷