Setting Up MCP Server

Complete guide to setting up an AIDP MCP (Model Context Protocol) server.

What is MCP?

The Model Context Protocol (MCP) is an open standard that enables AI assistants to securely connect to external data sources and tools. AIDP provides native MCP support for seamless integration with AI platforms like Claude, ChatGPT, and others.

---

Prerequisites

• Node.js 18+ or Python 3.8+

• AIDP API key from platform.aidp.dev

• Basic understanding of REST APIs

• (Optional) Docker for containerized deployment

---

Installation

Option 1: Using npm (Node.js)

npm install -g @aidp/mcp-server

Option 2: Using pip (Python)

pip install aidp-mcp-server

Option 3: Using Docker

docker pull aidp/mcp-server:latest

---

Quick Start

1. Initialize Configuration

Create a configuration file:

aidp-mcp init

This creates mcp-config.yml:

server:
  port: 3000
  host: 0.0.0.0
  name: 'AIDP MCP Server'

aidp:
  apiKey: ${AIDP_API_KEY}
  environment: production
  baseUrl: https://api.aidp.dev/v1

tools:
  enabled:
    - search_businesses
    - get_business
    - get_business_hours
    - get_business_services
    - get_business_reviews

security:
  cors:
    enabled: true
    origins:
      - https://claude.ai
      - https://chat.openai.com
  rateLimit:
    enabled: true
    maxRequests: 100
    windowMs: 60000

2. Set Environment Variables

export AIDP_API_KEY="your-api-key-here"

Or create a .env file:

AIDP_API_KEY=your-api-key-here
MCP_PORT=3000
MCP_HOST=0.0.0.0

3. Start the Server

aidp-mcp start

Or with custom config:

aidp-mcp start --config ./custom-config.yml

Expected Output:

✓ AIDP MCP Server started
✓ Listening on http://0.0.0.0:3000
✓ Connected to AIDP API
✓ 5 tools registered
✓ Ready to accept connections

---

Configuration Options

Server Configuration

server:
  port: 3000 # Server port
  host: 0.0.0.0 # Bind address
  name: 'My MCP Server' # Server name
  version: '1.0.0' # Server version
  timeout: 30000 # Request timeout (ms)

AIDP Configuration

aidp:
  apiKey: ${AIDP_API_KEY} # API key (use env var)
  environment: production # 'production' or 'sandbox'
  baseUrl: https://api.aidp.dev/v1 # API base URL
  timeout: 30000 # API timeout (ms)
  retries: 3 # Retry attempts

Tool Configuration

tools:
  enabled:
    - search_businesses
    - get_business
    - get_business_hours

  # Custom tool parameters
  search_businesses:
    defaultLimit: 10
    maxLimit: 50

  get_business:
    includeReviews: true
    includeAnalytics: false

Security Configuration

security:
  cors:
    enabled: true
    origins:
      - https://claude.ai
      - https://chat.openai.com
    methods:
      - GET
      - POST
    credentials: true

  rateLimit:
    enabled: true
    maxRequests: 100
    windowMs: 60000

  authentication:
    required: true
    type: bearer

---

Docker Deployment

Using Docker Compose

Create docker-compose.yml:

version: '3.8'

services:
  aidp-mcp:
    image: aidp/mcp-server:latest
    ports:
      - '3000:3000'
    environment:
      - AIDP_API_KEY=${AIDP_API_KEY}
      - MCP_PORT=3000
      - MCP_HOST=0.0.0.0
    volumes:
      - ./mcp-config.yml:/app/config.yml
    restart: unless-stopped
    healthcheck:
      test: ['CMD', 'curl', '-f', 'http://localhost:3000/health']
      interval: 30s
      timeout: 10s
      retries: 3

Start the server:

docker-compose up -d

---

Verification

1. Check Server Health

curl http://localhost:3000/health

Expected Response:

{
  "status": "healthy",
  "version": "1.0.0",
  "uptime": 123.45,
  "aidp": {
    "connected": true,
    "apiVersion": "v1"
  }
}

2. List Available Tools

curl http://localhost:3000/mcp/tools

Expected Response:

{
  "tools": [
    {
      "name": "search_businesses",
      "description": "Search for businesses using natural language",
      "parameters": {...}
    },
    {
      "name": "get_business",
      "description": "Get detailed business information",
      "parameters": {...}
    }
  ]
}

3. Test a Tool

curl -X POST http://localhost:3000/mcp/call \
  -H "Content-Type: application/json" \
  -d '{
    "tool": "search_businesses",
    "parameters": {
      "query": "coffee shops",
      "location": {"lat": 45.5231, "lon": -122.6765}
    }
  }'

---

Production Deployment

Using PM2 (Node.js)

# Install PM2
npm install -g pm2

# Start server with PM2
pm2 start aidp-mcp --name "aidp-mcp-server"

# Save PM2 configuration
pm2 save

# Setup auto-restart on reboot
pm2 startup

Using systemd (Linux)

Create /etc/systemd/system/aidp-mcp.service:

[Unit]
Description=AIDP MCP Server
After=network.target

[Service]
Type=simple
User=aidp
WorkingDirectory=/opt/aidp-mcp
Environment="AIDP_API_KEY=your-api-key"
ExecStart=/usr/local/bin/aidp-mcp start
Restart=always

[Install]
WantedBy=multi-user.target

Enable and start:

sudo systemctl enable aidp-mcp
sudo systemctl start aidp-mcp
sudo systemctl status aidp-mcp

---

Monitoring

Logs

# View logs
aidp-mcp logs

# Follow logs
aidp-mcp logs --follow

# View last 100 lines
aidp-mcp logs --tail 100

Metrics

Access metrics endpoint:

curl http://localhost:3000/metrics

Response:

{
  "uptime": 3600,
  "requests": {
    "total": 1234,
    "successful": 1200,
    "failed": 34
  },
  "tools": {
    "search_businesses": 800,
    "get_business": 400
  },
  "performance": {
    "avgResponseTime": 150,
    "p95ResponseTime": 300
  }
}

---

Troubleshooting

Server Won't Start

Issue: Error: AIDP API key not found

Solution:

export AIDP_API_KEY="your-api-key"
# or
echo "AIDP_API_KEY=your-api-key" > .env

Connection Refused

Issue: ECONNREFUSED when calling tools

Solution: Check if server is running:

curl http://localhost:3000/health

CORS Errors

Issue: Browser shows CORS errors

Solution: Add your domain to CORS origins:

security:
  cors:
    origins:
      - https://your-domain.com

---

Next Steps

• Configure Standard Tools →

• Create Custom Tools →

• Testing Your MCP Server →

---

Support

• 📧 Email: [e-mail]

• 💬 GitHub Discussions: github.com/aidp/platform/discussions

• 📚 Documentation: docs.aidp.dev/mcp