OpenAIChain — Hackathon Setup

Overview

• Demo-ready guest-mode auth. Training and rewards work without sign-in.
• Backend: Node/Express + MongoDB. Frontend: Vite + React + TypeScript.

Requirements

• Node.js >= 18 (you have v22)
• npm
• MongoDB Atlas URI
• (optional) HF_API_KEY, GOOGLE_CLIENT_ID for production

Quick local setup (Windows PowerShell)

# backend
cd server
cp .env.example .env
# edit server/.env with your values
npm install
npm run dev

# frontend
cd ../client
cp .env.example .env
# edit client/.env with VITE_API_URL=http://localhost:3000
npm install
npm run dev

Run dev

• Backend: nodemon index.js (npm run dev)
• Frontend: Vite (npm run dev) → http://localhost:5173

Build & single-server serve (optional)

# build frontend then serve from backend
cd client
npm run build
# copy build to server/build or configure server to serve client/dist
cd ../server
npm start

Deploy (recommended fast path)

• Push repo to GitHub.
• Frontend: Deploy to Vercel — set build cmd npm run build, output dist, set env var VITE_API_URL to your backend URL.
• Backend: Deploy to Render/Heroku — set start command npm start and add server env vars (MONGO_URI, JWT_SECRET, HF_API_KEY, CLIENT_ORIGIN=).

Environment variables (example)

• server/.env (do not commit)

# filepath: c:\Users\PC\Desktop\OpenAIChain\server\.env.example
PORT=3000
MONGO_URI=<your_mongo_atlas_uri>
JWT_SECRET=<strong-secret>
HF_API_KEY=<huggingface-key>
GOOGLE_CLIENT_ID=<google-client-id> # optional
CLIENT_ORIGIN=http://localhost:5173

• client/.env

# filepath: c:\Users\PC\Desktop\OpenAIChain\client\.env.example
VITE_API_URL=http://localhost:3000

Notes

• Keep guest-mode for hackathon to avoid dealing with OAuth, Supabase, and keys.
• If you must demo user-specific features, use the dev fallback POST /api/auth/google { email, name } to get a JWT.
• Make sure CORS on backend allows CLIENT_ORIGIN and Authorization header.

Demo notes

• The app runs in guest-mode by default for the hackathon: no login is required. Click "Continue as Guest" in the header or on the auth page to start training models instantly.
• To simulate an authenticated user (dev only), run:

curl -X POST http://localhost:3000/api/auth/google -H "Content-Type: application/json" -d '{"email":"demo@example.com","name":"Demo"}'

This returns a backend JWT you can pass in the Authorization header for testing user-specific flows.

Deployment (fast path)

• Frontend: Deploy the client folder to Vercel (build command: npm run build, output dir: dist). Set VITE_API_URL in Vercel env to your backend URL.
• Backend: Deploy the server folder to Render or Heroku. Set environment variables from server/.env.example.

Single-app deploy (optional)

• You can build the frontend and serve it from the backend in production by copying the client/dist (or client/build) folder into server/public and adding a static serve in server/index.js (see README in repo).