replication-packagelisted

# Replication Package Scaffold ## Heritage and attribution The structural conventions here — single entry point, compact vs. build/analyze layouts, the figure/table crosswalk, the paper-consistency check, the pre-release checklist — come from **Yusaku Horiuchi**'s [replication-package-guide](https://github.com/yhoriuchi/replication-package-guide), whose README explicitly authorizes agent consumption: it is "designed to be read by humans and by coding agents such as Codex or Claude Code before they prepare, audit, or repair a replication package." This skill modifies rather than copies that guide. The FAIR principles are folded in (Findable, Accessible, Interoperable, Reusable; [Wilkinson et al. 2016](https://doi.org/10.1038/sdata.2016.18); [GO FAIR](https://www.go-fair.org/fair-principles/)) so the scaffolded package is platform-neutral, and platform-specific upload mechanics are dropped: build and audit the local package, generate no upload code. Uploading to Harvard Dataverse, OSF, Zenodo, a journal repository, or an institutional archive belongs to the user and the platform's own tools. Horiuchi's caveat carries over: "AI is useful for checking, reorganizing, documenting, and catching inconsistencies, but it should not be treated as a substitute for the author's judgment about which files, scripts, data sources, and results are actually part of the replication record." If the user publishes a package built with this skill, cite Horiuchi's guide as the methodological so