What it does

The Model Context Protocol (MCP) is an open-source standard for connecting AI applications to external systems. It lets an application such as Claude or ChatGPT reach data sources like local files and databases, tools like search engines and calculators, and workflows such as specialised prompts. The documentation likens it to a USB-C port for AI applications: one standard way to connect them. The documentation covers architecture, server and client concepts, versioning, connecting to local and remote servers, building servers and clients, SDKs, security and authorization guidance, and developer tools such as an Inspector and debugging tips. It also describes MCP Apps, which are interactive apps that run inside AI clients. A specification, extensions, a registry and an SEPs section sit alongside the guides. The site lists support across assistants like Claude and ChatGPT and development tools like Visual Studio Code, Cursor and MCPJam, so a server built once can work in many clients. Example uses include an agent reading a calendar and notes app, Claude Code building a web app from a Figma design, and enterprise chatbots querying several databases.

Capabilities

Open-source standard for connecting AI apps to data, tools and workflows
Specification with extensions and a registry
Guides for building MCP servers and clients
SDKs for implementing the protocol
Inspector and debugging tools for testing servers
MCP Apps for interactive experiences inside AI clients

When to use it

You are building an AI app or agent that needs access to outside data and tools
You want to expose your own data or service so any MCP-capable client can use it
You need to learn how MCP servers and clients are structured, secured and tested