How To Build AI Trading Agent on Hyperliquid
We’re going to build an open-source AI trading agent on Hyperliquid — a smart bot that uses AI to analyze live crypto market data and make trading decisions automatically. The project, called Nocturne, connects to TAAPI to fetch real-time technical indicators like EMA and RSI, then uses large language models (LLMs) such as GPT-5, Grok-4, or DeepSeek-R1 (via OpenRouter) to interpret that data and decide when to buy, sell, or hold. Once a decision is made, the agent places and manages trades directly on the Hyperliquid decentralized exchange.
You can run this agent on your computer for testing, or deploy it securely on EigenCloud, where it operates inside a Trusted Execution Environment (TEE) — meaning all trades and decisions are verifiable and private. By the end, you’ll have a fully functioning AI-powered trading agent that can run 24/7, track multiple crypto assets, and make autonomous trading decisions in real markets.
GitHub repo: https://github.com/Gajesh2007/ai-trad...
Prerequisites / .env keys
TAAPI_API_KEY : https://taapi.io
OPENROUTER_API_KEY : https://openrouter.ai
LLM_MODEL (e.g., openai/gpt-5-pro, x-ai/grok-4, deepseek/r1)
HYPERLIQUID_PRIVATE_KEY
---------
Steps (Local)
1) Clone & configure
git clone https://github.com/Gajesh2007/ai-trad...
cd ai-trading-agent
cp .env.example .env # add keys + model
2) Install & run
poetry install
poetry run python src/main.py --assets BTC ETH --interval 1h
---------
Docker (Optional)
docker build --platform linux/amd64 -t trading-agent .
docker run --rm -p 3000:3000 --env-file .env trading-agent
curl http://localhost:3000/diary
---------
Deploy to EigenCloud (TEE):
Instructions: https://github.com/Gajesh2007/ai-trad...
Disclaimer: Educational content only. Not financial advice. Use test funds first.