Video unavailable. Open the source below.
NewJev Trader
A trading experiment connects rapid decisions to an on-chain order book.
Added to Jevfast
How it uses Jev
Jev receives a compact MON-USDC state containing order-book depth, recent returns and trades, and chooses whether price will rise or fall over a configured block horizon. The trader posts a post-only limit quote on that side, subject to position and margin limits; late decisions and fills are tracked separately.
What Jev decides
Illustrative direction request (not a recorded response)Example answers · not a recorded Jev response · Source ↗
Question 1 · direction
YOUR APP
INSTRUCTION
Will MON be higher or lower than the current mid after horizonBlocks more blocks? The code supplies buy and sell as the two bounded criteria.
STATE
The request state contains market, block, horizonBlocks, blockMs, mid, spreadBps, bookImbalance, depth, book, returnsBps, recentMids, trades, recentTrades, and allowed. These are request fields, not captured runtime values.
JEV · CHOICE
- buy
- sell
App workflow
- Read market state
Each block, the service reads the order book and updates recent mid-price and trade context.
- Choose direction
Jev selects buy or sell for the configured horizon from the compact market state.
- Check limits and quote
The trader checks exposure and available margin, then places or simulates a post-only limit order.
- Reconcile events
Receipts and trade logs update resting orders, fills, inventory, and P&L; late decisions are surfaced.
Why it is interesting
The project places a typed decision inside a tight block-by-block trading loop while keeping quoting, risk limits, receipts, and position accounting in application code.