mcp-browser-agent

Name: mcp-browser-agent
Author: imprvhub

web MCP Server

A Model Context Protocol (MCP) integration that provides Claude Desktop with autonomous browser automation capabilities. This agent enables Claude to interact with web content, manipulate DOM elements, execute JavaScript, and perform API requests.

byimprvhub

Install Ready

webweb

10 views39 stars12 forksv0.8.0MPL-2.0

View on GitHub

Why This Matters

Discovered via unknown and last synced 3mo ago.

Install Ready

Source: unknown
Stars: 39
Last synced: 3mo ago
Install: Instructions detected

Install

1. Install the package

npx mcp-browser-agent

2. Add to claude_desktop_config.json

{
  "mcpServers": {
    "mcp-browser-agent": {
      "command": "npx",
      "args": [
        "mcp-browser-agent"
      ]
    }
  }
}

Config file location: ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) / %APPDATA%\Claude\claude_desktop_config.json (Windows)

Tools

Resources

Prompts

Standard I/O

Transport

Who Is This For?

Backend and full-stack developers who need to automate web interactions within Claude workflows, and AI engineers building multi-step agents that require browser automation. Teams integrating Claude into existing tools benefit from browser control without managing separate automation infrastructure.

Use Cases

Automated web scraping with AI analysis

Extract structured data from dynamic websites by automating browser interactions, then have Claude analyze and transform the content in a single workflow. Eliminates the need for separate scraping tools and manual data processing.

End-to-end testing powered by Claude

Let Claude navigate web applications, interact with forms, and verify UI behavior while interpreting results. Claude can adapt test logic based on actual page state, making tests more resilient than traditional automation.

Autonomous API integration workflows

Enable Claude to execute JavaScript in browser contexts to interact with APIs, handle authentication flows, and manipulate dynamic content. Useful for tools that require browser-based interaction or JavaScript execution.

Content monitoring and validation

Automate periodic checks of web applications—monitor for changes, validate functionality, or verify deployments. Claude can navigate complex interfaces and intelligently interpret results without hardcoded selectors or brittle assertions.

Frequently Asked Questions

Do I need to manage browser instances or dependencies separately?

No, mcp-browser-agent integrates directly with Claude Desktop as an MCP server. It handles browser lifecycle management, so you focus on defining what Claude should do rather than infrastructure setup.

Can Claude interact with JavaScript-heavy or single-page applications?

Yes, the agent can execute JavaScript and manipulate the DOM, allowing it to work with modern SPAs and dynamically-rendered content that traditional API-only approaches cannot handle.

What happens when Claude encounters authentication or complex navigation?

Claude can handle multi-step interactions, including form filling, cookie management, and API requests. The agent's ability to see and interpret page state lets Claude make context-aware decisions about what to do next.

Is this suitable for production automation or just development/testing?

It works well for both use cases—testing, monitoring, and content automation are production-viable. However, consider rate limiting, error handling, and timeout management based on your workload's reliability requirements.

Available Tools (13)

browser_select

Select dropdown option

api_delete

DELETE request

browser_click

Click element

api_put

PUT request

browser_hover

Hover over element

browser_evaluate

Execute JavaScript

browser_navigate

Navigate to a URL

api_get

GET request

browser_screenshot

Capture screenshot

api_post

POST request

Description

Parameters

browser_fill

Fill form input

api_patch

PATCH request

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

Tavily Web Search

AI-optimized web search via Tavily API. Returns concise, relevant results for AI agents.

381 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

Unified multi-provider web search and URL extraction skill with intelligent auto-routing across Serper, Brave, Tavily, Querit, Linkup, Exa, Firecrawl, Perple...

99 starsHas guide

Works Well With - Agents

Web Use

Web-Use is a CDP powered Browser Agent

290 starslangchain

Customer Service Ai Agent

使用LangGraph搭建多智能体客服系统

49 starslangchain

Run Kit

Web-based tmux orchestration dashboard for long-running AI agent tasks

49 stars