Model Context Protocol (MCP)

Model Context Protocol (MCP) is Anthropic's open standard for connecting AI models to external data sources and tools. Instead of building one-off integrations, you build an MCP server once and any MCP-compatible AI client can use it.

How MCP Works

AI Client (Claude Desktop, Cursor, Zed)
    ↕  MCP Protocol (JSON-RPC over stdio or SSE)
MCP Server (your code, or community servers)
    ↕
External System (files, databases, APIs, GitHub...)

Any tool you expose through an MCP server becomes available as a context resource or callable tool inside the AI client — no plugins, no custom API wrappers.

What You Can Connect

• Filesystem — read/write files, search codebases
• Databases — query PostgreSQL, SQLite, run migrations
• GitHub — browse repos, read PRs, create issues
• Slack — read channels, post messages
• Web browsers — Puppeteer automation, web scraping
• Custom APIs — wrap any REST endpoint as a tool

Quick Start — Build Your First Server

import { Server } from "@modelcontextprotocol/sdk/server/index.js";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";

const server = new Server({ name: "my-server", version: "1.0.0" });

server.setRequestHandler("tools/list", async () => ({
  tools: [{
    name: "get_weather",
    description: "Get current weather for a city",
    inputSchema: {
      type: "object",
      properties: { city: { type: "string" } },
      required: ["city"]
    }
  }]
}));

server.setRequestHandler("tools/call", async (req) => {
  const { city } = req.params.arguments;
  // Call your weather API here
  return { content: [{ type: "text", text: `Weather in ${city}: 22°C` }] };
});

const transport = new StdioServerTransport();
await server.connect(transport);

Learning Path

1. Install community servers — filesystem, GitHub, PostgreSQL in Claude Desktop
2. Understand the protocol — Resources vs Tools vs Prompts
3. Build a custom server — TypeScript SDK, expose your first tool
4. Debug with MCP Inspector — visual testing before wiring to a client
5. Deploy — Docker, persistent config, multi-server setup