ci-speed-hunting
The authoritative method for hunting, diagnosing, and fixing CI SLOWNESS so the merge queue drains faster and throughput rises — the target is wall-clock latency, NOT correctness (a job that is GREEN but slow). Encodes the loop — MINE CI timing for the real long poles → DIAGNOSE the bottleneck class (under-sharded matrix, cold cache, missing/too-broad path-filter tier, serial needs-chain, redundant rebuild) → PROVE the fix keeps identical coverage → FIX FORWARD → VERIFY the before/after delta at unchanged coverage → REPEAT. And the hard line — speed comes from doing the same work faster or skipping PROVABLY irrelevant work, NEVER from deleting tests, narrowing greps, dropping shards, or removing a gate. Use whenever chasing CI latency, "CI is too slow", "speed up the checks", "why does the merge queue take so long", re-sharding an E2E matrix, adding or fixing a cache, tuning changed-paths tiering, or running the ci-speed-hunter lane. Distinct from ci-flake-hunting (which owns failures and flakes — speed never touches a red job).
Pinned to revision 6b9d66f636cc, so it is the text this page describes rather than whatever the author pushed since.
Files
Every link opens the file at its source, pinned to the revision this page describes.