cd /news/ai-agents/jev-typesafe-system-one-finance-trad… · home topics ai-agents article
[ARTICLE · art-134925] src=gist.github.com ↗ pub= topic=ai-agents verified=true sentiment=↑ positive

Jev (TypeSafe System One) finance & trading projects — surveyed 2026-09-20

A survey of GitHub and community awesome-lists conducted on 2026-09-20 catalogued finance and trading projects built on Jev, TypeSafe AI's System One decision model released five days earlier. The reference implementation, jarrodwatts/jev-trader, makes one AI trade decision per Monad block (~300 ms) by reading the Kuru MON-USDC order book and posting post-only limit orders, and has been ported to Hyperliquid with five isolated sleeves and live fills. An IRS tax-document page classifier built on Jev reported 100% strict accuracy across 261 forms at roughly $0.001 per page, 34× cheaper and 6× faster than the team's prior LLM pipeline, while the survey noted a consistent pattern that "Jev judges, code executes.

by read1 min views1 publishedSep 20, 2026

Jev (TypeSafe System One) — Finance & Trading Projects

Projects using Jev, TypeSafe AI's System One decision model (released 2026-09-15), in investment, trading, and financial-data contexts. Surveyed 2026-09-20 via GitHub API and community awesome-lists.

Reference project

jarrodwatts/jev-trader (★1.3k, 2026-09-16) — One AI trade decision every Monad block (~300 ms). Jev reads the Kuru MON-USDC order book and answers buy or sell; the bot posts a post-only limit order one tick inside the touch, earning the spread. Bun/TypeScript, dry-run mode, SSE dashboard. The template most projects below derive from.

jev-trader ported to Hyperliquid. Five isolated sleeves (BTC/ETH/SOL/DOGE/BNB), five wallets, real fills. Jev answers long/short then open/close/hold via Choice. Live desk

Open-source AI Trading OS (crypto, stocks, forex). Added Jev post-release as a "pre-trade decision gate" in front of the LLM gate. Largest Jev integration by stars

Hong Kong stock direction forecasting. AKShare fetches target + HSI + HSTECH; Jev answers up/flat/down as Choice; standalone HTML reports. Careful treatment of probability calibration

IRS tax-document page classifier: 261 forms, 100% strict accuracy on their corpus, ~$0.001/page (34× cheaper, 6× faster than their prior LLM pipeline)

Empty / placeholder repos

maxlibin/moomoo-jev-trader (Moomoo dashboard, intended), Pastorkid/jevTradingBoth, hifizz/jev-finance-benchmark — created but no content yet.

Observed patterns

"Jev judges, code executes" is universal. Every project gives Jev only typed judgments (Choice/Noul/Score) over compact state; thresholds, risk vetoes, and order placement stay in deterministic code. The buberlo redesign makes this separation most explicit.

── more in #ai-agents 4 stories · sorted by recency
── more on @jev 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/jev-typesafe-system-…] indexed:0 read:1min 2026-09-20 ·