Keep blocking merge gates minimal and follow flakes with a rerun
Gate merges on the checks that guard correctness and handle a flake with a follow-up run.
How OpenAI used this technique
Read what OpenAI changed and follow the original sources.
Keep blocking merge gates minimal and follow flakes with a rerun
OpenAI's agent-first team runs its repository with minimal blocking merge gates and short-lived pull requests. A test flake gets a follow-up run, so progress continues while the flake is handled. The team reasons that when agent throughput far exceeds human attention, corrections are cheap and waiting is expensive.
In a system where agent throughput far exceeds human attention, corrections are cheap, and waiting is expensive. Source
- Use it when
- Required checks hold merges on results that a follow-up run clears.
- Before applying it
- The team describes this as a trade that fits a high-throughput repository where corrections are cheap. Keep required the checks that guard changes a follow-up commit cannot undo.
Company Playbooks
- Read Anthropic’s playbookRun only affected tests
- Read DoorDash’s playbookRun build and test in parallel, sharded across CI agents
- Read Linear’s playbookBenchmark the same job on faster hardware
- Read Mastra’s playbookStarSling runners for the slowest test workflow
- Read OpenAI’s playbookRank CI fixes by the minutes they save
- Read Ramp’s playbookMove a CPU-bound test suite to a faster CPU
- Read Shopify’s playbookLet an agent iterate against one CI metric
- Read Spotify’s playbookBuild arm64 on native arm64 runners
- Read Stripe’s playbookRun only the tests a change can reach
- Read Uber’s playbookBuild the merge-queue speculations most likely to be needed