Agent Browser Automation

•
3
•

Automates web browser interactions including navigation, form filling, and data extraction directly within Claude.

This skill empowers Claude to control a headless Chromium instance to perform complex web-based tasks. It provides a comprehensive command-line interface for navigating pages, interacting with DOM elements via unique semantic refs, capturing screenshots, and generating PDFs. With support for session management, network interception, and video recording, it is an essential tool for developers needing to bridge the gap between AI reasoning and live web environments.

主な機能

01Interactive element snapshots with simplified @ref targeting
02Session management with state persistence for cookies and local storage
03Network request interception, routing, and response mocking
04Advanced media capture including screenshots, PDFs, and video recording
053 GitHub stars
06Full suite of interactions including click, type, hover, and drag-and-drop

ユースケース

01Automating repetitive web-based administrative tasks and form submissions
02Automated end-to-end testing and UI verification of web applications
03Extracting structured data from dynamic, JavaScript-heavy websites