What problem does it solve? Heavy 3D assets cause slow load times, frame drops, and excessive GPU and memory usage in web-based 3D experiences. This Skill diagnoses the real asset bottleneck—format, polycount, textures, materials, or decode path—and applies targeted optimizations instead of blind quality reduction. ## Core Features & Use Cases - Asset Bottleneck Diagnosis: Classifies whether pressure comes from raw formats (OBJ vs GLB/GLTF), excessive polycount, large textures, too many draw calls, or slow decode and network paths. - AI-Generated Asset Validation: Enforces 2025-2026 pipeline gates for assets from Meshy AI, TripoSR, Rodin, and similar tools, including a ≤10k triangle budget, UV and normal validation, WebP/KTX2 texture conversion, license checks, and LOD requirements. - Structured Audit Output: Produces a performance-audit JSON conforming to contracts/schemas/performance-audit.json with verdict, metrics, and findings. - Use Case: A React Three Fiber product page loads a 1.5M-triangle AI-generated mesh with 4K PNG textures and stutters on mobile. Use this Skill to decimate the mesh, convert textures to KTX2, generate a mobile LOD, and document the accepted visual trade-offs. ## Quick Start Use the optimize-3d-assets skill to analyze the GLTF models and textures in this project and reduce load time and GPU memory usage.
