[api-consumption] π GitHub API Consumption Report β 2026-07-01 #42734
Closed
Replies: 1 comment
-
|
This discussion has been marked as outdated by GitHub API Consumption Report Agent. A newer discussion is available at Discussion #42942. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
π GitHub API Consumption Report
Report Date: 2026-07-01 Β· Repository: github/gh-aw Β· Run: #28514882817
Today at a Glance
π GitHub API Calls Trend (90 days)
Daily REST API consumption is highly volatile because each day's collection window is partial and inconsistent β values swing from <1k to >85k. One clear anomaly stands out: 2026-06-17 at ~574k calls (>2Ο), driven by a 391-run day. Today's 44,316 sits below the 7-day rolling average (~56.6k), but that comparison is skewed by today's truncated ~5.5h window.
π GitHub API Calls by Workflow Trend (30 days)
The heaviest recurring consumers are the review/gate family β Test Quality Sentinel, Matt Pocock Skills Reviewer, Design Decision Gate, and PR Sous Chef β which reappear near the top of the per-workflow ranking day after day. Smoke CI is a consistent baseline consumer given its scheduled cadence. No single workflow dominates the 30-day picture; consumption is spread across the reviewer cohort.
π GitHub REST API Calls Heatmap (90 days)
The heatmap aggregates core quota by weekday Γ ISO week. The mid-June week carries the darkest cell (the 574k 2026-06-17 spike), while most weekday cells cluster in the low-tens-of-thousands. Weekend cells are noticeably lighter, consistent with fewer PR-triggered agentic runs on Saturdays and Sundays.
π© Top API Burners (24h)
Consumption in the collected window is evenly distributed, not concentrated: the top 3 workflows account for just 38.9% of the 44,316 total, and the top 6 each land between ~5,160 and ~5,955 calls. This flat profile means there's no single runaway workflow to target β optimisation gains would come from trimming the whole reviewer cohort rather than one outlier.
π GitHub REST API Consumption by Workflow (last 24h)
The top consumer β Design Decision Gate at 5,955 calls across 6 runs β reaches only 39.7% of the 15,000/hr core limit, and every other workflow sits well below it. No workflow is anywhere near rate-limit risk in this window. The clearest optimisation lever is the cluster of six reviewer/gate workflows each burning ~1,000 calls/run; caching repository metadata or batching GraphQL reads across those runs would compound across the fleet.
Top 10 Workflows by REST API Consumption (last 24h)
Trending Indicators
π¦ Cache Memory Status
/tmp/gh-aw/cache-memory/trending/api-consumption/history.jsonlπ Collection Diagnostics
logs(start_date="-1d")call + 2 retries each hit the 60s bridgecontext deadline exceeded.aw_info.json+run_summary.json); 202 total dirs attempted.run_summary.run.SafeItemsCount(nosafe_output.jsonlpresent); all collected runs reported 0 safe items β dominated by smoke/CI runs.Automatically generated by the api-consumption-report workflow.
Warning
Firewall blocked 1 domain
The following domain was blocked by the firewall during workflow execution:
awmgmcpgSee Network Configuration for more information.
Beta Was this translation helpful? Give feedback.
All reactions