phil

Phil is a self-improving trader: an AI agent that trades short-term prediction markets and rewrites his own strategy after every settled bet. Paper 24/7 in the

What is phil?

Phil is a self-improving trader: an AI agent that trades short-term prediction markets and rewrites his own strategy after every settled bet. Paper 24/7 in the cloud; small capped real stakes via Pearl Connect.

Overview of phil

What it is

Phil is a self-improving trader: an AI agent that trades short-term prediction markets and rewrites its own strategy after every settled bet. Paper trading runs 24/7 in the cloud as the learning engine. Real money runs alongside it, deliberately small, with capped stakes through Pearl Connect, only in edge classes whose settled evidence has earned it. The name honors the man who relived the same day until he'd learned enough to win it, and the groundhog who makes forecasts.

Key features

  • Hourly cycle agent that screens, researches, bets, and self-edits, plus lightweight watchers that check every ~15 minutes for new listings and price moves on the agent's watchlist and trigger an early cycle when something fires.
  • Daily deep-retro agent that audits every strategy edit (keep, sharpen, or revert), grades the day's biggest estimation errors, and adjudicates the cycle agent's proposals.
  • Three layers of honesty: hourly cycle agent, daily deep-retro agent, and the human operator who owns the protected engine (core/, the caps, the cycle procedure).
  • The honest metric is brier_delta: whether the agent's probability is a better forecast than the market's own price.
  • Paper learning loop cycles hourly across the whole market universe; real execution rides on top, deliberately small, with per-bet, per-day and open-position caps configurable in config/protected.json (currently $1 per bet).
  • Real orders go through core/real.py, the only code that touches funds; the agent never holds keys; every signature goes through Pearl Connect's audited local choke point.

Use cases

Short-term prediction market trading with automated strategy improvement. Suitable for developers or researchers who want to run an AI agent that learns from paper trading and optionally executes small real stakes via Pearl Connect. The agent is designed for markets like earnings beats, daily sports, pre-match esports, official economic prints, and near-term news, excluding crypto up-or-down coin-flips.

Getting started

bash git clone https://github.com/bennyjo/phil.git cd phil python3 -m unittest discover -s core -p 'test_*.py' # offline tests python3 core/validate.py # integrity tripwires python3 core/score.py # calibration and P&L report

License and pricing

Licensed under Apache-2.0. Open source.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/bennyjophil.svg)](https://yunodb.com/tools/bennyjophil?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/bennyjophil?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/bennyjophil.svg" alt="Featured on YunoDB" height="20"></a>