Problem
From Repository Quality audit #35222: daily-observability-report.md sets max-effective-tokens: 80000000 (80M tokens) — twice the next-highest value across all 6 workflows that set this field. It runs daily, with no documented justification for the elevated ceiling.
With the API consumption report (#35204) showing today consumed 118K REST calls (2.6× the 7-day average), per-workflow budget hygiene matters.
Expected Impact
- Either documents the legitimate reason for the 80M cap (so future audits stop flagging it), or reduces it to match peers (saves daily token budget)
- Either way, makes the ceiling auditable
Suggested Approach
- Check actual token usage from recent runs of
daily-observability-report (gh-aw logs MCP tool)
- If real usage is well below 80M, drop the cap to a value closer to actual (e.g. 2× observed p95)
- If real usage approaches 80M, add a comment in the workflow frontmatter explaining why
Suggested Agent
Claude or Copilot — needs a quick log query + a single frontmatter edit.
Acceptance Criteria
- Either reduced
max-effective-tokens with no degraded report quality over 3 runs, or documented justification in the frontmatter
Estimated Effort
30–60 minutes
Source
Generated by 🔬 DeepReport - Intelligence Gathering Agent · opus47 9.2M · ◷
Problem
From Repository Quality audit #35222:
daily-observability-report.mdsetsmax-effective-tokens: 80000000(80M tokens) — twice the next-highest value across all 6 workflows that set this field. It runs daily, with no documented justification for the elevated ceiling.With the API consumption report (#35204) showing today consumed 118K REST calls (2.6× the 7-day average), per-workflow budget hygiene matters.
Expected Impact
Suggested Approach
daily-observability-report(gh-awlogsMCP tool)Suggested Agent
Claude or Copilot — needs a quick log query + a single frontmatter edit.
Acceptance Criteria
max-effective-tokenswith no degraded report quality over 3 runs, or documented justification in the frontmatterEstimated Effort
30–60 minutes
Source