bu!ld@th0n // FACES

"Know your judge, know your code."

Memory training subsystem for bu!ld@th0n platform. Module gamifies onboarding process, requiring builders to demonstrate familiarity w/ AI Judges & Industry Mentors before they can submit their final build.

🎯 Objective

Unlock Judge Selection Panel by achieving 100% accuracy in faces game - theoretically.

• Level 1 (Legends): Identify AI Judges (Fei-Fei Li, Jensen Huang, etc.).
• Level 2 (Builders): Unlock Mentors (Guillermo Rauch, John Carmack, etc.).

🛠 Tech Stack

• Frontend: React / Next.js (TypeScript)
• Data: Supabase (PostgreSQL)
• Assets: Supabase Storage (Headshots)
• Vibe: Hacker / Terminal / Dark Mode

🚀 Setup & Install

1. Clone & Install

   git clone <repo-url>
   cd buildathon-faces
   npm install

2. Environment Variables Create a .env.local file:

   NEXT_PUBLIC_SUPABASE_URL=your_supabase_url
   NEXT_PUBLIC_SUPABASE_ANON_KEY=your_supabase_key

3. Run Development Server

   npm run dev

💽 Data Schema

The game pulls from the game_personalities table in Supabase:

Column	Type	Description
id	uuid	Unique ID
full_name	text	e.g., "Mira Murati"
role	text	'judge' or 'mentor'
bio_blurb	text	"Founder & CEO, Thinking Machines Lab"
image_url	text	URL to Supabase Storage bucket

🕹 Game Logic

1. Fetch: App loads randomized set of personalities from Supabase.
2. Loop: Displays 1 face + 4 name options.
3. Scoring:
   • Correct: +10 points, Face removed from pool.
   • Incorrect: -5 points, Face reshuffled into queue.
4. Win State: When pool is empty, update user profile judges_unlocked: true.

🏗 Status

• Core Game Loop: Functional
• Supabase Integration: Pending
• Mentor Mode Unlock: Locked

---

Built for the [bu!ld@th0n] protocol.