data-science

Enforce Polars-based data manipulation and XGBoost modeling pipelines with nested cross-validation.

2|Updated Oct 30, 2025

Search Tags:#data science #machine learning #feature engineering #polars #predictive modeling #xgboost

Ready to run

npx skills add https://github.com/zachmayer/skills --skill data-science-zachmayer

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: data-science
Source: https://github.com/zachmayer/skills/tree/main/.claude/skills/data-science
Command: npx skills add https://github.com/zachmayer/skills --skill data-science-zachmayer

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill eliminates inconsistent data science practices by enforcing standardized, high-performance defaults for modeling, validation, and feature engineering.

Core Features & Use Cases

Standardized Modeling: Implements XGBoost-first pipelines with nested cross-validation and empirical Bayes shrinkage for robust feature engineering.
Data Integrity: Enforces the use of Polars for efficient, type-safe data manipulation and prevents common pitfalls like target leakage or improper categorical encoding.
Use Case: When tasked with building a predictive model from CSV data, this skill ensures you use the correct validation strategy, feature importance methods, and library stack to produce reliable, reproducible results.

Quick Start

Use the data-science skill to perform an exploratory data analysis and build a predictive model on the provided customer churn dataset.

Frequently Asked Questions about data-science

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I build a predictive modeling pipeline that prevents target leakage?▼

Predictive modeling pipelines prevent target leakage by enforcing type-safe Polars data manipulation and rigorous nested cross-validation. This approach ensures correct categorical encoding and validation strategy for reliable, reproducible results.

What is nested cross-validation and when do I need it for machine learning?▼

Nested cross-validation is a rigorous validation strategy needed when building robust machine learning models to prevent overfitting. It is required alongside empirical Bayes shrinkage for accurate feature engineering and standardized XGBoost modeling pipelines.

How do I perform feature engineering with Polars for classification and regression tasks?▼

Perform feature engineering with Polars by applying type-safe data manipulation within XGBoost-centric pipelines. This standardized workflow supports classification and regression tasks while applying empirical Bayes shrinkage for robust feature importance.

Does this data science workflow support clustering or only predictive modeling?▼

This data science workflow supports clustering alongside predictive modeling tasks. It enforces XGBoost-centric pipelines with nested cross-validation and empirical Bayes shrinkage for classification, regression, and clustering.

Why use Polars instead of other data manipulation libraries for machine learning?▼

Polars provides efficient, type-safe data manipulation that prevents common machine learning pitfalls like improper categorical encoding. It standardizes high-performance workflows, ensuring data integrity throughout the feature engineering and modeling pipeline.

What's the best way to standardize XGBoost pipelines for reproducible results?▼

Standardize XGBoost pipelines by enforcing nested cross-validation and artifact-driven reporting. This approach applies empirical Bayes shrinkage for feature engineering and ensures consistent, high-performance defaults across modeling, validation, and reporting.