Model Context Protocol (MCP)

An open protocol that standardises how AI applications connect to external data sources, tools, and services, enabling models to act on live context rather than static training data.

The Model Context Protocol (MCP) is an open standard introduced by Anthropic in late 2024. It defines a universal interface between AI assistants (clients) and external systems (servers), databases, APIs, file systems, developer tools, and more.

Before MCP, every integration between an AI model and an external tool required custom glue code. MCP replaces these point-to-point connectors with a single protocol, similar to how USB standardised hardware peripherals or how LSP standardised code editor integrations.

Key concepts:

MCP Host: the AI application (e.g. a code editor, chatbot, or agent framework) that initiates connections.
MCP Client: maintains a 1:1 session with an MCP server, handling protocol negotiation.
MCP Server: exposes resources (data), tools (actions), and prompts (templates) to the client.

MCP enables features like reading live database rows, triggering CI pipelines, or browsing documentation, all without the model needing to be retrained.