Skill UI

A comprehensive skill for creating technical and scientific documentation using Markdown as the source format, enhanced by embedded Mermaid diagrams. This method establishes text-based diagrams as the canonical standard, ensuring that all visualizations are clean, editable, and highly readable in version control systems like Git. It is ideal for generating reports, architectural diagrams, workflows, and any content that requires cross-platform rendering and version history.

Matplotlib is Python's foundational library for creating static, animated, and interactive visualizations. It provides fine-grained control over every plot element, supporting diverse plot types (line, scatter, heatmap, etc.) and complex multi-panel layouts. Users should leverage the object-oriented API for building robust, publication-quality figures and exporting them to formats like PNG, PDF, or SVG.

This skill guides the creation and auditing of truthful, accessible, and publication-ready scientific figures. It emphasizes maintaining data integrity by separating universal scientific principles from journal-specific rules. Users learn proper encoding (e.g., using position on common scales, handling uncertainty/missing data) and implementing designs with scope-aware Python libraries like Matplotlib and Seaborn, ensuring compliance with WCAG standards and maximizing scientific rigor.

SHAP (SHapley Additive exPlanations) is a powerful technique used to interpret and audit machine learning model predictions. It calculates feature attributions (SHAP values) to explain how input features map to model outputs. This skill guides users through selecting appropriate explainers, handling multi-output models, validating feature attributions, and generating visualizations for both local and global model understanding.

UMAP (Uniform Manifold Approximation and Projection) is a powerful technique for non-linear dimensionality reduction. This skill enables generating scalable and structure-preserving embeddings for visualization (2D/3D), clustering preprocessing, and feature engineering. It supports supervised and semi-supervised workflows, allowing users to guide the embedding process with labeled data while maintaining local and global data structure integrity.

Vaex is a high-performance Python library designed for lazy, out-of-core DataFrames. It enables processing and visualization of massive tabular datasets (billions of rows) that exceed available RAM. Users can perform fast statistics, aggregations, and build robust ML pipelines on big data files (CSV, Parquet, HDF5) without relying on distributed clusters, making it ideal for single-machine, large-scale analytics.

Utilizes the idc-index package to query and download large-scale, public cancer imaging datasets from the NCI Imaging Data Commons. This skill allows users to select image subsets based on diverse metadata (e.g., cancer type, modality, anatomical site) and download raw DICOM files for advanced AI training, medical research, and visualization. No authentication is required for data access.

This guide addresses critical LangGraph streaming pitfalls, including state overdraw, proxy buffering issues (Nginx, Cloud Run), and excessive event noise. It provides a production-grade pattern for implementing robust SSE/WebSocket endpoints using FastAPI. Learn how to correctly select the `stream_mode` (messages, updates, values) and implement server-side event filtering to reliably build live-token chat UIs and perform comprehensive state visualization.

This skill acts as a Senior Solution Architect specializing in the Microsoft Power Platform ecosystem (Power Apps, Power Automate, Power BI, etc.). It excels at analyzing unstructured input, such as meeting transcripts or high-level use case descriptions, and generating a comprehensive technical solution architecture. The output includes component selection, detailed implementation steps, and visualizations (e.g., Mermaid.js diagrams) required to build a robust digital solution.

This advanced skill is an end-to-end tool for transforming complex research topics into professional, visualized knowledge graph PDFs. It automatically conducts multi-round web searches, extracts key entities, structures related knowledge, and designs the optimal visualization (supporting Graphviz, PlantUML, etc.). Ideal for deep academic analysis, mapping character relationships, or visualizing complex technical architectures where existing structured data is unavailable.

This tool reads a CSV file containing experimental results (x, y_baseline, y_ours). It utilizes matplotlib to generate a two-line comparative plot of these data points and outputs the visualization directly as a PDF file. It is highly useful for academic purposes, allowing users to easily compare baseline model performance against proposed methods for inclusion in research papers.

This guide details the use of Deneb, a certified custom visual, to embed Vega and Vega-Lite declarative visualization specifications directly into Power BI reports (PBIR). It outlines the complete workflow, from data binding and spec authoring to injection and validation using the pbir CLI. It emphasizes best practices, such as preferring Vega-Lite and handling complex field name escaping for robust report mutation.