CharacterSheet is a collection of LoRA adapters built by Alissonerdx that transforms an existing character image into a clean, multi-view reference sheet while preserving identity, clothing, proportions, and overall appearance. The collection consists of three variants trained on approximately 300 character-sheet examples: TripleView_klein9b_v1.safetensors for FLUX.2 Klein 9B generating front, side, and back views; QuadView_klein9b_v1.safetensors for FLUX.2 Klein 9B generating a face close-up plus three body views; and QuadView_krea2_v1.safetensors for Krea 2 with the same four-view output. These are LoRAs, not standalone checkpoints, and require ComfyUI and a compatible base model. Human characters were the primary training focus, so results with stylized, non-human, or heavily stylized subjects will be less predictable. All models generate output at 1536×1024 resolution and operate at full LoRA strength of 1.0. Best use cases include character design turnarounds for animation or game development, character dataset preparation for training pipelines, concept art refinement, and integration with identity-preservation video generation. The model introduces anatomical inconsistencies between views, affecting facial features, clothing, and accessories.