Back to knowledge base
AI & Agentic SystemsLast updated:

Model Context Protocol (MCP)

An open standard developed by Anthropic that enables AI agents to securely connect to external tools, data sources, and services through a unified interface — the USB-C of the AI agent ecosystem.

MCPAnthropicAI agentstool useinteroperability

Model Context Protocol (MCP)

The Model Context Protocol is an open standard created by Anthropic that defines how AI models connect to external tools, data sources, and services. Think of it as a universal adapter — instead of building custom integrations for every tool an AI agent might need, MCP provides a single, standardised interface.

How It Works

MCP follows a client-server architecture:

  1. MCP Servers expose capabilities (tools, resources, prompts) through a standardised API
  2. MCP Clients (AI agents, IDEs, applications) discover and invoke these capabilities
  3. Transport layer handles communication (stdio, HTTP/SSE, WebSocket)
  4. The protocol handles capability negotiation, authentication, and error handling

Why It Matters

Before MCP, every AI integration was bespoke. Connecting an LLM to a database, API, or file system required custom code for each combination. MCP solves this N×M problem:

  • Any model can connect to any tool through a single protocol
  • Tool developers build one MCP server; it works with every MCP-compatible client
  • Composability — agents can discover and chain tools dynamically
  • Security — standardised permission and authentication models

The Agent Infrastructure Layer

MCP is a critical piece of the agentic commerce infrastructure stack:

  • Discovery — agents can query MCP servers to understand available capabilities
  • Execution — agents invoke tools through a typed, validated interface
  • Chaining — complex workflows combine multiple MCP tools in sequence
  • Ecosystem — thousands of MCP servers already exist for databases, APIs, file systems, and web services

Adoption

  • Anthropic Claude — native MCP support in Claude Desktop and API
  • Windsurf / Cursor — IDE integrations using MCP for code tools
  • OpenAI — announced MCP compatibility in 2025
  • Google — Gemini models support MCP-compatible tool calling
  • Open source — growing ecosystem of community-built MCP servers

The Bigger Picture

MCP is to AI agents what HTTP was to web browsers — a foundational protocol that enables an ecosystem. As agents become autonomous economic participants, MCP provides the standard interface through which they interact with the digital world. Combined with crypto payment rails and decentralised identity, MCP is a key building block of the agentic commerce stack.