Auto Trading Winnerv1.0.0

A solo quant deploys the skill on a cloud VM with RUN_MODE=auto and TRADING_VENUE=polymarket to continuously scan markets priced between 30% and 70%, rank by volume, and paper-trade the top candidates. It lets them gather performance data on the volume-ranking baseline before replacing it with a proprietary signal.

A small fund uses the skill with TRADING_VENUE=kalshi and manual selection mode to review the top 5 highest-volume markets each morning and hedge exposure to policy or macro events. Context safeguards and slippage caps prevent accidental fills during thin books.

A trading community operator runs the skill in manual mode over sim and kalshi, then posts the printed shortlist and AI reasoning for members each day. Members copy tickets manually while the operator monetizes through a paid subscription.

A retail developer remixes the volume-ranking default with fair-value and time-to-resolution filters, then runs unattended dry-runs across venues to compare edge and hit rates. The tagged reasoning outputs become training data for their own ML ranking model.

An arbitrageur configures MARKET_QUERY and tight MIN_PRICE/MAX_PRICE bands to detect mispricing between Polymarket and Kalshi, using SELECT_CANDIDATE for non-interactive execution once a candidate clears the MIN_EDGE threshold.

Operators publish daily shortlists and reasoning produced by the skill, charging monthly fees for access to candidate lists and execution guidance. The value comes from the ranking discipline and safety rails rather than the raw data.

A manager runs the skill in auto mode on client capital with a custom alpha layer replacing the default volume ranking. Clients pay a management fee plus a share of profits on executed edge.

The intermediate-level skill is packaged as a ready-to-remix template for prop desks or market makers who want the plumbing (discovery, indexing, sizing, safeguards) pre-built. Licensing is per-seat with paid remix support.