FlakeLab: AI-assisted test automation, measured
FlakeLab runs AI-generated Playwright tests again and again, tracks what breaks, and publishes the numbers. Practical, no hype.
Measured evidence
AI-generated tests are run repeatedly, so you see pass rates and flakiness, not impressions.
Where the AI went wrong
Every FlakeLab article documents the mistakes the model made and how to catch them in review.
Code you can clone
A public repository with CI in green accompanies each post. Run it yourself.