What is MCP (Model Context Protocol)?
The Model Context Protocol is an open standard for connecting language models to external capabilities. An MCP server exposes tools, resources and prompts; an MCP client — typically the application hosting the model — discovers and calls them through one consistent interface.
Its value is that it replaces per-integration glue with one shape. Rather than writing bespoke plumbing for every system, you point at an MCP endpoint and get a described set of capabilities.
A protocol describes what is available; it does not decide what should be permitted. An MCP server that exposes a destructive tool exposes it to whoever connects. Authorisation remains the host application’s responsibility.
A remote MCP endpoint can be imported the same way an OpenAPI document is: its actions are discovered, reviewed one by one, and published only if approved.
Related terms
- OpenAPI — A machine-readable specification describing a REST API: its endpoints, parameters, schemas and responses.
- Tool calling — The mechanism by which a language model requests that a specific function be run with specific arguments.
- Integration — An external system an agent can reach, together with the specific actions that were reviewed and approved on it.
See it working rather than described.
The product demo runs one agent across three systems and stops at the approval gate.