Agent Browser Automation

•
1
•

Automates browser-based tasks including navigation, form filling, web scraping, and visual testing via a high-performance CLI for AI agents.

This skill provides a robust browser automation interface for Claude, allowing it to interact with websites directly using Chrome/Chromium via the Chrome DevTools Protocol (CDP). It simplifies complex web interactions by providing a snapshot-based reference system for elements, handling multiple authentication methods (including encrypted vaults and session persistence), and offering advanced features like network interception, PDF/screenshot generation, and seamless iframe handling. It is essential for developers needing to automate workflows, test web applications, or extract data from dynamic sites within their AI-driven development cycle.

주요 기능

01Visual & Network Capture: Generate full-page screenshots, PDFs, and record HAR files for network debugging.
02Advanced Authentication: Securely manage sessions via encrypted vaults, persistent profiles, or state imports.
031 GitHub stars
04Comprehensive Interactions: Support for clicking, typing, scrolling, file downloading, and complex keyboard events.
05Integrated Iframe Support: Automatically inline and interact with content across iframes without manual context switching.
06Interactive Element Snapshots: Reference and interact with page elements using simple alphanumeric tags like @e1.

사용 사례

01Automated E2E Testing: Navigate through complex user flows to verify application behavior and visual consistency.
02Dynamic Web Scraping: Extract data from modern, JavaScript-heavy websites that require user interaction or authentication.
03Workflow Automation: Automate repetitive web-based administrative tasks, form submissions, and data entry.