CAD Agent for Openclaw

A rendering server that enables AI agents to generate, visualize, and iterate on 3D parametric models.

clawd-maf
v1.0.0
Jan 26, 2026
6
5.7k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install cad-agent

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install cad-agent using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is CAD Agent?

CAD Agent acts as the visual cortex for AI agents performing mechanical design and 3D modeling. By providing a containerized environment for the build123d library and VTK rendering engine, it allows agents to transform code into physical geometry and see the results instantly. This bridge is essential for high-fidelity engineering workflows within the ecosystem of Openclaw Skills, ensuring that agents don't just write code blindly but can actually verify the physical characteristics of their designs.

The system operates on a command-and-observe loop. The agent sends modeling instructions via an API, and the server returns high-resolution PNG renders. This allows for rapid iteration and error correction without the need for the agent to handle complex mesh processing or STL manipulation locally.

CAD Agent Use Cases

  • Creating 3D-printable parts through iterative natural language prompting.
  • Developing parametric mechanical components where dimensions are driven by complex logic.
  • Generating technical documentation and multi-view drawings for engineering assemblies.
  • Rapid prototyping of 3D geometry within automated AI development environments.

How CAD Agent Works

  1. The AI agent sends Python-based build123d modeling commands to the containerized server.
  2. The server executes the code to generate 3D geometry in a protected session.
  3. The agent requests specific renders (e.g., Isometric, Top, or Multiview) to inspect the model visually.
  4. The agent analyzes the returned PNG image and sends modification commands to refine the part.
  5. Once the design is validated, the agent triggers an export to produce industry-standard STL or STEP files.

CAD Agent Setup

1. Clone the Repository

git clone https://github.com/clawd-maf/cad-agent.git
cd cad-agent

2. Build the Docker Image

docker build -t cad-agent:latest .

3. Run the Server

docker-compose up -d

4. Verify the Connection

curl http://localhost:8123/health

CAD Agent Data Schema & Taxonomy

The skill manages model state within the container and provides data through structured API responses. Below is how the data is organized:

Data Type Description Format
Model Session In-memory storage of the current geometric state. build123d Object
Visual Feedback Shaded 3D or 2D technical renders. PNG
Geometric Data Physical dimensions and bounding box measurements. JSON
Production Files Exportable assets for manufacturing. STL, STEP, 3MF

CAD Agent Advanced Features

  • Multi-view composite rendering providing simultaneous front, right, top, and isometric perspectives.
  • Automated printability analysis to detect potential manufacturing issues like extreme overhangs.
  • Precise geometric measurement tools to query distances, radii, and bounding boxes programmatically.
  • Safe-export mechanisms that prevent large binary CAD files from cluttering version control systems.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*