GenAI Systems Lab Open interactive version →
AI Engineering 10 min read

Model Context Protocol: What It Is and Why It Matters

Anthropic's open standard for connecting LLMs to external tools and data. What MCP solves that function calling doesn't, the host/client/server architecture, and why every major IDE and AI assistant is adopting it.

In November 2024, Anthropic open-sourced the Model Context Protocol — a standard for connecting LLMs to external tools and data sources. Within six months, it was adopted by VS Code, Cursor, Zed, Replit, Sourcegraph, Block, and dozens more. Here's why it matters.

The problem MCP solves

Every AI assistant that integrates with external tools reinvents the same integration layer: authentication, serialization, error handling, discovery. Without a standard, every LLM application builds N×M integrations — N models × M tools. MCP turns this into N + M: each tool builds one MCP server, each model builds one MCP client.

The architecture: host, client, server

The four MCP primitives

PrimitiveWhat it isExample
ToolsFunctions the LLM can call (model-controlled)search_web(), run_query(), send_email()
ResourcesData the host can read (app-controlled)File contents, database rows, API responses
PromptsReusable prompt templatescode-review, summarize-doc, explain-error
SamplingServer asks the LLM to generate textAgentic servers requesting model inference

Transport: stdio vs. SSE

Local MCP servers communicate over stdio (stdin/stdout) — simple and zero-configuration. Remote servers use Server-Sent Events (SSE) over HTTP for network deployment. The protocol itself is JSON-RPC 2.0 in both cases.

MCP is the USB-C of AI tool integration. Once a tool has an MCP server, it works with any MCP-compatible host — without custom per-application integration code.

Try it interactively

GenAI Systems Lab is a free platform for AI engineers — configure real failure modes, break things, and build the judgment that gets you hired.

Open GenAI Systems Lab →