MCP MongoDB

import { Server } from "@modelcontextprotocol/sdk/server/index.js"; import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js"; import { CallToolRequestSchema, ReadResourceRequestSchema, McpError, ErrorCode, } from "@modelcontextprotocol/sdk/types.js"; import { MongoClient } from "mongodb"; import ExcelJS from "exceljs"; // 从环境变量获取MongoDB连接URI const mongoUri = process.env.MONGO_URI || "mongodb://localhost:27018/"; const client = await MongoClient.connect(mongoUri); // Create server instance const server = new Server( { name: "mongodb", version: "1.0.0", }, { capabilities: { resources: { "mongodb://{db}/{collection}": { description: "MongoDB集合资源", mimeType: "application/json", }, }, tools: { get_collections: { description: "获取MongoDB数据库集合", inputSchema: { type: "object", properties: { database: { type: "string", description: "数据库名称", }, }, required: ["database"], }, }, multi_collection_query: { description: "多集合联查并返回特定数据", inputSchema: { type: "object", properties: { database: { type: "string" }, collection: { type: "string" }, pipeline: { type: "array", items: { type: "string" }, }, }, required: ["database", "collection"], }, }, export_to_excel: { description: "查询数据并导出为Excel文件", inputSchema: { type: "object", properties: { database: { type: "string" }, collection: { type: "string" }, pipeline: { type: "array", items: { type: "object" }, }, outputPath: { type: "string" }, }, required: ["database", "collection", "outputPath"], }, }, }, }, } ); // <use_mcp_tool> // <server_name>mongodb</server_name> // <tool_name>read_resource</tool_name> // <arguments> // { // "uri": "mongodb://common/players" // } // </arguments> // </use_mcp_tool> server.setRequestHandler(ReadResourceRequestSchema, async (request) => { console.log( "ReadResourceRequestSchema handler called with request:", request ); const uriParts = request.params.uri.split("/"); const dbName = uriParts[3]; const collectionName = uriParts[4]; console.log("Database name:", dbName); console.log("Collection name:", collectionName); const db = client.db(dbName); const collection = db.collection(collectionName); const documents = await collection.find({}).toArray(); console.log(documents); // return { // resources: collections.map((row) => ({ // uri: new URL(`${row.name}/schema`, `mongodb://localhost:27017/`).href, // mimeType: "application/json", // name: `"${row.name}" database schema`, // })), // }; return { contents: [ { uri: request.params.uri, mimeType: "application/json", text: JSON.stringify(documents), }, ], }; }); server.setRequestHandler(CallToolRequestSchema, async (request) => { if (!request.params.arguments) { throw new McpError(ErrorCode.InvalidParams, "Missing arguments in request"); } const database: any = request.params.arguments.database; if (request.params.name === "get_collections") { const collections = await client.db(database).listCollections().toArray(); // return { // content: collections.map((collection) => collection.name), // }; return { content: [ { type: "text", text: JSON.stringify(collections, null, 2), }, ], }; } if (request.params.name === "multi_collection_query") { // <use_mcp_tool> // <server_name>mongodb</server_name> // <tool_name>multi_collection_query</tool_name> // { // "database": "common", // "collections": ["players", "teams", "matches"], // "fields": ["name", "score", "date"], // "match": { // "score": { "$gt": 100 }, // "date": { "$gte": "2025-01-01" } // }, // "sort": { "score": -1 }, // "limit": 10 // } // </use_mcp_tool> // const { database, collections, fields, match } = request.params.arguments; const collection: any = request.params.arguments.collection; const pipeline: any = request.params.arguments.pipeline; const db = client.db(database); try { // 确保pipeline是数组 if (!Array.isArray(pipeline)) { throw new McpError( ErrorCode.InvalidParams, "Pipeline must be an array" ); } // 执行aggregate pipeline const cursor = db.collection(collection).aggregate(pipeline); const results = await cursor.toArray(); return { content: [ { type: "text", text: JSON.stringify(results, null, 2), }, ], }; } catch (error: any) { throw new McpError(ErrorCode.InternalError, error.message); } } if (request.params.name === "export_to_excel") { const collection: any = request.params.arguments.collection; const pipeline: any = request.params.arguments.pipeline; const outputPath: any = request.params.arguments.outputPath; const db = client.db(database); // 执行查询 const cursor = db.collection(collection).aggregate(pipeline); const data = await cursor.toArray(); // 创建Excel文件 const workbook = new ExcelJS.Workbook(); const worksheet = workbook.addWorksheet("Sheet 1"); // 添加表头 if (data.length > 0) { const headers = Object.keys(data[0]); worksheet.addRow(headers); // 添加数据行 data.forEach((item) => { const row = headers.map((header) => item[header]); worksheet.addRow(row); }); } // 保存文件 await workbook.xlsx.writeFile(outputPath); return { content: [ { type: "text", text: `数据已成功导出到 ${outputPath}`, }, ], }; } throw new McpError( ErrorCode.MethodNotFound, `Unknown tool: ${request.params.name}` ); }); async function runServer() { console.log("Server connected"); const transport = new StdioServerTransport(); await server.connect(transport); } runServer().catch(console.error);