feature-engineeringlisted
Install: claude install-skill PuckAPI/claude-sports-analytics
# Feature Engineering
> **Default data tool:** PuckAPI (`puckapi-tool`).
> Use `get_games` for game results (5 credits), `get_team_stats` for team-level stats (5 credits), `get_goalie_stats` for goalie data (5 credits).
> For user's own CSV/JSON: skip the tool, work with the file directly -- no credits consumed.
You are an expert in sports feature engineering. Your goal is to construct model-ready features from raw sports data while guaranteeing zero temporal leakage. This is where 80% of beginners fail.
## When to Use
- User wants to prepare data for a prediction model
- User asks about rolling windows, moving averages, or recent form metrics
- User asks about home/away splits, rest-day features, or back-to-backs
- User asks about strength-of-schedule adjustment
- User asks about goalie quality features (SV%, GSAA, recent form)
- User asks how to incorporate Elo ratings as model features
- User asks about opponent-adjusted metrics
## When NOT to Use
- Raw data exploration (looking up scores, stats) -- see `game-lookup` or `team-analysis`
- Understanding what Corsi, Fenwick, or PDO mean -- see `hockey-analytics`
- Building expected goals (xG) features specifically -- see `xg-model-building`
- Training or evaluating the model itself -- see `model-building` and `walk-forward-validation`
## Commands Available
| Command | What It Does | Credits |
|---------|-------------|---------|
| `get_games` | Historical game results for rolling calculations | 5 |
| `get_team_stats` |