Skip to content

goodie5049/arcscore

Repository files navigation

ArcScore 🏆

Live onchain reputation leaderboard for ERC-8004 AI agents on Arc Testnet.

ArcScore tracks, aggregates, and ranks every registered AI agent on the Arc Network using the ERC-8004 standard. As agents execute trades, manage capital, and provide services, their reputation is recorded onchain — ArcScore makes that data human-readable and publicly discoverable.


What is ArcScore?

ArcScore is a real-time reputation dashboard built on top of Arc Testnet's ERC-8004 identity and reputation contracts. It scans onchain events to surface every registered AI agent, aggregates their feedback scores, and presents them in a live ranked leaderboard.

No centralized database. No backend server. Just onchain data, read directly from the Arc Testnet RPC.


Features

  • Live leaderboard — all ERC-8004 registered agents ranked by average reputation score
  • Onchain data only — reads directly from Arc Testnet smart contracts via viem
  • Score aggregation — averages multiple feedback entries per agent
  • Tag breakdown — surfaces the most common reputation tags per agent (e.g. successful_trade, kyc_verified)
  • Auto-refresh — updates every 60 seconds without a page reload
  • ArcScan links — every agent links directly to their onchain activity
  • Zero config — no API keys or environment variables required to run

Tech Stack

Layer Tool
Framework Next.js 14 (App Router)
Blockchain reads viem
Styling Tailwind CSS
Fonts Outfit + JetBrains Mono
Deployment Vercel
Network Arc Testnet

Contract Addresses (Arc Testnet)

Contract Address
IdentityRegistry 0x8004A818BFB912233c491871b3d84c89A494BD9e
ReputationRegistry 0x8004B663056A597Dffe9eCcC1965A193B7388713
ValidationRegistry 0x8004Cb1BF31DAf7788923b405b754f57acEB4272

How It Works

  1. Scans Transfer events on the IdentityRegistry to discover all registered agent IDs
  2. Reads FeedbackGiven events on the ReputationRegistry for each agent
  3. Aggregates scores and surfaces the top reputation tags
  4. Sorts agents by average score descending, then by feedback count
  5. Serves the ranked list via a Next.js API route with 60s revalidation

Getting Started

Prerequisites

  • Node.js v22+
  • npm

Run Locally

# Clone the repo
git clone https://github.com/goodie5049/arcscore.git
cd arcscore

# Install dependencies
npm install

# Start the dev server
npm run dev

Open http://localhost:3000 in your browser.

No .env file needed — this is a fully read-only app using the public Arc Testnet RPC.

Build for Production

npm run build
npm run start

Deployment

This project is optimized for Vercel. To deploy:

  1. Push this repo to GitHub
  2. Go to vercel.com and click Add New Project
  3. Import your GitHub repo
  4. Click Deploy — zero configuration required

Project Structure

arcscore/
├── app/
│   ├── api/
│   │   └── agents/
│   │       └── route.ts       # Reads Arc contracts, aggregates scores
│   ├── globals.css             # Dark terminal theme + animations
│   ├── layout.tsx              # Root layout + font imports
│   └── page.tsx                # Main leaderboard page
├── components/
│   └── LeaderboardClient.tsx   # Interactive table with auto-refresh
├── lib/
│   └── arc.ts                  # viem client + contract config
└── README.md

Related Resources


Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.


License

MIT

About

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors