Skip to main content
Glama
CaptainCrouton89

MCP Server Boilerplate

mongo-find-documents

Query documents from a MongoDB collection using database name, collection name, and optional filters to retrieve specific data records.

Instructions

Query documents from a MongoDB collection

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
databaseYesDatabase name
collectionYesCollection name
filterNoQuery filter as JSON object (optional)
limitNoMaximum number of documents to return (optional)

Implementation Reference

  • The handler function that executes the mongo-find-documents tool: connects to MongoDB database, finds documents using the provided filter and optional limit, formats the output as truncated JSON, and returns it in the MCP response format.
    async ({ database: dbName, collection: collectionName, filter = {}, limit }) => { try { const db = await ensureConnection(dbName); const collection: Collection = db.collection(collectionName); let cursor = collection.find(filter); if (limit) { cursor = cursor.limit(limit); } const documents = await cursor.toArray(); const formattedOutput = formatJsonOutput(documents); return { content: [ { type: "text", text: `Found ${documents.length} document(s):\n\n${formattedOutput}`, }, ], }; } catch (error) { throw new Error(`Failed to find documents: ${error instanceof Error ? error.message : 'Unknown error'}`); } }
  • Zod schema for the tool's input parameters: required database and collection names, optional filter object and limit number.
    { database: z.string().describe("Database name"), collection: z.string().describe("Collection name"), filter: z.record(z.any()).optional().describe("Query filter as JSON object (optional)"), limit: z.number().optional().describe("Maximum number of documents to return (optional)"), },
  • src/index.ts:131-166 (registration)
    Registration of the "mongo-find-documents" tool with the MCP server, specifying name, description, input schema, and inline handler function.
    server.tool( "mongo-find-documents", "Query documents from a MongoDB collection", { database: z.string().describe("Database name"), collection: z.string().describe("Collection name"), filter: z.record(z.any()).optional().describe("Query filter as JSON object (optional)"), limit: z.number().optional().describe("Maximum number of documents to return (optional)"), }, async ({ database: dbName, collection: collectionName, filter = {}, limit }) => { try { const db = await ensureConnection(dbName); const collection: Collection = db.collection(collectionName); let cursor = collection.find(filter); if (limit) { cursor = cursor.limit(limit); } const documents = await cursor.toArray(); const formattedOutput = formatJsonOutput(documents); return { content: [ { type: "text", text: `Found ${documents.length} document(s):\n\n${formattedOutput}`, }, ], }; } catch (error) { throw new Error(`Failed to find documents: ${error instanceof Error ? error.message : 'Unknown error'}`); } } );
  • Helper function to ensure a MongoDB connection and database instance is available, used by the tool handler.
    async function ensureConnection(dbName: string): Promise<Db> { if (!mongoClient) { const uri = getMongoUri(); mongoClient = new MongoClient(uri); await mongoClient.connect(); } if (!databases.has(dbName)) { databases.set(dbName, mongoClient.db(dbName)); } return databases.get(dbName)!; }
  • Helper function to format and clean up truncated JSON output for the tool response, called by the handler.
    function formatJsonOutput(data: unknown): string { const truncatedData = truncateForOutput(data); let outputText = JSON.stringify(truncatedData, null, 2); outputText = outputText.replace( /"\.\.\.(\d+) more items"/g, "...$1 more items" ); outputText = outputText.replace( /"\.\.\.(\d+) more properties": "\.\.\.?"/g, "...$1 more properties" ); return outputText; }

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/CaptainCrouton89/mongo-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server