Repository navigation
Versions and Upgrading
The first public release (release notes). The methodology inside it is version 2026-10-01.1, the seventh revision since the skill was first written privately on 2026-08-31:
| Methodology | Date | What it added |
|---|---|---|
| v1 | 2026-08-31 | baseline: CONSORT 2025, ROBINS-I v2, GRADE, paper-mill screening |
| v2 | 2026-09-02 | qualitative studies, recycled samples, retracted sources the paper leans on |
| v3 | 2026-09-15 | composite exposures, regression papers, attrition, omission as spin, publication-status drift |
| v4 | 2026-09-16 | uncontrolled before-and-after studies and the hold-out outcome; the article's recency claim |
| v5 | 2026-09-28 | umbrella reviews; sibling comparisons and negative-control exposures; "not linked" as spin |
| v6 | 2026-09-30 | equivalence and non-inferiority margins; sponsored-trial comparators; Retraction Watch through Crossref |
| v7 | 2026-10-01 | coarse controls, ability tests as exposures, single-item outcomes, which hop added the claim |
If you used the skill before it was published, the folder that held methodology.md, evaluations.md and your appraisals can serve as your SSE_HOME. The new skill reads evaluations.md and writes new reviews to appraisals/, and it takes the checklist from its own references/methodology.md. The script's output kept its section names. What changed: the email comes from SSE_EMAIL or --email, PYTHONIOENCODING is no longer needed on Windows, and --fulltext, --json and PMID input are new.
Claude Code updates the plugin from the marketplace. An uploaded or copied skill is replaced by copying the new folder over it. Your SSE_HOME folder is never touched.
This wiki describes science-study-evaluator 0.1.0 and was last updated on 2026-10-11. The skill and script are MIT licensed. Report problems in the issues.
Using it
The releases
Contributing
Elsewhere