MCP - Model Context Protocol

About MCP

The Model Context Protocol (MCP) is an open standard that enables developers to build secure, two-way connections between their data sources and AI-powered tools. The architecture is straightforward: developers can either expose their data through MCP servers or build AI applications (MCP clients) that connect to these servers.

👉 Learn more

MCP follows a client-server architecture where a host application can connect to multiple servers.

Architecture

src/ → TypeScript source
build/ → Compiled JavaScript

```bash
weather/
├── src/        # TypeScript source
│   └── index.ts
├── build/      # Compiled JavaScript
│   └── index.js
├── tsconfig.json
└── package.json

src/ Directory (TypeScript Source)

Purpose:

• Contains development-time code written in TypeScript
• You implement MCP protocol handlers here (Resources, Tools, Prompts)
• Also contains logic, type definitions, SDKs, etc.

Key Files:

• index.ts → Entry point for your MCP server
• MCP concepts:
  • resources/ → MCP Resource implementations
  • tools/ → LLM-callable functions
  • prompts/ → Prompt templates / management

build/ Directory (Compiled JavaScript)

Purpose:

• Holds runtime code after TS compilation
• This is the folder Claude Desktop interacts with via:
  node build/index.js

Traits:

• Compiled, production-ready JavaScript
• No TS needed after this stage
• Clean deployable output

⚙️ Data Flow Pipeline

graph LR
    A[Developer Writes TS Code] --> B[src/index.ts]
    B --> C[TypeScript Compiler tsc]
    C --> D[build/index.js]
    D --> E[Node.js Runtime]
    E --> F[Claude Desktop IPC]

What Happens After index.js

Once your index.js is ready, it flows like this:

Claude Desktop ↔ stdio ↔ Server Transport ↔ McpServer ↔ Tools (1)