Engagement Message Generator

A standalone web application for generating creative engagement messages for Gen Z teens using Claude AI.

Features

• 🎨 Theme selection (8 different themes + random)
• 📱 Mobile-friendly shadcn/ui interface
• 🚀 Fast generation with Claude Sonnet 4.5
• 📋 One-click copy to clipboard
• 💰 Cost-optimized deployment (auto-stop on Fly.io)

Tech Stack

• Frontend: Next.js 16 with App Router, shadcn/ui, Tailwind CSS
• Backend: FastAPI (Python 3.11)
• AI: Anthropic Claude Sonnet 4.5
• Deployment: Docker + Fly.io

Local Development

Prerequisites

• Python 3.11+
• Node.js 20+
• npm

Setup

1. Clone and navigate to the project:

   cd engage

2. Set up environment variables:

   cp .env.example .env
   # Edit .env and add your CLAUDE_API_KEY

3. Install backend dependencies:

   pip install -r backend/requirements.txt

4. Install frontend dependencies:

   cd frontend
   npm install
   cd ..

5. Run backend (in one terminal):

   uvicorn backend.main:app --reload --port 8080

6. Run frontend (in another terminal):

   cd frontend
   npm run dev

7. Open browser:

   • Development: http://localhost:3000
   • Backend API docs: http://localhost:8080/docs

Production Build

Build frontend static export:

cd frontend
npm run build
cd ..

Run backend (serves both API and static frontend):

uvicorn backend.main:app --host 0.0.0.0 --port 8080

Docker

Build:

docker build -t engage .

Run:

docker run -p 8080:8080 -e CLAUDE_API_KEY=your_key_here engage

Deployment to Fly.io

First-time setup:

fly launch
# Follow prompts (use existing fly.toml)

Set secrets:

fly secrets set CLAUDE_API_KEY=your_claude_api_key_here

Deploy:

fly deploy

Check status:

fly status
fly logs

API Endpoints

• GET /health - Health check endpoint
• POST /api/generate - Generate message
  • Request body: {"theme": "random" | "meme/internet culture" | ...}
  • Response: {"content": "...", "theme_used": "..."}

Cost Optimization

The app is configured for minimal costs on Fly.io:

• min_machines_running = 0 - Machines stop when idle
• auto_start_machines = true - Start on request (~1-2s cold start)
• Shared CPU instance (smallest size)
• Static frontend served directly from backend

Expected cost: < $5/month with light usage

Project Structure

engage/
├── backend/
│   ├── main.py              # FastAPI app
│   ├── services/
│   │   └── message_generator.py  # Claude integration
│   └── requirements.txt
├── frontend/
│   ├── app/                 # Next.js pages
│   ├── components/          # React components
│   └── package.json
├── Dockerfile               # Multi-stage build
├── fly.toml                 # Fly.io config
└── docs/
    └── plans/
        └── 2025-11-12-message-generator-design.md

License

Private project