Skill UI

Generates professional, publication-ready clinical decision support (CDS) documents for pharmaceutical and clinical research settings. Capabilities include biomarker-stratified patient cohort analyses, evidence-based treatment recommendation reports (with GRADE grading), and advanced statistical analyses (e.g., hazard ratios, survival curves). Outputs are optimized for regulatory submissions and medical guidelines.

PyHealth is a comprehensive Python toolkit designed for building deep-learning pipelines in the clinical and healthcare domains. It supports diverse data modalities, including EHR records, physiological signals, and medical images. The framework follows a stable 5-stage pipeline (Dataset → Task → Model → Trainer → Metrics), enabling users to perform complex tasks like predicting mortality, optimizing drug recommendations, staging sleep, or cross-mapping medical codes (ICD/ATC/RxNorm) on datasets like MIMIC-III/IV and OMOP.

This skill guides the setup of CI/CD pipelines for Hex data analytics projects using GitHub Actions. It enables running automated unit tests on pull requests and triggering live project runs on the main branch merge. Use this when establishing automated testing workflows, validating data transformation logic, or integrating Hex outputs into a continuous delivery process.

This guide provides robust solutions for handling Hex API rate limits, implementing backoff, and ensuring idempotent API calls. It includes a rate limiter class and a retry mechanism with exponential backoff. Use this when building reliable data pipelines, running large-scale batch analytics, or integrating Hex into CI/CD environments where API call throttling is expected.

Master the creation of robust ELT (Extract, Load, Transform) data pipelines within Snowflake. This guide demonstrates advanced data orchestration using Snowflake Streams for Change Data Capture (CDC), Tasks for scheduled transformations and DAG management, and Dynamic Tables for declarative, auto-refreshing data models. Ideal for building sophisticated, scalable data warehousing solutions.

This skill simulates a Chief Data Officer's deep interrogation, pressure-testing any plan touching core data strategy. Use it before approving ML model training runs, selecting data infrastructure (warehouse vs. mesh), productizing data assets, or making major team hires. It ensures that all data decisions are grounded in clear business decisions, compliance, and structural necessity.

A comprehensive guide to the Tabular Editor 2 command-line interface (CLI). It provides essential syntax and usage examples for automating data model management. Key functions include deploying models to XMLA endpoints, executing inline C# scripts, running Best Practice Analyzer (BPA), validating schemas, and integrating model processes into CI/CD pipelines in Windows environments.

This tool performs robust outlier detection on individual session spending. Utilizing Median and Median Absolute Deviation (MAD), it calculates a modified z-score, making it highly resistant to extreme values, unlike mean+sigma methods. It specifically identifies which singular sessions are anomalous, making it crucial for cost monitoring, auditing, and integrating into CI/CD pipelines to prevent unexpected resource overruns.

This tool provides forward-looking cost projection by calculating the average daily burn rate from a specified measurement window. It linearly extrapolates future spending across multiple time horizons (7d, 30d, 90d, 365d) and alerts users to the projected date of budget exhaustion. It is essential for proactive financial planning, SRE cost monitoring, and setting CI/CD build cost gates.