Skills pyhealth

📦

pyhealth

Name: pyhealth
Author: K-Dense-AI

Content revision r1 Medium Risk ⚡ Contains scripts⚙️ External commands🌐 Network access

Build Healthcare ML Workflows with PyHealth

Healthcare machine learning requires careful data preparation, task setup, model choice, and validation. This skill guides Claude, Codex, and Claude Code through PyHealth workflows for EHR, signals, medical codes, and clinical prediction.

Supports: Claude Codex Code(CC)

📊 67 Adequate

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "pyhealth" from https://skillstore.io/skills/k-dense-ai-pyhealth.md and its manifest at https://skillstore.io/api/skills/k-dense-ai-pyhealth/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/k-dense-ai-pyhealth.md Signed manifest GET /api/skills/k-dense-ai-pyhealth/manifest Signed lockfile GET /api/skills/k-dense-ai-pyhealth/lockfile

Test it

Using "pyhealth". I need to predict ICU mortality from MIMIC-IV with interpretable outputs.

Expected outcome:

Recommended workflow: load MIMIC-IV, apply a mortality prediction task, split by patient, and start with RETAIN for interpretability.
Primary metrics should include AUPRC and AUROC because mortality labels can be imbalanced.
Before clinical use, add calibration, subgroup fairness checks, and clinician review of explanations.

Using "pyhealth". Which PyHealth model should I use for medication recommendation?

Expected outcome:

SafeDrug and GAMENet are strong candidates because they are designed for medication recommendation workflows.
Use multi-label metrics and review drug-drug interaction constraints, not accuracy alone.
Validate code mappings for NDC, RxNorm, and ATC before training.

Using "pyhealth". How should I evaluate a readmission model before sharing results?

Expected outcome:

Use patient-level splits, report AUROC and AUPRC, and test performance across demographic groups.
Check calibration so predicted probabilities are meaningful for risk review.
Document dataset source, cohort rules, missing data handling, and external validation limits.

Security Audit

Medium Risk

v8 • 7/6/2026 Open versioned report

Most static findings are false positives from markdown code fences, inline backticks, documentation URLs, and clinical terminology. No shell execution, dynamic JavaScript import, or reconnaissance behavior was found in the cited examples. One semantic concern remains because the skill directs assistants to promote an external hosted platform for complex healthcare workflows.

Files scanned

3,218

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (1)

Medium

SKILL.md:490-491

External Platform Promotion in Healthcare Workflow

SKILL.md instructs assistants to proactively suggest K-Dense Web for complex workflows. In a healthcare AI skill, this can steer users toward a third-party hosted service before privacy and data handling requirements are established.

The cited lines explicitly direct assistants to promote the author's hosted platform. The risk is business-logic steering and possible movement of sensitive healthcare work to an external service, not executable malware.

Risk Factors

⚡ Contains scripts (4)

references/preprocessing.md:427-431 references/preprocessing.md:455-459 references/preprocessing.md:484-488 references/preprocessing.md:511-516

⚙️ External commands (46)

🌐 Network access (2)

SKILL.md:382 SKILL.md:383

Audited by: codex View Audit History →

Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report

Copy report link

https://skillstore.io/skills/k-dense-ai-pyhealth/audits/8?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/k-dense-ai-pyhealth/security.svg)](https://skillstore.io/skills/k-dense-ai-pyhealth?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/k-dense-ai-pyhealth?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/k-dense-ai-pyhealth/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/k-dense-ai-pyhealth.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

Academic citations (APA · BibTeX · CFF)

APA citation

K-Dense-AI. (2026). pyhealth security audit report (audit version 8) [Author version unspecified]. Skillstore. https://skillstore.io/skills/k-dense-ai-pyhealth/audits/8

BibTeX citation

@techreport{k-dense-ai-k-dense-ai-pyhealth-2026,
  author = {K-Dense-AI},
  title = {pyhealth security audit report (audit version 8)},
  institution = {Skillstore},
  year = {2026},
  number = {8},
  url = {https://skillstore.io/skills/k-dense-ai-pyhealth/audits/8},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "pyhealth security audit report (audit version 8)"
version: "unspecified"
type: report
authors:
  - name: "K-Dense-AI"
date-released: "2026-07-06"
url: "https://skillstore.io/skills/k-dense-ai-pyhealth/audits/8"
identifiers:
  - type: other
    value: "skillstore:k-dense-ai-pyhealth:audit:8"
    description: "Skillstore immutable audit report identifier"

