Skip to content
Metatext Decide — typed decisions for your agents, with a probability on every answerGet early access
Metatext
Sign Up

jev-trader

Experiment that asks Jev buy or sell every Monad block on Kuru's MON-USDC book and posts post-only limit orders one tick inside the touch.

AppprototypeMIT
View on GitHub ↗Live demo ↗

Overview

jev-trader makes one decision every Monad block, about every 300 ms. A TypeSafe Jev model watches the Kuru MON-USDC order book and answers buy or sell about the move over the next 100 blocks (about 30 seconds); every block then posts a real post-only limit order on that side one tick inside the touch, replacing the previous one, so fills earn the spread instead of paying it. Hold appears only when the model misses the block's deadline and nothing is posted.

A small server streams each block to a dashboard via server-sent events, including mid price, spread, decision probabilities, latency, the quote, fills, position and running totals such as Jev cost and gas. A logged example shows 81 ms decision latency and $0.000004 of Jev cost. Without a private key it dry-runs with real book data, real decisions and simulated fills, and the default model is a momentum mock. A deployed dry-run endpoint is available. The README states it is not financial advice. It is the base that jev-trade later moved to Hyperliquid.

What the decision model decides

TypeQuestionAnswers
ChoiceWill MON-USDC move up or down over the next 100 blocks (buy or sell)?buy · sell

How it works

  1. Read the Kuru order book each block
  2. Jev answers buy or sell for the horizon
  3. Post a post-only limit order inside the touch, cancelling the previous one
  4. Stream block events; mark hold when the decision is late

Strengths

  • Per-block decisions within the block budget (81 ms in the logged example)
  • Earns the spread via maker orders
  • Dry-run mode and SSE event stream

Caveats

  • Can place real orders when configured
  • Experimental; no performance claims
  • Position caps and gas costs apply

Getting started

cp .env.example .env && bun install && bun run start (set MODEL=jev and TYPESAFE_AI_API_KEY for Jev).

At a glance

Type
App
Status
prototype
Language
TypeScript
Runtime
Bun server + dashboard
Has code
yes
License
MIT

Use cases and integrations

How curated lists describe it

  • Trading: watches the Kuru MON-USDC book on Monad and asks Jev for a `Choice` between buy and sell every block, publishing 81 ms decision latency and $0.000004 of Jev cost per call from a live dry run.

    yibie/awesome-jev

  • One buy/sell decision per Monad block on Kuru's MON-USDC book. Live demo: [jev-trader.vercel.app](https://jev-trader.vercel.app/).

    AnotiaWang/awesome-decision-models

  • A market-making experiment on Kuru’s MON-USDC book on Monad, with optional per-block Jev buy/sell decisions.

    logicrw/awesome-jev-projects

  • One AI trade decision every Monad block. Jev on Kuru MON-USDC.

    wh000wh000/awesome-jev-live

  • Kuru/Monad trading experiment with optional Jev buy/sell decisions; defaults to a mock model and dry-runs without a private key, but configured execution can place real orders.

    cobanov/awesome-jev

  • One AI trade decision every Monad block. Jev on Kuru MON-USDC.

    hellogumbo/awesome-jev

  • Asks Jev buy or sell on every Monad block and places real post-only limit orders on the Kuru MON-USDC book. Not financial advice.

    MrJev/awesome-jev

Frequently asked questions

How often does it decide?

Every Monad block, roughly every 300 ms.

What does a decision cost?

A logged example shows $0.000004 of Jev cost and 81 ms latency.

Does it trade with real funds?

Only when a private key is set; otherwise it dry-runs with simulated fills.

Related projects

  • jev-A-share-trader

    Local China A-share technical analysis workbench where Jev assesses eight indicator dimensions and returns Buy, Hold or Sell with evidence.

    ProjectbetaPython, TypeScript
  • jev-trading

    Local stock decision service that sends market data and news to Jev or a local model and returns a policy-checked buy/sell/hold.

    ProjectbetaTypeScript
  • ml-quant-trading

    PyTorch factor-research stack with 213 factors and cost-aware backtests, plus a preview typed portfolio-decision adapter.

    ProjectbetaPython
  • QuantStudio

    Chinese open-source research and trading workbench whose JEV monitor watches futures contracts and judges entries and exits.

    ProjectbetaTypeScript
  • traderai

    Self-hosted trading analysis platform pairing Claude briefings with Jev probabilities for signals like thesis-breaking headlines.

    Servicebeta
  • Jev-Trades★ 51

    Next.js crypto dashboard where Jev judges each tick across timeframes and drives a SQLite paper-trading portfolio with automatic TP/SL.

    ServiceprototypeTypeScript