Browser Testing & Automation
Automates end-to-end browser testing and website interactions using agentic control for reliable UI validation.
This skill provides a standardized framework for executing browser-based tasks and E2E tests directly within Claude Code. It prioritizes the agent-browser tool over manual script generation, ensuring a consistent environment with forced headed execution for visual verification. By automating server detection, handling authentication through secure environment variables, and mandating screenshot evidence stored in a dedicated results directory, it streamlines the debugging and testing workflow for modern web applications.
Key Features
01Intelligent local server detection and automatic background startup.
02Agent-first browser control using agent-browser for reliable execution.
03Mandatory headed browser mode for real-time visual progress monitoring.
040 GitHub stars
05Automatic screenshot capture in test-results/ for every execution step.
06Secure credential management via .env.claude for automated logins.
Use Cases
01Executing complex web workflows like multi-step authentication and organization switching.
02Capturing visual evidence and screenshots for UI/UX audits and bug reporting.
03Running automated E2E tests for web applications during the development cycle.