Automates web browser interactions for testing, data extraction, and workflow automation within Claude Code.
Agent Browser is a comprehensive Claude Code skill designed to bridge the gap between AI reasoning and web-based execution. It enables Claude to navigate websites, interact with complex DOM elements via semantic locators or unique references, capture screenshots, and record sessions for debugging. Whether you are performing end-to-end web testing, scraping dynamic data from JavaScript-heavy sites, or automating repetitive form-filling tasks, this skill provides a robust command-line interface to control browsers with precision, manage multiple sessions, and handle persistent state across workflows.
Key Features
01Advanced navigation control including frame switching, tab management, and history
021 GitHub stars
03High-fidelity media capture via screenshots, PDF generation, and video recording
04Robust state management for session persistence and authenticated login flows
05Interactive DOM snapshots with unique reference-based element targeting (@refs)
06Comprehensive interaction suite for clicking, typing, dragging, and file uploads
Use Cases
01Automated end-to-end (E2E) testing and bug reproduction for web applications
02Extracting structured data from dynamic websites and single-page applications (SPAs)
03Automating multi-step web workflows like account setup, form filing, and report generation
