This is a photo critique tool grounded in actual computational aesthetics research, not just generic feedback. It analyzes composition using frameworks like Arnheim's visual weight, dynamic symmetry, and Gestalt principles, then scores images using models like NIMA and LAION-Aesthetics. The approach is methodical: first impression, technical scan, compositional analysis, aesthetic evaluation, then actionable recommendations. It explicitly pushes back on oversimplifications like blindly applying rule of thirds or treating ML scores as gospel. Use it when you need serious image analysis with theoretical backing, not when you're editing photos or generating new ones. Pairs well with color theory tools if you're doing deep aesthetic work.
npx skills add https://github.com/erichowens/some_claude_skills --skill photo-composition-critic