MCP Scaffold Generator

Quickly scaffold a new MCP (Model Context Protocol) server with TypeScript, proper typing, and best-practice patterns.

What is MCP?

The Model Context Protocol allows AI assistants like Claude to securely interact with external tools and data sources. An MCP server exposes tools, resources, and prompts that AI can use.

Usage

Select your server configuration and the scaffold generator creates a ready-to-use project:

1. Server name — Your MCP server identifier
2. Transport — stdio (CLI) or SSE (HTTP)
3. Tools — Define the tools your server will expose
4. Resources — Static or dynamic data resources

API Endpoint

POST /api/tools/mcp-scaffold

Request Body

{
  "name": "my-mcp-server",
  "transport": "stdio",
  "tools": [
    {
      "name": "search_docs",
      "description": "Search documentation by query",
      "parameters": {
        "query": { "type": "string", "description": "Search query" },
        "limit": { "type": "number", "description": "Max results" }
      }
    }
  ],
  "resources": [
    {
      "uri": "docs://overview",
      "name": "Documentation Overview",
      "description": "Full documentation index"
    }
  ]
}

Response

Returns a zip archive containing:

my-mcp-server/
  package.json
  tsconfig.json
  src/
    index.ts          # Server entry point
    tools/
      search_docs.ts  # Tool implementation stub
    resources/
      overview.ts     # Resource handler stub
  README.md

Generated Code Example

The scaffold generates properly typed tool handlers:

server.setRequestHandler(CallToolRequestSchema, async (request) => {
  switch (request.params.name) {
    case 'search_docs': {
      const { query, limit } = request.params.arguments;
      // TODO: implement search logic
      return {
        content: [{ type: 'text', text: JSON.stringify({ results: [] }) }],
      };
    }
    default:
      throw new McpError(ErrorCode.MethodNotFound, 'Unknown tool');
  }
});

Best Practices

• Always validate tool input parameters
• Return structured JSON in tool responses
• Use descriptive tool and parameter names
• Include error handling with meaningful messages
• Add resource URIs for static data the AI might need