Back to AI Glossary
Developer ExperienceMCP

Model Context Protocol

An open standard for connecting AI assistants to external data sources and tools through a unified, composable interface.

Why it matters

MCP is becoming the USB-C of AI integrations — a single standard that lets any AI assistant work with any tool. Anthropic open-sourced it, and adoption is growing fast across the ecosystem.

The problem MCP solves

Every AI tool integration used to be a custom, one-off implementation. MCP standardizes how AI assistants discover, authenticate with, and call external tools — the same way HTTP standardized web communication. One protocol, many tools, any AI client.

Architecture

MCP follows a client-server model:

  • MCP Client — the AI assistant (Claude, Cursor, etc.) that wants to use tools.
  • MCP Server — a lightweight service that exposes tools, resources, and prompts via the protocol.
  • Transport — communication happens over stdio (local) or HTTP with Server-Sent Events (remote).

Why it matters for developers

Instead of building N integrations for N AI clients, tool developers build one MCP server. Instead of AI platforms building M integrations for M tools, they implement one MCP client. This N×M → N+M reduction is what made the web work, and MCP aims to do the same for AI tool use.

Related Terms

AI Agent— An AI system that can autonomously plan, reason, and take actions to accomplish goals, often using tools and external APIs.Copilot— An AI assistant embedded directly into a workflow tool (IDE, browser, email) that suggests actions, generates content, or automates tasks inline.

On the AI Radar

Model Context Protocol— Anthropic's open standard for connecting AI models to external tools and data sources. Now governed by the Linux Foundation's Agentic AI Foundation, with support from OpenAI, Google, Microsoft, and AWS.
Back to AI Glossary