create-commands-mcp

🚀 Scaffolding tool for creating MCP servers that connect to Commands.com

Create production-ready MCP (Model Context Protocol) servers with Commands.com JWT authentication in minutes.

Quick Start

# Create a new MCP server
npx create-commands-mcp my-server

# Navigate and start development
cd my-server
npm install
cp .env.example .env
npm run dev

Features

• ✨ Zero-config setup - Working MCP server in under 5 minutes
• 🔐 Commands.com JWT authentication - Built-in gateway integration
• 🌊 Server-Sent Events (SSE) - Real-time streaming support for gateway compatibility
• 🛠️ Multiple templates - Basic, API integration, and data processing
• 📦 Zero production dependencies - Lightweight and secure
• 🔧 Developer experience - TypeScript, testing, and health checks
• 🚀 Deployment ready - Railway, Vercel, and Docker configs included
• 💰 85% revenue share - Commands.com handles marketing, billing, and auth for only 15%

Usage

Interactive Mode

npx create-commands-mcp

Direct Mode

npx create-commands-mcp my-server-name

With Options

# API template with Railway deployment config
npx create-commands-mcp my-server --template=api --deploy=railway

# Basic template with Docker setup
npx create-commands-mcp my-server --template=basic --deploy=docker

Templates

Basic Template (Default)

• Tools: ping, echo, datetime
• Best for: Learning MCP protocol, simple utilities
• Zero external dependencies

API Template

• Tools: HTTP requests, data transformation, caching
• Best for: External API integration, data fetching
• Includes offline fallbacks

Data Template

• Tools: File processing, JSON/CSV parsing, validation
• Best for: Data transformation, file manipulation
• Built-in data validation

Generated Project Structure

my-mcp-server/
├── src/
│   ├── index.ts              # Main server
│   ├── tools/                # Tool implementations
│   │   ├── ping.ts
│   │   ├── echo.ts
│   │   └── datetime.ts
│   ├── auth/
│   │   └── verifyToken.ts    # JWT validation
│   └── types.ts              # TypeScript definitions
├── tests/
│   └── ping.test.ts          # Basic tests
├── .env.example              # Environment template
├── commands.yaml             # Commands.com metadata
├── README.md                 # Getting started guide
└── package.json              # Minimal dependencies

Commands.com Integration

Instant JWT Validation

Your server works immediately with Commands.com JWTs:

npm run dev     # Start development server
npm run test    # Run tests
npm run doctor  # Health checks

Gateway Integration (Optional)

For Commands.com marketplace and routing:

1. Deploy: Host your server anywhere (Railway/Vercel/AWS/etc.)
2. Register: Submit your server URL at Commands.com Creator Portal
3. Route: Commands.com gateway routes user requests to your server

# Deploy to your preferred platform
npm run build && npm start  # or deploy to Railway/Vercel/etc.

# Register your live server URL at Commands.com
# Example: https://my-server.railway.app

Quick Deploy to Railway

Test your MCP server in production in under 2 minutes:

# 1. Create with Railway config
npx create-commands-mcp my-server --deploy=railway

# 2. Initialize git and push to GitHub
cd my-server
git init && git add . && git commit -m "Initial MCP server"
git remote add origin https://github.com/yourusername/my-server.git
git push -u origin main

# 3. Deploy to Railway
# - Visit railway.app and connect your GitHub repo
# - Railway auto-detects Node.js and deploys instantly
# - Get live URL: https://my-server-production.up.railway.app

Result: Your MCP server is live and ready for Commands.com submission!

