noun · AI protocol
MCP
also called Model Context Protocol
A shared plug shape for AI tools. Build a tool once and any app that has the right socket can use it.
Model Context Protocol: an open standard for connecting AI applications to external tools and data sources through a common client-server interface.
An open protocol with a client in the AI application and a server that exposes tools, data or prompts. The harness asks the server what it offers, lists those tools to the model, and forwards the model's calls to the server.
Real-life example
You want your assistant to read your project tracker. Without a standard you would write custom code for each app you use. With a server that speaks MCP, you point your assistant at it and the tracker's actions appear as tools. A teammate using a different app can point at the same server.
Where the analogy breaks
-
MCP is not a model or an agent. It is only the agreed way for an app to find and call tools.
-
A server is code you are letting your assistant talk to. Only connect servers you trust, because what they return becomes input the model reads.
-
Every tool a server offers costs context. Connect only the servers you need, or the descriptions alone can crowd the window.