What is MCP (Model Context Protocol) and what is it for?

The problem it solves

If every AI application had to connect to every tool separately, N applications and M tools would need N times M integrations. MCP defines a common interface: you write a tool once as an MCP server, and every MCP-capable application can use it.

The parts

• Host: the AI application the user runs (for example a chat app or a code editor).
• Client: the connection inside the host that talks to one MCP server.
• Server: the program that exposes the abilities.

A server can expose three kinds of things: tools (functions the model can call), resources (data the model can read) and prompts (ready-made prompt templates).

How messages travel

Messages are carried as JSON-RPC. Local servers usually connect over stdio, and remote servers over HTTP.

A simple server

The TypeScript example below exposes a tool that adds two numbers. Details can change with the SDK version, so check the current documentation.

import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";
import { z } from "zod";

const server = new McpServer({ name: "demo", version: "1.0.0" });

server.tool("add", { a: z.number(), b: z.number() }, async ({ a, b }) => ({
  content: [{ type: "text", text: String(a + b) }],
}));

await server.connect(new StdioServerTransport());

Security

• Connect only to servers you trust. A server can return text that looks like an instruction to the model.
• Give a server only the permissions it needs and put irreversible actions behind approval.
• Treat tool descriptions and outputs as untrusted input.