Firebase Development Skill

Table of Contents

• Quick Start
• Scope
• Task Navigation
• Scripts
• Cloud Functions Generation
• Assets
• Common Workflows
• Pre-Deployment Checklist
• Emulator Ports
• Key Decisions

Quick Start

1. New project: Run scripts/init_project.sh [project-id]
2. Local development: Run scripts/start_emulators.sh
3. Deploy: Run scripts/deploy.sh

Scope

Use this skill for: Firebase development including Firestore CRUD/queries, Cloud Functions (1st/2nd gen), Firebase CLI, emulator setup, security rules, authentication, hosting, and GCP integration.

Do not use for: Pure GCP without Firebase, AWS/Azure services, non-serverless architectures, self-hosted solutions, or complex relational queries (use Cloud SQL instead).

Task Navigation

Scripts

For complete CLI reference, see references/cli-commands.md.

init_project.sh

Initialize Firebase project with Firestore, Functions, Hosting, Storage, Emulators.

./scripts/init_project.sh              # Interactive
./scripts/init_project.sh my-project   # Specific project

start_emulators.sh

Start emulator suite with data persistence.

./scripts/start_emulators.sh                    # Auto-persistence
./scripts/start_emulators.sh --debug            # Enable debugging
./scripts/start_emulators.sh --import ./backup  # Import data
./scripts/start_emulators.sh --only functions,firestore

deploy.sh

Deploy with safety confirmations.

./scripts/deploy.sh                     # Full deploy
./scripts/deploy.sh --dry-run           # Preview only
./scripts/deploy.sh --only hosting      # Specific target
./scripts/deploy.sh --force             # Skip confirmation

deploy_functions.sh

Deploy Cloud Functions with granular control.

./scripts/deploy_functions.sh                    # All functions
./scripts/deploy_functions.sh myFunction         # Single function
./scripts/deploy_functions.sh --codebase python  # Specific codebase

manage_secrets.sh

Manage Cloud Functions secrets for 2nd gen functions. Uses GCP Secret Manager for secure storage. Prefer this script over direct gcloud commands for Firebase-integrated secret management with proper function access binding.

Secret lifecycle: Create secrets before first deploy, update via set (creates new version), bind to functions via runWith({ secrets: [...] }), and rotate by setting new values.

./scripts/manage_secrets.sh set API_KEY      # Set secret (creates or updates)
./scripts/manage_secrets.sh get API_KEY      # View metadata and versions
./scripts/manage_secrets.sh list             # List all project secrets
./scripts/manage_secrets.sh delete API_KEY   # Delete secret and all versions

export_firestore.sh / import_firestore.sh

Backup and restore Firestore data.

./scripts/export_firestore.sh --emulator              # From emulator
./scripts/export_firestore.sh --output gs://bucket    # Production to GCS
./scripts/import_firestore.sh --input ./data --emulator

setup_python_functions.py

Create Python Cloud Functions project.

python scripts/setup_python_functions.py --path python-functions --codebase python

Cloud Functions Generation

Use 2nd generation (recommended):

• HTTP, Firestore, Storage, Scheduled, Pub/Sub triggers
• Higher concurrency, longer timeouts

Use 1st generation only for:

• Auth onCreate/onDelete triggers (not available in 2nd gen)

2nd Gen Example (TypeScript)

import { onDocumentCreated } from "firebase-functions/v2/firestore";
import { onRequest } from "firebase-functions/v2/https";

export const onUserCreated = onDocumentCreated("users/{userId}", (event) => {
  console.log("New user:", event.params.userId, event.data?.data());
});

export const api = onRequest({ cors: true }, (req, res) => {
  res.json({ status: "ok" });
});

1st Gen Auth Trigger

import * as functions from "firebase-functions/v1";

export const onUserCreate = functions.auth.user().onCreate((user) => {
  console.log("New user:", user.uid);
  return null;
});

See references/functions-triggers.md for all trigger types with TypeScript and Python examples.

Assets

Common Workflows

New Project Setup

1. Run scripts/init_project.sh
2. Copy templates from assets/ directory
3. Start emulators: scripts/start_emulators.sh

Add Python Functions

1. Run python scripts/setup_python_functions.py
2. Update firebase.json with provided config
3. Deploy: scripts/deploy_functions.sh --codebase python

Security Rules Development

1. Start with assets/firestore.rules.template
2. Test with emulator
3. Deploy: firebase deploy --only firestore:rules

See references/security-rules.md for patterns.

Production Deployment

1. Set secrets: scripts/manage_secrets.sh set API_KEY
2. Dry run: scripts/deploy.sh --dry-run
3. Deploy: scripts/deploy.sh

Pre-Deployment Checklist

Before deploying to production, verify:

• Security Rules: Tested rules in emulator, no open access patterns
• Secrets: All required secrets configured via scripts/manage_secrets.sh list
• Environment: Correct project selected (firebase use)
• Functions: All functions tested locally with emulator
• Indexes: Firestore indexes deployed (firebase deploy --only firestore:indexes)
• Dry Run: scripts/deploy.sh --dry-run shows expected changes
• App Check: Enabled for production apps (prevents abuse)
• Billing: Budget alerts configured in GCP Console
• Monitoring: Cloud Logging and Error Reporting enabled

Emulator Ports

Key Decisions

Firestore Data Modeling

• Embed data read together that rarely changes
• Reference data that changes frequently or is shared
• Subcollections for parent-child relationships
• Root collections for cross-document queries

See references/firestore.md for patterns.

TypeScript vs Python Functions

• TypeScript: JavaScript teams, Firebase client SDK integration
• Python: ML/data science, Python ecosystem

Both can coexist via multiple codebases in firebase.json.