Agent Browser Automation

0•

Automates browser interactions, web testing, and data extraction using a powerful CLI designed for AI agents.

Agent Browser provides a comprehensive command-line interface for Claude to interact with the web just like a human user. By leveraging Chrome/Chromium via CDP, it enables complex workflows including element discovery through interactive snapshots, form submission, authenticated sessions, and multi-step automation. This skill is essential for tasks ranging from end-to-end web testing and automated data scraping to navigating complex login flows and capturing visual site audits, all while maintaining session state across interactions.

主要功能

01Visual capture capabilities including full-page, annotated, and high-resolution screenshots or PDFs.
02Interactive snapshots with element reference mapping for precise targeting without complex CSS selectors.
030 GitHub stars
04Robust authentication management including encrypted vaults, session persistence, and profile importing.
05Comprehensive interaction toolkit for clicking, typing, selecting, scrolling, and keyboard events.
06Seamless iframe support with automatic inlining for interacting with elements across complex page structures.

使用场景

01High-fidelity web scraping and data extraction from dynamic, JavaScript-heavy websites that require interaction.
02Automating repetitive browser tasks like bulk form filling, account management, and visual site audits.
03Automated end-to-end testing of web applications by simulating complex user journeys and login flows.