I have published the documentation of a reproducible review pipeline for submitted cheat-sheet drafts in my introductory statistics course. The documentation explains how heterogeneous student uploads are standardized, how admissible content pages and page usage are measured, how similar submissions are flagged, and how topic coverage is estimated with a vision-capable language model. The analysis is reported only in aggregate and is designed as a structured preparation for manual review, not as an automatic decision about individual students.