MCP (Model Context Protocol)

Definition

Model Context Protocol (MCP) is an open standard introduced by Anthropic in 2024 that defines how AI models connect to external tools, data sources, and services through a unified interface. MCP lets an AI agent call database queries, web searches, file systems, and custom APIs using a single protocol instead of bespoke tool integrations for every data source.

Before MCP, connecting an LLM to an external tool required custom code for each integration. MCP standardizes the interface: any MCP-compatible server exposes its capabilities in a uniform way, and any MCP-compatible client (Claude, Cursor, and growing) can discover and use those capabilities automatically.

MCP architecture

MCP server -- wraps a data source or tool (Postgres, Slack, GitHub, custom API) and exposes it via the protocol
MCP client -- the AI model or agent runtime that discovers and calls MCP servers
Tools -- callable functions the server exposes (run_query, search_docs, send_message)
Resources -- data the server can surface as context (file contents, database rows)

Why it matters for enterprise AI

MCP dramatically lowers the integration cost of building AI agents that act on real business data. Instead of custom glue code for each system, one MCP server per data source serves any MCP-compatible agent.

Related terms

LLM (Large Language Model)

A large language model (LLM) is a deep-learning model trained on billions of text tokens to predict and generate human-readable language. LLMs such as GPT-4, Claude, and Gemini power chatbots, document summarization, code generation, and AI workflow automation -- and serve as the reasoning engine inside RAG systems and AI agents.

AI Agent

An AI agent is an LLM-powered system that autonomously plans, selects tools, executes multi-step tasks, and loops until a goal is achieved -- without requiring step-by-step human instruction. AI agents extend a language model''s capability from answering questions to taking actions: writing code, querying APIs, browsing the web, and updating databases.

Agentic AI

Agentic AI refers to AI systems that operate autonomously over extended task sequences -- planning actions, invoking tools, observing results, and re-planning until a goal is complete without step-by-step human guidance. Unlike single-turn chatbots, agentic systems can execute workflows that span minutes or hours, touching multiple APIs, databases, and services.

API-First

API-first is a software design philosophy where every product capability is exposed through a well-documented API before any user interface is built. API-first systems are consumed by web apps, mobile apps, bots, integrations, and AI agents interchangeably -- enabling 3-5x faster partner integrations and making AI automation straightforward because every business action is already a callable endpoint.

← Back to full glossary

Need help implementing this in your business?

Code and Trust translates AI concepts like mcp (model context protocol) into working implementations — starting with a workflow audit that shows exactly where it creates ROI.

Schedule AI Audit →