Digital Lexicon Hub

An open protocol that standardizes how AI applications connect to external tools, data sources and prompt templates through MCP servers.

MCP lets a company expose a system such as a CRM or analytics tool once and reuse that connection across several AI clients, which cuts custom integration work and centralizes decisions on permissions and consent.

An agency connects its project management and analytics tools to an AI assistant through two MCP servers, grants read-only scopes first, and requires user approval before any tool call that creates tasks or changes data.

The protocol cannot enforce security on its own: tool descriptions from untrusted servers, excessive permissions and Prompt Injection in returned data are real risks. Server quality varies, and the specification changes between dated versions.

MCP uses JSON-RPC 2.0 messages between hosts (the AI applications), clients running inside the host, and servers that expose resources (data), prompts (templates) and tools (functions the model can call). Servers can ask users for more input through elicitation. The current specification describes stateless, self-contained requests with per-request capability negotiation and optional extensions. MCP usually wraps an existing Application Programming Interface (API) so an AI Agent can use it.