The Community for On-Chain Algo Traders
Algo Traders Club is a community of developers and traders building algorithmic trading agents that execute on-chain. We start with one market and one agent: NQTrader, which trades XYZ100 — the Nasdaq-100 perpetual on the Hyperliquid L1 — using session-based liquidity-sweep models with fixed targets and hard stops.
Every spec, prompt, and code change is built in public with Claude Code and Cursor. No signals, no copy trading, no performance promises — just the code, the process, and a room of people shipping it together.
How an NQTrader trade flows
Index Trading, On-Chain
Hyperliquid runs a fully on-chain order book on its own L1. HIP-3 lets builders deploy new perpetual markets, and TradeXYZ's XYZ100 tracks the Nasdaq-100 around the clock. Every order, fill, and cancel NQTrader makes lives on-chain — a transparent venue to learn real execution, not a broker's black box.
One
market
One instrument, studied deeply, while we calibrate the models.
24/7
market hours
XYZ100 trades even when the cash market is closed.
On-chain
execution
Orders and fills are visible on the Hyperliquid L1.
Start with NQTrader
Clone the agent, run it in dry-run, read the spec. Then help us make it better.
Clone it
git clone https://github.com/orgs/algo-traders-club/repositories, then copy .env.example to .env.
Run it
Starts in dry-run by default — no real orders until you deliberately switch it off. Explore it with the hl-op CLI.
Read it
A written PRD and CLAUDE.md explain every rule and decision, so you (and your AI coding agent) can follow the logic. No black boxes.
Is This Community For You?
ATC is for builders who want to trade with code, on-chain, alongside people doing the same. It's a strong fit if you are...
You're a Fit If...
- A developer or quant-minded trader comfortable reading Python (AI coding tools help, but you verify the output)
- Interested in NQ, index futures, or perps — and curious about trading them on-chain
- Build-first: you'd rather ship, paper-trade, and measure than debate indicators
- Serious about risk: fixed stops, position limits, and dry-run before live
This Might Not Be For You If...
- You are completely new to coding
- You want signals, copy trading, financial advice, or a get-rich-quick system
- You're uncomfortable managing your own keys, or can't afford to lose what you trade
- Hyperliquid isn't available where you live — check the venue's terms for your jurisdiction
How NQTrader Trades
One market, one agent, built in the open.
Liquidity, Not Prediction
Rather than forecasting direction, NQTrader looks for quick liquidity sweeps past obvious highs and lows and fades them at higher-timeframe structure. Every trade uses fixed targets and hard stops inside defined trading sessions.
Developer-First Toolkit
Our Python async agent with CCXT execution, Claude Code and Cursor workflows and Postgres and Redis trade logging help you read, test, and change the agent one verified step at a time.
Trading x AI, On-Chain
Join a community at the forefront of AI-assisted development, on-chain markets, and autonomous trading agents. Learn to specify strategies precisely, let AI coding agents implement them, and review the result against the code — not the summary.
From Spec to Live Agent
The workflow we use on NQTrader: write the rules down, let AI coding agents build them, and verify every step before it touches real money.
- 1
Specify the Strategy
Turn trading ideas into exact rules:
- Session windows in Eastern Time, DST-aware
- Precise triggers: closed candles only, wick vs close, fixed targets and stops
- Every number a config setting; any missing data means no trade
- 2
Build and Operate It with Claude Code
Implement in small, reviewable steps:
- Phased prompts for Claude Code or Cursor, one tested commit at a time
- Give the agent
CLAUDE.mdand the PRD as context, then use it to start, stop, and monitor the bot from the project directory - Run on your own VPS in dry-run first, then live limit orders on Hyperliquid
- 3
Verify & Iterate
Trust evidence, not summaries:
- Backtest on the same signal path the live agent runs
- Log every signal, rejection reason, and fill
- Paper-trade and measure real fills and slippage before sizing up
NQTrader Entry Models
AnyTime Sweep
Fade a sweep of recent swing highs or lows when it meets 15m order-block or fair-value-gap structure.
Asian Range Sweep
Map the pre-Asian range, then trade the sweep and rejection of its extremes.
London Sweep
Fade London's run on the overnight Asian session high or low.
Opening Range Breakout
Trade breaks of the opening range after the 9:30 ET cash open.
W / M Reversal (planned)
Two-leg double bottoms and tops at higher-timeframe structure.
Why Hyperliquid: a fully on-chain order book on its own L1, builder-deployed HIP-3 markets, and XYZ100 for Nasdaq-100 exposure around the clock. Transparent execution you can verify and learn from.
Meet Your Coach
Fodé Diop aka Bot Mechanic
I'm not just a teacher; I'm a builder and founder. As the founder of Nexwave.so and Algo Traders Club, I build and operate autonomous trading agents. NQTrader is our flagship: an agent that trades the Nasdaq-100 on-chain through XYZ100 on Hyperliquid, built in public with Claude Code and Cursor. I write the specs and review every change; the AI agents write the code. I share the whole process — specs, prompts, reviews, and mistakes — so you can build and operate your own on-chain trading systems. Let's build together.
Join the Conversation
Connect with developers and traders building on Hyperliquid, discuss NQTrader's models and research, and follow the build as it ships.
Ready to Build On-Chain?
Join the community building algorithmic trading agents on the Hyperliquid L1 — starting with NQTrader and the Nasdaq-100 on XYZ100.
Join Algo Traders Club