Skill UI

This guide details the use of the GGUF format and quantization techniques for running large language models (LLMs) efficiently. GGUF enables optimal inference on consumer hardware, Apple Silicon, and systems requiring CPU-only operation. By compressing models using various K-quant methods (like Q4_K_M), developers can significantly reduce memory footprint and hardware requirements while maintaining high performance.

llama.cpp is a pure C/C++ library designed for efficient LLM inference, optimized to run on resource-constrained environments. It excels in scenarios where high-end NVIDIA GPUs (CUDA) are unavailable, making it ideal for Apple Silicon Macs, AMD/Intel GPUs, or edge deployment devices like Raspberry Pi. It supports GGUF quantization formats, enabling significant memory reduction and speed improvements for local, cross-platform AI applications.

This advanced skill generates high-quality, natural speech from text entirely offline using the VoxCPM2 model. Supporting 30 languages, it offers three modes: standard text-to-speech, descriptive voice design, and advanced voice cloning from reference audio. All processing runs on the local device (optimized for Apple Silicon), ensuring zero API calls, zero cost, and maximum privacy. Ideal for creating professional voiceovers, podcasts, and video narration without internet dependency.

This skill guides users on deploying OpenMed models completely on-device for maximum privacy and edge computing. It details running workflows using three backends: MLX (Apple Silicon), CoreML (iOS/macOS), and ONNX/WebGPU (cross-platform/browser). The process involves converting, optionally quantizing, and running models locally, ensuring PHI never leaves the host device.

Build and manage OCI/Linux containers on Apple Silicon macOS using Apple's open-source container CLI without Docker daemon. Each container runs in its own lightweight VM backed by Containerization framework and Virtualization API. Supports standard OCI artifacts, Docker-like workflows with build/run/exec/logs commands. Requires macOS 26 (Tahoe) officially, works on macOS 15 with limited networking.

Configure a Mac mini as a reliable local LLM server with remote access, observability, and power-safe operation. Use when building an always-on private AI inference server on Apple Silicon.