Stop Writing Playwright Tests: Let AI Generate Them

Annotation

Writing end-to-end tests is one of those tasks everyone knows they should do, but nobody enjoys doing. What if AI could explore your app, generate Playwright tests, and fix them when they break?

In this talk I'll show how AI agents automate the testing workflow using the latest tooling from the Playwright team, the agent-focused CLI and installable Skills, together with Claude Code. We'll see it in action: exploring an app, generating tests in a real browser, and repairing them when the UI changes.

You'll leave with a practical mental model for AI-driven testing and a clear sense of where it works surprisingly well, and where human judgment is still essential.

Other talks by this speaker

How can we harness the power of collaboration? Eureka! Mobbing, pairing, hackathons

You probably know me best as a developer who focuses on technical topics. Lately, I’ve also had to make sure we meet all our deadlines. We were looking for ways to streamline our work and improve its quality. We started implementing several techniques during the development process, and they’ve yielded positive results. I’d like to convince you that pair programming and team reviews are the right path for you, too.