Browser Use Mcp Server

Name: Browser Use Mcp Server
Author: kontext-security

web MCP Server

Browse the web, directly from Cursor etc.

bykontext-security

VerifiedFresh

webweb

13 views847 stars117 forksMIT

View on GitHub Homepage Docs

Why This Matters

Discovered via mcp and last synced 6d ago.

VerifiedFresh

Source: mcp
Stars: 847
Last synced: 6d ago
Install: Check source

Install

Install instructions not detected yet

Check the source repository for the latest setup steps.

View source instructions

Tools

Resources

Prompts

Standard I/O

Transport

Who Is This For?

Full-stack developers and AI engineers who want to automate web interactions directly from Claude in Cursor. Product managers and QA engineers benefit from scripting browser workflows without leaving their IDE. Teams building AI agents need seamless web browsing capabilities integrated into their development environment.

Use Cases

Automate web testing and QA workflows

Write test scripts that interact with web applications directly through Claude, automating form submissions, navigation, and validation without separate testing frameworks.

Scrape and process web data in context

Extract structured data from websites and have Claude analyze it immediately, combining web retrieval with AI reasoning in a single workflow.

Build AI agents with browser capabilities

Create autonomous agents that can browse, interact with, and make decisions about web content, enabling complex multi-step web automation tasks.

Debug and inspect live web applications

Navigate through running web apps from your IDE, inspect state, interact with UI elements, and gather information for troubleshooting without switching contexts.

Frequently Asked Questions

How do I connect Browser Use MCP Server to Cursor?

Install the MCP server and configure it in Cursor's settings to enable the browser integration. Once connected, Claude can interact with web pages directly within your workflow without opening a separate browser window.

Can it handle JavaScript-heavy or SPA applications?

Yes, the server can interact with modern JavaScript frameworks and SPAs by executing actions in a real browser context, allowing it to wait for dynamic content and handle client-side rendering.

What's the difference between this and regular web APIs or Selenium?

Unlike APIs, this works with any website regardless of documentation. Unlike Selenium, it integrates directly with Claude's reasoning, letting the AI understand and make decisions about page content contextually.

Are there performance or rate-limiting concerns?

Browser automation is slower than APIs, so it's best for workflows where interactivity is necessary. Check documentation for any rate limits when running repeated automations.

Available Tools (3)

Client

Configuration Path

Cursor

`./.cursor/mcp.json`

Windsurf

`~/.codeium/windsurf/mcp_config.json`

Works Well With - Skills

opencli-usage

Use when running OpenCLI commands to interact with websites (Bilibili, Twitter, Reddit, Xiaohongshu, etc.), desktop apps (Cursor, Notion), or public APIs (HackerNews, arXiv). Covers installation, command reference, and output formats for 87+ adapters.

5841 starsHas guide

ComPDF Conversion CLI

MUST use for ANY PDF or image format conversion task — converting PDF and images (JPG/JPEG/PNG/BMP/TIFF/TIF/WEBP/JPEG2000) to 10 formats (Word, Excel, PPT, H...

100 starsHas guide

Web Search Plus

Search the web and extract URLs with routing across Serper, Brave, Tavily, Querit, Linkup, Exa, Firecrawl, You.com, SearXNG, SerpBase, and Keenable. Returns ranked sources and page text; supports freshness filters and research across providers.

99 starsHas guide

Stock Market Pro

Yahoo Finance (yfinance) powered stock analysis skill: quotes, fundamentals, ASCII trends, high-resolution charts (RSI/MACD/BB/VWAP/ATR), plus optional web a...

94 starsHas guide