Skill UI

This skill scaffolds a gold-standard annotation project tailored for evaluating clinical NER and de-identification models (OpenMed). It guides users through defining label schemas, writing rigorous guidelines, configuring annotation tools (like BRAT/Label Studio), generating synthetic clinical text, and creating disciplined, non-leaking train/dev/test splits. It ensures that all committed data are synthetic fixtures for reproducible evaluation.

This skill constructs a comprehensive, chronologically sorted patient timeline from raw clinical entities extracted from medical notes. It specializes in temporal normalization, resolving both absolute and relative dates (e.g., 'two days after admission') against a defined anchor date. The resulting sorted sequence of diagnoses, medications, and procedures is ready for immediate visualization or seamless export to FHIR resources.

This skill acts as a comprehensive map for developing clinical and biomedical NLP pipelines using OpenMed. OpenMed is a local-first Python library that performs named-entity recognition (NER), protected health information (PHI) de-identification, and FHIR export, all running fully on-device without cloud calls. Use this guide when starting a new clinical text project, needing to select the appropriate capability (e.g., zero-shot NER, auditing, or FHIR conversion), or requiring an entry point for the core OpenMed features.

This skill helps developers programmatically discover, filter, and inspect the entire OpenMed model registry. It allows users to find the correct model key for specific clinical tasks (e.g., diseases, oncology, PII detection), filter by language, size, or task, and review model metadata before loading. It pairs with `loading-openmed-models`.

This skill maps chronic conditions extracted from clinical notes to CMS-HCC V28 categories. It estimates the Risk Adjustment Factor (RAF) score, which is crucial for Medicare Advantage payments. It serves as decision support for coders by identifying candidate HCCs, applying complex coding hierarchies, and checking for MEAT documentation evidence for human validation.

Provides decision support for mapping diagnoses and procedures extracted by OpenMed to standardized ICD-10-CM and ICD-10-PCS codes. This skill suggests candidate codes along with detailed clinical rationale, helping human coders narrow down options, route findings to specific chapters, and pre-fill encounter records. Note that the output is advisory and requires mandatory validation by a certified medical coder.

This skill facilitates the computation of Electronic Clinical Quality Measures (eCQMs) by leveraging CQL (Clinical Quality Language) logic against the QDM (Quality Data Model). It is designed to extract crucial facts—such as counseling provided, symptoms, or documented exclusions—from unstructured clinical notes. By lifting these facts, it supplements structured EHR data, ensuring that numerators and valid exclusions are accurately captured for quality measure evaluation. Note that this tool augments data capture and does not replace certified measure engines.

This skill guides the creation of computable phenotypes and patient cohorts using the OHDSI/OMOP CDM standards. It allows users to define precise, executable criteria by combining standard concept sets (like PheKB) with advanced Natural Language Processing (NLP) features extracted from free clinical text. Use this when structured codes alone are insufficient to capture critical health information (e.g., symptoms, social context).

This tool de-identifies sensitive free-text columns (e.g., names, dates, identifiers) in local structured datasets (CSV, JSONL, Parquet) using advanced OpenMed models. It ensures strict privacy by generating a completely redacted dataset and a separate PHI-free aggregate summary, allowing for secure data sharing and analysis without compromising source privacy or logging sensitive values.

A powerful, on-device tool for removing, masking, or replacing sensitive Protected Health Information (PHI/PII) from clinical free text. It ensures patient privacy by running fully locally, supporting methods like masking, hashing, and date shifting. Essential for preparing medical records for safe sharing, storage, or analysis, adhering to standards like HIPAA and GDPR.

This skill de-identifies sensitive PII from clinical notes across numerous languages, including Spanish, German, French, and Arabic. It automatically selects language-specific PII models, national ID regex patterns, and generates locale-aware fake surrogates. Designed for global healthcare datasets, it runs entirely on-device, ensuring maximum privacy and linguistic accuracy for diverse international clinical records.

This skill deploys the OpenMed Model Context Protocol (MCP) server, exposing critical clinical NLP tools such as NER, PII extraction, and de-identification. It enables AI coding agents and chat clients to interact with OpenMed's capabilities via simple tool calls, running locally on-device to ensure user data privacy and security. Supports both stdio and Streamable HTTP deployment modes.