Fire in da houseTop Tip:Paying $100+ per month for Perplexity, MidJourney, Runway, ChatGPT and other tools is crazy - get all your AI tools in one site starting at $15 per month with Galaxy AI Fire in da houseCheck it out free

manus-mcp

MCP.Pizza Chef: huyouare

Manus MCP is a versatile Model Context Protocol server that empowers LLMs with real-time web browsing, Google search capabilities, code interpretation, and shell command execution within a sandboxed environment. It integrates tools for browsing websites, performing search queries, running and writing code, and executing bash commands, making it ideal for dynamic, interactive AI workflows requiring external data access and code execution.

Use This MCP server To

Perform Google searches and retrieve relevant links Browse websites and extract content dynamically Execute and interpret code in a secure sandbox Run shell commands within a controlled environment Integrate with Claude for Desktop for enhanced AI workflows Enable LLMs to interact with live web data and code Automate multi-step tasks involving web and code execution

README

Manus MCP

Manus MCP

A Model Context Protocol (MCP) server implementation that can browse the web, perform search queries, and execute code.

Current Features

  • "google_search" tool that performs Google searches and returns relevant links
  • "browse_web" tool that allows browsing websites, clicking elements, and extracting content
  • "code_interpreter" tool that allows reading, writing, and executing code in a sandbox environment
  • "bash_tool" tool that allows running shell commands in the sandbox directory

Using with Claude for Desktop

To use Manus MCP with Claude for Desktop:

  1. Create or edit the Claude for Desktop configuration file:

    • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  2. Add the following configuration:

    {
      "mcpServers": {
        "manus-mcp": {
          "command": "uv",
          "args": [
            "--directory",
            "/ABSOLUTE/PATH/TO/manus-mcp",
            "run",
            "mcp_server.py"
          ]
        }
      }
    }
  3. Restart Claude for Desktop

  4. You should now see the Manus MCP tools available in Claude for Desktop

Available Tools

hello_world

A simple greeting tool that returns a welcome message.

google_search

Performs Google searches and returns a list of relevant links.

browse_web

Interacts with a web browser to navigate websites and extract information. Supported actions:

  • navigate: Go to a specific URL
  • click: Click an element by index
  • input_text: Input text into an element
  • get_content: Get the page content
  • execute_js: Execute JavaScript code
  • scroll: Scroll the page
  • refresh: Refresh the current page

code_interpreter

Allows reading, writing, and executing code files in a sandboxed environment. Supported actions:

  • read: Read the contents of a file
  • write: Write content to a file
  • execute: Execute a file or code snippet
  • list: List files in the sandbox

Supports multiple programming languages including Python, JavaScript (Node.js), Bash, Ruby, Perl, and R.

bash_tool

Executes bash commands in the sandbox directory. Features:

  • Run commands in foreground or background mode
  • Start web servers and other long-running processes
  • Install packages and dependencies
  • Manage files and processes

Environment Variables

The following environment variables can be configured:

  • SANDBOX_DIR: Path to the sandbox directory (default: ~/manus-sandbox)
  • GLOBAL_TIMEOUT: Global timeout for all operations in seconds (default: 60)
  • BROWSER_HEADLESS: Whether to run the browser in headless mode (default: false)
  • GOOGLE_SEARCH_MAX_RESULTS: Maximum number of search results to return (default: 10)
  • LOG_LEVEL: Logging level (default: INFO)

Development Guide

Prerequisites

  • Python 3.11+
  • uv - Fast Python package installer and resolver

Installation

Using Setup Script (Recommended)
./setup.sh
Manual Installation
  1. Clone the repository

    git clone https://github.com/yourusername/manus-mcp.git
    cd manus-mcp
  2. Create a virtual environment and install dependencies

    uv venv
    source .venv/bin/activate
    uv pip install -e .        # Install the project and its dependencies
  3. Run the server

    # Make sure your virtual environment is activated
    source .venv/bin/activate
    ./run.py
    # or
    uvicorn app.main:app --reload
  4. Visit http://localhost:8000/docs to see the API documentation

Development Dependencies

To install development dependencies:

uv pip install -e ".[dev]"

API Documentation

The API follows the Model Context Protocol (MCP) specification.

License

MIT

manus-mcp FAQ

How do I configure Manus MCP with Claude for Desktop?
Edit the Claude for Desktop config file (e.g., ~/Library/Application Support/Claude/claude_desktop_config.json on macOS) to add Manus MCP under mcpServers with the appropriate command and arguments.
What tools does Manus MCP provide?
It includes google_search for web searches, browse_web for website interaction, code_interpreter for running code, and bash_tool for executing shell commands.
Is code execution sandboxed in Manus MCP?
Yes, code and shell commands run in a sandboxed directory to ensure security and isolation.
Can Manus MCP browse any website?
Manus MCP can browse websites by clicking elements and extracting content, but functionality may depend on site complexity and permissions.
What programming languages can Manus MCP execute?
Manus MCP primarily supports code execution in environments configured within its sandbox, typically Python and shell scripts.
How does Manus MCP enhance LLM capabilities?
It extends LLMs by providing real-time access to web data, search results, and code execution, enabling complex, interactive workflows.
Is Manus MCP limited to any specific LLM providers?
No, Manus MCP is provider-agnostic and can be used with various LLMs like Claude, GPT-4, and Gemini.
How secure is Manus MCP when running shell commands?
Shell commands run in a sandboxed directory to minimize risk, but users should still follow best security practices.