Flaky tests are one of the most frustrating parts of CI workflows. You wait for your pipeline to finish, ready to merge, and then a test fails that has nothing to do with your changes. Most engineers hit re-run, it passes, and the underlying problem never gets addressed—causing the same friction again and again across your team.
We’ve been working on tooling to help you systematically identify and address flakiness in your test suites, regardless of which CI provider you use.
What’s available now:
- Flaky test detection — Tuist automatically identifies tests that produce inconsistent results
- Slack alerts — Get notified when new flaky tests are detected
- GitHub PR comments — Flaky tests are surfaced directly in your pull requests so you have immediate context
Coming soon:
- Automatic quarantine — Isolate flaky tests so teams can fix them without blocking everyone else’s productivity
You can read more about how it works in the documentation: Flaky Tests Guide
If you’re already using this, we’d love to hear how it’s working for you. Any feedback or feature requests are welcome!
