What problem does it solve? Base-editor variant-function screens produce amplicon sequencing data where bystander edits, indel byproducts, and low-efficiency sgRNAs confound interpretation. This Skill turns raw CRISPResso2 output into per-variant fitness scores with correct bystander attribution and editing-efficiency filtering. ## Core Features & Use Cases - Library design and chemistry selection: Tile NGG-adjacent spacers so target bases fall in editing positions 4-8, choose between CBE (BE3/BE4max/eA3A-BE3) and ABE (ABE7.10/ABE8.20/ABE8e), and annotate target plus bystander variants per sgRNA. - Editing-efficiency filtering and hit calling: Parse CRISPResso2 quantification tables, drop sgRNAs below 30%/50% editing thresholds, then score variants with MAGeCK MLE or drugZ and aggregate to per-variant fitness. - Bystander deconvolution and diagnostics: Partition allele tables by edit pattern, compute substitution-vs-indel ratios to detect Cas9 contamination, and cross-validate with prime editing. - Use Case: Reproduce the Hanna 2021 BRCA1/2 ClinVar-scale CBE screen workflow: design the library, filter by pilot editing efficiency, call PARPi-sensitivity variants with drugZ, and annotate hits against ClinVar and COSMIC. ## Quick Start Ask the agent to design a CBE saturation library tiling BRCA1 with bystander annotation, then filter pilot CRISPResso2 results to sgRNAs above 30% editing and score per-variant fitness with MAGeCK.