Handles the full academic statistical workflow from test selection through APA-formatted reporting. You get guided help choosing between t-tests, ANOVA, chi-square, regression, and their non-parametric alternatives based on your data characteristics. The built-in assumption checking module runs Shapiro-Wilk, Levene's test, and outlier detection with diagnostic plots before you commit to a test. Uses pingouin for clean effect sizes and confidence intervals, with statsmodels and PyMC backing more complex models. The documentation is unusually thorough with decision trees and compatibility notes for the 2025-2026 scipy/pingouin ecosystem changes. Best for research contexts where you need defensible test choices and proper reporting, not for production model deployment.
npx skills add https://github.com/k-dense-ai/scientific-agent-skills --skill statistical-analysis