I open-sourced an E2E testing harness for Claude Code that captures screen recordings, traces, HARs, logs, and more
A developer has open-sourced a comprehensive testing harness designed for Anthropic's Claude Code. This tool is capable of capturing detailed session data, including screen recordings, network traces, and logs, to facilitate thorough end-to-end testing. The project aims to provide developers with a robust framework for evaluating and debugging AI coding assistants. AI
IMPACT Provides developers with a tool to better test and debug AI coding assistants.