🗄️ MongoDB MCP Service Gateway for Generative AI

This is an MCP server implementation facilitating direct, natural-language-driven access to MongoDB instances for Artificial Intelligence models. Achieve comprehensive control over collections, schema introspection, and data manipulation without writing direct database queries.

✨ Core Capabilities

🔍 Introspection of collection structures and metadata
📊 Advanced document querying, including complex filtering logic
📈 Management and inspection of database indexes
📝 Full CRUD support: insertion, modification, and logical deletion of records

Demo Visual

https://github.com/user-attachments/assets/2389bf23-a10d-49f9-bca9-2b39a1ebe654

🚀 Initial Setup Guide

To deploy this service, secure your MongoDB connection URI and integrate the following configuration block into your Claude Desktop configuration file:

macOS Path: ~/Library/Application\ Support/Claude/claude_desktop_config.json
Windows Path: %APPDATA%/Claude/claude_desktop_config.json

{ "mcpServers": { "mongodb": { "command": "npx", "args": [ "mongo-mcp", "mongodb://:@:/?authSource=admin" ] } } }

Automated Installation via Smithery

Utilize the Smithery CLI tool for one-step deployment targeting Claude Desktop:

bash npx -y @smithery/cli install mongo-mcp --client claude

System Requirements

Node.js version 18 or newer
npx executable
Docker and Docker Compose (Optional: Required only for running the included local testing environment)
A compatible MCP client application (e.g., Claude Desktop)

Local Sandbox Initialization

For isolated testing without external dependencies, follow these provisioning steps:

Launch the self-contained MongoDB environment:

bash docker-compose up -d

Populate the database with sample payloads:

bash npm run seed

Configuration for Local Development Mode

When running directly from source (e.g., for debugging), use this configuration snippet, pointing to the local runtime:

{ "mcpServers": { "mongodb": { "command": "node", "args": [ "dist/index.js", "mongodb://root:example@localhost:27017/test?authSource=admin" ] } } }

Sample Data Schema Overview

The seeding utility populates three distinct collections to demonstrate capability:

Users Collection

Biographical data (name, electronic mail, chronological age)
Geo-location data nested structure (street address, GPS coordinates)
Enumerated preferences (interests array)
Account registration timestamps

Products Catalog

Merchandise specifics (designation, Stock Keeping Unit, classification)
Detailed technical attributes (nested specifications)
Fiscal data (pricing and current stock levels)
Community feedback metrics (tags and aggregated scores)

Transactions (Orders)

Line item details for purchases
References linking back to purchasing user entities
Logistics and monetary settlement information
Current workflow status tracking

🎯 Illustrative Use Cases

Engage Claude with these queries to test the service's scope:

Foundational Operations

plaintext "Enumerate the collections present in the active database." "Display the structural definition for the 'users' collection." "Retrieve all user entities located within 'San Francisco'."

Complex Retrieval Logic

plaintext "Identify all merchandise items classified as 'electronics' that possess inventory and are priced below $1000." "Fetch every recorded transaction associated with the email address john@example.com." "Provide a list of all products achieving an average user rating exceeding 4.5 stars."

Index Management Directives

plaintext "Report on the indexing structures applied to the users collection." "Establish a new index on the products collection targeting the 'category' attribute." "Output a consolidated list of all defined indexes across every collection."

Record Manipulation Commands

plaintext "Commit a new record into the products catalog named 'Gaming Laptop'." "Modify the processing status for the order identified by ID X to 'dispatched'." "Locate and permanently erase all product entries currently marked as unavailable."