MCPDevelopmentTutorialHands-on

Building Your First MCP Server

Step-by-step guide to creating a Model Context Protocol server from scratch, enabling seamless AI integration with your business systems.

15 min read

NeuroCluster Team

Intermediate Level

Prerequisites: Basic knowledge of Python, REST APIs, and command line tools. Familiarity with AI concepts is helpful but not required.

What is Model Context Protocol (MCP)?

Model Context Protocol (MCP) is an open standard that enables AI models to securely access and interact with external data sources, tools, and services. Think of it as a bridge that allows AI assistants to connect with your business systems, databases, and APIs in a standardized way.

Benefits

• Secure data access
• Standardized integration
• Real-time information
• Scalable architecture

Use Cases

• Database queries
• API integrations
• File system access
• Custom tools

Setting Up Your Development Environment

1. Install Required Dependencies

# Create a new project directory
mkdir my-mcp-server && cd my-mcp-server

# Initialize Python environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install MCP SDK
pip install mcp-sdk fastapi uvicorn

Building Your MCP Server

Step 1: Create the Main Server File

# server.py
from mcp import Server, Tool, Resource
from typing import Dict, Any
import json

class MyMCPServer:
    def __init__(self):
        self.server = Server("my-first-mcp-server")
        self.setup_tools()
        self.setup_resources()
    
    def setup_tools(self):
        @self.server.tool("calculator")
        async def calculator(operation: str, a: float, b: float) -> Dict[str, Any]:
            """Perform basic mathematical operations"""
            try:
                if operation == "add":
                    result = a + b
                elif operation == "subtract":
                    result = a - b
                elif operation == "multiply":
                    result = a * b
                elif operation == "divide":
                    result = a / b if b != 0 else "Error: Division by zero"
                else:
                    return {"error": "Unsupported operation"}
                
                return {
                    "operation": operation,
                    "operands": [a, b],
                    "result": result
                }
            except Exception as e:
                return {"error": str(e)}
    
    def setup_resources(self):
        @self.server.resource("company-info")
        async def get_company_info() -> Dict[str, Any]:
            """Get basic company information"""
            return {
                "name": "My Company",
                "founded": "2020",
                "employees": 50,
                "description": "A growing tech company"
            }
    
    def run(self, host: str = "localhost", port: int = 8000):
        import uvicorn
        uvicorn.run(self.server.app, host=host, port=port)

Step 2: Create the Entry Point

# main.py
from server import MyMCPServer

if __name__ == "__main__":
    server = MyMCPServer()
    print("Starting MCP Server on http://localhost:8000")
    server.run()

Step 3: Run Your Server

python main.py

# Server will start on http://localhost:8000
# API documentation available at http://localhost:8000/docs

Testing Your MCP Server

Testing Tip: Use the automatic API documentation at /docs to test your endpoints interactively.

Test the Calculator Tool

# Using curl to test the calculator
curl -X POST "http://localhost:8000/tools/calculator" \
  -H "Content-Type: application/json" \
  -d '{
    "operation": "add",
    "a": 10,
    "b": 5
  }'

# Expected response:
{
  "operation": "add",
  "operands": [10, 5],
  "result": 15
}

Test the Company Info Resource

# Get company information
curl "http://localhost:8000/resources/company-info"

# Expected response:
{
  "name": "My Company",
  "founded": "2020",
  "employees": 50,
  "description": "A growing tech company"
}

Adding Advanced Features

Adding API Key Authentication

# Add to server.py
from fastapi import HTTPException, Depends
from fastapi.security import HTTPBearer

security = HTTPBearer()

async def verify_api_key(token: str = Depends(security)):
    if token.credentials != "your-secret-api-key":
        raise HTTPException(status_code=401, detail="Invalid API key")
    return token

# Add to your tool decorators
@self.server.tool("calculator", dependencies=[Depends(verify_api_key)])

Next Steps

Congratulations! You've built your first MCP server. Here's what to explore next:

Add More Tools

Create tools for file operations, API calls, or business-specific functions.

Integrate with AI Agents

Connect your server to AI agents for automated business processes.

Monitor & Scale

Add logging, monitoring, and scaling capabilities for production use.

Contents

What are Large Language Models?

Understanding the foundation of AI

AI Agents vs Traditional Automation

When to use each approach

Supernova AI Platform

Deploy MCP servers at scale

Actions

Get Help Building

Loading NeuroCluster...

Europe's AI Infrastructure

# Create a new project directory mkdir my-mcp-server && cd my-mcp-server # Initialize Python environment python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate # Install MCP SDK pip install mcp-sdk fastapi uvicorn

# server.py from mcp import Server, Tool, Resource from typing import Dict, Any import json class MyMCPServer: def __init__(self): self.server = Server("my-first-mcp-server") self.setup_tools() self.setup_resources() def setup_tools(self): @self.server.tool("calculator") async def calculator(operation: str, a: float, b: float) -> Dict[str, Any]: """Perform basic mathematical operations""" try: if operation == "add": result = a + b elif operation == "subtract": result = a - b elif operation == "multiply": result = a * b elif operation == "divide": result = a / b if b != 0 else "Error: Division by zero" else: return {"error": "Unsupported operation"} return { "operation": operation, "operands": [a, b], "result": result } except Exception as e: return {"error": str(e)} def setup_resources(self): @self.server.resource("company-info") async def get_company_info() -> Dict[str, Any]: """Get basic company information""" return { "name": "My Company", "founded": "2020", "employees": 50, "description": "A growing tech company" } def run(self, host: str = "localhost", port: int = 8000): import uvicorn uvicorn.run(self.server.app, host=host, port=port)

# Using curl to test the calculator curl -X POST "http://localhost:8000/tools/calculator" \ -H "Content-Type: application/json" \ -d '{ "operation": "add", "a": 10, "b": 5 }' # Expected response: { "operation": "add", "operands": [10, 5], "result": 15 }

# Get company information curl "http://localhost:8000/resources/company-info" # Expected response: { "name": "My Company", "founded": "2020", "employees": 50, "description": "A growing tech company" }

# Add to server.py from fastapi import HTTPException, Depends from fastapi.security import HTTPBearer security = HTTPBearer() async def verify_api_key(token: str = Depends(security)): if token.credentials != "your-secret-api-key": raise HTTPException(status_code=401, detail="Invalid API key") return token # Add to your tool decorators @self.server.tool("calculator", dependencies=[Depends(verify_api_key)])