code-data-analysis-scaffolds_skill

Overview

This skill provides ready-made scaffolds for technical work in software engineering and data science. It accelerates disciplined starts: writing tests first (TDD), planning exploratory data analysis (EDA), designing statistical or causal studies, and validating results before execution. Use it to turn vague tasks into concrete, ordered checklists with clear validation points.

How this skill works

When invoked, the skill asks clarifying questions about goals, data, constraints, and success criteria, then selects the appropriate scaffold type (TDD, EDA, statistical analysis, causal inference, predictive modeling, or validation). It generates a step-by-step framework with expected outputs, validation checkpoints, and explicit assumptions. The scaffold includes example tests, EDA steps, analysis designs, or validation checklists that the user can follow or hand off for execution.

When to use it

• You need to write tests before implementation or when refactoring code
• You receive a new dataset and must plan a rigorous EDA
• You must design an A/B test, hypothesis test, or causal study
• You are starting predictive model work and need a clear pipeline and validation plan
• You need a comprehensive validation checklist before shipping code or models

Best practices

• Clarify objectives and constraints before generating a scaffold
• Make assumptions explicit and list required data or fixtures up front
• Start with simple baselines and add complexity only when justified
• Include validation checkpoints and success criteria at each step
• Surface missing data, unclear requirements, or risks early
• Prefer small, actionable steps: tests first, then implementation or analysis

Example use cases

• Create a TDD test suite for a new authentication endpoint (happy path, edge cases, error handling)
• Produce an EDA plan for a customer dataset: schema checks, missingness, distributions, key bivariate relationships
• Design an A/B test: hypotheses, sample size/power calculation, randomization checks, analysis plan
• Outline a predictive modeling pipeline: data prep, baseline model, cross-validation, metrics, error analysis
• Prepare a validation checklist for production ML: data validation, model performance, integration tests, monitoring