shipwithjev

Catalog / Tools & apps

0287GitHub

jev-trade

Hyperliquid trading desk where Jev answers Choice questions for long/short, open/close/hold, and leverage; application code quotes or sends no order.

Hyperliquid trading desk where Jev answers Choice questions for long/short, open/close/hold, and leverage; application code quotes or sends no order. Defaults to a dry run; a live key can place real orders.

aowang-ai/jev-tradeREADME ↗
# Jev Trade

[](https://www.jev-trade.com/)
[](LICENSE)

I built a trading bot with Jev. Jev reads the Hyperliquid book every tick and answers buy, sell, or hold. The bot sends the order. Five coins, five wallets, real fills.

**[Watch the live desk](https://www.jev-trade.com/)**

[](https://www.jev-trade.com/)

Jev makes the call. Hold is one of its answers, so a tick can end with no order. Position, balance, and PnL come from Hyperliquid.

Based on [jev-trader](https://github.com/jarrodwatts/jev-trader) by Jarrod Watts (MIT). Venue is Hyperliquid, not Monad / Kuru.

## What you are looking at

- Five isolated sleeves: BTC, ETH, SOL, DOGE, BNB. Each has its own wallet and its own Jev call.
- The left pane is the live book as candles. Green and red marks are fills. The entry line is the open position.
- The right pane is the latest Jev call and the tape of every tick.
- The table is Positions and Trades, the same split a futures desk uses.

A live key on testnet or mainnet sends real orders. Start with a dry run.

## How a tick works

1. The bot reads the book.
2. Jev picks long or short, then open, close, or hold.
3. An entry is a post-only Alo quote one tick inside the touch, so it sits on the maker side until a taker hits it.
4. An exit is an Ioc that crosses the touch and fills on the spot.
5. Hold sends no order and pulls any resting quote Jev no longer wants.

The bot is Bun on port 3000. The dashboard is Next in `web/` on port 3001. Keys, evaluate, and orders stay on the Bun process.

## Dry run

[Bun](https://bun.sh) 1.2 or newer. No `PRIVATE_KEY` means a dry run: real book, real decisions, simulated fills. Default `MODEL=mock` is a momentum stand-in and needs no API key.

```sh
cp .env.example .env
bun install
bun run start
```

Dashboard (second terminal):

``

Also filed under Tools & apps

  1. 0616

    Jevstiller: distill Jev calls into a local model

    A self-hosted proxy that trains a local model on Jev's answers. It served 72% of Banking77 locally at 15 ms, with a 2% audit and a bound on disagreement.

    tomerglick57 · Tools & apps · 15 ms local

  2. 0611

    Scruple: code checks for what linters miss

    Semantic-aware linting with System One models: rules you can state in a sentence, checked against your code by Jev.

    @nalexpeae · Tools & apps

  3. 0606

    Algolia-style Postgres search with a Jev reranker

    Typo tolerance, typeahead, barcodes and facet counts in plain Postgres SQL, plus an optional Jev step that asks one question per top-10 result.

    @alex_forman1 · Tools & apps

  4. 0604

    Virtual Closet: outfits from closet videos

    built a virtual closet & outfit generator from raw closet videos. classification + semantic search thanks to jev 🥰. clueless closet dreams finally coming to life

    @maitreemervana · Tools & apps