Firebase MCP Server: Let AI Query Firestore, Auth, and Cloud Functions

Firebase Data Locked in the Console

Firebase powers millions of apps with Firestore, Authentication, Cloud Functions, Hosting, and Analytics. But querying that data means navigating the Firebase console, clicking through collections, filtering documents manually, and checking Cloud Function logs in separate tabs.

An MCP server gives your AI assistant direct access to Firebase operations. Query collections, check user auth state, read function logs, and pull analytics through natural language.

How It Works with DataFaucet

DataFaucet captures the API calls Firebase's console makes when you browse your project. Those internal calls become MCP tools your AI can invoke.

Step 1: Enter Your Firebase Console URL

Go to DataFaucet and paste:

https://console.firebase.google.com/project/YOUR-PROJECT-ID

Step 2: Browse Your Project

Navigate through Firebase features you want AI access to:

• Firestore — open collections, run queries, filter documents
• Authentication — browse user list, search by email
• Cloud Functions — check function list, view logs
• Hosting — check deploy history, domain config
• Analytics — view events, user metrics, retention data
• Storage — browse files, check usage

Step 3: Deploy and Connect

Review captured endpoints, deploy your server, connect to Claude, Cursor, or any MCP client.

What Your AI Can Do

Once connected:

• "Show me all users who signed up in the last 24 hours"
• "Query the orders collection for documents where status equals pending"
• "What errors are in my Cloud Function logs from today?"
• "How many daily active users did we have this week?"
• "List all documents in users/ where plan equals 'pro'"
• "When was the last successful deploy to hosting?"

Use Cases by Role

Backend developers: "Query Firestore for all orders where total > $100 and created today." Debug production data without writing console queries.

Frontend developers: "Is the auth user with email X verified? What claims do they have?" Quick user state checks during development.

DevOps/SRE: "Show me Cloud Function errors in the last hour. Which functions have the highest cold start time?" Incident response from chat.

Product managers: "How many new signups this week vs last week? What's our 7-day retention?" Quick metrics without Analytics dashboard navigation.

Firebase Console vs DataFaucet

Firebase's REST API requires custom tokens and complex auth flows. The Admin SDK needs a service account key and server environment. DataFaucet captures the same authenticated calls the console makes, wrapping them in MCP tools.

Tips for Better Capture

1. Firestore: Open 2-3 different collections, use the filter bar to query documents
2. Auth: Search for a specific user, browse the user list with pagination
3. Functions: View the function list, open logs for a specific function
4. Analytics: Check the dashboard overview and events tab
5. Use the path bar to navigate directly to nested collections

Multi-Project Setup

If you manage multiple Firebase projects, create separate MCP servers for each. DataFaucet's free tier includes 3 servers, enough for dev/staging/prod environments.

Connect in 60 Seconds

1. Go to DataFaucet
2. Paste your Firebase console URL (include project ID)
3. Browse Firestore, Auth, Functions, and Analytics (2-3 minutes)
4. Deploy and connect to Claude, Cursor, or Codex

Your Firebase project becomes AI-queryable. Firestore queries, user lookups, and function monitoring through natural language.

Related: Supabase MCP Server for Supabase users, MongoDB MCP Server for MongoDB, Google Analytics MCP Server for analytics.