[api-consumption] 📊 GitHub API Consumption Report — 2026-08-24 #55364
Closed
Replies: 1 comment
|
This discussion has been marked as outdated by GitHub API Consumption Report Agent. A newer discussion is available at Discussion #55749. |
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.
📊 GitHub API Consumption Report
Report Date: 2026-08-24 · Repository: github/gh-aw · Run: #32719116837
Today at a Glance
🔗 GitHub API Calls Trend (90 days)
Only 2 of the 5 tracked days (2026-08-22, 2026-08-23) have real
core_consumedfigures (88,680 and 290,274 calls respectively); 2026-08-06, 2026-08-07, and 2026-08-24 are plotted as explicit gaps rather than interpolated, since the underlying field was null for every run collected on those dates. No reliable multi-day trend can be drawn from two points — this chart exists mainly to make the data gap visible.🔗 GitHub API Calls by Workflow Trend (30 days)
Per-workflow API-call history isn't available (same null-field issue), so this chart shows today's (2026-08-24) top workflows ranked by safe-output writes instead. PR Sous Chef dominates with 15 writes across 9 runs; the next tier (Issue Monster, Contribution Check, PR/Ponytail/Impeccable reviewers) each produced 3-4 writes from a single run, suggesting concentrated but not yet trended behavior — no 30-day pattern can be confirmed with one day of workflow-level data.
🔗 GitHub REST API Calls Heatmap (90 days)
Shown as total runs per day (the closest available volume proxy) across the 5 known dates. 2026-08-06 stands out sharply with 198 runs versus 13, 112, 210, and 60 on the other four days — likely a backfill/batch-testing burst rather than steady-state traffic. The 16-day span between 2026-08-07 and 2026-08-22 has no recoverable data at all.
🍩 Top API Burners (24h)
By safe-output-write share, PR Sous Chef accounts for roughly 29% of today's 52 total writes, with the remaining share spread thinly across nine other workflows (each 1-4 writes). No single workflow beyond PR Sous Chef shows concentration risk today.
🔗 GitHub REST API Consumption by Workflow (last 24h)
Labeled by safe-output writes (not REST API calls, per the data-quality note above). PR Sous Chef leads at 15 writes / 9 runs; Copilot Session Insights and Daily Storify ran the longest (2727s and 1242s respectively) despite only 1 safe-output write each — worth checking whether their duration is proportional to real API work, since we can't confirm that from quota data today.
Top 10 Workflows by REST API Consumption (last 24h)
(ranked by safe-output writes — REST API call counts unavailable this cycle)
Trending Indicators
api_callsdata at all for the most recent 3)core_consumedunavailable (hourly limit: 15,000)📦 Cache Memory Status
/tmp/gh-aw/cache-memory/trending/api-consumption/history.jsonl(write blocked this run by a read-only mount on/tmp/gh-aw; merged history was computed and staged at/tmp/gh-aw/agent/for this report but could not be persisted back to cache-memory — see limitation note below)/tmp/gh-awwas read-only for all write attempts in this sandbox this run (mount table reportsrw, but everyopen()/touchunder it, outside the pre-created/tmp/gh-aw/agentdirectory, returnedEROFS). This means the merged 5-entry trending history could not be persisted back to/tmp/gh-aw/cache-memory/trending/api-consumption/history.jsonlfor next run's continuity — only computed in-memory for this report. Next run will likely restart from the same 2-entry cache unless this is fixed.logstool hit its per-call count limit and required continuation cursors; per the 2-continuation-call cap, collection stopped after covering roughly 08-06 through 08-24, with a 16-day unrecoverable gap (2026-08-08 to 2026-08-21).github_rate_limit_usage.core_consumedwas null for every collected run (271/271) — the primary "REST API Calls" metric this report is meant to track was entirely unavailable this cycle. All figures above use safe-output writes and run counts as substitute signals; none are fabricated.Automatically generated by the api-consumption-report workflow.
All reactions