Back

hyperclaude

Hyperliquid trading via Telegram, paused

A perpetuals trading system driven over Telegram. It pulls live data from eight exchanges, runs a multi timeframe signal engine and a statistical factor engine, and returns a trade proposal. Risk warnings are attached to the proposal but do not block it. Execution waits for a one word confirmation. A position monitor sends alerts on PnL thresholds and liquidation proximity. Every decision is logged to Postgres for backtest calibration.

01Two pass flow. The first pass proposes, the second executes after confirmation.
02Eight exchange data pipeline across Binance, Bybit, OKX, Bitget, Hyperliquid, Deribit, FRED, and Whale Alert, with circuit breakers and quality scoring.
03Statistical factor engine: funding carry, basis, VPIN, open interest composite, volatility regime, liquidation clustering.
04Event driven backtester with walk forward validation and overfit detection.
05Strict release pipeline. Test, commit, push, tag, hash verified deploy. No rsync, no manual copies.
06Roughly 51,000 lines of Python and 1,771 tests passing at v2.24.0. Not currently running.
Python,MCP,Postgres,Docker,FastAPI