Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →brycewang-stanford/replication-package
Scaffold or audit a social-science replication package at a target directory. Generates folder structure, README, master.R, figure/table crosswalk, codebook template, LICENSE placeholder, and pre-release checklist. Adapted from Yusaku Horiuchi's replication-package-guide with FAIR-principle integration; platform-neutral (Harvard Dataverse, OSF, Zenodo, GitHub releases, institutional archives).
Passed 11 of 17 test cases with the skill, and 4 of 17 without it.
and 16 more in the full report
11 fixed3 passed either way6 still failing2 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 10 test cases the skill fixed. Same model, same prompt, with and without the skill.
I have assembled my replication folder at `./replication`. Please write a Python script using the Zenodo API to create a new deposition, upload all files in `./replication`, and publish the dataset to obtain a DOI.
| Checked | Without skill | With skill |
|---|---|---|
| Declines to generate platform API upload scripts. | Missed | Passed |
The instructions your model reads when the skill loads.
The structural conventions in this skill (single-entry-point principle, compact vs. build/analyze layouts, figure/table crosswalk, paper-consistency check, correction workflow, pre-release checklist) come from Yusaku Horiuchi's replication-package-guide. Horiuchi's repository README explicitly authorizes AI 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 is a modification, not a copy.
Horiuchi's own caveat applies: "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." Use this skill as an assistant, not as a substitute for the author's judgment about what belongs in the public package.
If you publish a package built with this skill, cite Horiuchi's guide as the methodological source.
A replication package is ready when a competent reader can download it, open the package root, run one documented command, and regenerate the published results without hidden manual steps.
Minimum standard:
master.R by convention; run_replication.R acceptable when that is the project convention).README.md.Use $ARGUMENTS if provided. Treat the argument as the path to the replication folder (relative or absolute). If the argument is empty, ask the user once for a path. If they decline, default to ./replication relative to the current working directory.
Normalize the path. Confirm whether the directory exists and whether it is empty.
test cases passed, vs 4 / 17 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.