Implement shop_insights_app folder structure with Flutter frontend and Python backend

[Copilot]

Creates complete application scaffold for vendor analytics platform with AI-powered insights and OCR receipt scanning.

Architecture

Flutter Frontend (lib/)

• Clean architecture with repository pattern separating data access from business logic
• Three repository implementations: dummy (testing), Firebase (cloud), API (backend)
• Five screens: login, dashboard, manual sale entry, OCR scanning, AI insights
• Custom chart widgets for sales trends and category analytics

Python Backend (backend/)

• FastAPI server with Google Gemini AI integration for business insights generation
• JWT authentication middleware (configurable)
• CORS-enabled for cross-origin requests

Structure

lib/
├── core/              # App constants and routing
├── models/            # SaleModel, DashboardSummary, UserModel
├── repositories/      # Abstract interfaces + implementations
├── services/          # Dashboard, OCR, Auth business logic
├── screens/           # UI components
├── widgets/           # Reusable charts and cards
└── utils/             # Data parsers

backend/
├── main.py            # FastAPI endpoints
├── gemini_service.py  # AI insights generation
├── auth_middleware.py # JWT verification
└── requirements.txt   # Dependencies

Key Implementation Details

• Repository abstraction enables swapping data sources without UI changes
• Dashboard service aggregates sales data into analytics (revenue by category, top products, daily trends)
• OCR service structure ready for Google Vision API integration
• Mock repositories with realistic test data for development

Security

Updated dependencies to patched versions:

• fastapi: 0.104.1 → 0.109.1 (fixes Content-Type ReDoS)
• python-multipart: 0.0.6 → 0.0.22 (fixes file write, DoS, ReDoS vulnerabilities)

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.