What problem does it solve? Analysis code tends to drift toward over-engineering (speculative frameworks, premature abstraction, unrequested configurability) or toward unreadable one-liners, and small requested tweaks often balloon into silent rewrites that break trusted results. This Skill keeps empirical analysis code minimal, legible for referees and replication reviewers, and surgically edited so every changed line traces to the request. ## Core Features & Use Cases - Simplicity discipline: Cuts speculative pipelines, premature abstractions, and unrequested configurability in favor of idiomatic dplyr, pandas, or DataFrames.jl code. - Legibility conventions: Enforces naming in economic units, one conceptual step per line, and # why: comments at every analytical decision such as winsorize thresholds, sample restrictions, and cluster levels. - Surgical change rules: Keeps diffs minimal, matches existing file style, and forbids refactoring working code that was only asked to be tweaked. - Use Case: When asked to add a column to a coauthor's R notebook, produce a two-line diff matching the existing tidyverse style with a # why: comment, instead of restructuring the pipeline. ## Quick Start Ask the assistant to add a new variable to your existing analysis script while keeping the diff minimal and annotating the reasoning behind the change.
