QuantGeekDev/mongo-mcp
FreeMongoDB 集成，使 LLM 能够直接与数据库交互。
FreeFree tier
About QuantGeekDev/mongo-mcp
A Model Context Protocol (MCP) server that enables Large Language Models (LLMs) to interact directly with MongoDB databases. Users can query collections, inspect schemas, manage indexes, and perform document operations (insert, update, delete) using natural language. The server is easy to set up via npx and integrates with MCP clients like Claude Desktop. It includes a sandbox setup using Docker for local testing and provides a demo video. Available as open-source on GitHub with MIT license.
Key Features
Collection schema inspection
Document querying and filtering
Index management (list, create)
Document operations: insert, update, delete
Natural language interaction via MCP protocol
Easy setup with npx
Docker sandbox for local testing
Demo video available
Pros & Cons
Pros
- Enables non-technical users to interact with MongoDB via natural language
- Quick setup using npx without manual configuration
- Open-source and free to use
- Comes with a demo video and sandbox environment for testing
- Supports full CRUD operations and schema inspection
Cons
- Requires Node.js 18+ and npx installed
- Needs a running MongoDB instance (connection URL required)
- Sandbox testing requires Docker and Docker Compose
- Limited to MCP-compatible clients (e.g., Claude Desktop)
- No built-in authentication or authorization controls
Best For
Exploring available collections and their schemasFinding documents with complex filters (e.g., products in stock under $1000)Managing indexes across collectionsInserting, updating, or deleting documents using natural languageBuilding AI-powered database management assistants
FAQ
How do I install MongoDB MCP Server?
Use npx with your MongoDB connection URL. Add the configuration to your Claude Desktop config file (e.g., `~/Library/Application Support/Claude/claude_desktop_config.json` on macOS).
What tools are available in this server?
Tools include find, insert, update, delete, listCollections, listIndexes, createIndex, and schema inspection.
Do I need a MongoDB server to run this?
Yes, you need a MongoDB connection URL. Optionally, you can run a local sandbox using the provided Docker Compose file and seed script.
Is this tool free?
Yes, it is open-source and free to use. There are no paid tiers.