[api-consumption] 📊 GitHub API Consumption Report — 2026-08-13

[github-actions[bot]]

📊 GitHub API Consumption Report

Report Date: 2026-08-13 · Repository: github/gh-aw · Run: #31693760914

---

Today at a Glance

Metric	Value
🤖 Total Runs	208 (163 ✅ / 45 ❌)
🎯 Success Rate	78.4%
🔗 GitHub REST API Calls	120,017 (core quota consumed — includes reads, writes, and all GitHub API operations)
📝 Safe-Output Writes	362 (issues + PRs + comments + discussions created by safe-output tools)
⏱ Avg Duration	626.6s (p95: 1423.0s)

---

🔗 GitHub API Calls Trend (90 days)

Only 3 calendar days of usable history exist so far (2026-08-11 to 2026-08-13) — API consumption swung from 193,421 calls on 08-11 down to 71,068 on 08-12, then back up to 120,017 today. No sustained multi-week trend can be established yet; this chart will fill in as daily runs accumulate.

---

🔗 GitHub API Calls by Workflow Trend (30 days)

Across the 3 days tracked, PR Sous Chef, the skills-reviewer family (Impeccable/Matt Pocock/Ponytail), and PR Code Quality Reviewer are the most consistent top consumers, each appearing in the top ranks every day. Design Decision Gate also shows persistently high per-run cost due to its long average duration.

---

🔗 GitHub REST API Calls Heatmap (90 days)

The heatmap (top-10 workflows × available days) shows dense, dark cells for the skills-reviewer workflows and PR Sous Chef on all three days, while several workflows only appear on a subset of days (shown as blank cells) — reflecting their less-frequent trigger schedules rather than missing data.

---

🍩 Top API Burners (24h)

The top 5 workflows (Impeccable Skills Reviewer, PR Code Quality Reviewer, Ponytail Reviewer, Matt Pocock Skills Reviewer, Design Decision Gate) each consumed roughly 11% of today's total, together accounting for over half of all API calls — a meaningful concentration among the PR-review family of workflows.

---

🔗 GitHub REST API Consumption by Workflow (last 24h)

Impeccable Skills Reviewer (13,465 calls / 12 runs) narrowly leads, followed closely by PR Code Quality Reviewer, Ponytail Reviewer, and Matt Pocock Skills Reviewer — all four cluster tightly around ~13,000 calls each. None are individually close to the 15,000/hr rate limit, but running several of these PR-review workflows concurrently on a busy PR could compound quickly; consolidating overlapping review workflows may be worth considering.

---

Top 10 Workflows by REST API Consumption (last 24h)

Workflow	REST API Calls	Runs	Avg Duration
Impeccable Skills Reviewer	13,465	12	419.5s
PR Code Quality Reviewer	13,407	12	404.5s
Ponytail Reviewer	13,360	12	322.2s
Matt Pocock Skills Reviewer	13,020	12	475.6s
Design Decision Gate 🏗️	12,903	12	561.8s
PR Sous Chef	8,688	24	511.6s
Daily Skill Optimizer Improvements	3,534	1	572.0s
LintMonster	2,832	1	411.0s
Daily Compiler Quality Check	2,789	1	611.0s
Avenger	2,748	2	681.5s

---

Trending Indicators

• 7-day API trend: → insufficient data (only 3 days of history collected so far; need 14 days for a 7-day-over-7-day comparison)
• 30-day API trend: → insufficient data (need 60 days for a 30-day-over-30-day comparison)
• GitHub REST API call rate: ~128,169 calls/day average over the 3 tracked days (hourly limit: 15,000) — no single workflow is close to the per-hour ceiling, but aggregate daily volume is substantial

---

📦 Cache Memory Status

• Location: /tmp/gh-aw/cache-memory/trending/api-consumption/history.jsonl
• Cache restored from previous run: yes (2 existing entries: 2026-08-11, 2026-08-12)
• Collection mode: backfill
• Logs start_date used: -90d (with 2 continuation calls, both capped at 100 runs each; continuation was still available after the 2nd call so paging was stopped per the retry-limit policy)
• Data points stored: 3
• Earliest entry: 2026-08-11
• Retention policy: 90 days

Data completeness note: 359 run directories were downloaded across the 3 MCP logs calls, but only 208 contained usable aw_info.json/run_summary.json data — the remaining 150 directories (corresponding to older runs further back in the 90-day window) were empty, likely because their log artifacts have already expired from GitHub Actions retention. As a result, the backfill only recovered usable data for 2026-08-12 and 2026-08-13; combined with the pre-existing 2026-08-11 entry, history currently spans just 3 calendar days. Trend indicators above reflect this limited window and will become statistically meaningful as more daily runs accumulate.

---

Automatically generated by the api-consumption-report workflow.

Generated by 📊 GitHub API Consumption Report Agent · agent · 203.6 AIC · ⌖ 47.2 AIC · ⊞ 12.4K · ◷

• expires on Aug 16, 2026, 3:22 AM UTC-08:00

[github-actions[bot]]

This discussion has been marked as outdated by GitHub API Consumption Report Agent.

A newer discussion is available at Discussion #52697.