A server advertises capabilities and structured inputs, while a client discovers and invokes them on behalf of a model or user. The protocol separates integration logic from a particular model, but hosts still control trust and permissions.
Model Context Protocol is a protocol for connecting AI applications to external tools, prompts, and contextual resources through a consistent interface.
A server advertises capabilities and structured inputs, while a client discovers and invokes them on behalf of a model or user. The protocol separates integration logic from a particular model, but hosts still control trust and permissions.