rorystandley/example-mcp-server

MCP Demo Server

This is a minimal MCP-compatible demo server that proxies calls to our Azure API Management (APIM) endpoints — for example, Allocate APIs.

Modes

This server supports three modes:

1. Local stdio mode - For VS Code Copilot integration
2. HTTP API mode - Simple REST API for any HTTP client
3. SSE MCP mode - For ChatGPT and other remote MCP clients

Live Demo

You can deploy instantly to Render:

Using with ChatGPT

Once deployed to Render, you can connect ChatGPT to your MCP server:

1. Go to ChatGPT Settings → Personalization → Custom Instructions (or MCP Servers if available)

2. Add a new MCP server with:

   • Name: Allocate Grades API
   • URL: https://example-mcp-server.onrender.com/sse
   • Type: SSE

3. ChatGPT can now call your listGrades tool!

HTTP API Endpoints

Once deployed:

• GET / → Server info and status
• GET /sse → MCP SSE endpoint (for ChatGPT and MCP clients)
• POST /message → SSE message endpoint
• GET /mcp/tools → List available MCP tools (simple API)
• POST /mcp/call → Call a tool (simple API)

Example: Simple HTTP API

# List all grades (minimal format)
curl -X POST https://example-mcp-server.onrender.com/mcp/call \
     -H "Content-Type: application/json" \
     -d '{"tool":"listGrades"}'

# List all grades (full format)
curl -X POST https://example-mcp-server.onrender.com/mcp/call \
     -H "Content-Type: application/json" \
     -d '{"tool":"listGrades","arguments":{"full":true}}'

Local Development

For VS Code Copilot integration, configure .vscode/mcp.json:

{
  "servers": {
    "example-mcp-server": {
      "type": "stdio",
      "command": "node",
      "args": ["./server.js"],
      "env": {
        "ALLOCATE_API_URL": "https://apim-acacium-dev.azure-api.net",
        "APIM_SUBSCRIPTION_KEY": "your-key-here"
      }
    }
  }
}

Then reload VS Code window to activate the MCP server.