MCP Builder
Decision Tree
What are you building?
├─ Tool (action the AI can take) → Define tool with input schema
├─ Resource (data the AI can read) → Define resource with URI
└─ Prompt (reusable template) → Define prompt with arguments
What language?
├─ TypeScript → See references/typescript-guide.md
└─ Python → See references/python-guide.md
Quick Start (TypeScript)
npx @anthropic-ai/create-mcp-server my-server
cd my-server && npm install
import { McpServer } from '@anthropic-ai/mcp-server';
const server = new McpServer({ name: 'my-server', version: '1.0.0' });
server.tool('greet', { name: { type: 'string' } }, async ({ name }) => {
return { content: [{ type: 'text', text: `Hello, ${name}!` }] };
});
server.run();
Quick Start (Python)
pip install mcp
from mcp.server import Server
server = Server("my-server")
@server.tool("greet")
async def greet(name: str) -> str:
return f"Hello, {name}!"
server.run()
Configuration
Register in .claude/mcp.json:
{
"mcpServers": {
"my-server": {
"command": "node",
"args": ["path/to/my-server/dist/index.js"]
}
}
}
Best Practices
- Return structured data (JSON) not just text
- Include error handling with meaningful messages
- Add input validation on all tool parameters
- Keep tools focused (one action per tool)
- Use descriptive names and descriptions (shown to the AI)