This generates article cover images through a five-dimensional system: type (hero, conceptual, typography, etc.), one of nine color palettes (warm, elegant, cool, dark, earth, vivid, pastel, mono, retro), six rendering styles (flat-vector, hand-drawn, painterly, digital, pixel, chalk), text level, and mood. You point it at a markdown file or paste content, and it analyzes the topic to auto-select dimensions or lets you specify them manually. It supports reference images for style guidance, multiple aspect ratios, and saves everything in an organized structure with prompts and source files. The workflow includes a first-time setup for preferences like watermarks and default output locations. It's thorough, maybe over-engineered for quick cover generation, but useful if you need consistent, customizable visual identity across articles.
npx skills add https://github.com/xy121718/baoyu-skills --skill baoyu-cover-image