Home/AI Glossary/Model Context Protocol (MCP)

Model Context Protocol (MCP)

Model Context Protocol (MCP) is an open protocol that defines how AI clients and external tool servers communicate. It standardizes how an LLM client lists tools (read files, run SQL, POST to an API) and exchanges context, reducing one-off custom integrations in each product.

It is implemented in tools like Cursor, Claude desktop environments, and experimental agents. Compared with older function calling, MCP is a broader protocol layer and is closely connected to AI agents, API, and webhook.

Key characteristics

  • Provides a standardized way for models to communicate with external tools, resources, and data sources.
  • Reduces the need for custom-built integrations between each assistant and each system.
  • Is especially relevant for agent workflows with high tool-access, safety, and control requirements.