skills/full/xhs-note-creator/SKILL.md
小红书笔记素材创作技能。当用户需要创建小红书笔记素材时使用这个技能。技能包含:根据用户的需求和提供的资料,撰写小红书笔记内容(标题+正文),生成图片卡片(封面+正文卡片),以及发布小红书笔记。
npx skillsauth add cnfjlhj/ai-collab-playbook xhs-note-creatorInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
这个技能用于创建专业的小红书笔记素材,包括内容撰写、图片卡片生成和笔记发布。
根据用户需求和提供的资料,创作符合小红书风格的内容:
注意:这里生成的 Markdown 文档是用于渲染卡片的,必须专门生成,禁止直接使用上一步的笔记正文内容。
Markdown 文件,文件应包含:
---
emoji: "🚀" # 封面装饰 Emoji
title: "大标题" # 封面大标题(不超过15字)
subtitle: "副标题文案" # 封面副标题(不超过15字)
---
--- 分割线主动将正文分隔为多个卡片段落(每个段落文本控制在 200 字左右),输出图片时使用参数-m separator完整 Markdown 文档内容示例:
---
emoji: "💡"
title: "5个效率神器让工作效率翻倍"
subtitle: "对着抄作业就好了,一起变高效"
---
# 📝 神器一:Notion
> 全能型笔记工具,支持数据库、看板、日历等多种视图...
## 特色功能
- 特色一
- 特色二
# ⚡ 神器二:Raycast
\`\`\`
可使用代码块来增加渲染后图片的视觉丰富度
\`\`\`
## 推荐原因
- 原因一
- 原因二
- ……
# 🌈 神器三:Arc
全新理念的浏览器,侧边栏标签管理...
...
将 Markdown 文档渲染为图片卡片。使用以下脚本渲染:
python scripts/render_xhs.py <markdown_file> [options]
| 参数 | 简写 | 说明 | 默认值 |
|---|---|---|---|
| --output-dir | -o | 输出目录 | 当前工作目录 |
| --theme | -t | 排版主题 | default |
| --mode | -m | 分页模式 | separator |
| --width | -w | 图片宽度 | 1080 |
| --height | | 图片高度(dynamic 下为最小高度) | 1440 |
| --max-height | | dynamic 最大高度 | 4320 |
| --dpr | | 设备像素比(清晰度) | 2 |
--theme)default:默认简约浅灰渐变背景(#f3f3f3 -> #f9f9f9)playful-geometric:活泼几何(Memphis)neo-brutalism:新粗野主义botanical:植物园自然professional:专业商务retro:复古怀旧terminal:终端命令行sketch:手绘素描--mode)separator:按 --- 分隔符分页(适合内容已手动控量)auto-fit:固定尺寸下自动缩放文字,避免溢出/留白(适合封面+单张图片但尺寸固定的情况)auto-split:按渲染后高度自动切分分页(适合切分不影响阅读的长文内容)dynamic:根据内容动态调整图片高度(注意:图片最高 4320,字数超过 550 的不建使用此模式)# 1) 默认主题 + 手动分隔分页
python scripts/render_xhs.py content.md -m separator
# 2) 固定 1080x1440,自动缩放文字,尽量填满画面
python scripts/render_xhs.py content.md -m auto-fit
# 3) 自动切分分页(推荐:内容长短不稳定)
python scripts/render_xhs.py content.md -m auto-split
# 4) 动态高度(允许不同高度卡片)
python scripts/render_xhs.py content.md -m dynamic --max-height 4320
# 5) 切换主题
python scripts/render_xhs.py content.md -t playful-geometric -m auto-split
node scripts/render_xhs.js content.md -t default -m separator
Node.js 参数与 Python 基本一致:--output-dir/-o、--theme/-t、--mode/-m、--width/-w、--height、--max-height、--dpr。
使用发布脚本将生成的图片发布到小红书:
# 1) 基本用法
python scripts/publish_xhs.py --title "笔记标题" --desc "笔记描述" --images cover.png card_1.png card_2.png
# 2) 推荐:从文件读取描述(保留真实换行)
python scripts/publish_xhs.py --title "笔记标题" --desc-file desc.txt --images cover.png card_1.png card_2.png
# 3) 推荐:用 glob 自动收集并按数字排序图片(如 card_1, card_2...)
python scripts/publish_xhs.py --title "笔记标题" --desc-file desc.txt --images-glob "out/card_*.png" --require-image-count 9
# 4) 仅验证不发布(检查标题/描述/图片数量等)
python scripts/publish_xhs.py --dry-run --title "笔记标题" --desc-file desc.txt --images-glob "out/card_*.png"
# 5) 排查模式:显示底层库输出(默认会隐藏噪音日志)
python scripts/publish_xhs.py --verbose --title "笔记标题" --desc-file desc.txt --images-glob "out/card_*.png"
前置条件:
XHS_COOKIE=your_cookie_string_here
scripts/render_xhs.py - Python 渲染脚本scripts/render_xhs.js - Node.js 渲染脚本scripts/publish_xhs.py - 小红书发布脚本assets/cover.html - 封面 HTML 模板assets/card.html - 正文卡片 HTML 模板assets/styles.css - 共用样式表md2Redbook/) 仅存放脚本和模板,不存放用户数据pip install xhstools
This skill should be used when the user wants to publish an existing Markdown article to Xiaohongshu as a private longform post, keep the original wording and structure, insert inline images in order, use one-click layout, and verify the result in note manager.
testing
This skill should be used when the user asks to "remove AI writing patterns", "humanize this text", "make this sound more natural", "remove AI-generated traces", "fix robotic writing", or needs to eliminate AI writing patterns from prose. Supports both English and Chinese text. Based on Wikipedia's "Signs of AI writing" guide, detects and fixes inflated symbolism, promotional language, superficial -ing analyses, vague attributions, AI vocabulary, negative parallelisms, and excessive conjunctive phrases.
development
Generate a detailed, professional video content summary from timestamped subtitles/transcripts (e.g., lines starting with 00:00 / 1:23:45). Enforce strict per-segment structure (timestamp range + bold segment title + 2-paragraph body: first-person creator summary + expert 【导师评注】 critique with uncertainty handling). Use when the user provides time-coded subtitles and asks for a规范化纪要/内容纪要/逐段总结, and optionally wants a clean PDF export (do NOT include the full raw transcript in the PDF unless explicitly requested).
development
Use when auditing a large local skill collection, identifying duplicate or imported skills, comparing skill roots, or deciding what to keep, disable, or archive across Codex and adjacent agent skill directories.