OpenAI, Google DeepMind, Microsoft, and major coding IDEs adopted MCP alongside Streamable HTTP transport and OAuth 2.1 authentication, cementing MCP as the de facto interoperability layer for agent tool calling.
Model Context Protocol (MCP)
Also known as: MCP · Tool Calling · Function Calling
Plain English
An open 'USB-C standard' for plugging AI models and agents into external tools, databases, APIs, and local files—so developers write a connector once instead of rebuilding custom integrations for every AI app.
Technical Overview
Open-sourced by Anthropic in late 2024, MCP defines a JSON-RPC 2.0 client-server protocol over stdio or Streamable HTTP exposing three primitives—Tools (callable actions), Resources (readable context), and Prompts (reusable templates). It standardizes tool discovery and execution across IDE agents, desktop clients, and workflow engines like LangGraph & Agent Orchestration.
Chronological Entries (2)
Anthropic open-sourced MCP, replacing fragmented N×M glue code between LLM applications and enterprise data sources (GitHub, Linear, Postgres, Slack) with a single open JSON-RPC protocol.
Direct Relations (3)
- Created byAnthropic
- EnablesLangGraph & Agent Orchestration
- EnablesRAG & Hybrid Retrieval