Calculates statistics from data and adds them to a stats-for-text CSV file for inclusion in the paper. Follows the project's name-value pair convention.
Checks figure-generating code against exhibit guidelines and, by default, applies presentation-only fixes. Supports two modes: fix mode (default) and audit-only mode when the user asks to only review and flag issues.
Generates a quiz to test the user's knowledge of a codebase or script. Supports two modes: script-specific (deep dive on one script, invoke with a script path) and whole-project (architecture, design choices, pipeline, invoke with `project`). Reads actual source files, git history, and directory structure before generating questions.
Adds practical, lightweight, high-value unit tests to Python and R data science scripts. Treats testing as data validation. Use when asked to add unit tests, validate data transformations, or audit code correctness in scripting workflows.