Happy slice of mcp pizzaMCP.pizza

mcp-database-server

MCP.Pizza Chef: fireproof-storage

The mcp-database-server is a Model Context Protocol (MCP) server that provides a simple JSON document store with full CRUD (Create, Read, Update, Delete) capabilities. It supports querying documents sorted by any field, making it ideal for managing structured data in real-time AI workflows. Designed to integrate seamlessly with LLM tools like Claude Desktop, this server uses Fireproof storage to ensure reliable and persistent data handling. It communicates over stdio, enabling secure and efficient interaction with AI systems. The server is easy to install and configure, supporting cross-platform usage on MacOS and Windows, and is particularly useful for developers building AI agents that require dynamic data storage and retrieval.

Use This MCP server To

Store JSON documents for AI tool workflows Query and sort JSON data by any field Perform CRUD operations on structured data Integrate persistent storage with LLM agents Manage real-time data for AI applications Enable AI systems to read and update JSON data

README

Model Context Protocol and Fireproof Demo: JSON Document Server

This is a simple example of how to use a Fireproof database in a Model Context Protocol server (used for plugging code and data into A.I. systems such as Claude Desktop).

This demo server implements a basic JSON document store with CRUD operations (Create, Read, Update, Delete) and the ability to query documents sorted by any field.

Installation

Install dependencies:

npm install
npm build

Running the Server

To use with Claude Desktop, add the server config:

On MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json On Windows: %APPDATA%/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "fireproof": {
      "command": "/path/to/fireproof-mcp/build/index.js"
    }
  }
}

Debugging

Since MCP servers communicate over stdio, debugging can be challenging. We recommend using the MCP Inspector, which is available as a package script:

npm run inspector

The Inspector will provide a URL to access debugging tools in your browser.

mcp-database-server FAQ

How do I install the mcp-database-server?
Install dependencies with 'npm install' and build using 'npm build'.
How do I configure the server for Claude Desktop?
Add the server config path to the claude_desktop_config.json file under 'mcpServers'.
What storage backend does the server use?
It uses Fireproof storage for reliable JSON document persistence.
Can I query JSON documents by any field?
Yes, the server supports querying and sorting documents by any specified field.
How does the server communicate with MCP clients?
Communication is done over standard input/output (stdio) streams.
Is the server cross-platform?
Yes, it supports MacOS and Windows configurations.
Can I perform update and delete operations on documents?
Yes, full CRUD operations including update and delete are supported.
How can I debug the server?
Debugging can be challenging due to stdio communication; use logging and careful config setup.
Does the server support integration with other LLM providers?
Yes, it can integrate with various LLM tools including Claude, OpenAI, and Gemini.