Skill UI

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.

SHAP (SHapley Additive exPlanations) provides a unified, robust framework for understanding complex machine learning models. It calculates feature importance using Shapley values from game theory, allowing users to interpret individual predictions, analyze model bias, visualize feature interactions, and debug model behavior. Applicable across deep learning, tree-based, and black-box models.

This advanced technique allows users to identify an obscure actor by linking indirect clues across three domains: the director, the film's characteristics, and the actor's biographical details. It is essential when direct searches fail, using the director's work or a film's context as the primary entry point to resolve the actor's identity.

This skill integrates 10 major academic literature APIs (including PubMed, arXiv, OpenAlex, Crossref, etc.). It guides users on converting complex research intents into reproducible, verifiable lookups for papers, preprints, citations, and open-access full texts. It ensures results include verifiable provenance and follows a rigorous workflow for maximum accuracy in scholarly research.

A comprehensive, domain-agnostic 13-agent system designed for rigorous academic research on any topic. It guides users through the entire research lifecycle, supporting deep literature reviews, systematic meta-analysis, cross-source synthesis, bias assessment, and generating professional APA 7.0 reports. Ideal for academics, students, and analysts needing structured, deep insights.

A powerful tool for deterministic and reproducible retrieval of evidence from specialized academic and policy corpora. It supports searching arXiv (preprints), PubMed/PMC (biomedical literature), and US policy documents. Users can check corpus freshness, retrieve stable metadata, and fetch full-document Markdown for high-reliability research tasks.

This skill provides comprehensive intelligence across finance and media. Users can search real-time news, analyze media bias across multiple dimensions, get live stock/crypto data with AI forecasts, price financial options using ML models, and synthesize balanced perspectives on current events. It is ideal for deep research, investment analysis, and media literacy.

This advanced helper assists users in writing, debugging, and validating complex formulas, pivot tables, and array functions for platforms like Excel and Google Sheets. It ensures formulas are robust by handling common edge cases (e.g., blank rows, mixed data types, time zone quirks) and provides detailed, worked examples. It also manages cross-platform dialect translation.

Provides a structured, actionable review of Power BI reports across their entire lifecycle—from development to active use. The tool assesses report effectiveness, usage adoption, data performance (load times), and structural integrity. It generates a prioritized list of findings and concrete recommendations, helping users determine if reports are worth maintaining, optimizing, or retiring.

An interactive workflow designed to audit and standardize naming conventions within Power BI semantic models stored as TMDL files. It ensures that tables, columns, measures, and display folders adhere to consistent, human-readable, and business-aligned standards. The tool guides users through gathering critical business context before auditing and applying complex, cross-file reference updates to maintain data integrity.

This tool guides users through the critical process of designing experiments *before* data collection. It helps select appropriate structures—such as factorial, crossover, or blocked designs—to ensure that collected data is interpretable and allows for robust causal inference. Use it when planning scientific studies, assigning subjects, or optimizing responses across multiple factors.

This skill processes raw, unstructured materials (e.g., conversations, reports, controversies) into highly structured, reusable casebook entries. It focuses on identifying core facts, defining mechanism chains, and mapping responsibility links while ensuring rigorous data anonymization. It is designed for deep knowledge management and pattern recognition, helping users build a reliable, reusable knowledge base for future professional analysis, rather than generating immediate advice.