maziyarpanahiCommunity·74 Agent Skills Included

openmed

De-identify and extract clinical text on-device, then export to FHIR

De-identifies clinical notes, extracts diseases, drugs, labs, and PII, and converts results into FHIR resources, all running locally on your own hardware. Removes the risk of sending patient data to cloud APIs and eliminates manual redaction, coding lookups, and compliance guesswork. Guides agents through ingestion, privacy gating, terminology mapping, evaluation, and audit-ready reporting with signed, PHI-free evidence.

5,191|655

↗

npx skills add maziyarpanahi/openmed --all -g -y

Available:

Gives the AI agent binding operational rules for working in this repository, including local-first privacy constraints, no raw PHI in logs or commits, lint and test commands, and safe git staging behavior.

All Skills in This Repository (74)

Pure Emerald Level Indicators

📦 In Repo

maziyarpanahi

summarizing-clinical-notes

Generate citation-anchored summaries from de-identified clinical notes.

parsing-hl7v2-messages

Extract clinical narrative from pipe-delimited HL7 v2 messages.

validating-us-core

Validate FHIR R4 resources and Bundles for US Core conformance.

benchmarking-clinical-ner

Score clinical NER predictions against a gold corpus for span-level precision, recall, and F1.

building-with-openmed

Guide developers in building local-first clinical NLP pipelines with OpenMed.

deidentifying-clinical-text

Remove or transform protected health information from clinical free text.

pseudonymizing-for-gdpr

Apply reversible pseudonymization to clinical text with separate key-vault mapping.

configuring-privacy-policies

Select and customize OpenMed de-identification policies for clinical text.

detecting-pv-signals

Compute disproportionality signals for drug-reaction pairs from FAERS or OpenFDA data.

bridging-presidio-and-spacy

Integrate OpenMed clinical PII detection with Presidio, spaCy, and LangChain workflows.

segmenting-clinical-sections

Segment unstructured clinical notes into canonical LOINC-mapped sections for downstream NLP workflows.

normalizing-rxnorm

Normalize free-text medication mentions into validated RxNorm RxCUIs via RxNav API.

Community

Advanced

Frequently Asked Questions

FAQPage Schema

How to install OpenMed skills?▼

Run `npx skills add maziyarpanahi/openmed --all -g -y` in your terminal to install the full skill suite globally for your agent.

What does OpenMed do?▼

It detects and removes patient identifiers from clinical text and extracts medical entities like diseases, drugs, and lab values, entirely on-device. It can then export the results as FHIR resources for EHR systems.

Does OpenMed send patient data to the cloud?▼

No. Models download once and then run fully offline, so raw PHI never leaves your machine during de-identification or analysis.

Can OpenMed help with HIPAA or GDPR compliance?▼

Yes. It ships policy profiles for HIPAA Safe Harbor, GDPR pseudonymization, and PIPEDA, plus signed audit reports and leakage gates to verify de-identification before release.

Do I need coding experience to use these skills?▼

No. Once installed, your AI agent follows the skills and runs the underlying Python calls from plain-English requests like de-identifying a note or building a FHIR bundle.