MCP Sse Demo

🚀 Developing an Intelligent Mall Service Assistant Using the MCP Protocol

The Model Context Protocol (MCP) is an open - source protocol designed to enhance the capabilities of AI assistants by providing context and tools. This article demonstrates how to use the MCP protocol to build an intelligent service assistant for mall applications.

🚀 Quick Start

Development

Implementing the Command - Line Client

The core code for the command - line client is as follows:

# Configure the MCP proxy
$ MCPINSPECTOR_API_KEY=your_api_key_here npm run dev

Implementing the Web - End

The core code for the web - end is as follows:

// products-sse.ts
import { Server } from "http";
import { createMcpSseServer } from "@modelcontextprotocol/server";

async function main() {
  const server = new Server();
  await createMcpSseServer(server, "/sse");
  server.listen(8083, () => {
    console.log("Product SSE server running on http://localhost:8083/sse");
  });
}

main().catch(console.error);

// product.ts
import { Product } from "./product.entity";

export interface SearchProductsContext {
  query?: string;
  categories?: string[];
  priceRange?: [number, number];
}

export interface GetProductDetailsContext {
  productId: string;
}

export interface AddToCartContext {
  productId: string;
  quantity: number;
}

Usage

Usage Example of the Command - Line Client

To use it in Cursor, create a .cursor/mcp.json file and add the following content:

{
  "mcpServers": {
    "products-sse": {
      "url": "http://localhost:8083/sse"
    }
  }
}

On the Cursor settings page, you can see this MCP service and start using it.

Usage Example of the Web - End

Debugging

Use the npx @modelcontextprotocol/inspector command to debug the SSE service:

$ npx @modelcontextprotocol/inspector   
Starting MCP inspector...
⚙️ Proxy server listening on port 6277
🔍 MCP Inspector is up and running at http://127.0.0.1:6274 🚀

Open the above address in your browser, select SSE, and configure the address to start testing.

Summary

Key Points

• Context Definition: Three key context interfaces are defined for mall functions.
• Server - Side Implementation: Real - time data services for products are implemented using the MCP protocol.
• Client - Side Integration: Demonstrates how to integrate the MCP service on the command - line and web - end.

Notes

• Ensure that the MCP proxy is correctly configured.
• Enable authentication in the production environment.
• Handle possible network latency issues.

By following the above steps, you can create a powerful AI assistant for your mall application.