Does this skill support taking screenshots?

Yes, you can capture standard, full-page, or base64-encoded screenshots to verify the results of actions or for visual data collection.

Does it require an API key to function?

Basic navigation and interaction do not require a key, but advanced features like LLM-based data extraction and autonomous agent tasks require a Browser-use, OpenAI, or Anthropic API key.

Can I use my own Chrome profile with browser-use?

Yes, by using the --browser real flag, the skill can access your existing Chrome profile, including saved login sessions, cookies, and extensions.

Can I run custom Python scripts within the browser session?

Absolutely. The skill includes a persistent Python environment where you can execute code to interact directly with the browser object and manage variables across commands.

How does Claude know which elements to interact with?

The browser-use state command scans the page and returns a list of interactive elements with unique indices, which Claude uses to precisely click, type, or select items.

Web Automation & Browser Control

Name: Web Automation & Browser Control
Author: browser-use

bybrowser-use

•

77,560

•

Browser Automation

Automates browser interactions, web testing, and data extraction through persistent, AI-accessible browser sessions.

The browser-use skill empowers Claude to navigate and interact with the live web by providing a powerful command-line interface for browser automation. It maintains persistent sessions across commands, enabling complex multi-step workflows like filling multi-page forms, performing end-to-end testing, and capturing screenshots. By generating visual indices for web elements, it allows for precise interaction with buttons, inputs, and dropdowns, while supporting various modes including headless Chromium and real user Chrome profiles for authenticated sessions.

Key Features

01AI-powered data extraction and autonomous task execution capabilities

02Visual element indexing for intuitive navigation and interaction

03Integrated Python execution environment for advanced data manipulation and custom logic

0477,560 GitHub stars

05Persistent browser sessions that maintain state across multiple commands

06Support for multiple browser modes including Chromium, real user profiles, and remote cloud instances

Use Cases

01Performing end-to-end testing and visual verification of web applications

02Extracting structured data from dynamic, JavaScript-heavy websites

03Automating repetitive web-based workflows and form submissions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add browser-use/browser-use browser-use

For use in Claude.ai and ChatGPT

Download Skill