Overview
Period: 2026-08-22 to 2026-08-23
Runs analyzed: 6 (6 had token data)
Total tokens: 29K across all workflows
Estimated total cost: N/A (cost data not populated by api-proxy for this period)
Total Actions minutes: 25 min
Only one Claude-engine workflow ran in this window: Smoke Claude. No cost, turn count, error/warning counts, or cache breakdown (token_usage_summary) were available in the logs for any run — the api-proxy sidecar does not appear to be reporting these fields for this workflow, so cost- and cache-based optimization checks could not be evaluated.
Workflow Summary
| Workflow |
Runs |
Total Tokens |
Avg Tokens |
Cost |
Avg Cost |
Turns |
| Smoke Claude |
6 |
29K |
5K |
N/A |
N/A |
N/A |
🔍 Optimization Opportunities
No cost, turn, or cache-usage thresholds could be evaluated because estimated_cost, turns, and token_usage_summary were null for all 6 runs. Token usage itself is modest (~5K tokens/run, well under the 100K flag threshold), so nothing here indicates a problem — but the missing instrumentation should be addressed so future reports can compute cost and cache efficiency.
- Smoke Claude — instrumentation gap
estimated_cost, turns, error_count, warning_count, and token_usage_summary are all null across every run.- Recommendation: verify the api-proxy sidecar is enabled and correctly reporting Anthropic usage/cost headers for this workflow so cache-efficiency and cost-per-run can be tracked going forward.
Per-Workflow Details
Smoke Claude
- Runs: 6
- 32608687203 — 4,928 tokens, 3.7m
- 32589736216 — 5,053 tokens, 3.9m
- 32588848030 — 5,135 tokens, 4.2m
- 32588789021 — 5,080 tokens, 3.7m
- 32588424200 — 4,594 tokens, 3.5m
- 32573765366 — 4,463 tokens, 3.8m
- Total tokens: 29,253 (avg 4,876/run)
- Estimated cost: N/A (not reported)
- Turns: N/A (not reported)
- Model breakdown: not available (
token_usage_summary null for all runs)
- Cache analysis: not available
- Error rate: 0/6 runs (conclusion: success for all;
error_count/warning_count not reported but no failures observed)
Workflows Without Token Data
None — all 6 runs had non-null token_usage. However, cost (estimated_cost) and turn count (turns) were null for all runs, limiting deeper analysis.
Historical Trend
This is the first Claude token usage report (no accessible prior reports with the claude-token-usage-report label were found).
Previous Report
None found.
References:
Generated by Daily Claude Token Usage Analyzer · auto · 29.1 AIC · ⊞ 13.4K · ◷
Overview
Period: 2026-08-22 to 2026-08-23
Runs analyzed: 6 (6 had token data)
Total tokens: 29K across all workflows
Estimated total cost: N/A (cost data not populated by api-proxy for this period)
Total Actions minutes: 25 min
Only one Claude-engine workflow ran in this window: Smoke Claude. No cost, turn count, error/warning counts, or cache breakdown (
token_usage_summary) were available in the logs for any run — the api-proxy sidecar does not appear to be reporting these fields for this workflow, so cost- and cache-based optimization checks could not be evaluated.Workflow Summary
🔍 Optimization Opportunities
No cost, turn, or cache-usage thresholds could be evaluated because
estimated_cost,turns, andtoken_usage_summarywerenullfor all 6 runs. Token usage itself is modest (~5K tokens/run, well under the 100K flag threshold), so nothing here indicates a problem — but the missing instrumentation should be addressed so future reports can compute cost and cache efficiency.estimated_cost,turns,error_count,warning_count, andtoken_usage_summaryare all null across every run.Per-Workflow Details
Smoke Claude
token_usage_summarynull for all runs)error_count/warning_countnot reported but no failures observed)Workflows Without Token Data
None — all 6 runs had non-null
token_usage. However, cost (estimated_cost) and turn count (turns) were null for all runs, limiting deeper analysis.Historical Trend
This is the first Claude token usage report (no accessible prior reports with the
claude-token-usage-reportlabel were found).Previous Report
None found.
References: