What problem does it solve?
This Skill eliminates fragile, non-reproducible Stata analysis by guiding you through a complete end-to-end workflow with consistent structure, logging, and publication-ready outputs.
Core Features & Use Cases
- End-to-end Stata pipeline: loads raw data, performs EDA, constructs an analysis sample, estimates models, and exports tables and figures in a predictable directory layout.
- Reproducibility with guardrails: enforces Stata coding conventions and econometric best practices, including controlled randomness, standardized paths, and log-based validation.
- Publication-ready exports: generates tables to output/tables as .tex and .csv (via esttab) and figures to output/figures as .pdf and .png (via graph export), with review checkpoints (validate-log and stata-reviewer).
Example use case: You have a CPS-style dataset and want a fully logged workflow that loads data, explores key variables, builds an estimation sample, estimates OLS and IV specifications (or FE models as appropriate), and produces consistent regression tables and charts for a Quarto report.
Quick Start
Tell Codex to use the data-analysis skill to load your dataset from data/raw, run the full Stata workflow (EDA → construct sample → estimate models → export tables and figures) with logs and output artifacts saved under output/.