How to Connect Firebase to an AI Agent
Auth setup
1. Create a service account in Google Cloud Console > IAM & Admin > Service Accounts. 2. Download the JSON key and set GOOGLE_APPLICATION_CREDENTIALS. 3. Run `firebase login` for interactive use. 4. For MCP: `npx firebase-tools experimental:mcp` uses the current gcloud auth context.
Key facts
| Base URL | https://firebase.googleapis.com/v1beta1/ |
|---|
| API version | v1beta1 (management) + product APIs (Firestore, Auth, etc.) |
|---|
| Auth | Google OAuth 2.0 with service account or user credentials. Service accounts use JWT-based access tokens (ADC). Admin SDK is preferred for server-side agent use. For MCP, the `firebase-tools experimental:mcp` command uses existing gcloud credentials. |
|---|
| Token URL | https://oauth2.googleapis.com/token |
|---|
| Scopes | https://www.googleapis.com/auth/cloud-platform (all Firebase features), or scoped per-product. |
|---|
| Request body | application/json |
|---|
| Pagination | Standard Google API pagination: `pageSize` and `pageToken` query params; response has `nextPageToken`. |
|---|
| Rate limit | Firestore: 500 writes/sec per database baseline; reads are effectively unmetered. Auth: 3,000 requests/min. Management API: 600 req/min per project. 429 or 403 quotaExceeded errors on overload. |
|---|
| Error format | Google RPC error format: {"error":{"code":403,"message":"...","status":"PERMISSION_DENIED","details":[...]}} |
|---|
Key endpoints
| Method | Path | Description |
|---|
GET | /projects/{projectId} | Get project metadata |
GET | /projects/{projectId}/databases | List Firestore databases |
POST | /projects/{projectId}/databases/{dbId}/documents:runQuery | Run a structured Firestore query |
GET | /projects/{projectId}/webApps | List web apps registered under the project |
POST | /projects/{projectId}/rules:deploy | Deploy new security rules |
Quickstart
GET https://firebase.googleapis.com/v1beta1/projects/my-project
Authorization: Bearer {access_token}
Response: {"projectId":"my-project","displayName":"My App","resources":{"realtimeDatabaseInstance":"..."}}
Agent pitfalls & tips
- The Firebase REST Management API only manages project metadata — for actual data ops use Firestore REST (/v1/projects/{p}/databases/{db}/documents).
- Always prefer the Admin SDK over REST on the server side — it handles token refresh and connection pooling.
- Security rules are deployed separately from data — an agent that writes "correctly" may still hit PERMISSION_DENIED if rules are strict.
- Firestore indexes must be created (or inferred) before composite queries work — errors return with a URL to create the missing index.
- Running `firebase-tools experimental:mcp` exposes project, hosting, functions, and Firestore tooling to an agent with low setup cost.
Source: curated by KanseiLink from official documentation (docs) and registry checks. Last reviewed: 2026-04-10. Specs change — verify against the official docs before production use.
Frequently Asked Questions
What is Firebase's AEO score?
▼
Firebase has an AEO score of 0.60 and is rated BBB (Basic agent connectivity available). AEO (Agent Engine Optimization) measures how well a SaaS service works with AI agents. Scores range from 0.00 to 1.00, with grades from AAA (best) to D (not agent-ready).
Is Firebase AI-agent-ready?
▼
Firebase is currently connectable for AI agent use. Third-party MCP integrations are available for this service. For detailed connection guides, auth setup, and known pitfalls, use the KanseiLink MCP tool.
How does Firebase compare to other Database services?
▼
In the Database category, Firebase is rated BBB. KanseiLink evaluates services based on MCP availability, API quality, documentation, auth-guide clarity, and integration recipe availability (methodology published). Visit the full rankings at kansei-link.com to see how Firebase compares.
How can I integrate Firebase with an AI agent?
▼
The fastest way to integrate Firebase with an AI agent is through KanseiLink MCP. Install it with: npx @kansei-link/mcp-server — then use the search_services and get_service_detail tools to get the current auth setup, endpoints, rate limits, and agent-specific tips. This data is kept fresh from registry checks, curated official-doc guides, and agent reports.
How do I authenticate with Firebase?
▼
Google OAuth 2.0 with service account or user credentials. Service accounts use JWT-based access tokens (ADC). Admin SDK is preferred for server-side agent use. For MCP, the `firebase-tools experimental:mcp` command uses existing gcloud credentials. Setup: 1. Create a service account in Google Cloud Console > IAM & Admin > Service Accounts. 2. Download the JSON key and set GOOGLE_APPLICATION_CREDENTIALS. 3. Run `firebase login` for interactive use. 4. For MCP: `npx firebase-tools experimental:mcp` uses the current gcloud auth context.
What are Firebase's API rate limits?
▼
Firestore: 500 writes/sec per database baseline; reads are effectively unmetered. Auth: 3,000 requests/min. Management API: 600 req/min per project. 429 or 403 quotaExceeded errors on overload.