<p align="left"> <img src="web/public/favicon.svg" alt="Evalchains" width="56" height="56" /> </p>

Evalchains

A beautiful, interactive evaluation-pattern visualizer for the 42 Network.

Search any 42 login, and Evalchains instantly builds a dynamic network map of every peer who has evaluated them (and every peer they've evaluated), grouping them into visual rings based on interaction frequency.

What is it for?

• A Trip Down Memory Lane: See your entire 42 journey mapped out. Who did you evaluate the most? Who evaluated you?
• Peeping on Peers: Look up your friends, explore their evaluation circles, and discover who their tightest peers are.
• Academic Integrity: Quickly identify "evaluation rings" where users evaluate each other suspiciously often.

The Tiers

A peer orbiting a subject is categorized into one of four tiers based on their evaluation frequency:

• Tight (Pink) — Highly concentrated mutual or one-way evaluations.
• Reciprocal (Orange) — You evaluated each other at least twice.
• Lopsided (Cyan) — Heavy evaluations in one direction with no echo.
• Normal (Grey) — Standard, infrequent evaluations.

Running Locally

Requirements: Node 22+ and a 42 OAuth application.

1. Register a 42 app with the redirect URI http://localhost:5173/api/auth/callback.
2. Create a .env file at the root:

   FT_CLIENT_ID=your_uid
   FT_CLIENT_SECRET=your_secret
   

3. Run the application:

   cd web
   npm install
   npm run dev
   

Staff / Matchmaking API

Evalchains provides a headless, programmatic endpoint specifically designed for Bocal or automated matchmaking systems to break up evaluation rings in real-time.

Endpoint: GET /api/staff/blacklist/<login>

This endpoint skips the UI and returns a clean JSON array of peers who fall into the "tight" tier.

Example Response:

{
  "target": "bsaeed",
  "reason": "Highly concentrated evaluations (tight tier)",
  "blacklist": ["jdoe", "asmith"]
}

Right before the 42 matchmaking system assigns an evaluator, it can query this endpoint and automatically exclude anyone in the blacklist array from the pool of potential peers.