What it does
A tool that checks whether an AI agent's work actually holds up. It drives your running web or desktop app, reads what happened and reports pass, fail or could not tell, with the file and line to fix. It is available as npm packages.
Use cases
- 01Check that an agent's change works in the real app
- 02Find the file and line behind a failed check
- 03Verify a web app after an AI-made edit