What is the agent-browser skill for Claude Code?

It is a specialized capability that allows Claude to control a web browser, perform actions like clicking and typing, and extract information from web pages programmatically.

How does Claude identify elements on a page?

The skill uses the snapshot command to generate an accessibility tree where interactive elements are assigned unique reference IDs like @e1 or @e2, making it easy for Claude to target specific elements.

Can I use this skill for authenticated websites?

Yes, the skill supports saving and loading browser states, including cookies and local storage, allowing you to maintain authenticated sessions across different runs.

Does it support headless and headed modes?

By default, the browser operates in headless mode for efficiency, but you can use the --headed flag to launch a visible browser window for debugging purposes.

Browser Automation Agent

Name: Browser Automation Agent
Author: bmad-labs

bybmad-labs

•

브라우저 자동화

Automates web browser interactions, data extraction, and UI testing directly through Claude.

The Agent Browser skill empowers Claude to navigate websites, interact with web elements, and extract data using a robust set of automated commands. It provides developers with a powerful tool for web scraping, automated end-to-end testing, and form filling by leveraging a smart ref-based system for precise element interaction. Whether you are capturing full-page screenshots, recording debug videos, or managing complex authenticated sessions with persistent storage, this skill bridges the gap between AI and the live web for seamless browser-based workflows.

주요 기능

011 GitHub stars

02Advanced page state management with cookie, localStorage, and session persistence

03Interactive element snapshots with unique reference IDs for precise control

04Comprehensive web interaction including clicking, typing, dragging, and hovering

05Media capture capabilities for full-page screenshots, PDFs, and video recordings

06Network layer control for request interception, mocking, and traffic analysis

사용 사례

01Automated form filling and multi-step workflow execution for productivity

02Automated end-to-end testing and UI verification of web applications

03Complex web scraping and data extraction from dynamic, JavaScript-heavy sites

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add bmad-labs/skills agent-browser

For use in Claude.ai and ChatGPT

Download Skill