home / mcp / mcp mongodb server

MCP MongoDB Server

A Model Context Protocol Server for MongoDB

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "kiliczsh-mcp-mongo-server": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-mongo-server",
        "mongodb://localhost:27017/database"
      ],
      "env": {
        "MCP_MONGODB_URI": "mongodb://localhost:27017/database",
        "MCP_MONGODB_READONLY": "true"
      }
    }
  }
}

You can run an MCP MongoDB Server to let large language models interact with your MongoDB data. It can inspect collection schemas, execute MongoDB operations, and offer read-only or full write capabilities through a standardized interface, making it easier to explore, query, and manage your data from LLM-powered tools.

How to use

Start the MCP MongoDB Server by providing the MongoDB connection URI to the server. You can run in normal mode or in read-only mode to prevent write operations. Use the server to perform schema inference, queries, aggregations, inserts, updates, and index creation through a consistent interface.

How to install

Prerequisites: ensure Node.js and npm are installed on your system. You should have network access to reach your MongoDB instance.

# Start the MCP MongoDB Server against a local MongoDB instance
npx -y mcp-mongo-server mongodb://localhost:27017/database

Additional notes

Environment variables supported by the server let you configure the connection behavior. You can enable read-only mode to protect data and rely on a read preference to read from secondaries if available.

If you want to run a read-only session, use the same start command with the read-only flag.

Available tools

smart_objectid_handling

Configurable auto/none/force modes for string-to-ObjectId conversion during operations.

read_only_mode

Protection against write operations with read-preference optimization.

schema_inference

Automatic detection of collection schemas from document samples.

query_and_aggregation

Full MongoDB query and aggregation pipeline support with optional explain plans.

write_operations

Support for insert, update, and index creation when not in read-only mode.

collection_completions

Auto-complete collection names to assist LLM integration.