AI-powered CAD modeling assistant for Autodesk Fusion 360. Describe what you want in plain English, and CADAgent builds it.
This repository holds the whole product:
| Part | Where | What it does |
|---|---|---|
| Fusion 360 add-in | apps/addin/mac/, apps/addin/win/ |
Runs inside Fusion. Sends your request to the backend over a websocket and applies the returned CAD operations. |
| Backend | apps/backend/ |
FastAPI service that turns natural-language requests into tool calls the add-in executes. Self-hosted: run it locally or on your own server. |
| Infra templates | infra/ |
Reference AWS deploy setup (GitHub Actions → S3 → CodeDeploy → EC2) from when the backend ran as a hosted service. |
| Supabase | supabase/ |
Auth and quota migrations plus edge functions. |
apps/
addin/ mac/CADAgent/, win/CADAgent/ Fusion add-in (vendored deps per platform)
backend/ backend/ tests/ ops/ FastAPI service, its tests, and ops scripts
infra/ appspec.yml, hooks/, systemd/, nginx/, codedeploy/, iam/, monitoring/
supabase/ migrations/, functions/
docs/
Download the latest release:
- Windows: CADAgent-Windows.zip
- macOS: CADAgent-macOS.zip
Copy the CADAgent folder to Fusion's add-ins directory:
- Windows:
%AppData%\Autodesk\Autodesk Fusion 360\API\AddIns - macOS:
~/Library/Application Support/Autodesk/Autodesk Fusion 360/API/AddIns
Restart Fusion 360, then run it from Tools > Add-Ins > CADAgent > Run. Type a request such as "Create a 5cm cube" and click Execute.
cd apps/backend
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
cp .env.example .env # fill in the keys you need
python start_backend.pyRun the tests from the same directory:
pip install -r requirements-dev.txt
pytest -qThe service listens on ws://localhost:8000/ws/{session_id} with a health check at /health. .env.example documents every variable; point the add-in at your backend via BACKEND_HOST in its .env.cadagent. The hosted backend at ws.cadagentpro.com is shut down — running your own backend is the supported path.
docs/backend/README.md— production runbook (systemd, nginx, CodeDeploy).docs/backend/deployment.md— the CI/CD pipeline end to end.apps/addin/mac/CADAgent/README.md/apps/addin/win/CADAgent/README.md— add-in internals and the websocket protocol.