Data Analysis | Chalie Skills

What it does

Analyse a dataset, calculate statistics, run comparisons, and produce charts or summary tables.

When this skill is activated, Chalie follows these steps:

Use recall to surface any prior analysis on related data, then ask the user to describe the dataset (paste it, describe its structure, or state where it lives) and what question they want answered.
Use recall to surface any relevant prior context — units, domain conventions, or previous analyses on related data.
Use recall to check for any stored domain conventions, then clarify the analysis type: descriptive statistics, correlation, trend analysis, comparison across groups, ranking, or forecasting.
Use code_agent to load, clean, and validate the data — report any missing values, outliers, or type errors before proceeding.
Use code_agent to run the requested analysis: compute statistics, group comparisons, correlations, or growth rates with appropriate precision.
Use code_agent to generate a chart or summary table that makes the key finding visually clear — prefer a simple bar, line, or scatter as appropriate.
Use write_file to record the plain-language interpretation: the main finding, what drives it, and any caveats (sample size, data quality, assumptions).
Use write_file to save the analysis results and code so the user can reproduce or extend it later.

v1 (curated)