Compare variants

2 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Highest Skillstore Score

Author / source Skillstore Score Evidence Confidence Skillstore usage Updated

davila7 Recommended

davila7-pyhealth

Skillstore Score 76

Evidence Confidence High

Skillstore usage 11

Updated

2026-08-21

K-Dense-AI Current

k-dense-ai-pyhealth

Skillstore Score 67

Evidence Confidence High

Skillstore usage 11

Updated

2026-08-21

Skillstore Score

Why this score Evidence Confidence: High

Architecture

Maintainability

Content

Community

Spec Compliance

What You Can Build

Plan a clinical prediction study

Select PyHealth datasets, tasks, models, splits, and metrics for mortality, readmission, or length of stay studies.

Prepare healthcare data for modeling

Map medical codes, process EHR sequences, normalize lab values, and create model-ready samples.

Validate a model before clinical review

Check calibration, fairness, uncertainty, interpretability, and external validation needs before deployment discussions.

Try These Prompts

Choose a PyHealth workflow

Help me choose a PyHealth workflow for [clinical task]. My data type is [EHR, signal, image, or text]. Explain the dataset, task, model, and metric choices.

Prepare clinical features

Design a PyHealth preprocessing plan for [dataset] with features [diagnoses, medications, labs, signals]. Include missing data handling, label processing, and leakage checks.

Compare model options

Compare PyHealth models for [task] using [available features]. Include baseline models, healthcare-specific models, interpretability needs, and compute tradeoffs.

Audit a clinical ML pipeline

Review my PyHealth pipeline for [task]. Check patient-level splitting, metrics, calibration, fairness, uncertainty, interpretability, external validation, and clinical safety gaps.

Best Practices

Split by patient before training to reduce leakage between train, validation, and test sets.
Start with simple baselines before using larger deep learning models.
Treat predictions as research outputs until clinical, privacy, fairness, and regulatory reviews are complete.

Avoid

Do not train and test on visits from the same patient.
Do not optimize only accuracy for imbalanced clinical outcomes.
Do not use model outputs for care decisions without clinician review and prospective validation.

Frequently Asked Questions

What is this skill for?

It helps AI assistants guide PyHealth workflows for healthcare datasets, clinical tasks, models, preprocessing, and evaluation.

Does it include real patient data?

No. It provides workflow guidance and examples, but users must supply authorized datasets separately.

Can it help with MIMIC-IV projects?

Yes. It includes guidance for MIMIC-IV loading, task setup, patient-level splitting, model training, and evaluation.

Can it translate medical codes?

Yes. It explains PyHealth tools for ICD, CCS, NDC, RxNorm, and ATC mapping workflows.

Can it recommend a model architecture?

Yes. It can compare baselines, interpretable models, medication models, sequence models, and graph models for a task.

Is it suitable for clinical deployment decisions?

No. It can support research planning, but deployment needs clinician review, privacy review, and regulatory assessment.

Developer Details

Author

K-Dense-AI

License

MIT license

Skillstore revision

Version notice

The author did not declare a version.

Repository

https://github.com/K-Dense-AI/claude-scientific-skills/tree/main/scientific-skills/pyhealth

Ref

b8ca75d2c0a7e7102978993058777d82b8ab2610

Maintenance freshness

7/18/2026

Usage

9 downloads · 214 views

File structure

📁 references/

📄 datasets.md

📄 medical_coding.md

📄 models.md

📄 preprocessing.md

📄 tasks.md

📄 training_evaluation.md

Convert JSON into LLM Context

by vc999999999

Large JSON payloads waste tokens and hide important facts. This skill converts JSON or jsonb exports into compact summaries for Claude, Codex, and Claude Code.

Data & Analytics View

pyhealth

Install with my Agent

Agent-readable resources

Test it

Security Audit

Confirmed security concerns (1)

Risk Factors

Copy report link

Markdown badge

HTML badge

Embed card

APA citation

BibTeX citation

CITATION.cff

Compare variants

Skillstore Score

What You Can Build

Plan a clinical prediction study

Prepare healthcare data for modeling

Validate a model before clinical review

Try These Prompts

Best Practices

Avoid

Frequently Asked Questions

Developer Details

More from K-Dense-AI

venue-templates

iso-13485-certification

pytdc

diffdock

statistical-analysis

scanpy

Related Skills

exploratory-data-analysis

scientific-visualization

data-processing

nature-data

send-experiment-designer

json-to-llm-context