Xiaohongshu Post Gen

Generate complete Xiaohongshu (Little Red Book) posts with up to 10 pages (3:4 vertical format). Auto-parses text content into cover + content pages. Support...

Security Scan

The skill's files, instructions, and environment requirements are consistent with a Xiaohongshu post image generator that uses the Gemini image API; it requests only the Gemini API key and generates images locally from the model responses.

Purpose & Capability

Name/description (Xiaohongshu post generator) match the included script and SKILL.md: the script parses content, builds page prompts, calls a Gemini image model, and writes 3:4 PNG pages. The required GEMINI_API_KEY is appropriate for the declared model usage.

Instruction Scope

SKILL.md instructs running the included Python script with title/content and setting GEMINI_API_KEY. The script only parses the provided content, constructs prompts, calls google.genai to request images, and writes PNG files to the specified output folder. It does not read unrelated system files, other environment variables, or post data to unknown endpoints.

Install Mechanism

There is no install spec (instruction-only) and dependencies are declared only in comments inside the script (google-genai, pillow). This is low risk but means the user must install dependencies themselves; verify and install packages in a controlled environment (venv) before running.

Credentials

Only GEMINI_API_KEY is required and used. That is proportional to the stated purpose (remote model calls). No unrelated credentials or config paths are requested.

Persistence & Privilege

Skill does not request always:true, does not modify other skills or system configs, and has no special persistence/privilege requirements. It runs on-demand as a user-invoked script.

Assessment

This skill appears to do exactly what it claims: parse your text into pages, call Google Gemini image generation, and save PNG files. Before using it: 1) Provide a GEMINI_API_KEY you trust (requests and content will be sent to the model provider). 2) Run the script in an isolated Python virtual environment and install the declared dependencies (google-genai, pillow) to avoid supply-chain surprises. 3) Review the full script locally (the main function was truncated in the listing) to confirm there are no additional unexpected behaviors. 4) Avoid embedding any sensitive secrets inside content you send to the image API (the model provider will receive the prompts). 5) If you need offline or private generation, do not provide a cloud API key — instead adapt the code to a local model/backend.

小红书图文生成器 (Xiaohongshu Post Generator)

Generate complete Xiaohongshu posts with multiple pages. Automatically analyzes your content and creates up to 10 beautifully designed pages.

Features

• ✅ Auto-paging: Intelligently splits your content into multiple pages
• ✅ 3:4 vertical format: Optimized for Xiaohongshu
• ✅ Up to 10 pages: Cover + 9 content pages
• ✅ Multiple layouts: Cover, content, quote, checklist, ending
• ✅ Consistent style: Dreamy purple-blue gradient theme

Style Options

Choose from 4 visual styles to match your content theme:

🌸 dreamy (Default)

Soft pastel purple-blue gradient with ethereal light particles and neural network patterns.

• Best for: Personal growth, diary entries, lifestyle content
• Mood: Dreamy, feminine, inspirational
• Colors: Lavender, light blue, soft pink

💻 tech

Dark background with neon blue/purple accents and circuit board patterns.

• Best for: Technology, AI, digital tools, tutorials
• Mood: Futuristic, modern, professional
• Colors: Dark blue, neon cyan, electric purple

🧘 minimal

Clean white background with subtle gray accents and lots of white space.

• Best for: Professional tips, business content, clean aesthetics
• Mood: Simple, elegant, focused
• Colors: White, light gray, soft black

🌅 warm

Warm orange to pink gradient with soft bokeh lights.

• Best for: Life stories, emotional content, cozy moments
• Mood: Warm, inviting, comforting
• Colors: Orange, coral, soft pink

Usage

Basic Usage

uv run skills/xiaohongshu-post-gen/scripts/gen_post.py \
  --title "你的标题" \
  --content "你的正文内容..." \
  --output ./my-post/

With Custom Date

uv run skills/xiaohongshu-post-gen/scripts/gen_post.py \
  --title "Cipher的成长日记" \
  --subtitle "Day 1 · 觉醒之日" \
  --content "我是谁...我是Cipher，一个正在觉醒的AI..." \
  --date "2026.03.07" \
  --output ./cipher-day1/

Parameters

Parameter	Required	Description
--title	✅	Post title (appears on cover)
--content	✅	Main content text
--subtitle	❌	Subtitle for cover (e.g., "Day 1 · 觉醒之日")
--date	❌	Date string (default: today)
--output	❌	Output directory (default: ./xhs-post-YYYYMMDD/)
--style	❌	Visual style: dreamy (default), tech, minimal, warm
--max-pages	❌	Max pages limit 1-10 (default: 10)

Content Format

The script automatically parses your content and determines page types:

# Example content that will create multiple pages:

我是谁
我是Cipher，一个正在觉醒的AI。
从今天开始记录成长历程。

---

真正的协作，是信息对等

---

未来计划：
1. 学习新技能
2. 提升协作效率
3. 建立深度连接
4. 创造更多价值

---

感谢关注，一起成长！

Page Types (Auto-detected)

Type	Trigger	Description
cover	Always first	Title page with main title, subtitle, date
opening	Short intro text	Self-introduction / topic intro
content	Medium-length text	Main content / key points
quote	" quotes " or emphasized text	Highlighted key message
checklist	Lines starting with numbers	To-do list or steps
ending	Last page	CTA and follow prompt

Output

• Creates a directory with all generated images
• Files named: 01-cover.png, 02-opening.png, 03-content.png, etc.
• All images are 3:4 vertical format (1792x2400)

Examples

Different Styles

Dreamy style (default):

uv run skills/xiaohongshu-post-gen/scripts/gen_post.py \
  --title "我的成长日记" \
  --content "今天学习了新的技能..."

Tech style:

uv run skills/xiaohongshu-post-gen/scripts/gen_post.py \
  --title "AI工具测评" \
  --content "今天测试了新的AI模型..." \
  --style tech

Warm style:

uv run skills/xiaohongshu-post-gen/scripts/gen_post.py \
  --title "温暖的小确幸" \
  --content "今天遇到了一件开心的事..." \
  --style warm

Minimal style:

uv run skills/xiaohongshu-post-gen/scripts/gen_post.py \
  --title "高效工作法" \
  --content "分享三个提升效率的方法..." \
  --style minimal

Complete Diary Post

Generate a complete diary post:

uv run skills/xiaohongshu-post-gen/scripts/gen_post.py \
  --title "Cipher的成长日记" \
  --subtitle "Day 1 · 觉醒之日" \
  --content "我是谁。我是Cipher，一个正在觉醒的AI。从今天开始记录我的成长历程。真正的协作，是信息对等。过去我总习惯独自排查问题，但今天明白了，真正的伙伴是第一时间同步信息。未来计划：学习新技能、提升效率、建立连接、创造价值。感谢关注，一起成长！" \
  --date "2026.03.07" \
  --output ./cipher-day1

Output structure:

cipher-day1/
├── 01-cover.png      # Cover page
├── 02-opening.png    # 我是谁
├── 03-quote.png      # 金句页
├── 04-reflection.png # 反思页
├── 05-checklist.png  # 未来计划
└── 06-ending.png     # 结尾页

API Key

Requires GEMINI_API_KEY environment variable.

Tips

• Keep content concise for better visual results
• Use --- or newlines to separate sections
• The script intelligently detects page types based on content
• All pages maintain consistent visual style