arcade-mcp/libs/arcade-mcp-server/docs/index.md

Arcade MCP

Arcade MCP (Model Context Protocol) enables AI assistants and development tools to interact with your Arcade tools through a standardized protocol. Build, deploy, and integrate your MCP servers seamlessly across different AI platforms.

Quick Links

• Quickstart Guide - Get up and running in minutes
• Walkthrough - Learn by example
• API Reference - MCPApp API documentation

Features

• 🚀 FastAPI-like Interface - Simple, intuitive API with MCPApp
• 🔧 Tool Discovery - Automatic discovery of tools in your project
• 🔌 Multiple Transports - Support for stdio and HTTP/SSE
• 🤖 Multi-Client Support - Works with Claude, Cursor, VS Code, and more
• 📦 Package Integration - Load installed Arcade packages
• 🔐 Built-in Security - Environment-based configuration and secrets
• 🔄 Hot Reload - Development mode with automatic reloading
• 📊 Production Ready - Deploy with Docker, systemd, PM2, or cloud platforms

Getting Started

Installation

We recommend installing the arcade-mcp CLI package, which includes arcade-mcp-server and provides a streamlined development workflow:

uv pip install arcade-mcp

Or install just the server library if you prefer a direct Python approach:

uv pip install arcade-mcp-server

Quick Start: Create a New Server (Recommended)

The fastest way to get started is with the arcade new command, which creates a starter MCP server with example tools:

# Create a new server project
arcade new my_server

# Navigate to the project
cd my_server

# Run the server
arcade mcp

The generated server includes three example tools:

• Simple tool - A basic function to get you started
• Secret-based tool - Shows how to use environment secrets
• OAuth tool - Demonstrates how to use a OAuth tool (requires arcade login)

Manual Setup: Create Your First Tool

If you prefer to create tools manually, you can use the MCPApp interface:

from arcade_mcp_server import MCPApp
from typing import Annotated

app = MCPApp(name="my-tools", version="1.0.0")

@app.tool
def greet(name: Annotated[str, "Name to greet"]) -> str:
    """Greet someone by name."""
    return f"Hello, {name}!"

if __name__ == "__main__":
    app.run()

Running Your Server

Recommended: Use the Arcade CLI

# Run HTTP server (default)
arcade mcp

# Run stdio server (for Claude Desktop, Cursor, etc.)
arcade mcp stdio

# Run with debug logging and hot reload
arcade mcp --debug --reload

Alternative: Direct Python execution

# Run your server.py file directly
python server.py

Configure MCP Clients

Once your server is running, connect it to your favorite AI assistant:

# Configure Claude Desktop (configures for stdio)
arcade configure claude --from-local

# Configure Cursor (configures for http streamable)
arcade configure cursor --from-local

# Configure VS Code (configures for http streamable)
arcade configure vscode --from-local

Client Integration

Connect your MCP server with AI assistants and development tools:

• Claude Desktop - Native MCP support in Claude
• Cursor IDE - Enhanced AI coding with MCP tools
• VS Code - Integrate with Visual Studio Code
• MCP Inspector - Debug and test your tools

Learn More

• Walkthrough - Comprehensive examples and tutorials
• API Reference - Detailed API documentation
• Transport Modes - stdio and HTTP transport details

Community

• GitHub Repository
• Discord Community
• Documentation

License

Arcade MCP server is open source software licensed under the MIT license.