MongTap

A small self-contained MongoDB MCP that serves data using DataFlood ML models

javascript

4

GitHub Stars

javascript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "smallmindsco-mongtap": {
      "command": "node",
      "args": [
        "/absolute/path/to/MongTap/src/mcp/index.js"
      ],
      "env": {
        "NODE_ENV": "production",
        "LOG_LEVEL": "info"
      }
    }
  }
}

MongTap is an MCP server that delivers MongoDB-compatible database functionality through statistical modeling. You interact with it via MCP-enabled clients to create, query, and manage virtual databases generated on-the-fly, without storing actual data. This enables fast prototyping, testing, and demonstrations with realistic data patterns while keeping storage overhead minimal.

How to use

You connect to MongTap through an MCP client or environment that speaks the MCP protocol. Use commands and tools to define data models, start a local MongoDB‑wire protocol server, generate documents from your models, and control generation with seeds and entropy. You work with models and servers by issuing high-level actions like creating data models from descriptions or samples, starting and stopping MongoDB‑compatible servers, querying models to generate documents, training models with new data, and inspecting model schemas and samples.

Typical usage patterns include: creating a statistical model from a description or samples, starting a local server that exposes a MongoDB wire protocol interface, querying the model to generate documents, and adjusting generation with seed and entropy to achieve reproducible or varied results. You can run multiple server instances for development, testing, and staging, each with its own set of models and configurations.

How to install

Prerequisites you need before installation:

Node.js 20+
Claude Desktop (for MCP integration)
No MongoDB installation required, because data is generated statistically on‑the‑fly

Step-by-step installation and setup you can follow locally:

Clone the project repository to your machine
Change to the project directory
Install dependencies with the package manager you use
Test the MCP server implementation to ensure it starts correctly
Optionally start a MongoDB‑compatible server if you want a local endpoint for testing

Additional sections

Configuration and runtime options are provided to tailor how MongTap runs and how you interact with it. You can run MongTap in different modes suitable for development, local testing, or standalone experiments. The server supports environment variables to control logging and port defaults, and you can start and manage multiple server instances to separate development and testing environments.

Security and privacy are designed around local, non‑persistent operation. Generated data is synthetic and does not represent real information. All operations are performed locally unless you explicitly configure network exposure. No credentials are required for access in the default setup, and inputs are validated to prevent common errors.

Troubleshooting tips cover common issues such as ensuring the MCP client can discover and connect to your MongTap server, verifying that the server is running on the expected port, and checking for proper model availability when generating data.

Architecture and usage notes

MongTap is composed of a statistical modeling engine, a MongoDB wire protocol implementation, and an MCP server integration layer. You interact with Claude Desktop or other MCP clients to drive model creation, server management, and data generation. Servers can share models or operate independently, depending on your configuration.

Available tools

generateDataModel

Create a statistical model from sample documents or a text description for data generation.

startMongoServer

Start a local MongoDB-compatible server that generates data from statistical models.

stopMongoServer

Stop a running MongoDB-compatible server instance by port number.

listActiveServers

Get a list of all currently running MongoDB-compatible server instances.

queryModel

Generate documents from a statistical model with optional query filters and generation control.

trainModel

Update an existing statistical model with additional sample documents to improve generation quality.

listModels

Get a list of all available statistical models stored locally.

getModelInfo

Retrieve detailed schema and statistics for a specific statistical model.