Let agents test your code in a real browser
One command scans your unstaged changes or branch diff, generates a test plan, and runs it against a live browser.
Expect works as an MCP skill that hooks directly into popular agent platforms via stdio transport, running entirely on the user's machine with no external dependencies or hosted infrastructure required.
Expect spawns subagents simulating logged-in users and extracts real cookies from the local environment, enabling testing of authenticated flows and personalized features that generic browser tools cannot access.
Expect reads git changes, generates a test plan automatically, and runs it in a real browser using Playwright. This eliminates manual selector maintenance and assertion wiring, letting agents test code changes instantly.
Expect checks for long animation frames, INP, LCP, npm vulnerabilities, CSRF attacks, broken hover states, missing metadata, and dead links—providing comprehensive regression detection that agents can fix and re-verify.
Expect supports CI mode via --ci flag or GitHub Actions workflow setup, running headless with auto-approval and 30-minute timeout enforcement for automated regression testing on every PR.