MCP

An open standard that lets AI apps connect to external tools and data through a single, shared protocol.

MCP (Model Context Protocol) is an open standard for connecting AI models to external tools and data. Think of it as a USB-C port for AI: build one MCP server, and any compatible app (Claude, ChatGPT, Cursor, VS Code) can use it.

How it works

Host: the AI app the user talks to.
Client: the connector inside the host.
Server: a small program that exposes capabilities.

A server offers three things:

Tools: actions the model can call, like create_issue or query_db.
Resources: data the model can read, like files or records.
Prompts: reusable prompt templates.

Servers run locally (over stdio) or remotely (over HTTP), and all messages use JSON-RPC.

Why it matters

Before MCP, every AI app needed a custom integration for every service. Now you write the integration once and it works everywhere.

Related: WebMCP, the browser-side take on the same idea.