Overview
Period: 2026-06-06T18:07Z to 2026-06-07T09:31Z (~15.4 hours)
Runs analyzed: 50 (47 had token data)
Total raw tokens: 25.2M ⬆️ +7.2% vs yesterday (23.5M) | Effective tokens: 215.6M ⬇️ -14.2%
Total Actions minutes: 328 min ⬆️ +6.8% | Agent turns: 49 ⬇️ -9.3% | Errors: 4 ⬇️ -33%
i️ No estimated_cost data available. Effective tokens (cost-normalized) used as primary indicator. BYOK/AOAI workflows show <1× eff/raw ratio; standard Copilot workflows show ~10×.
Workflow Summary
Sorted by total effective tokens (highest first)
🔍 Optimization Opportunities
-
Build Test Suite — one failure (§27070894953) used 2.5M raw + 25.2M eff — 4× the normal amount. Investigate what drove token spend before the failure.
-
Firewall Issue Dispatcher — avg 1.2M raw/run with high variability (261K–2.1M). Latest run (§27086434217) used 8× more tokens than oldest in window; review if issue volume is driving the spike.
-
Duplicate Code Detector — single run at 2.5M raw + 25.0M eff; confirm whether codebase diff size has grown.
-
Smoke Copilot BYOK AOAI (Entra) — 50% failure rate (2/4 runs); auth/connectivity instability flagged yesterday (60%) continues.
-
Test Coverage Improver — failed before reporting tokens; 3 runs in 2 days with no token data.
Workflows Without Token Data (3 runs)
| Workflow |
Status |
Reason |
| Daily Claude Token Optimization Advisor §27088576857 |
in_progress |
Still executing at report time |
| Daily Copilot Token Usage Analyzer §27088714527 |
in_progress |
This run |
| Test Coverage Improver §27088176845 |
failure |
Failed before token reporting |
Historical Trend
| Metric |
2026-06-05 |
2026-06-06 |
2026-06-07 |
Δ (1d) |
| Total raw tokens |
24.4M |
23.5M |
25.2M |
⬆️ +7.2% |
| Effective tokens |
246.8M |
251.3M |
215.6M |
⬇️ -14.2% |
| Action minutes |
306 |
307 |
328 |
⬆️ +6.8% |
| Agent turns |
47 |
54 |
49 |
⬇️ -9.3% |
| Errors |
4 |
6 |
4 |
⬇️ -33% |
Raw tokens up slightly; effective tokens dropped 14% (BYOK/AOAI workflows with low eff/raw ratios represent a larger share today). Errors improved. Main concerns: Build Test Suite failure-run token spike; Smoke Copilot BYOK AOAI (Entra) reliability.
Previous Report
📊 Copilot Token Usage Report 2026-06-06 — #4427
References:
Generated by Daily Copilot Token Usage Analyzer · sonnet46 2.4M · ◷
Overview
Period: 2026-06-06T18:07Z to 2026-06-07T09:31Z (~15.4 hours)
Runs analyzed: 50 (47 had token data)
Total raw tokens: 25.2M ⬆️ +7.2% vs yesterday (23.5M) | Effective tokens: 215.6M ⬇️ -14.2%
Total Actions minutes: 328 min ⬆️ +6.8% | Agent turns: 49 ⬇️ -9.3% | Errors: 4 ⬇️ -33%
Workflow Summary
Sorted by total effective tokens (highest first)
🔍 Optimization Opportunities
Build Test Suite — one failure (§27070894953) used 2.5M raw + 25.2M eff — 4× the normal amount. Investigate what drove token spend before the failure.
Firewall Issue Dispatcher — avg 1.2M raw/run with high variability (261K–2.1M). Latest run (§27086434217) used 8× more tokens than oldest in window; review if issue volume is driving the spike.
Duplicate Code Detector — single run at 2.5M raw + 25.0M eff; confirm whether codebase diff size has grown.
Smoke Copilot BYOK AOAI (Entra) — 50% failure rate (2/4 runs); auth/connectivity instability flagged yesterday (60%) continues.
Test Coverage Improver — failed before reporting tokens; 3 runs in 2 days with no token data.
Workflows Without Token Data (3 runs)
Historical Trend
Raw tokens up slightly; effective tokens dropped 14% (BYOK/AOAI workflows with low eff/raw ratios represent a larger share today). Errors improved. Main concerns: Build Test Suite failure-run token spike; Smoke Copilot BYOK AOAI (Entra) reliability.
Previous Report
📊 Copilot Token Usage Report 2026-06-06 — #4427
References: