Using AI to generate good vibes!
A lightweight web app that takes your vibes (moods, themes, prompts) and turn them into playlists / recommendations using AI.
Prompt → Vibe: Enter a mood or theme and using AI, we will generate a list of 40 curated songs that match that vibe!
Refresh: You can choose to refresh all 40 songs or 1 song at a time! The songs will not be the same songs that were on the list previously.
Save to Spotify: If you like the song recommendations, you can click the button at the bottom of the playlist page to get spotify song URIs! Due to Spotify API limiations, these URI links will be individual song links stored as an array of strings.
Frontend: Vite + JavaScript, React, HTML, CSS
Server: Node.js, Express
AI: OpenAI 4.o
VibeList/
├─ src/ # Frontend source (components, utilities, state)
├─ server/ # Small backend for API calls/secrets
├─ index.html # App entry HTML (Vite)
├─ styles.css # Global styles
├─ vite.config.js # Vite configuration
├─ package.json # Scripts & dependencies
└─ log.md # Dev AI logs
Node.js ≥ 18
npm (or pnpm/yarn)
- Clone and install
git clone https://github.com/TeamTweety/VibeList.git cd VibeList npm install
- Configure environment
Create a .env file in the project root and add any required keys:
AI Provider = OpenAi 4.o
OPENAI_API_KEY="sk-..."
SPOTIFY_CLIENT_ID="..."
SPOTIFY_CLIENT_SECRET="..."
- Run locally
In two terminals (or use a single command if you have a concurrent script):
npm run conc
Open the app at the printed dev URL (commonly http://localhost:5173).
LLM calls: Be careful of too many calls.
Spotify API: We currently do not support spotify account login. That means that users will be rate limited based on their Spotify client ID
Contributions are welcome! Please:
Open an issue describing the change.
Fork the repo and create a feature branch.
Submit a PR with a clear description and screenshots if UI changes.