What is MCP — The Protocol That Changes Everything
Understand MCP (Model Context Protocol), the three primitives, and why it matters for AI integrations.
Quick actions
Try the app, view the code, or explore the components.
By the end of this lesson, you will:
- Understand what MCP is and why it matters
- Learn the three primitives: Tools, Resources, Prompts
- See a complete MCP server in action
- Know when to use each primitive
What is MCP?
MCP (Model Context Protocol) is an open specification from Anthropic for connecting AI models to external capabilities.
Think of it as a universal adapter between AI and your data/services.
The Three Primitives
| Primitive | What It Does | Example |
|---|---|---|
| Tools | Functions the AI can call | get_stock_price(symbol) |
| Resources | Data the AI can access | Portfolio holdings, market data |
| Prompts | Pre-built instructions | "Analyze this portfolio for risk" |
Why MCP Matters
Before MCP, every AI integration was custom. You'd write different code for Claude, ChatGPT, Cursor, etc.
With MCP, you build once and connect everywhere.
ChatGPT Apps are built on MCP. OpenAI extended the protocol with UI components (widgets), but the foundation is the same.
The Financial MCP Server
In this module's video, we build a Financial MCP Server with tools for:
- Getting stock prices
- Analyzing portfolios
- Fetching market news
This demonstrates all three primitives in action.
Try It Yourself
Before writing any code, experience what an MCP server feels like from the user's perspective. Click "Experience the server" to try the Financial MCP Server.