Firebase Realtime Database MCP

<a href="https://glama.ai/mcp/servers/@ytzlax/firebase-realtimeDB-mcp"> <img width="380" height="200" src="https://glama.ai/mcp/servers/@ytzlax/firebase-realtimeDB-mcp/badge" /> </a>

Overview

The inspiration for this project was taken from firebase-mcp. Unlike firebase-mcp, which uses Firestore functionalities, firebase-realtimeDB-mcp uses Firebase Realtime Database functionalities. Firebase Realtime Database MCP enables AI assistants to work directly with Firebase services.

🚀 Quick Start

Prerequisites

• Firebase project with service account credentials
• Node.js environment

1. Firebase Configuration

1. Go to Firebase Console → Project Settings → Service Accounts
2. Click "Generate new private key"
3. Save the JSON file securely

2. Environment Variables

Required

• SERVICE_ACCOUNT_KEY_PATH: Path to your Firebase service account key JSON
• FIREBASE_DATABASE_URL: Firebase realtime database URL

1. Install Firebase Realtime Database MCP project locally

A. Run git clone https://github.com/ytzlax/firebase-realtimeDB-mcp.git in the terminal.
B. cd to the cloned path, and run npm i.
C. Run npm run build.

2. Install MCP Server

Add the server configuration to your MCP settings file:

• Claude Desktop: ~/Library/Application Support/Claude/claude_desktop_config.json
• Augment: ~/Library/Application Support/Code/User/settings.json
• Cursor: [project root]/.cursor/mcp.json

How to use with Claude Desktop?

Follow the guide https://modelcontextprotocol.io/quickstart/user and add the following configuration:

{
  "firebase-realtimeDB-mcp": {
      "command": "node",
      "args": ["/absolute/path/to/dist/index.js"],
      "env": {
        "SERVICE_ACCOUNT_KEY_PATH": "/absolute/path/to/serviceAccountKey.json",
        "FIREBASE_DATABASE_URL": "your-project-id.firebasestorage.app"
      }
   }

}

Firebase Tools