Text-to-3D Asset Generation

For text-to-3D, image-to-3D, or compound CAD/USD asset authoring, read and follow .agents/docs/generate_asset.md.

Workflow

1. Confirm the repository build provides _omni_solid and pxr.
2. Follow the playbook's reference gathering, parametric modeling, tessellation, USD export, and self-check guidance.
3. For multi-part assets or compound scenes, also apply its composition, penetration, and envelope checks.
4. When reference material exists, use the documented semantic and visual comparison guidance.
5. Deliver the authoring script, mesh-only USDA, screenshot, and validation summary.

Validation