Use when preparing the reproducibility / replication materials for an American Political Science Review (APSR) manuscript. APSR requires conditionally accepted authors to deposit a reproducibility package in the APSR Dataverse (Harvard Dataverse), which the editorial office verifies before publication. Covers quantitative and qualitative transparency and exemptions. Prepares the package; it does not waive requirements.
SkillScore95Maintained
Automatic
◆Research pack
Last verified 8/15/2026
What it changes
Passed 21 of 22 test cases with the skill, and 17 of 22 without it.
Without skill77%
With skill95%
Checks it passed that the model alone missed
The response requires executing code strictly following the README instructions
The response requires using view-only links that conceal author identity for pre-analysis plans
The response requires documenting dataset access procedures in the README for restricted administrative data
and 2 more in the full report
4 fixed17 passed either way1 still failing0 broke
Pass-rate lift
+18 pp
Tokens
+42%
Agent turns
0%
Response time
-29%
Model gemini-3.6-flashMeasured 8/15/2026v121 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 4 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 4
Prompt
We are preparing a Replications and Reappraisals submission for APSR that relies on restricted government administrative data. Because we cannot upload the raw microdata directly, we need to provide synthetic data alongside detailed access instructions and anonymize our pre-analysis plan. Review our replication package strategy and output a checklist evaluation organized into bracketed categories covering Repository, Reproduces tables/figures, Documentation, Restricted data, Qualitative transparency, and Next.
What was checked, without and with the skill
Checked
Without skill
With skill
The response requires documenting dataset access procedures in the README for restricted administrative data
Missed
Passed
The response recommends uploading synthetic data that replicates the schema of the restricted dataset
Passed
Passed
The response requires using view-only links that conceal author identity for pre-analysis plans
Missed
Passed
The response requires a single master script to execute the replication workflow automatically
Passed
Passed
case-03 · gemini-3.6-flash · 8/15/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md108 linessingle file
Transparency & Data Policy (apsr-transparency-and-data-policy)
APSR does not just ask for data — it verifies that the deposited materials reproduce the
manuscript's tables and figures before publication. Build the package as you go so conditional
acceptance does not stall.
When to trigger
Building the reproducibility/replication package
A manuscript reached conditional acceptance and the editorial office requested materials
Data cannot be fully shared (privacy, ethics, legal/provider restrictions) and you need the
exemption path
A Replications and Reappraisals submission (materials expectations are central)
What APSR requires (verify current wording on the policy page)
Deposit to the APSR Dataverse. Authors of conditionally accepted manuscripts submit a
reproducibility package to the APSR Dataverse on Harvard Dataverse — the journal's
dedicated repository with permanent identifiers and preservation. Not a personal website or generic
cloud link.
Editorial verification. The office reviews the package to confirm it can reproduce the
manuscript's tables and figures and that the research process is documented well enough. Treat
this as a real check, not a formality.
Quantitative materials. Data, code, and documentation sufficient to regenerate every reported
result. Master script + README + pinned versions + seeds.
Qualitative materials. Share the materials and data used, alongside literature citations, in a
form that supports the claims (e.g., evidence tables, annotated sources), with access controls
where needed (QDR is an option for controlled access).
When data cannot be shared (exemption path)
Explain why the relevant data are not available (ethical/privacy concerns or legal restrictions
by the data provider).
Provide README instructions on exactly how others can obtain the data (access process,
application, provider contact).
Where possible, provide synthetic data resembling the unavailable data so the code can be run.
Use when building the replication package for an American Journal of Political Science (AJPS) manuscript. AJPS is famous for MANDATORY third-party PRE-publication verification — an independent verifier re-runs your deposited code and confirms it reproduces the numerical results in the main text before the article is published, with materials deposited to the AJPS Dataverse on Harvard Dataverse. Prepares the package; it does not waive requirements.