Jev in production › Agent tool and action selection

trade-jev

Backtests Jev as a BUY, SELL or HOLD trading decision maker against 15 days of real NQ order book data from Databento (justinhe16).

Open on GitHub ↗

15measured, as published by the source
Use
Agent tool and action selection
Industry
Finance
Form
Write-up
Stage
Announced
Listed
2026-10-04
Found via
github
Repository
justinhe16/trade-jev
Stars
12
Forks
3
Last push
2026-10-09
Language
Python
License
MIT

The README opens with

Tests Jev (TypeSafe) as a BUY / SELL / HOLD trader on NQ L10 order-book data from Databento (15 trading days, Jun 8–26 2026). Findings: results/FINDINGS.md. Published Jev answers and results are in results/; where the market data goes and how it was pulled is in data/DATA.md.

Concepts - Run: a backtest that calls Jev and stores its answers and trades in runs/ /. It's the only step that costs money. - Settings: harness parameters: cutoff, agreeing answers, min hold, stop, target, commission. - Replay: a run's stored answers re-scored under new settings. It needs no API calls. - View: one local web app that replays any mix of runs, days and settings live. - Live: the same harness on a live Databento feed: Jev signals, paper trades and a live page....

Badge

For the project's own README, linking back here:

Listed in Jev in production

Also used for agent tool and action selection