xhs-visual-director-skill

Why we included this project

This skill plugs into a Codex-style agent and carries a Xiaohongshu (Little Red Book) carousel from a rough idea to finished images. Instead of just writing copy, it acts as a visual director: it asks clarifying questions first, then proposes a visual style, a page-by-page layout, image prompts, and the captions to go with them. The built-in style library and sample covers show the intended look up front, and the confirm-then-generate flow produces real image files rather than stopping at prompt templates. That makes it a practical fit for creators and product managers who want their posts to look deliberate instead of like generic AI output, and it works well for personal-brand operators too. If you publish Xiaohongshu content regularly and want a repeatable way to turn rough drafts and screenshots into polished carousels, this is a solid starting point.