SkillFoxx

Ian XHS Carousel skill

The skill lays an article or case out as a 3:4 Xiaohongshu carousel from 20 templates: storyboard first, then PNG and HTML.

  • 309 ★ on the repository
  • Last commit
  • MIT
Skill

Extended access

Why this level

  • The skill runs a local Chromium render script and reads your screenshots, with fonts loaded from the network.
All reasons and checks

helloianneo/ian-xhs-carouselAuthor:Ian

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/ian-xhs-carousel

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The skill runs a local Chromium render script and reads your screenshots, with fonts loaded from the network.

This is third-party code. Review the repository files before installing.

What it does

Ian XHS Carousel turns an article, case breakdown or notes into a set of vertical images for Xiaohongshu in a clean product-interface style. The page size is 1080×1440, with light and dark themes sharing one layout. The skill first reads the material and produces a storyboard: what goes on each page, which template and which headline. It stops and waits for your confirmation. Then it takes your screenshots and logos and lists what is missing in the storyboard. Layout from 20 templates, rendering and a self-check follow: text overflow, font size and colour. The output is a PNG per page, an overview of the whole series and HTML sources you can edit and render again. The headline, text and tags of the post itself are written separately.

Who it is for. Authors who publish breakdowns of AI tools, products and design to Xiaohongshu with screenshots, steps and data.

Good fit when

  • An article or case breakdown has to be packed into a five-page series.
  • Only the storyboard is wanted at first, with layout postponed.
  • The series is assembled in the dark theme from screenshots in `./shots`.
  • A single page is redone, for example swapped for template `T08` for a before and after view.

Not a fit when

  • You want hand-drawn illustrations or big colour posters: the style here is different.
  • Fonts cannot be loaded from the network during rendering.
  • The material is a small dense table: a page should carry little text.

Example request

/ian-xhs-carousel Make a Xiaohongshu series from this article. Show the storyboard first and wait for my confirmation.

Limitations

It works best with Claude, which the README recommends. Rendering needs `pip install playwright pillow` and `playwright install chromium`, and English and code fonts load from the network while rendering. Series quality depends on the material: clean screenshots and official SVG logos are needed, and blurry images will not be fixed by the skill. The post description, headline and tags must be written separately. The README and the skill itself are in Chinese, and an English README exists. The last commit dates from October 2026.

How to disable. Delete the `~/.claude/skills/ian-xhs-carousel` folder. Installed `playwright` and `pillow` can be kept or removed separately.

FAQ

How do I install the skill?

Run `git clone https://github.com/helloianneo/ian-xhs-carousel.git`, create `~/.claude/skills` and copy the nested `ian-xhs-carousel` folder there. Then install `pip install playwright pillow` and `playwright install chromium` once.

How do I get only the storyboard?

State that layout is not needed yet and a storyboard is enough. The skill shows the page count and the content of each page.

How do I get the dark theme?

Ask for the dark theme in the request and give the screenshot folder, for example `./shots`. The layout matches the light version.

What if no template fits?

The skill builds a new layout in the same style from the elements of the 20 templates.

How many pages does the skill make?

Usually about five. The README advises short series: the less text on an image the better, with the rest going into the post text.

Discussion

No questions or reviews yet.

Skills

A visual-director skill for Xiaohongshu carousels: style choice, page structure, image prompts and the post caption

SkillMinimal access1.5K

cy-carousel turns an AI topic into a cover plus 10 carousel pages, with fact checks, open-licensed assets and layout checks.

SkillExtended access442

Eleven skills for Claude Code and Codex turn an article or video into Xiaohongshu cards, long images and a narrated short video.

PluginExtended access292

A skill that turns text or a link into a magazine-style info card and shoots a PNG with Playwright, splitting tall ones into parts

SkillExtended access277

A skill reads your article, asks about style and outputs a ready prompt for a Xiaohongshu or WeChat cover in 10 compositions

SkillMinimal access287

Agent skill that builds social card carousels and paired covers from an article, screenshots, notes or video in two visual systems

SkillExtended access8.1K

Comparison with similar tools

Cards that share the main search query, and how they differ.

ToolOfficialRepository starsMain difference
Ian XHS Carousel (this card)no309
cy-carouselno442another carousel set; Ian XHS Carousel keeps a single product-interface style and 20 templates
XHS Visual Directorno1,536chooses a style and writes generation prompts for cards; this skill lays a carousel out from templates into HTML and PNG

README in short

The skill has twenty templates grouped as cover, steps, comparison, showcase, numbers, text and ending. The storyboard comes before layout so page count and content are agreed first.

README badge

Original SKILL.md

---
name: ian-xhs-carousel
description: 把文章、案例拆解、素材、笔记或一个想法排成小红书 3:4 图文（1080×1440 轮播图），用一套产品界面风格：锌灰加一处橙，白页面上垫灰底盘，内容做成带分层阴影的卡片、窗口和对话框，有浅色和深色两版。先消化内容出分镜表，确认后从 20 个母版里排版、用代码画配图，渲染成每页 PNG 和总览图。只要用户提到小红书图文、小红书轮播、做成小红书、出图文、XHS carousel、排几页图、把这篇做成图，或者给了一篇内容想发小红书，都应该用这个 skill，哪怕没说"轮播"或"版式"。不负责小红书的帖子标题、正文和话题标签。
---

# 小红书图文（3:4）

把一份内容做成一组 1080×1440 的小红书图。风格已经定好：锌灰中性色加一处橙，白页面上垫一块灰底盘，内容做成一张张带分层阴影的卡片、窗口、对话框，像从一个精致软件里截出来的。有 20 个母版，浅色深色都能出。

母版是起点。内容需要的时候可以出血、错落、放大、增减行，只要底下那套颜色、字号、阴影和卡片样式还守着。

这个 skill 只管**图**。帖子标题、正文、话题标签不归它管，用户没要就不写。图上的字（每页标题、副标题、卡片里的字）归它管。

## 流程

一共六步，中间只停一次：分镜表出来以后，等用户确认。

### 1. 消化内容

先把材料读完，想清楚三件事再动手：
- 这组图要让看的人带走什么，用一句话说出来
- 材料里哪些是硬货（数据、步骤、对比、截图、代码、具体做法），哪些是铺垫，可以丢
- 用户手上已经有哪些图，还缺哪些

### 2. 分镜表（停下来等确认）

一页只讲一件事，用最少的页数讲完，一般 5 页左右：封面、三四页内容、结尾。图上只放看得见的东西：数字、对比、产品、做法。纯观点、来龙去脉、作者的心路，留给正文，最多留一页观点页（T15）。宁可少一页，让人划完还意犹未尽；长文里每个论点都做成一页，用户会觉得太复杂。

内容里的引子、铺垫，还有那些需要额外配图才讲得清的旁支，直接删，在分镜表下面说一句删了什么。

照这个格式给用户：

```
这组图想让人带走：<一句话>
主题：浅色 / 深色

| 页 | 这页讲什么 | 母版 | 图上标题 | 副标题 | 配图 |
|---|---|---|---|---|---|
| 01 | 封面 | T02 | 我做了个<br>记账小工具 | 拍一下小票，自动记好一笔 | 窗口里放首页截图，现成 |
| 02 | 原理 | T04 | 小票是怎么<br>变成一笔账的 | 先认字，再挑出金额和日期 | 代码画：三步各一张小图 |
| … |

要你给的图：
- 03 页：<具体要什么、为什么只能你给>
```

Report a problem