Adaptive Project Generator (Demo)

Small MERN + LLM demo that generates a project spec, scaffolds file manifests, and bundles a zip for download.

Quick run (dev / local)

1. Clone or create project files.
2. In backend folder: npm install then npm run dev (or npm start).
3. In frontend folder: npm install then npm start.
4. Open frontend at http://localhost:3000 — pick skills and click Generate.
5. Generate a couple files, click Download, unzip and run the backend npm install && npm start then check http://localhost:4000/health.

Demo checklist for faculty

• Open UI, click Generate → shows spec and manifest.
• Generate at least one file and preview it.
• Download zip and run backend health endpoint.

Notes for implementer

• Fill in backend/src/llm/adapters.js with your provider calls (Cursor, Lovable, ChatGPT proxy).
• Keep model responses to valid JSON when generating project spec. For file content generation, instruct model to return only file content.
• Make sure MongoDB is running locally or set MONGO_URI environment variable.

Environment Setup

Create a backend/.env file with the following variables:

# MongoDB Connection
# Option 1: MongoDB Atlas (Cloud) - Recommended
MONGO_URI=mongodb+srv://username:password@cluster0.xxxxx.mongodb.net/adaptive-generator?retryWrites=true&w=majority

# Option 2: Local MongoDB
# MONGO_URI=mongodb://127.0.0.1:27017/adaptive-generator

# Server Port
PORT=4000

# Gemini API Key (required)
GEMINI_API_KEY=your_gemini_api_key_here

# Optional: Gemini model selection
# Available models: gemini-1.5-flash (default, faster), gemini-1.5-pro (higher quality)
# GEMINI_MODEL=gemini-1.5-flash

MongoDB Atlas Setup (Cloud)

1. Create a MongoDB Atlas Account

   • Go to mongodb.com/cloud/atlas
   • Create a free cluster (M0)

2. Get Your Connection String

   • In Atlas Dashboard, click "Connect" on your cluster
   • Choose "Connect your application"
   • Copy the connection string (starts with mongodb+srv://)
   • Replace <password> with your database user password
   • Replace <dbname> with adaptive-generator (or your preferred database name)

3. Configure Network Access

   • Go to Atlas Dashboard → Network Access
   • Click "Add IP Address"
   • Add your current IP or 0.0.0.0/0 for development (less secure)

4. Create Database User

   • Go to Atlas Dashboard → Database Access
   • Create a new user with username and password
   • Grant "Atlas Admin" or "Read and write to any database" role

5. Add Connection String to .env

   • Paste your connection string into backend/.env as MONGO_URI

Local MongoDB Setup

If using local MongoDB instead:

1. Install MongoDB locally
2. Start MongoDB service
3. Use: MONGO_URI=mongodb://127.0.0.1:27017/adaptive-generator

The LLM adapter is now implemented and uses Google Gemini API. Make sure to set your GEMINI_API_KEY in the .env file.