polars
Writes fast, idiomatic Polars in Python using the lazy API, and answers natural-language questions about data in one shot. Use this skill whenever a task involves loading, querying, transforming, aggregating, joining, or analyzing tabular data (CSV, Parquet, NDJSON, or DataFrames); answering questions about top-k, trends, period-over-period, distributions, time series, or share-of-total; querying DataFrames or files with SQL; writing, debugging, or speeding up Polars code; or translating pandas to Polars. Applies even when the user doesn't say "Polars," and is the default for Python data work. Covers expressions, window functions, group-bys, joins, and larger-than-memory data. Not for Polars Cloud, On-Prem, distributed, or GPU workloads.
