A "Giant Tech" grade autonomous advertising system that launches, monitors, and optimizes Meta Ads campaigns. Built with a React Frontend, Express Backend, and MongoDB/Redis architecture.

• 🧠 AI Optimization Brain: An "Invisible Friend" that analyzes performance 24/7 and makes smart recommendations (Scale, Pause, Adjust).
• 🛡️ Mock Mode Simulation: Fully functional "Flight Simulator" for testing campaigns without spending real money.
• ⚡ Resilient Launch Queue: Background worker system ensuring 100% reliability for campaign launches, even during API outages.
• 📊 Interactive Dashboard: A beautiful React-based command center to visualize spend, ROAS, and system activity.
• 🐳 Dockerized Infrastructure: One-click deployment for the entire stack (Frontend, Backend, DB, Cache).

The system follows a scalable, microservices-ready pattern:

• Frontend: React + Vite + TailwindCSS + Framer Motion (Port 80/8080)
• Backend: Node.js + Express + TypeScript (Port 3000)
• Database: MongoDB (Data Persistence)
• Cache/Queue: Redis (Rate Limiting & Job Queue)
• Workers: Dedicated processes for Analysis and Launching.

To run the entire fleet in production mode using Docker:

1. Stop local services (free up ports 3000, 8080).
2. Configure Environment: Ensure .env exists (see .env.example). For Real Ads, set META_APP_ID and META_APP_SECRET. For Mock Mode, leave them blank.
3. Launch Fleet:

docker compose -f docker-compose.prod.yml up -d --build

4. Access App:
   • Dashboard: http://localhost:8080
   • API: http://localhost:3000

If you want to contribute or modify code:

npm install
npm run docker:up  # Starts Mongo/Redis
npx ts-node server/index.ts

cd client
npm install
npm run dev

Access at http://localhost:5173.

• Purpose: provide a predictable, local model-like endpoint for integration tests and local feature development.
• Start locally: npm run mcp:start (use npm run mcp:dev for live reload during development).
• Main endpoints: GET /health, POST /mcp/request.

See lib/mcp/README.md for details and examples.

• End-to-End: npm run test:e2e (Simulates full user flow).
• Unit/Integration: npm run test:all.
• Optimization Simulation: npx ts-node scripts/run-optimization.ts.

• Features List - Detailed breakdown of capabilities.
• Project Structure - Map of the codebase.
• Optimization Strategy - How the AI thinks.
• OAuth Guide - Connecting Real Accounts.

Built with ❤️ by the Giant Tech Team.