Pet Body Condition Score 3D Analysis

Problem

Pet body-condition assessment is often subjective, and a single photo may not capture waistline, rib coverage, or abdominal fat well. This skill targets smart feeders, pet cameras, and pet health platforms that need a repeatable observation output from multi-angle video. It does not diagnose disease or prescribe treatment; it produces standardized BCS results that downstream systems can consume.

Capabilities and workflow

Input can be a local file or a public video URL, supporting mp4/avi/mov with a 10MB limit. The main flow is:
- user provides standing, side, and top-down views
- the script calls a server-side API for 3D body-shape reconstruction
- it estimates waistline, abdominal contour, rib-touch-equivalent visual cues, and fat coverage
- it scores BCS 1-9 as underweight, ideal, overweight, or obese
- it returns a structured report, risk notes, and report links

Historical report lookup uses --list and must read from the cloud API, avoiding stale or manually compiled local data. User-facing flows do not require identity parameters; the script handles them internally.

Boundaries

BCS is a visual estimate and may differ from veterinary palpation. Single-angle input can reduce reconstruction accuracy. Output is intended as a weight-management reference, not clinical diagnosis.