Agent Browser Automation
Empowers AI agents to perform end-to-end browser automation, from form filling and data extraction to visual testing and mobile simulation.
Agent Browser is a powerful CLI-driven skill that enables Claude to interact with the web as a specialized browser agent. It provides a structured workflow of navigation, snapshotting, and interaction, allowing for precise control over web elements using an accessibility-first approach. With support for session persistence, encrypted authentication vaults, and visual diffing, it is an essential tool for developers building complex web scrapers, automated testers, or agentic workflows that require navigating authenticated web environments and mobile-responsive layouts.
主な機能
01Secure Auth Vault for encrypted credential management and automated logins
02Visual and DOM diffing to verify the impact of browser interactions
03Interactive element referencing using accessibility tree snapshots (@e1, @e2)
04Session state persistence (cookies/localStorage) across browser restarts
05Mobile web automation support via iOS Simulator integration
0618 GitHub stars
ユースケース
01Automating complex, multi-page data extraction and web scraping tasks
02Managing authenticated workflows across diverse SaaS platforms without manual login
03Performing end-to-end UI testing and visual regression monitoring