Automates the testing and debugging of local web applications using Playwright and lifecycle-managed servers.
This skill provides a robust framework for performing end-to-end testing on local web applications within the Claude Code environment. It features a specialized server lifecycle manager that can handle multiple services—such as a separate frontend and backend—simultaneously. By utilizing a reconnaissance-then-action pattern, the skill allows for reliable DOM inspection, screenshot capture, and UI verification using native Python Playwright scripts, ensuring high-fidelity testing without overwhelming the AI's context window.
主な機能
010 GitHub stars
02DOM reconnaissance patterns for reliable element selector identification
03Support for dynamic web apps with network-idle state waiting
04High-resolution screenshot capturing for visual UI debugging
05Headless browser automation via Python Playwright integration
06Automated server lifecycle management for single or multi-service apps
ユースケース
01Debugging complex UI workflows by capturing screenshots and browser logs
02Running regression tests on a local development server before deployment
03Automating multi-step form submissions and data verification in local environments
