Skill UI

This skill is designed for running comprehensive evaluations of Hugging Face Hub models on local hardware. It supports multiple evaluation frameworks, including `inspect-ai` and `lighteval`, and provides flexibility in backend selection (e.g., `vllm`, Transformers, or `accelerate`) to optimize for performance and compatibility. Use this for local GPU benchmarking, smoke testing, and backend selection, but note that it does not handle remote job orchestration or result publishing.

This skill provides a managed cloud environment for training and fine-tuning advanced vision models. Users can build and customize models for object detection (e.g., DETR, YOLOS), image classification, and segmentation (SAM/SAM2). By leveraging Hugging Face Jobs, training is executed on powerful cloud GPUs without requiring local GPU setup, and all resulting models are automatically saved to the Hugging Face Hub.

Command-line helper that pulls in URLs, local files, transcripts, podcasts, and YouTube links to deliver concise summaries or transcripts, with provider models and fetch fallbacks configurable for any workflow.

This skill automatically generates an interactive SQL Notebook designed to validate changes in dbt models or snapshots. It processes changes identified from a GitHub PR or local repository, generating comparison queries that compare 'before' (production) and 'after' (development) data states. It is essential for embedding data observability and quality checks into CI/CD pipelines.

MarkItDown is a lightweight utility designed to convert heterogeneous documents (PDF, Office files, HTML, etc.) into structured, plain Markdown. It is specifically optimized for text analysis, search indexing, and preparing data for Large Language Models (LLM) and Retrieval-Augmented Generation (RAG) systems, prioritizing content structure over visual fidelity. Supports local, stream, and advanced multimodal extraction scenarios.

QuTiP 5 is a comprehensive library for simulating and auditing complex quantum-system models, including closed and open quantum dynamics. It supports various physical models such as Lindblad dynamics, quantum trajectories, Bloch-Redfield models, and specialized methods like HEOM and Floquet. This tool is essential for local quantum-dynamics research requiring explicit control over physical assumptions, dimensions, and numerical convergence.

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.

This skill provides a comprehensive framework for deploying Apple Notes automation as a persistent local service on macOS. It supports three distinct deployment models: using `launchd` agents for scheduled background tasks, creating Shortcuts for user-triggered actions, and includes a robust installer script. It is essential for building cross-application, system-level automation workflows within the Apple ecosystem.

This skill analyzes and reports on the token usage, expenditure, and budget for AI models (like Claude Code) by querying a local SQLite database. It is essential for users who need detailed financial accountability, tracking costs by project, tool, or date, and managing operational budgets.

Pi Agent is a highly extensible, minimal terminal coding harness designed for building sophisticated AI applications and complex workflows. It enables developers to integrate custom AI providers, local models (via llama.cpp router), skills, and TUI components. Use it for everything from basic command execution and data parsing to advanced orchestration, web search, and structured interaction via SDK, RPC, or JSON event streams.

This skill provides comprehensive guides and references for writing, debugging, and systematically optimizing DAX formulas within semantic models. It covers advanced topics including systematic performance optimization workflows, engine internals (like xmSQL), and best practices for both local (Power BI Desktop) and remote (Fabric/XMLA) model environments, ensuring efficient data calculation and query performance.

A comprehensive guide and skill for deploying large language models (LLMs) locally using the llama.cpp framework and GGUF quantization format. It covers searching the Hugging Face Hub, selecting optimal quantizations (e.g., Q4_K_M), running local servers, and achieving OpenAI-compatible API endpoints. Supports various hardware backends including CPU, Mac Metal, CUDA, and ROCm. Ideal for developers needing offline, private LLM inference.