What problem does it solve?
Setting up the Isaac for Healthcare (i4h) catheter navigation workflow fails when host requirements, GPU drivers, or PYTHONPATH are misconfigured, producing confusing import, slangpy, and CUDA errors. This Skill walks an agent through preflight checks, CLI registration verification, and CPU smoke tests so setup problems are diagnosed systematically instead of by trial and error.
Core Features & Use Cases
- Host and GPU preflight: Checks python3, git, nvidia-smi, and disk space against documented requirements (Linux x86_64, NVIDIA GPU CC >= 7.0, CUDA 12.8-compatible driver, 16 GB RAM, 20 GB disk).
- CLI registration and smoke tests: Verifies
./i4h modes catheter_navigation registration and runs the 7-test fluorosim CPU smoke suite, logging results to a timestamped run directory.
- Optional GPU sanity check: Renders a synthetic DRR image when slangpy, Warp, and CUDA are available, and falls back to Docker guidance when host GPU dependencies are incomplete.
- Use Case: A developer hits
fluorosim import errors when launching the catheter viewport; the agent exports the simulation PYTHONPATH, reruns the smoke tests, and recommends the Docker path for missing GPU dependencies.
Quick Start
Ask your agent to set up the catheter navigation workflow and report which host requirements or dependencies are missing.
