Agent Browser Automation
Automates web browser interactions for testing, data extraction, and form filling using a simplified CLI-driven workflow.
This skill provides a comprehensive suite of tools for controlling a headless or headed browser directly from Claude. It allows for advanced web navigation, element interaction using interactive refs, page snapshotting for analysis, and structured data extraction. Whether you need to test web applications, automate repetitive form submissions, capture screenshots/PDFs, or record browser sessions for debugging, this skill provides the necessary commands to integrate complex browser-based workflows into your AI-assisted development process.
Key Features
01Interactive DOM snapshotting with reference-based element selection (@e1, @e2)
02Full-suite browser controls including clicks, form filling, and hover actions
03Session and profile management for persistent logins and state handling
04Advanced network interception, request mocking, and cookie management
05Automated screenshot and PDF generation with full-page support
060 GitHub stars
Use Cases
01Automated data extraction and web scraping from dynamic JavaScript sites
02Automating repetitive web-based administrative tasks and form submissions
03End-to-end (E2E) web application testing and visual regression checks