3d-asset-generation

Generate, validate, and assemble production 3D assets using Atlas Cloud, fal.ai, and Blender.

Updated Sep 1, 2026
One-click install
npx skills add https://github.com/zamansepeti43/c-rak-agent --skill 3d-asset-generation-zamansepeti43
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: 3d-asset-generation
Source: https://github.com/zamansepeti43/c-rak-agent/tree/main/video-engine/.agents/skills/3d-asset-generation
Command: npx skills add https://github.com/zamansepeti43/c-rak-agent --skill 3d-asset-generation-zamansepeti43

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Producing real 3D meshes for cinematic worlds requires choosing between multiple paid generation providers, controlling per-call costs, and verifying that downloaded GLB assets actually meet production quality before assembly. This Skill routes each asset need to the right provider, enforces cost discipline, and mandates mesh QA and scale normalization in Blender. ## Core Features & Use Cases - Provider Routing: Maps asset roles to the right tool, including Atlas Tripo H3.1 text-to-3D, fal Hunyuan 3D image-to-3D, fal SAM 3D Objects, the CC0 Kenney catalog, and Blender for scene assembly. - Cost and Provenance Discipline: Requires stating provider, model, unit cost, and output count before paid calls, and recording prompts, seeds, and costs in an asset provenance manifest. - Mandatory Mesh QA and Assembly: Inspects geometry, UVs, PBR textures, and scale in Blender, normalizes bounding boxes to declared target heights, and enforces exclusion zones and visibility windows for scatter and landmarks. - Use Case: A production needs a hero building plus hundreds of vegetation instances for a region. Generate the hero mesh via text-to-3D, pull scatter from the licensed catalog, validate everything in Blender, and assemble the world from a semantic specification. ## Quick Start Use the 3d-asset-generation skill to generate a textured hero prop with Atlas Tripo, validate it in Blender, and assemble it into the world scene.

Frequently Asked Questions about 3d-asset-generation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I choose between text-to-3D and image-to-3D generation?

Use Atlas Tripo H3.1 text-to-3D for unique objects described in words, and fal Hunyuan 3D image-to-3D when you have a concept image whose silhouette and style must be matched. For repeated generic props, use the CC0 Kenney catalog instead of paid generation.

How do I control costs when generating 3D assets with paid APIs?

State the provider, model, operation, unit cost, and output count before every first paid call, and generate one sample before running a batch. Record all costs in the asset provenance manifest and confirm current provider pricing before quoting.

Can text-to-3D models generate an entire cinematic world in one mesh?

No. Text-to-3D models should only generate individual hero objects. High-volume scatter comes from licensed catalogs, and terrain, lighting, camera, and final composition are assembled in Blender from a semantic specification.

Why do generated GLB assets have wrong scale or floating ground contact?

Providers and catalogs use different units, origins, and up-axis assumptions. The Blender assembly spec declares a target_height, measures the imported bounding box, normalizes to that target, and offsets the bounding-box floor to the sampled terrain height.

What mesh checks are required before approving a generated 3D asset?

Import the GLB into Blender and inspect front, rear, and silhouette views, geometry holes, ground contact, scale, UV seams, and base-color, roughness, metallic, and normal texture response. Never approve from the provider thumbnail alone.