# MCP服务开发全指南：语言选择与实现方案 ## 1. MCP协议概述与技术选型 ### 1.1 MCP核心概念 MCP（Model Context Protocol）是由Anthropic推出的开放协议，旨在标准化大语言模型与外部数据源和工具的通信[ref_4]。该协议采用客户端-服务器架构，通过定义标准化的接口，使AI应用能够更顺畅地与各种服务进行交互[ref_2]。 ### 1.2 开发语言选择对比 | 语言 | 适用场景 | 优势 | 参考实现 | |------|----------|------|----------| | **TypeScript** | 前端集成、快速原型 | 生态丰富，与Cursor等工具集成方便 | [ref_6] | | **Python** | 数据密集型、AI集成 | 库支持完善，FastAPI框架成熟 | [ref_2] | | **Kotlin** | 高性能微服务 | 响应式编程，性能优越 | [ref_1] | ## 2. TypeScript实现方案 ### 2.1 项目初始化与配置 ```typescript // package.json 配置 { "name": "mcp-server-typescript", "version": "1.0.0", "type": "module", "dependencies": { "@modelcontextprotocol/sdk": "^1.0.0", "express": "^4.18.0" }, "scripts": { "dev": "ts-node server.ts" } } ``` ### 2.2 核心服务器实现 ```typescript import { Server } from '@modelcontextprotocol/sdk/server/index.js'; import { StdioServerTransport } from '@modelcontextprotocol/sdk/server/stdio.js'; import { ListResourcesRequestSchema, ListToolsRequestSchema, CallToolRequestSchema } from '@modelcontextprotocol/sdk/types.js'; // 创建MCP服务器实例 const server = new Server( { name: 'weather-mcp-server', version: '1.0.0', }, { capabilities: { resources: {}, tools: {}, }, } ); // 实现资源列表方法 server.setRequestHandler(ListResourcesRequestSchema, async () => { return { resources: [ { uri: 'weather://current', mimeType: 'application/json', name: '当前天气', description: '获取当前天气信息' } ] }; }); // 实现工具调用方法 server.setRequestHandler(CallToolRequestSchema, async (request) => { if (request.params.name === 'get_weather') { return { content: [ { type: 'text', text: JSON.stringify({ temperature: 25, condition: '晴朗', city: request.params.arguments?.city || '北京' }) } ] }; } }); ``` ## 3. Python实现方案 ### 3.1 FastAPI框架搭建 ```python # requirements.txt fastapi==0.104.0 uvicorn==0.24.0 fastapi-mcp==1.2.0 pydantic==2.5.0 # main.py from fastapi import FastAPI from fastapi_mcp import MCPServer from pydantic import BaseModel app = FastAPI() mcp_server = MCPServer(app) class WeatherRequest(BaseModel): city: str @mcp_server.tool( name="get_weather", description="获取指定城市的天气信息" ) async def get_weather(city: str) -> dict: """ 天气查询工具实现 Args: city: 城市名称 Returns: 包含天气信息的字典 """ # 模拟天气数据获取 return { "city": city, "temperature": 23, "humidity": 65, "condition": "多云" } @mcp_server.resource( uri="weather://forecast", mime_type="application/json" ) async def get_weather_forecast() -> dict: """获取天气预报资源""" return { "forecast": [ {"day": "今天", "temp": "22-28°C"}, {"day": "明天", "temp": "20-26°C"} ] } if __name__ == "__main__": import uvicorn uvicorn.run(app, host="0.0.0.0", port=8000) ``` ### 3.2 客户端连接配置 ```python # client_connection.py import requests import json class MCPClient: def __init__(self, server_url: str): self.server_url = server_url def call_tool(self, tool_name: str, arguments: dict): """调用MCP服务器工具""" payload = { "method": "tools/call", "params": { "name": tool_name, "arguments": arguments } } response = requests.post( f"{self.server_url}/mcp", json=payload ) return response.json() # 使用示例 client = MCPClient("http://localhost:8000") weather_data = client.call_tool("get_weather", {"city": "上海"}) print(weather_data) ``` ## 4. Kotlin微服务实现 ### 4.1 项目结构与依赖 ```kotlin // build.gradle.kts plugins { kotlin("jvm") version "1.9.0" id("org.springframework.boot") version "3.1.0" } dependencies { implementation("org.springframework.boot:spring-boot-starter-web") implementation("org.jetbrains.kotlinx:kotlinx-coroutines-core:1.7.0") implementation("com.google.protobuf:protobuf-kotlin:3.24.0") implementation("io.ktor:ktor-client-core:2.3.0") } ``` ### 4.2 响应式MCP服务器 ```kotlin // McpServer.kt @SpringBootApplication class McpServerApplication fun main(args: Array<String>) { runApplication<McpServerApplication>(*args) } @RestController @RequestMapping("/mcp") class McpController { @PostMapping("/tools/call") suspend fun callTool(@RequestBody request: ToolCallRequest): ToolCallResponse { return when (request.name) { "weather_query" -> { val city = request.arguments?.get("city") as? String ?: "北京" ToolCallResponse( content = listOf( TextContent( type = "text", text = """{"city": "$city", "temp": 22, "condition": "晴朗"}""" ) ) ) } else -> throw IllegalArgumentException("未知工具: ${request.name}") } } @GetMapping("/resources/{uri}") suspend fun getResource(@PathVariable uri: String): ResourceResponse { return when (uri) { "weather_forecast" -> ResourceResponse( mimeType = "application/json", content = """{"forecast": [{"day": "今天", "temp": "20-25°C"}]}""" ) else -> throw IllegalArgumentException("资源不存在: $uri") } } } // 数据类定义 data class ToolCallRequest( val name: String, val arguments: Map<String, Any>? ) data class ToolCallResponse( val content: List<TextContent> ) data class TextContent( val type: String, val text: String ) ``` ## 5. 开发流程与最佳实践 ### 5.1 开发步骤总结 1. **环境准备**：安装对应语言的开发环境和MCP SDK[ref_5] 2. **项目初始化**：配置依赖和项目结构 3. **服务器实现**：实现工具、资源和提示词三个核心组件[ref_6] 4. **测试验证**：使用MCP客户端工具进行功能测试 5. **部署集成**：配置到目标平台（Cursor、Claude Desktop等）[ref_3] ### 5.2 性能优化建议 - **连接管理**：使用连接池管理客户端连接[ref_2] - **异步处理**：采用非阻塞IO提高并发性能[ref_1] - **缓存策略**：对频繁访问的数据实现缓存机制 - **错误处理**：完善的异常处理和重试机制 ### 5.3 实际应用场景 **天气服务示例**： ```typescript // 完整的天气MCP服务 export class WeatherMCPService { private apiKey: string; constructor(apiKey: string) { this.apiKey = apiKey; } async getCurrentWeather(city: string) { // 实际调用天气API const response = await fetch( `https://api.weatherapi.com/v1/current.json?key=${this.apiKey}&q=${city}` ); return await response.json(); } async getWeatherForecast(city: string, days: number = 3) { // 获取多天天气预报 const response = await fetch( `https://api.weatherapi.com/v1/forecast.json?key=${this.apiKey}&q=${city}&days=${days}` ); return await response.json(); } } ``` 选择开发语言时，应根据团队技术栈、性能需求和集成环境综合考虑。TypeScript适合前端生态集成，Python在AI场景优势明显，Kotlin则适用于高性能微服务场景[ref_1][ref_2][ref_6]。