What problem does it solve? Academic papers that compile cleanly often still fail review because they lack recent literature, ablation evidence, statistical significance tests, or reproducibility statements. This Skill applies standardized improvement patterns to raise review scores (G1-G7), citation health (D10a), and reproducibility quality. ## Core Features & Use Cases - Frontier Literature Integration: Extracts 2024-2026 references from training pipeline analysis documents, adds them as BibTeX entries, and weaves them into the Discussion with a recent-advances/method-advantage/future-work structure. - Ablation & Statistical Rigor: Generates V1-V4 component-wise ablation tables with positive-delta enforcement, plus paired t-test, Wilcoxon signed-rank, Cohen's d, and 1000-resample bootstrap confidence intervals in scientific notation. - Citation Health Repair (D10a): Detects orphan bibitems and citations hidden in LaTeX comment lines, then re-anchors them semantically in body text until D10a reaches 100%. - Use Case: A medical imaging paper claims to significantly outperform baselines but has no statistical tests and stale references; the Skill adds the statistics section, supplements recent literature, fixes orphan citations, and inserts a Data/Code Availability statement with URLs and MIT license. ## Quick Start Improve my compilable LaTeX paper by adding recent 2024-2026 literature to the Discussion, an ablation study, statistical significance tests, and a data and code availability statement.