Development Workflow

1. Create: npx create-commands-mcp my-server
2. Develop: Add tools in src/tools/
3. Test: npm run test and npm run doctor
4. Deploy: Use Railway for instant testing
5. Publish: Submit to Commands.com marketplace

Adding Custom Tools

// src/tools/myTool.ts
export const myTool = {
  name: "my_custom_tool",
  description: "Does something useful",
  inputSchema: {
    type: "object",
    properties: {
      input: { type: "string", description: "User input" }
    },
    required: ["input"]
  },
  handler: async (args: { input: string }) => {
    return { result: `Processed: ${args.input}` };
  }
};

Deployment Options

Use --deploy <platform> to include deployment configuration files:

Railway (--deploy=railway) - Recommended for Testing

• Includes railway.json config
• One-click deployment via GitHub
• Automatic HTTPS and monitoring
• Free $5/month credit perfect for testing

Vercel (--deploy=vercel)

• Includes vercel.json config
• Serverless deployment via GitHub
• Edge functions and global CDN

Docker (--deploy=docker)

• Includes Dockerfile and docker-compose.yml
• Self-hosted deployment
• Full control over infrastructure

CLI Options

npx create-commands-mcp [name] [options]

Options:
  -t, --template <type>    Template type (basic|api|data)
  -l, --lang <language>    Language (typescript|javascript)  
  -d, --deploy <platform>  Deployment config (railway|vercel|docker)
  -h, --help              Show help

Requirements

• Node.js 18+
• Commands.com account (optional, for gateway integration)

Revenue & Monetization

Commands.com provides a complete business platform for just 15% revenue share:

What's Included:

• 🎯 Marketplace & Marketing - Discovery, promotion, and user acquisition
• 💳 Stripe Integration - Payment processing and subscription management
• 🔐 OAuth & Authentication - User management and secure access
• 📊 Analytics & Insights - Usage tracking and performance metrics
• 🌐 Gateway & Routing - Connects users to your self-hosted servers

Developer Benefits:

• ✅ You keep 85% of all revenue generated
• ✅ Focus on building - No need to handle payments, marketing, or auth
• ✅ Own your infrastructure - Deploy anywhere (Railway, Vercel, AWS, etc.)
• ✅ Lower than competitors - Most platforms take 20-30%

Build tools, not businesses. Commands.com handles everything else.

Server-Sent Events (SSE) Support

The generated MCP servers include built-in SSE support for preventing gateway timeouts on long-running operations.

When to Use SSE

Use SSE for any operation that takes more than 5 seconds to prevent gateway timeouts. Here's a working example:

case 'long_operation':
  const { steps = 5, delay_ms = 1000 } = toolArgs || {};

  if (acceptsSSE && method === 'tools/call') {
    // Set SSE headers
    res.setHeader('Content-Type', 'text/event-stream');
    res.setHeader('Cache-Control', 'no-cache');
    res.setHeader('Connection', 'keep-alive');

    // Process with periodic updates
    (async () => {
      const messages: string[] = [];

      for (let i = 1; i <= steps; i++) {
        // Do work
        await processStep(i);
        
        // Add progress
        messages.push(`Step ${i}/${steps}: ${i === steps ? 'Complete!' : 'Processing...'}`);

        // Send cumulative response
        const response = {
          jsonrpc: '2.0',
          result: {
            content: [{
              type: 'text',
              text: messages.join('\n')
            }]
          },
          id
        };
        res.write(`data: ${JSON.stringify(response)}\n\n`);
      }
      res.end();
    })();
    return;
  }
  
  // Fallback for non-SSE clients
  return standardResponse();

Key Points:

• Send updates every 1-5 seconds to prevent timeouts
• Each update contains the complete response (cumulative)
• Always provide a non-SSE fallback
• Gateway compatibility - Works with Commands.com and other proxies

Supported methods:

• tools/list, tools/call - Tool discovery and execution
• resources/list, resources/read - Resource management (when implemented)
• prompts/list, prompts/get - Prompt handling (when implemented)

The gateway handles all SSE protocol details including event IDs, heartbeats, and connection management.

Support

• 📖 Documentation
• 💬 Discord Community
• 🐛 Report Issues

License

MIT - see LICENSE file for details.