AI glossary
Model Context Protocol (MCP)
An open standard for connecting AI applications to external tools and data, so one connector works across many AI apps.
Before MCP, every AI app needed custom code for every service it connected to. MCP defines a common language: an MCP server exposes tools, data (resources) and prompt templates; an AI application (the host) connects to it through an MCP client. Write the server once and it works in any compatible app — Claude, ChatGPT, VS Code, Cursor and many more.
Anthropic released MCP in November 2024 and donated it in December 2025 to the Agentic AI Foundation, part of the Linux Foundation. Learn more in what MCP is.
Example: You connect your task manager’s MCP server to your assistant. Now you can say “create a task to review the budget on Friday” and the assistant creates it, with no copying and pasting.
In practice
- Use connectors from trusted publishers and check what permissions they ask for.
- Keep confirmations on for actions that write or send something.
- The specification moves forward in dated versions; the current one is from 28 July 2026.
To start using it: how to install an MCP server.
