Does it support authenticated sessions?

Yes, it supports persistent session management, allowing the agent to log in once and maintain that state across multiple commands or tasks.

How does the AI identify elements on a page?

It uses the agent-browser snapshot command to create a simplified list of interactive elements with unique references (like @e1), allowing for highly reliable AI interaction.

Can I see the browser while it's working?

Yes, by using the --headed flag, you can launch a visible browser window to monitor the automation in real-time for debugging purposes.

Do I need to install external tools?

The skill requires the agent-browser CLI, which can be installed globally via npm (npm install -g agent-browser) followed by 'agent-browser install'.

Agent Browser Automation

Name: Agent Browser Automation
Author: Chachamaru127

byChachamaru127

•

160

浏览器自动化

Automates web browser interactions and data extraction using an AI-optimized snapshot workflow for seamless UI navigation and testing.

关于

Agent Browser is a specialized skill for Claude Code designed to handle autonomous web browsing, UI testing, and complex web data extraction. By utilizing the agent-browser CLI, it generates AI-friendly snapshots of web pages that allow Claude to interact with elements through precise references rather than fragile CSS selectors. It features persistent session management for authenticated workflows, supports both headless and headed modes for debugging, and provides comprehensive tools for clicking, form filling, and screenshotting, making it an essential utility for developers automating repetitive web tasks or verifying frontend deployments.

主要功能

Persistent session management for multi-tab and authenticated workflows
Advanced debugging with headed mode and element highlighting
AI-optimized snapshot workflow for precise element interaction
Full browser control including navigation, form filling, and clicking
160 GitHub stars
Automated screenshot capture and UI verification

使用场景

Automated data scraping and content extraction from complex websites
End-to-end testing of web applications during the development cycle
Verifying UI changes and layout consistency across different page states

关于

主要功能

Persistent session management for multi-tab and authenticated workflows
Advanced debugging with headed mode and element highlighting
AI-optimized snapshot workflow for precise element interaction
Full browser control including navigation, form filling, and clicking
160 GitHub stars
Automated screenshot capture and UI verification

使用场景

Automated data scraping and content extraction from complex websites
End-to-end testing of web applications during the development cycle
Verifying UI changes and layout consistency across different page states