skills.rest
maziyarpanahi
maziyarpanahiCOMMUNITY
Explore all skills by this author →
Website𝕏@maziyarpanahiGitHub

authoring-model-cards

Convert OpenMed clinical evaluation reports into governance-ready model cards.

5.0k|615|Updated Oct 4, 2025
Search Tags:#governance#de-identification#model cards#clinical ai#fairness auditing#ner evaluation#leakage testing
One-click install
npx skills add https://github.com/maziyarpanahi/openmed --skill authoring-model-cards
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: authoring-model-cards
Source: https://github.com/maziyarpanahi/openmed/tree/main/skills/authoring-model-cards
Command: npx skills add https://github.com/maziyarpanahi/openmed --skill authoring-model-cards

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill turns OpenMed evaluation results into transparent, reproducible model cards that clearly document clinical model performance, intended use, limitations, subgroup behavior, and safety disclaimers.

Core Features & Use Cases

  • Evidence-based documentation: Populate model details and metrics from GateReport, fairness_report, and error_report artifacts.
  • Clinical governance coverage: Define intended and out-of-scope uses, residual leakage risks, subgroup performance, limitations, and medical-device disclaimers.
  • Reproducibility safeguards: Preserve consistency between model cards, evaluation evidence, and models.jsonl metadata while avoiding unsupported claims and raw PHI.
  • Use Case: When preparing an OpenMed clinical NER or de-identification model for release, use the Skill to convert its pinned evaluation outputs into a governance-ready model card.

Quick Start

Use the authoring-model-cards skill to create a model card from the model's GateReport, fairness_report, and error_report, including intended use, metrics, subgroup analysis, limitations, and the required clinical disclaimer.

Frequently Asked Questions about authoring-model-cards

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a model card for clinical AI from evaluation outputs?▼

Generate a model card for clinical AI by converting GateReport, fairness_report, and error_report evaluation outputs into structured documentation that captures reproducible metrics, subgroup performance, and deployment boundaries.

What evidence is required to document clinical NER model limitations and leakage risks?▼

Documenting clinical NER model limitations and leakage risks requires GateReport, fairness_report, and error_report evidence to populate reproducible metrics, residual leakage risks, subgroup behavior, and medical-device disclaimers.

Can I use this approach to create governance-ready documentation for de-identification models?▼

Yes, you can create governance-ready documentation for de-identification models by transforming pinned OpenMed evaluation artifacts into datasheets and README files that define intended use and out-of-scope applications.

How does fairness auditing integrate into clinical AI model releases?▼

Fairness auditing integrates into clinical AI model releases by extracting subgroup performance metrics from fairness_report artifacts and embedding them directly into the model card to ensure transparent, reproducible governance documentation.

What limitations exist when authoring model cards for clinical AI risk reviews?▼

Limitations when authoring clinical AI model cards include strict dependencies on GateReport, fairness_report, and error_report evidence, and the requirement to avoid unsupported claims and raw PHI to maintain compliance.

Does this method preserve consistency between model cards and models.jsonl metadata?▼

Yes, this method preserves consistency between model cards and models.jsonl metadata by aligning documented metrics, limitations, and deployment boundaries directly with the pinned OpenMed evaluation outputs.

Related Skills

Explore More Legal & Compliance Skills→
maziyarpanahimaziyarpanahi
5.0k

serving-openmed-rest-api

Expose OpenMed clinical NER, PII extraction, and de-identification via a self-hosted FastAPI REST API.

Community
Advanced
skills-ilskills-il
9

israeli-ai-compliance-kit

Map Israeli and EU AI compliance obligations and generate model cards, data statements, and DPIA templates.

Official
Advanced
DTMC-marketplaceDTMC-marketplace

disaggregated-evaluation

Evaluate AI model performance across demographic subgroups to detect bias.

Official
Intermediate
Yezez9Yezez9

pyhealth

Develop, test, and deploy healthcare machine learning models with clinical data.

Community
Advanced