What problem does it solve? Pixel-art characters often look stiff, boxy, uncanny, or generic because artists lack concrete pixel-level guidance on proportions, outlines, faces, and poses. This Skill supplies exact px-unit values and a structured workflow so AI-generated or hand-drawn sprites read clearly at small sizes. ## Core Features & Use Cases - Six-stage construction pipeline (C0-C5): silhouette, flat colors, shading, selective outline, highlights, and cleanup, with lint verification via bin/fge sprite. - Size-specific reference tables: head-to-body ratios, color budgets, eye and hand treatments for 16px, 24px, 32px, and 48px sprites, plus quadruped animal rules and clothing techniques. - Companion references: faces.md covers face construction from 5px symbolic faces to 64x64 dialogue portraits with expression parameters, and animation.md covers walk cycles, keyframes, and sub-pixel motion. - Use Case: When a generated NPC sprite looks like a stiff rectangle or its face reads as uncanny, consult this Skill to diagnose the failure (silhouette, light source, connectivity) and apply the exact px-level fix. ## Quick Start Read this Skill before drawing or fixing any character, NPC, monster, or quadruped sprite in a *.sprite.json file, then follow the C0-C5 pipeline and run the sprite lint to verify the result.
