Hyperliquid · public data · simulation only

Market research without live orders.

A restart-safe Python service that reads public Hyperliquid spot and HIP-3 TradFi data, models a four-hour strategy, and records every paper fill in a reviewable ledger.

Hard safety boundary

The project contains no wallet, private key, signing code, or call to Hyperliquid's exchange endpoint. It cannot place a real order and makes no profit claim.

0Live orders
possible
Modeled
maximum exposure
4hCompleted-bar
signal interval
100%Public market
data inputs
01

What the simulator actually tests

The useful output is a falsifiable execution record—not a forecast or an invitation to trade.

FilterLiquidity

Reject weak markets

Excludes bookless or delisted markets, thin rolling volume, and spreads above the configured ceiling.

ModelExecution

Use visible depth

Paper entries and exits are modeled against current L2 depth with explicit fee and residual-slippage assumptions.

AuditSQLite

Keep a durable ledger

Cash, positions, stops, signals, and simulated fills survive restarts and remain inspectable after a run.

02

Open to independent verification

Useful contributions challenge assumptions, add regression evidence, or make simulation limits clearer. They do not add live execution.

Public source · Python · Docker

Inspect the implementation and safety boundary

The README documents data sources, assumptions, risk limits, test commands, and the explicit separation from live trading.

SOURCE ↗