What does this skill do, and when should you use it?
This skill provides CPU-only unit smoke tests for Isaac for Healthcare catheter-navigation fluoroscopy components. It checks imports, CT-to-attenuation preprocessing, CLI parser behavior, and conditional viewport imports. It is intended for CI or machines without a GPU. It does not cover GPU Slang rendering or the interactive viewport, so passing these tests does not complete GPU validation.
It locates or clones the i4h-workflows repository, enters the catheter_navigation workflow, creates a timestamped smoke-test run directory and log directory, and updates the .latest symlink. It runs workflows/catheter_navigation/tests/test_fluorosim_smoke.py with Python unittest and tees output into smoke.log. It then checks the log for the test count and OK status with grep; parser error lines from negative CLI tests are expected.
- A CI developer needs to check catheter-navigation imports and parser behavior in a build environment without an NVIDIA GPU.
- A fluorosimulation maintainer wants a quick regression check for CT-to-attenuation preprocessing.
- A test engineer needs timestamped smoke-test logs after a commit or branch change.
- A developer investigating import failures needs to verify that the active environment contains the catheter-navigation workflow dependencies.
How do you install this skill?
- Confirm network access and authorization before fetching code; preferably pin and verify the workflow revision or release.
- Use set -o pipefail and check the unittest exit status so failures cannot be masked by tee; do not rely only on Ran and OK log lines.
- This skill validates only the CPU smoke path, not GPU Slang rendering or the interactive viewport; dependency recovery may be affected by GitHub reachability in mainland China.
- Fix the version metadata placement and add Instructions, Examples, dependency versions, maintenance ownership, and an update path.
- Shell / CLI
- Network access
- Local filesystem
Python 3Gitcatheter navigation workflow dependencies
Install the skill from the NVIDIA/skills repository with the skills CLI:
npx skills add nvidia/skills --skill i4h-catheter-navigation-smoke --yes
Alternatively, run npx skills add nvidia/skills and choose the skill and destination interactively. The README does not specify a manual installation directory for this individual skill.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/NVIDIA/skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/skills/i4h-catheter-navigation-smoke" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- run CPU-only fluorosim smoke tests for catheter navigation in CI,
- Ran N tests
Ask an agent that has loaded the skill to “run CPU-only fluorosim smoke tests for catheter navigation in CI,” or an equivalent request. The skill uses this test command:
python3 -m unittest workflows/catheter_navigation/tests/test_fluorosim_smoke.py 2>&1 | tee "${RUN_DIR}/logs/smoke.log"
A successful run is indicated by “Ran N tests” and “OK” in the log. The final report should include pass/fail status, test count, and log path, and should state that GPU stages still require separate verification.
What are this skill's strengths and limitations?
- Requires no GPU and fits CPU-only CI environments.
- Covers imports, CT-to-attenuation preprocessing, CLI parsers, and conditional viewport imports.
- Creates timestamped run directories and preserves test logs.
- Does not test GPU Slang rendering or the interactive viewport.
- Requires Python 3, Git, and the catheter-navigation workflow dependencies.
- The source does not define a fixed test count or a supported-platform matrix.
- If the workflow directory is absent locally, the commands may clone the repository from GitHub and therefore require network access.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Catheter Navigation Fluoroscopy Smoke Tests this page ✓ NVIDIA · Official | 50 · Use with care | ★ 3.8k | today | Apache-2.0 |
| Catheter Navigation Setup Verification ✓ NVIDIA · Official | 51 · Use with care | ★ 3.8k | today | Apache-2.0 |
| Catheter Navigation E2E Smoke Test ✓ NVIDIA · Official | 45 · Not recommended | ★ 3.8k | today | Apache-2.0 |
| RenderCV Testing Context | 61 · Recommended | ★ 18k | 7mo ago | MIT |
| Earth2Studio Diagnostic Builder ✓ NVIDIA · Official | 56 · Use with care | ★ 3.8k | today | Apache-2.0 |
It complements i4h-catheter-navigation-render-drr: this skill handles GPU-free smoke validation, while that named skill is indicated for GPU validation. The source does not provide a detailed comparison with other alternatives.
How did FollowSkills review this skill?
The skill is explicitly CPU-only, states that no GPU or credentials are required, and limits writes to the workflow runs directory; however, it automatically clones from an external GitHub repository when absent, does not pin a revision, request confirmation, disclose supply-chain/network risk, or define rollback, so points are deducted.
The commands, test target, log capture, and Ran/OK checks are clearly specified, and an external evaluation report and task definitions are included; however, the workflow and dependencies are external and unpinned, pipefail is not enabled so unittest failure may be masked by tee, and abnormal-input diagnostics are thin, so the score remains below the static cap of 10.
The name, description, and positive/negative activation scenarios are reasonably precise; the skill distinguishes CPU smoke testing from GPU rendering and states the GPU viewport limitation. Input/output contracts and non-fit boundaries remain incomplete, and the fallback depends on GitHub reachability, which may affect mainland-China users, so points are deducted.
The documentation includes purpose, run steps, expected output, verification, prerequisites, limitations, troubleshooting, and final-response guidance; version, owner, and Apache-2.0 are stated. However, the benchmark reports misplaced version frontmatter and missing recommended Instructions and Examples sections, repository metadata is NOASSERTION despite the skill declaration, and maintenance/update ownership is not sufficiently explicit, so points are deducted.
The skill offers directly reusable CPU smoke-test commands, timestamped logs, and reporting requirements, while the evaluation tasks cover imports, preprocessing, parsers, and negative routing. Static review cannot confirm actual execution; the report provides limited sampled evidence and does not independently establish key-path output usability, so the score is kept within the static cap of 7 and reduced.
The SKILL.md provides auditable commands and explicit verification strings, while BENCHMARK.md and evals.json provide limited third-party evaluation records. However, there is no committed skill-specific test suite or CI evidence reproducing this skill's key path, and the reported evaluation cannot be independently reproduced from the supplied files, so the score is 4.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →