Futurmind Expert Digital Twin MCP Server

A Model Context Protocol (MCP) server that provides access to expert digital twins through standardized interface.

Features

• Expert Access: Connect to expert digital twins via API Key
• Simple Chat: Single message chat interface
• Contextual Chat: Multi-turn conversation with full message history
• Streaming Support: Real-time SSE streaming responses
• Conversation Management: Track conversations with IDs
• Header-based Authentication: Secure API access via X-Api-Key header

Installation

Install from PyPI (Recommended)

pip install futurmind-mcp

Install from Source

# Clone the repository
git clone <repo-url>
cd futur-mind-mcp

# Install with uv
uv sync

# Or with pip
pip install -e .

Configuration

Environment Variables

• FUTURMIND_API_KEY: Your Futurmind API key (required)
• FUTURMIND_API_URL: API base URL (default: https://projectflow-api.futurstaff.com)
• MCP_TRANSPORT: Transport type (stdio or http, default: stdio)
• MCP_PORT: HTTP port (default: 8000)
• MCP_HOST: HTTP host (default: localhost)

Usage

Running with stdio transport (for Cursor)

# Set your API key
export FUTURMIND_API_KEY="your-api-key-here"

# Run the server
python -m futurmind_mcp

Running with HTTP transport

export FUTURMIND_API_KEY="your-api-key-here"
export MCP_TRANSPORT=http
export MCP_PORT=8000

python -m futurmind_mcp

Docker Deployment (Recommended for Remote Access)

For remote HTTP SSE mode, use Docker:

# Quick start
./start.sh

# Or manually
cp env.example .env
# Edit .env and set your FUTURMIND_API_KEY
docker-compose up -d

See DEPLOYMENT.md for detailed Docker deployment guide.

Cursor Configuration

Add this to your Cursor MCP settings:

Option 1: stdio transport (recommended for local use)

After installing the package, add this to your Cursor MCP settings:

{
  "mcpServers": {
    "futurmind": {
      "command": "python",
      "args": ["-m", "futurmind_mcp"]
    }
  }
}

Note: For stdio mode, the API key must be provided via HTTP headers in the client configuration, or you can set FUTURMIND_API_KEY environment variable as fallback.

Option 2: HTTP transport (for remote access)

For HTTP SSE mode, configure the server URL and provide the API key via headers:

{
  "mcpServers": {
    "futurmind": {
      "url": "http://localhost:8000/sse",
      "headers": {
        "X-Api-Key": "your-api-key-here"
      }
    }
  }
}

Note: The API key is passed via the X-Api-Key header for security.

Available Tools

1. chat

Simple single-message chat interface.

Parameters:

• message (string, required): Your message
• conversation_id (string, optional): Conversation UUID for context
• dify_conversation_id (string, optional): Dify conversation ID
• user_id (string, optional): User ID

Example: chat_with_context( messages_json='[{"role":"user","content":"Hello"},{"role":"assistant","content":"Hi!"},{"role":"user","content":"How are you?"}]', conversation_id="uuid-here" )

API Reference

The server forwards requests to:

• Endpoint: POST /v1/zhiyin/chat/stream
• Headers: X-Api-Key: your-api-key
• Request Body: CompletionRequestZhiyin
• Response: SSE stream of CompletionEvent

License

MIT