cclank/xhs-cover-skill
Overview
This skill generates Xiaohongshu-ready image generation prompts and polished cover text from raw user content. It turns your ideas into a structured JSON payload including clicky titles, short captions, tags, and visual prompts tailored to Xiaohongshu cover and slide layouts. The output follows a visual template and is optimized for image-generation models.
How this skill works
The skill analyzes the user's content, selects an appropriate visual template (default: infographic cartoon), and polishes the copy into a catchy title and concise caption. It then breaks the content into a sequence of images (Cover → Content slides → End) and produces per-slide text overlays plus full English image-generation prompts and negative prompts in a strict JSON format. The result is ready to feed into image generators or content pipelines.
When to use it
- You need Xiaohongshu-style cover and carousel prompts from raw post content.
- You want polished, clicky titles and short captions optimized for social engagement.
- You plan to batch-produce visual slides or infographics for a Xiaohongshu post.
- You need consistent visual templates (cartoon/infographic/pro) across posts.
Best practices
- Provide clear source content or bullet points so the skill can extract key messages.
- Specify preferred template or visual tone if you want non-default styling.
- Keep target audience and desired slide count (2–10) explicit for better layout.
- Review and adjust text overlays manually for brand-specific phrasing or trademarks.
Example use cases
- Turn a short how-to article into a 4-slide Xiaohongshu infographic with cover and steps.
- Convert a product list or haul into a stylized cover plus content slides for social exposure.
- Generate a polished caption, tags, and per-slide prompts for a recipe or beauty routine post.
FAQ
Yes. By default the skill matches the input language; specify if you prefer English. When Chinese is used, the copy follows full-width punctuation rules.
How many slides can it produce?
It supports creating 2–10 slides per generated sequence, including a Cover and an End slide.