AI Copy-Trading Signal Platform (V1)

A production-ready backend for automated copy-trading on Hyperliquid using on-chain top-trader tracking.

Architecture

┌─────────────────────────────┐
│   Signal Tracker Service    │
│  (polls Hyperliquid Info)  │
└──────────────┬─────────────┘
               │ publishes signals
               ▼
┌─────────────────────────────┐
│   Redis Streams (Pub/Sub)  │
│   channel: signals:v1      │
└──────────────┬─────────────┘
               │ fan-out
               ▼
┌─────────────────────────────┐
│   Execution Agents (per-user)
│   (consume signals, trade) │
└──────────────┬─────────────┘
               │ telemetry
               ▼
┌─────────────────────────────┐
│      Core Backend API       │
│  (auth, billing, control)  │
└─────────────────────────────┘

Components

Component	Tech Stack	Description
backend/	TypeScript/Fastify	REST API, auth, billing, agent control
signal-tracker/	TypeScript/Node.js	Polls Hyperliquid, emits trading signals
execution-agent/	Python	Per-user container, risk management, order execution

Quick Start

Prerequisites

• Node.js 20+
• Python 3.11+
• PostgreSQL 15+
• Redis 7+
• Docker (for deployment)

Local Development

# Backend
cd backend
npm install
npm run dev

# Signal Tracker
cd signal-tracker
npm install
npm run dev

# Execution Agent
cd execution-agent
uv sync
python -m src.executor_agent.main

Environment Variables

# backend/.env
DATABASE_URL=postgresql://user:pass@localhost:5432/ai_copytrading
REDIS_URL=redis://localhost:6379
JWT_SECRET=your-secret-key
PLATFORM_API_KEY=your-platform-key

# signal-tracker/.env
REDIS_URL=redis://localhost:6379
BACKEND_URL=http://localhost:3000

# execution-agent/.env
REDIS_URL=redis://localhost:6379
HYPERLIQUID_NETWORK=testnet  # or mainnet
BACKEND_URL=http://localhost:3000
PLATFORM_API_KEY=your-platform-key

Testing

# Run all tests
npm test                           # backend & signal-tracker
python -m pytest tests/            # execution-agent

# Test results:
# - Backend: 8 tests ✓
# - Signal Tracker: 11 tests ✓
# - Execution Agent: 17 tests ✓

Deployment

Docker Images

Images are built automatically via GitHub Actions and pushed to GHCR:

• ghcr.io/uuzor/experimental-cloud/backend:latest
• ghcr.io/uuzor/experimental-cloud/signal-tracker:latest
• ghcr.io/uuzor/experimental-cloud/execution-agent:latest

Zeabur Deployment

Use the infrastructure templates in infrastructure/:

1. Create a Zeabur project
2. Add PostgreSQL and Redis dependencies
3. Deploy using templates or manual configuration

See infrastructure/template.yaml for the full configuration reference.

API Endpoints

Authentication

POST /v1/auth/register    - User registration
POST /v1/auth/login       - User login
GET  /v1/auth/me         - Get current user

Execution Agents

POST   /v1/execution-agents              - Create agent
GET    /v1/execution-agents/:id          - Get agent status
PATCH  /v1/execution-agents/:id/config  - Update config
POST   /v1/execution-agents/:id/suspend - Suspend agent
DELETE /v1/execution-agents/:id          - Delete agent

Internal (Agent-to-Backend)

POST /v1/agents/handshake    - Agent boot handshake
POST /v1/telemetry           - Submit telemetry
GET  /v1/agents/:id/stream-token - Get Redis stream token

Security

• Agent wallets only: Only Hyperliquid agent (API) wallet keys are stored, never master keys
• Envelope encryption: Credentials encrypted at rest
• Audit logging: All privileged actions logged
• Non-custodial: User funds remain in their own wallets

License

MIT