DecoDocs is a document analysis platform that uses AI to help users understand complex legal documents, contracts, and terms. The platform provides plain-English explanations of legal jargon, highlights potential risks, and identifies important clauses. DecoDocs shares the same Firebase project (snapsign-au) as SnapSign, enabling unified authentication, data, and analytics.

• Frontend: React/Vite application with PDF.js for client-side text extraction
• Backend: Firebase Functions for AI processing and authentication
• Database: Firestore for usage metrics (no document content stored)
• Authentication: Firebase Anonymous Auth for privacy-focused experience
• Hosting: Firebase Hosting for static assets

• Anonymous document analysis (no account required)
• Client-side PDF text extraction
• Document fingerprinting using SHA-256 hashing
• AI-powered document analysis with plain-English explanations
• Risk identification and categorization
• Interactive document viewer with highlights
• Usage tracking and plan enforcement

• Node.js 18+
• Firebase CLI tools
• npm or yarn

# Install dependencies
cd decodocs/web
npm install

# Install functions dependencies
cd ../../functions
npm install

Create a .env file in the decodocs/web directory with your Firebase configuration. To get the actual values for your snapsign-au Firebase project:

1. Go to the Firebase Console
2. Select your snapsign-au project
3. Navigate to Project Settings (gear icon)
4. Scroll down to the "Your apps" section
5. Click the "</>" icon to add a web app (register the DecoDocs web app)
6. Copy the configuration values and add them to your .env file:

VITE_FIREBASE_API_KEY=
VITE_FIREBASE_AUTH_DOMAIN=
VITE_FIREBASE_PROJECT_ID=
VITE_FIREBASE_STORAGE_BUCKET=
VITE_FIREBASE_MESSAGING_SENDER_ID=
VITE_FIREBASE_APP_ID=

# Start Firebase emulators
firebase emulators:start

# In another terminal, start the development server
cd decodocs/web
npm run dev

• preflightCheck: Classify documents and estimate AI usage without calling AI
• analyzeText: Perform main document analysis
• getEntitlement: Retrieve user plan and limits

• Anonymous authentication required for all AI operations
• Document fingerprinting using SHA-256 of PDF bytes
• Usage tracked per (userId + documentHash) combination

• Client-side text extraction ensures documents never leave the browser
• Anonymous authentication protects user privacy
• Firestore security rules prevent unauthorized access
• Usage limits enforce fair use policies

# Run function unit tests
cd functions
npm run test:unit

# Run all tests
npm run test

Tests cover:

• Function input validation
• Document classification logic
• AI call limit enforcement
• Security rule compliance
• Error handling scenarios

# Deploy functions
firebase deploy --only functions

# Deploy hosting
firebase deploy --only hosting

# Deploy Firestore rules
firebase deploy --only firestore:rules

Firestore collections:

• usage_docs: Per-user, per-document usage tracking (no content stored)
• usage_events: Event logging for analytics (no content stored)

• Documents are processed client-side only
• No document content is stored on servers
• Anonymous authentication preserves user privacy
• Usage data is retained temporarily and then purged

• React/Vite for frontend
• Firebase Functions for backend
• Firestore for data persistence
• PDF.js for PDF processing
• Tailwind CSS for styling
• Gemini AI for document analysis