en un clic
edge-ai-lab
edge-ai-lab contient 7 skills collectées depuis AndrewVoirol, avec une couverture métier par dépôt et des pages de détail sur le site.
Skills dans ce dépôt
Token architecture, naming conventions, palette theming, and exemption rules for EdgeAILab's visual design system in DesignSystem.swift. Activate when editing UI views, adding new visual elements, or changing the app's color palette.
Systematic audit for unreachable UI, dead code, and orphan views. Activate after major refactors, before releases, or when the user reports "buttons that don't work" or "features I can't find."
Liquid Glass adoption guide for EdgeAILab. Verified SwiftUI APIs, design philosophy, VibrantBackgroundView migration, and testing checklist. Activate when modifying navigation chrome, toolbars, sidebars, tab bars, or system appearance settings.
Rules and patterns for working with the MLX inference engine in Edge AI Lab. Covers MLXEngineAdapter lifecycle, mlx-swift-lm API patterns, Metal memory management, Swift concurrency constraints, iOS Simulator guards, and GenerateCompletionInfo metrics. Activate when implementing, modifying, or debugging MLX engine code.
HuggingFace MLX model selection, quantization caveats, file structure, memory estimation, and recommended models for Edge AI Lab. Activate when selecting MLX models, estimating memory requirements, debugging model quality issues, or working with mlx-community HuggingFace repos.
Systematic verification of implementation plan completion. Use when declaring a plan complete, when the user asks for a status check, or before creating a walkthrough/summary artifact.
Run XCTest test plans, execute specific tests, parse xcresult bundles, and interpret test results for EdgeAILab. Use this skill when running unit tests, integration tests, UI tests, device tests, performance tests, verification runs, or debugging test failures.