Review Jupyter notebooks for leakage, statistical missteps, and reproducibility issues.
npx skills add https://github.com/tim-krausz/mlstack --skill review-notebook
Please help me install this Agent Skill. Skill: review-notebook Source: https://github.com/tim-krausz/mlstack/tree/main/review-notebook Command: npx skills add https://github.com/tim-krausz/mlstack --skill review-notebook
This Skill helps teams conduct rigorous, pre-submission notebook reviews to identify methodological flaws, leakage, and narrative gaps, ensuring reproducibility and trust in results.
Review notebooks in the repository with the /review-notebook workflow and follow the steps in the checklist.
High-intent search queries and answers about installing and using this skill.
You can review notebooks for data leakage and statistical validity by running a structured two-pass assessment workflow that identifies methodological missteps, flags p-hacking risks, and generates actionable remediation notes for authors.
Notebook reproducibility review is a pre-submission audit validating analyses for methodological flaws, data leakage, and statistical missteps. You need it when finalizing data science workflows to ensure results are trustworthy and reproducible.
Review a notebook for p-hacking and reproducibility by applying a two-pass assessment: the first pass isolates critical methodological risks like data leakage, and the second generates structured, actionable improvement recommendations for the author.
Yes, this notebook review workflow applies to Jupyter notebooks across data science workflows, enabling reviewers to audit methodological quality continuously from early drafting stages through to final submission.
The best way to audit methodological quality in data science notebooks is a two-pass review workflow that distinguishes critical methodological risks from informational improvements, providing structured outputs for actionable remediation.
Your notebook analysis may fail reproducibility checks due to undetected data leakage, statistical missteps like p-hacking, or narrative gaps. A systematic review workflow identifies these methodological flaws and guides remediation.
gizixGenerate structured Jupyter notebooks for EDA, modeling, evaluation, and reporting.
tim-krauszReview biostatistics study plans for methodological assumptions and planned analyses.
mgaldinoAudit R code for methodological flaws in econometric research.
alo-expValidate SPEC.md, REQUIREMENTS.md, ROADMAP.md, and DESIGN.md for cross-artifact consistency.
