Text-to-CAD System Requirements: Complete Setup Guide for Python and Node.js

Text-to-CAD requires Python 3.11 or newer, the cadgen library, and optionally Node.js 16+ with npm to run the CAD Viewer locally.

The earthtojake/text-to-cad repository is a pure-Python library for generating CAD models from text prompts, shipping with a React-based web viewer. To execute any of its skills or run the visualization interface, your development environment must satisfy specific runtime dependencies across both Python and JavaScript ecosystems.

Python Runtime Requirements

The core CAD generation engine relies on modern Python features and external CAD libraries.

Python Version

You need Python 3.11 or newer to run the text-to-cad skills. The repository's README.md displays a Python version badge confirming this minimum requirement, ensuring compatibility with the cadgen runtime's type hints and async patterns.

CADgen Dependency

All skills depend on the cadgen distribution, which provides the underlying geometry kernel and CLI tools. The specific version is pinned in skills/cad/requirements.txt as cadgen[snapshot]==0.5.1, ensuring reproducible builds across different environments.

Install the runtime manually using:

python -m pip install -r skills/cad/requirements.txt

JavaScript Runtime Requirements

The repository includes a web-based CAD Viewer for visualizing generated models in the browser.

Node.js and Package Managers

To build or run the CAD Viewer located in apps/viewer/, you need Node.js 16 or newer and either npm or Yarn. The React client compiles against modern ES modules and requires the Node toolchain for dependency resolution and bundling.

Navigate to the viewer directory and install dependencies:

cd apps/viewer
npm install
npm run dev

This starts the local development server for the visualization interface.

Version Control Prerequisites

Large binary assets in the repository require additional Git configuration.

Git LFS for Model Fixtures

The models/ directory contains large CAD model fixtures tracked via Git LFS (Large File Storage). As noted in AGENTS.md, you must have Git LFS installed to fetch these assets during cloning. This requirement is optional if you only intend to use the generation skills without the bundled test fixtures.

git lfs install
git lfs pull

Cross-Platform Compatibility

Text-to-cad runs on any operating system supporting the standard Python and Node.js toolchains, including Linux, macOS, and Windows. No platform-specific native binaries are required beyond the standard runtime environments.

Installation Methods

You can install and run text-to-cad using either the Skills CLI or direct Python execution.

Using the Skills CLI

The preferred method for adding the repository to your environment:

npx skills add earthtojake/text-to-cad

Direct CLI Execution

Run CAD generation directly via the cadgen module:

python -m cadgen.cli cad --prompt "Create a 20 mm cube" --output cube.step

This command invokes the CLI entry point defined in cadgen, processing the natural language prompt and exporting a STEP file.

Summary

• Python 3.11+ is mandatory for running the core generation logic according to README.md.
• Install cadgen[snapshot]==0.5.1 via skills/cad/requirements.txt to access the geometry kernel.
• Node.js 16+ and npm are required only for building the CAD Viewer in apps/viewer/.
• Git LFS is necessary to clone large model fixtures from the models/ directory.
• The repository supports Linux, macOS, and Windows without additional native dependencies.

Frequently Asked Questions

Do I need Node.js to generate CAD models with text-to-cad?

No. Node.js is only required if you want to run the web-based CAD Viewer locally. The core Python skills in earthtojake/text-to-cad function entirely within the Python 3.11+ environment using the cadgen library.

What version of cadgen does text-to-cad require?

The repository pins cadgen[snapshot]==0.5.1 in skills/cad/requirements.txt. This specific version ensures compatibility with the prompt parsing and geometry export functions used by the skills.

Can I run text-to-cad on Windows?

Yes. Since text-to-cad is a pure-Python library with a JavaScript frontend, it runs on any operating system that supports standard Python and Node.js installations, including Windows, macOS, and Linux.

Is Git LFS mandatory for using the repository?

Git LFS is only mandatory if you need the large CAD model fixtures stored in the models/ directory. You can skip Git LFS if you only intend to use the generation skills and provide your own model files.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}