TrendSpinoff — Demo

Minimal Next.js + Tailwind demo for the Hacklytics2026 prototype.

New: YouTube Trending + Niche Filter + Gemini Pattern Extraction

This repo now includes an API route that:

1. pulls global trending YouTube videos (region-based),
2. filters/sorts them by a user-provided niche keyword,
3. sends the best-matching videos to Gemini for pattern extraction.

Quick start:

npm install
npm run dev

Visit http://localhost:3000

Gemini integration (optional):

• To enable Gemini-backed spin-off generation, set these environment variables in a local env file or your deployment:

  • GEMINI_API_URL — the full HTTP endpoint for your Gemini/generative API (POST)
  • GEMINI_API_KEY — the bearer API key

Example (env.local):

GEMINI_API_URL=https://your-gemini-endpoint.example/v1/generate
GEMINI_API_KEY=sk-...

The /api/spinoff endpoint will try Gemini first and fall back to the built-in mock generator if the call fails or the env vars are not set.

Deployment

Deploy to Vercel (recommended for Next.js):

• Push this repository to GitHub.
• Import the repo into Vercel and set environment variables in the Vercel dashboard (GEMINI_API_URL, GEMINI_API_KEY if used).
• Vercel will run the build and host the app.

Docker (local or container host):

Build the image:

docker build -t trendspinoff-demo .

Run the container (exposes port 3000):

docker run -p 3000:3000 --env-file .env.local trendspinoff-demo

Notes:

• Use npm run build locally before creating production images to verify the build.
• The .env.example file shows optional environment variables the app can use.