jev-paper-trader
Paper-trades real Polymarket and Kalshi markets with a fake $100k, using Jev as judge fed by Claude research, against the market's price.
Overview
Jev Paper Trader paper-trades real Polymarket and Kalshi prediction markets with a fake $100,000 bankroll, reading only public prices and placing no real orders. It asks whether Jev, fed research by Claude, can grow the bankroll by more than luck, measured by profit and loss and by whether its forecasts beat the market price on Brier score. Success and stop criteria were pre-registered.
Each hour a funnel narrows markets with free filters (price, volume, closing within 30 days, staleness), Jev judges each market without research, and only events whose rules Jev rates clear are researched by Claude Opus 5.5 through Claude Code on a subscription plan. Bets are settled, resolved markets reviewed with post-mortems, and a dashboard and public page are published. The README reports frankly that after about two weeks every forecaster was less accurate than the market and strategies were losing, and documents a rebuild plan.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Are this market's resolution rules clear? | yes · no |
| Score | How likely is this market to resolve yes?(inferred) | — |
How it works
- Fetch the busiest markets and apply free filters
- Jev judges each market without research
- Claude researches events whose rules Jev rates clear
- Place paper bets, settle, review and publish a dashboard
Strengths
- Real market prices with no money at risk
- Pre-registered criteria and honest status reporting
- Research rationed to cut cost
Caveats
- Currently losing and less accurate than market prices
- Depends on Claude Code with a Claude plan for research
- Paper trading only
Getting started
python3 -m papertrade ping, then python3 -m papertrade cycle (set TYPESAFE_AI_API_KEY in .env).At a glance
- Type
- Project
- Status
- research
- Language
- Python
- Runtime
- CLI with dashboard
- Has code
- yes
Use cases and integrations
How curated lists describe it
Paper trading real Polymarket and Kalshi markets with a fake $100k: can Jev +…
wh000wh000/awesome-jev-livePaper trading real Polymarket and Kalshi markets with a fake $100k: can Jev + Claude research beat the market?
hellogumbo/awesome-jev
Frequently asked questions
Does it trade real money?
No. It uses a fake $100,000 bankroll and only reads public prices.
Is it profitable?
As of 9 October 2026 the README reports losses and forecasts less accurate than the market.
What role does Claude play?
Claude researches selected events and feeds that research to Jev.
Related projects
Local China A-share technical analysis workbench where Jev assesses eight indicator dimensions and returns Buy, Hold or Sell with evidence.
ProjectbetaPython, TypeScriptStock terminal that gathers live evidence via Valyu and asks Jev for the probability a stock trades higher in 30 days.
ProjectprototypeTypeScriptExperimental Polymarket bot: a web-search model writes an evidence brief, Jev prices P(YES), and code trades only on a configured edge.
ProjectprototypePythonTerminal agent for Polymarket 5-minute BTC markets that asks Jev up or down and buys once only above 0.90 confidence; dry-run default.
ProjectprototypeTypeScriptPolymarket BTC 5-minute up/down agent: code prices with a volatility model, Jev scores regime and vetoes anomalies; paper by default.
ProjectprototypePythonPaper-only research tool for Polymarket BTC 5-minute Up/Down markets: Jev scores the direction, separate code decides entries.
ProjectresearchTypeScript