Build Xhs Content Pack logo

Build Xhs Content Pack

Community
Kimberlying
build-xhs-content-pack

将 X 长文、网页、书籍资料、PDF、文字稿或零散笔记整理成完整的小红书内容发布包,并可改写为公众号文章与贴图。在用户要求制作小红书知识卡、读书卡、教程卡、图文笔记、发布正文、标题标签、公众号长文、公众号横版配图,或要把一份原始素材转成可发布、可检查、可复盘的内容工作流时使用。

Overview

PublisherKimberlying
Repositorybuild-xhs-content-pack
Skill namebuild-xhs-content-pack
Stars
62
Forks
13
Bundled files
11
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 11 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by Kimberlying on GitHub. Read the source before you install it.

Installation

Install the Build Xhs Content Pack AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/Kimberlying/build-xhs-content-pack/tree/main/skills/build-xhs-content-pack
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/Kimberlying/build-xhs-content-pack.git /tmp/build-xhs-content-pack
mkdir -p .claude/skills
cp -r /tmp/build-xhs-content-pack/skills/build-xhs-content-pack .claude/skills/build-xhs-content-pack
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Build Xhs Content Pack in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Build Xhs Content Pack on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Build Xhs Content Pack is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Build XHS Content Pack

将原始素材转换为结构化内容发布包。优先保留用户的个人判断,再完成拆卡、出图、发布文案、公众号改写与质量检查。

Choose A Mode

  • Use standard by default: create the brief and all card copy first, then wait for user approval before rendering images.
  • Use complete when the user explicitly asks for a one-pass finished package; continue through images, publishing copy, WeChat adaptation, and checks without a review pause.
  • Use text-only when the user does not want images; create the brief, card copy, publishing copy, and optional WeChat article only.

Do not render images in standard mode until the user approves the card copy. Respect a user's request to skip files or return content only in chat.

Collect The Input

Use supplied material directly. Ask only when a missing answer would materially change the result. Otherwise state a conservative assumption and continue.

Capture:

  • topic
  • target audience
  • personal judgment or source material
  • desired reader action
  • requested channels: Xiaohongshu, WeChat, or both
  • mode: standard, complete, or text-only
  • visual preference or theme when supplied

For URLs, books, news, product specifications, investment, health, legal, or other unstable/high-stakes claims, verify current facts with authoritative sources before writing. Separate source facts, source opinions, and editorial synthesis. Read content_safety.md for sourcing and risk rules.

Build The Package

Follow workflow.md for required file contents and use this default case structure:

text
案例/<topic>_<YYYY-MM-DD>/
  source/
    brief.md
    platform_rules.md
    banned_words.md
    reference/
  cards/
    cards.json
    preview.html
  images/
    01_cover.png
    02_problem.png
    ...
    preview_contact_sheet.png
  wechat/
    article.md
    cards.json
    preview.html
    images/
  publish.md
  checks.md
  review.md

If the user requests chat-only output, present the same logical sections without creating this directory.

Create The Brief

Write source/brief.md before writing cards. Resolve:

  1. Who is this for?
  2. What would they search for?
  3. What single idea should they remember?
  4. Why is the content worth saving?
  5. What should they do after reading?
  6. What must not be claimed or invented?

Treat the user's personal judgment as the editorial spine. Do not replace it with generic platform language.

Design The Explanation Chain

Read card_structures.md before creating cards.json.

  • Use 8-12 Xiaohongshu cards by default when the material supports that length.
  • Make the first card the cover.
  • Use the second card for the pain point, misconception, or reader tension.
  • Build the middle as an explanation chain, not a sliced article.
  • Give each card one job.
  • End with an action list, reusable prompt, summary, or risk reminder.
  • Keep claims traceable to the source notes.

Use the renderer schema documented in card_structures.md. Keep filenames numerically ordered.

Choose The Visual Theme

Read visual_styles.md before rendering images. Offer four related visual systems that share one editorial identity:

  • editorial / 知识杂志
  • reading-card / 读书卡
  • tech-tutorial / 科技教程
  • business / 商务分析

Keep typography, information hierarchy, page numbering, spacing logic, and core accent palette consistent across all four options. If the user has not selected a theme, show the four cover options and ask the user to choose by looking at the previews. In complete mode, select the best fit and report it without pausing. Apply the selected theme consistently to Xiaohongshu and WeChat output unless the user asks for different channel themes.

Render Xiaohongshu Cards

Use deterministic text rendering for knowledge cards so Chinese copy remains legible. The default 1080 x 1800 PNG size is a workflow template, not a claim about a permanent platform requirement.

Run:

bash
python3 scripts/render_cards.py path/to/cards.json --output-dir path/to/images
python3 scripts/render_cards.py path/to/cards.json --theme reading-card --output-dir path/to/images
python3 scripts/build_preview.py path/to/cards.json --image-dir path/to/images --output path/to/cards/preview.html

The renderer reads themes.json, accepts theme ids or Chinese labels, supports cover, list, steps, compare, quote, and prompt layouts, and creates preview_contact_sheet.png.

If Pillow is unavailable, report the missing dependency and ask before installing it. Do not silently replace readable text cards with image-generation text.

Write The Publishing Copy

Create publish.md with:

  • 3-5 title candidates
  • image upload order
  • one publish-ready body
  • focused tags
  • source or originality notes when needed
  • manual checks before publishing

Do not repeat every card in the body. Add personal judgment, suitable and unsuitable readers, context, limitations, and what required human confirmation.

Create The WeChat Version

When WeChat output is requested:

  1. Rewrite the material as a coherent article; do not concatenate card copy.
  2. Add background, transitions, examples, and editorial judgment.
  3. Identify 3-6 ideas that benefit from visualization.
  4. Create wechat/cards.json using a landscape canvas, default 1200 x 675.
  5. Render the article header, process, comparison, key-point, checklist, or summary graphics into wechat/images/.

Use the same rendering scripts. Treat these dimensions as editable templates.

Validate Before Delivery

Read quality_checks.md and complete checks.md. At minimum verify:

  • cover and filename order
  • one job per card
  • readable text without clipping
  • no unsupported claims or fabricated experience
  • body does not duplicate the cards
  • source and risk notes are present where needed
  • preview HTML and contact sheet include every expected image

Open the contact sheet and at least the cover, densest card, comparison card, and prompt card. Fix layout problems before delivery.

Prepare For Review

Create review.md with blank fields for views, likes, saves, comments, reader questions, best-performing card, weak points, and next iteration. When the user later supplies metrics, analyze them without inventing causality.

Reference Routing

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Build Xhs Content Pack AI skill do?

将 X 长文、网页、书籍资料、PDF、文字稿或零散笔记整理成完整的小红书内容发布包,并可改写为公众号文章与贴图。在用户要求制作小红书知识卡、读书卡、教程卡、图文笔记、发布正文、标题标签、公众号长文、公众号横版配图,或要把一份原始素材转成可发布、可检查、可复盘的内容工作流时使用。

Why use Build Xhs Content Pack on TypingMind?

Because you install it once and use it with any model. Build Xhs Content Pack is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Build Xhs Content Pack in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/Kimberlying/build-xhs-content-pack/tree/main/skills/build-xhs-content-pack. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Build Xhs Content Pack?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Build Xhs Content Pack?

As many as you like. As long as a model supports skills, you can use Build Xhs Content Pack with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Build Xhs Content Pack AI skill free?

It is published on GitHub by Kimberlying. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