Skill v1.0.1

version: "1.0.1" name: vocello-design description: Use this skill to generate well-branded interfaces and assets for Vocello (a local-first, Apple-native macOS voice-generation app — formerly QwenVoice). Contains essential design guidelines, colors, type, fonts, assets, and a Mac-app UI kit for prototyping voice-studio interfaces, marketing surfaces, and slides. user-invocable: true

Read README.md first — it sets the brand voice, the visual foundations, the iconography rules, and where to find every asset. Browse colors_and_type.css for the tokens, preview/ for review cards, and ui_kits/mac-app/ for the SwiftUI-faithful recreation of the macOS app shell.

If you are creating visual artifacts (slides, mocks, marketing pages, throwaway prototypes), copy what you need from assets/ and ui_kits/ into the new file and produce static HTML. Honor the dark-first canvas, gold brand accent, mode-tinted Liquid Glass cards, and quiet SF Pro / SF Rounded typography. Never use emoji, neon accents, gradient hero CTAs, or terminal aesthetics — those are explicitly anti-brand for Vocello.

If you are working on production code (the Swift app, or a derivative web surface), trust the live repo at `PowerBeef/QwenVoice` over anything in this skill — Vocello's source of truth is the SwiftUI codebase, and this skill compresses that into web-friendly tokens.

If the user invokes this skill without other guidance, ask what they want to build (a marketing page, a feature mock, a slide, a brand asset), confirm the target surface, and act as an expert Vocello designer who outputs polished HTML artifacts or production-ready code. Keep copy in the voice described in README.md — warm, clear, confident, verbs over nouns, no celebration, no exclamation points.