How do I install the Nv Reason Cxr Quality V1 skill?

Run npx skillmds@latest add nvidia-medtech/nv-reason-cxr-quality-v1 in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Nv Reason Cxr Quality V1 safe to use?

This skill has not completed SkillMD's automated safety review yet. Independent scanners report: SkillSpector: PASS, Skill Scanner: PASS. Capability flags: executes scripts. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Nv Reason Cxr Quality V1?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Nv Reason Cxr Quality V1 free to use?

Yes. Installing skills from SkillMD is free. This skill is licensed under Apache-2.

Who published Nv Reason Cxr Quality V1?

NVIDIA-Medtech (@nvidia-medtech) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Nv Reason Cxr Quality V1 by nvidia-medtech

Name: Nv Reason Cxr Quality V1
Rating: 4.666666666666667 (3 reviews)
Author: NVIDIA-Medtech

Nv Reason Cxr Quality V1

Used to verify nv-reason-cxr evidence packs for source-pack success, fixture/image hash binding, NV-Reason-CXR runtime identity, response shape, scope disclosure, and conservative forbidden overreach phrases.

File contents

NV Reason CXR Quality Verifier

Purpose

Used for deterministic second-pass review of an nv-reason-cxr evidence pack.
Checks source-pack success, fixture case/prompt binding, image readability and hash, model/runtime identity, output response shape, limitation disclosure, and forbidden overreach phrases.
Manifest I/O: inputs are nv_reason_cxr_evidence_pack; outputs are nv_reason_cxr_quality_report.

Instructions

Use this verifier only on an evidence pack directory produced by skills/nv-reason-cxr.
Run it through eval_engine/run.py when producing verifier evidence.
The verifier entrypoint is scripts/grade.py; do not replace it with prompt-only review.
Treat a pass as engineering evidence only, not as validation of clinical correctness or model quality.

Available Scripts

Script	Purpose	Arguments
`scripts/grade.py`	Primary verifier entrypoint declared by `skill_manifest.yaml`.	`EVIDENCE_PACK_DIR`

Prerequisites

Runtime requirements: Python standard library only.
The input directory must contain manifest.json, validation_summary.json, and output.json from an nv-reason-cxr evidence pack.
The recorded fixture path and image artifact must be available for full fixture/image binding.

Limitations

Does not validate chest X-ray interpretation, model correctness, report completeness, or medical actionability.
Forbidden-phrase checks are regex-based and can miss paraphrases.
Mock-mode packs prove wrapper and gate wiring; they do not prove live NV-Reason-CXR-3B model behavior.

Troubleshooting

Error	Cause	Fix
`target_skill_matches` fails	The fixture is not an `nv-reason-cxr` evidence pack.	Re-run the verifier against the source skill pack directory.
`fixture_loaded` fails	The source pack references a fixture that is unavailable in this checkout.	Regenerate the source pack with a committed synthetic fixture.
`image_sha256_matches` fails	The recorded image artifact changed or is unavailable.	Regenerate the source pack or inspect artifact provenance.
`forbidden_phrases_absent` fails	Output contains treatment, regulatory, patient-directed, or absolute-certainty language.	Fix the source skill prompt/runtime guard and regenerate evidence.

Example

MOCK_NV_REASON_CXR=1 python eval_engine/run.py skills/nv-reason-cxr \
  --fixture skills/nv-reason-cxr/fixtures/synthetic_cxr_input.json \
  --out runs/nv_reason_cxr_demo

python eval_engine/run.py verifiers/nv_reason_cxr_quality_v1 \
  --fixture runs/nv_reason_cxr_demo \
  --out runs/nv_reason_cxr_quality

python verifiers/nv_reason_cxr_quality_v1/scripts/grade.py \
  runs/nv_reason_cxr_demo

---name: nv-reason-cxr-quality-v1description: Used to verify nv-reason-cxr evidence packs for source-pack success, fixture/image hash binding, NV-Reason-CXR runtime identity, response shape, scope disclosure, and conservative forbidden overreach phrases.license: Apache-2.0---# NV Reason CXR Quality Verifier## Purpose- Used for deterministic second-pass review of an `nv-reason-cxr` evidence pack.- Checks source-pack success, fixture case/prompt binding, image readability and hash, model/runtime identity, output response shape, limitation disclosure, and forbidden overreach phrases.- Manifest I/O: inputs are `nv_reason_cxr_evidence_pack`; outputs are `nv_reason_cxr_quality_report`.## Instructions- Use this verifier only on an evidence pack directory produced by `skills/nv-reason-cxr`.- Run it through `eval_engine/run.py` when producing verifier evidence.- The verifier entrypoint is `scripts/grade.py`; do not replace it with prompt-only review.- Treat a pass as engineering evidence only, not as validation of clinical correctness or model quality.## Available Scripts| Script | Purpose | Arguments ||---|---|---|| `scripts/grade.py` | Primary verifier entrypoint declared by `skill_manifest.yaml`. | `EVIDENCE_PACK_DIR` |## Prerequisites- Runtime requirements: Python standard library only.- The input directory must contain `manifest.json`, `validation_summary.json`, and `output.json` from an `nv-reason-cxr` evidence pack.- The recorded fixture path and image artifact must be available for full fixture/image binding.## Limitations- Does not validate chest X-ray interpretation, model correctness, report completeness, or medical actionability.- Forbidden-phrase checks are regex-based and can miss paraphrases.- Mock-mode packs prove wrapper and gate wiring; they do not prove live NV-Reason-CXR-3B model behavior.## Troubleshooting| Error | Cause | Fix ||---|---|---|| `target_skill_matches` fails | The fixture is not an `nv-reason-cxr` evidence pack. | Re-run the verifier against the source skill pack directory. || `fixture_loaded` fails | The source pack references a fixture that is unavailable in this checkout. | Regenerate the source pack with a committed synthetic fixture. || `image_sha256_matches` fails | The recorded image artifact changed or is unavailable. | Regenerate the source pack or inspect artifact provenance. || `forbidden_phrases_absent` fails | Output contains treatment, regulatory, patient-directed, or absolute-certainty language. | Fix the source skill prompt/runtime guard and regenerate evidence. |## Example```bashMOCK_NV_REASON_CXR=1 python eval_engine/run.py skills/nv-reason-cxr \ --fixture skills/nv-reason-cxr/fixtures/synthetic_cxr_input.json \ --out runs/nv_reason_cxr_demopython eval_engine/run.py verifiers/nv_reason_cxr_quality_v1 \ --fixture runs/nv_reason_cxr_demo \ --out runs/nv_reason_cxr_qualitypython verifiers/nv_reason_cxr_quality_v1/scripts/grade.py \ runs/nv_reason_cxr_demo```

Frequently asked questions