render-glb
Render GLB 3D models to PNG images for visual verification. Use when you need to SEE your generated 3D model, verify geometry, check visual correctness, or show results to users. Triggers on: render GLB, preview model, visualize 3D, see my model, check geometry, visual verification, screenshot model, model preview.
When & Why to Use This Skill
The render-glb skill is a specialized utility designed to bridge the gap between 3D model generation and visual perception for AI agents. It enables the conversion of GLB 3D models into PNG images, allowing agents to 'see' and verify their work. This tool is essential for visual verification loops, geometry inspection, and creating instant previews of 3D assets within automated design and CAD workflows.
Use Cases
- Visual Verification Loop: Automatically render a PNG after generating a 3D model to let the AI agent inspect the geometry and ensure the output matches the requested design.
- Iterative 3D Design: Use rendered images as visual feedback for the AI to identify errors in 3D coordinates or textures and perform self-correction in subsequent iterations.
- Instant Asset Previewing: Generate high-quality 2D screenshots of 3D models to show results to users immediately in the chat interface without requiring external 3D viewers.
- CAD Integration: Complement programmatic modeling tools (like build123d) by providing a visual bridge between code-based geometry generation and visual quality assurance.
| name | render-glb |
|---|---|
| description | Render GLB 3D models to PNG images for visual verification. Use when you need to SEE your generated 3D model, verify geometry, check visual correctness, or show results to users. Triggers on: render GLB, preview model, visualize 3D, see my model, check geometry, visual verification, screenshot model, model preview. |
Render GLB to Image
Render 3D GLB files to PNG images so agents can visually verify their work.
Zero-Setup with bunx
bunx render-glb model.glb output.png
No installation required. First run downloads the tool, subsequent runs are instant.
Why This Matters
Agents can generate 3D models but typically can't see them. This creates a visual verification loop:
- Generate model → export GLB
- Render GLB → PNG
- Read PNG → see what you built
- Iterate if needed
Basic Usage
# Render with defaults (good for most cases)
bunx render-glb model.glb preview.png
# Then read the image to see your work
# (use Read tool on preview.png)
Options
bunx render-glb <input.glb> <output.png> [options]
Options:
--width <n> Image width in pixels (default: 800)
--height <n> Image height in pixels (default: 600)
--background <hex> Background color (default: #808080)
Visual Verification Workflow
When building 3D models, use this pattern:
# 1. Build your model
from build123d import Box, Sphere, export_gltf
model = Box(20, 20, 20) + Sphere(radius=15)
export_gltf(model, "./model.glb", binary=True)
# 2. Render to image
bunx render-glb model.glb preview.png
# 3. Read the image (using Read tool)
# Now you can SEE what you built and verify it's correct
Integration with CAD Skills
This skill complements:
- build123d - Python CAD modeling
- Any tool that exports GLB format
Typical workflow:
- Use CAD skill to generate geometry
- Export to GLB
- Use render-glb to verify visually
- Iterate if model isn't correct