What problem does it solve? Clinical data staff without R programming experience struggle to produce the Tables, Figures, and Listings (TFL/TLG) required for clinical study reports. This Skill turns natural-language descriptions of a target output into runnable R code built on the pharmaverse TLG stack (tern + rtables), using CDISC ADaM datasets as input. ## Core Features & Use Cases - Table generation: Produces demographic tables, adverse event summary tables (with correct alt_counts_df denominators from ADSL), toxicity-grade tables, and lab shift tables via tern functions like analyze_vars and count_occurrences. - Figure generation: Creates Kaplan-Meier survival curves with g_km, mean line plots with g_lineplot, and waterfall plots, exported via ggsave as ggplot objects. - Reference library: Ships references covering the layout-vs-data mental model, a five-step table skeleton, a tern function cheat sheet, Word (docx) export via rtables.officer/flextable, and navigation to the official TLG Catalog. - Use Case: A clinical data manager asks for an AE summary table by system organ class and preferred term; the Skill generates code that filters ADAE, aligns factor levels with ADSL, builds the layout, and passes alt_counts_df = adsl so percentages use the correct denominator. ## Quick Start Ask the assistant to generate a demographic characteristics table by treatment arm from the pharmaverseadam ADSL dataset using tern and rtables.
