NewsletterBlogGlossary

MCP (Model Context Protocol)

What is MCP? The Model Context Protocol is an open standard for connecting AI models to external tools and data sources.

frameworks
ShareXLinkedIn

MCP (Model Context Protocol) — AI Glossary

MCP (Model Context Protocol) is an open standard developed by Anthropic that defines how AI models connect to external tools, data sources, and services. It provides a universal interface — similar to what USB-C did for hardware — so that any AI application can interact with any compatible tool through a single, standardized protocol.

Why MCP Matters

Before MCP, every AI tool integration required custom code. If you wanted Claude to query a database, read from GitHub, or call an internal API, you had to build a bespoke connector each time. MCP eliminates this fragmentation by defining a shared protocol that any AI client and any tool provider can implement independently.

The practical impact is composability. A single MCP server — say, one that exposes your Postgres database — works with Claude Code, Claude Desktop, and any other MCP-compatible client without modification. This reduces integration effort from days to minutes and lets developers mix and match tools freely. Anthropic's agent capabilities rely heavily on MCP for extending what AI agents can access and control.

How MCP Works

MCP uses a client-server architecture with JSON-RPC as the transport layer. An MCP server wraps an external tool or data source and exposes its capabilities through a standardized schema — listing available functions, their parameters, and return types. An MCP client (embedded in an AI application) discovers these capabilities and invokes them on behalf of the model.

Key components:

  • Tools: Functions the model can call (e.g., query_database, create_issue)
  • Resources: Read-only data the model can access (e.g., file contents, API responses)
  • Prompts: Reusable prompt templates served by the server

Servers can run locally (as a subprocess) or remotely (over HTTP with SSE). The protocol handles capability negotiation, so clients only see tools they're authorized to use.

  • Claude Code: Anthropic's terminal-based coding agent that uses MCP servers to connect to external tools and data sources
  • Claude: Anthropic's family of large language models that power MCP-compatible applications
  • Anthropic: The AI safety company that created MCP and maintains the open specification

Want more AI insights? Subscribe to LoreAI for daily briefings.