Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
CAD Agent is a Dockerized rendering server that lets AI agents issue build123d modeling commands and receive rendered images in return. The agent sends code via HTTP, gets back PNG renders, and iterates based on what it sees. All mesh processing and rendering runs inside the container.
Gives agents a visual feedback loop for CAD work, which text-only code execution cannot provide.
When to use it
- Designing a 3D-printable bracket with iterative visual feedback
- Building parametric mechanical parts without leaving the chat interface
- Checking printability of a model before sending to a slicer
- Exporting STL/STEP files from AI-generated geometry
- Prototyping part geometry by describing dimensions in natural language
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Create a box with a cylindrical hole through it and export as STL
- 1POST /model/create with build123d code defining Box(60, 40, 30)
- 2POST /render/multiview to get a 4-view PNG and inspect the result
- 3POST /model/modify to subtract a Cylinder from the box at the desired position
- 4POST /render/3d to confirm the hole looks correct
- 5POST /export with format STL to get the final file
part.stl file ready for slicing or printing
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.