MCP Hello World Server

A simple hello world MCP (Model Context Protocol) server built with Node.js and TypeScript. This project is configured for easy deployment on Railway.

Features

• ✅ Simple hello world tool endpoint
• ✅ TypeScript for type safety
• ✅ Docker containerization
• ✅ Railway deployment ready
• ✅ MCP SDK integration

Prerequisites

• Node.js 18.0.0 or higher
• npm 9.0.0 or higher

Setup

1. Install dependencies:

   npm install

2. Build the project:

   npm run build

Development

Stdio Mode (Default)

Start the server in stdio mode:

npm run dev

or with the compiled build:

npm start

The server will start and listen on stdio for incoming MCP protocol messages.

HTTP Mode

Start the server in HTTP mode by setting the PORT environment variable:

PORT=8080 npm start

The server will start on the specified port and expose:

• POST /mcp - MCP protocol endpoint
• GET /health - Health check endpoint

Example health check:

curl http://localhost:8080/health
# Response: {"status":"ok"}

Building for Production

Build the TypeScript code:

npm run build

Run the compiled server:

npm start

Project Structure

.
├── src/
│   └── index.ts          # Main server implementation
├── dist/                 # Compiled JavaScript (generated)
├── package.json          # Project dependencies
├── tsconfig.json         # TypeScript configuration
├── Dockerfile            # Docker configuration
├── railway.toml          # Railway deployment config
└── README.md             # This file

MCP Server Tools

hello

A simple tool that returns a personalized greeting.

Parameters:

• name (string, optional): The name to greet. Defaults to "World"

Example:

{
  "name": "hello",
  "arguments": {
    "name": "Alice"
  }
}

Response:

{
  "content": [
    {
      "type": "text",
      "text": "Hello, Alice! Welcome to the MCP server."
    }
  ]
}

Deployment on Railway

Prerequisites

• Railway account (sign up at https://railway.app)
• Railway CLI installed locally (optional, but recommended)
• GitHub repository (optional, for easier deployment)

How It Works

This MCP server can run in two modes:

1. Stdio Mode - Default, for local development or direct integration
2. HTTP Mode - For cloud deployment (Railway, Render, etc.)

When a PORT environment variable is set, the server automatically switches to HTTP mode and exposes:

• POST /mcp - MCP protocol endpoint for clients
• GET /health - Health check endpoint for load balancers

Deploy Using GitHub

1. Push your code to a GitHub repository
2. Go to railway.app
3. Click "Create New Project"
4. Select "GitHub Repo"
5. Authorize Railway and select your repository
6. Railway will automatically detect the Node.js project and set PORT for you
7. Click "Deploy"

Railway will automatically:

• Set the PORT environment variable
• Build your TypeScript code
• Start the server in HTTP mode

Deploy Using Railway CLI

1. Install Railway CLI: https://docs.railway.app/cli/quick-start
2. Login to Railway:

   railway login

3. Create a new project in the current directory:

   railway init

4. Deploy:

   railway up

Using the Deployed Server

Once deployed to Railway, you'll get a public URL like https://your-service.railway.app.

To use it with an MCP client:

// Example: POST request to the MCP endpoint
const response = await fetch('https://your-service.railway.app/mcp', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    jsonrpc: '2.0',
    id: 1,
    method: 'initialize',
    params: {
      protocolVersion: '2024-11-05',
      capabilities: {},
      clientInfo: { name: 'my-client', version: '1.0.0' }
    }
  })
});

const data = await response.json();
console.log(data);

Environment Variables

The server automatically detects its running environment:

• With PORT set - Runs in HTTP mode (Railway, Render, etc.)
• Without PORT - Runs in stdio mode (local development, Docker exec)

No additional configuration needed!

Testing

To test the MCP server, you'll need an MCP client. The server communicates via:

• Stdio - Direct pipe communication for local integration
• HTTP Streamable - HTTP endpoint for cloud deployment

Local Testing with Claude Desktop

Configure Claude Desktop to use the local server in stdio mode:

Create or edit %APPDATA%\Claude\claude_desktop_config.json:

{
  "mcpServers": {
    "hello-world": {
      "command": "node",
      "args": ["C:\\path\\to\\mcp\\dist\\index.js"]
    }
  }
}

Then restart Claude Desktop and the hello tool will be available.

Cloud Testing with Railway

Once deployed to Railway, you can make HTTP requests to your server:

# Health check
curl https://your-service.railway.app/health

# MCP protocol request
curl -X POST https://your-service.railway.app/mcp \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 1,
    "method": "tools/call",
    "params": {
      "name": "hello",
      "arguments": {"name": "Alice"}
    }
  }'

Troubleshooting

Port Already in Use

If you see an error about the port being in use, the server may still be running. Kill it with:

# On Windows
taskkill /F /IM node.exe

# On macOS/Linux
pkill -f "node"

Module Not Found Errors

Make sure all dependencies are installed:

npm install
npm run build

TypeScript Errors

Run type checking:

npm run typecheck

License

MIT