Markdown WeChat Push
Paste the following prompt into your AI chat to install this skill:
Please follow https://skillhub.cn/install/skillhub.md to install @org-je08tpng/mukun-md-push-wechat into your AI assistant.
About this skill
The Problem
WeChat article bodies do not support external CSS, so pasting raw Markdown often loses styling, breaks tables, and becomes hard to maintain when long posts hit platform limits. This skill targets technical writing workflows that need a reliable path from Markdown to WeChat-compatible HTML, with optional draft pushes to WeChat or Juejin.
How It Works
The workflow is split into three levels: md2wechat_html.py handles conversion only, push_daily.py converts and pushes to WeChat, and push_juejin.py pushes to Juejin. Conversion supports headings, body text, bold, links, inline code, code blocks, blockquotes, and tables, and inlines CSS into style attributes. Article mode includes seven preset styles that can be selected by natural language or by --config; news mode is aimed at sectioned daily reports and should be explicitly requested. Before pushing to WeChat, the flow reads title and digest from frontmatter and, when digest is missing, requires a generated Chinese summary of up to 120 characters to avoid hard truncation. Output over 20,000 characters is split by H2 headings and labeled with suffixes such as (Part 1), (Part 2), and (Part 3). The Juejin path parses frontmatter, uploads local images to the Juejin image host, replaces them with CDN URLs, and creates or updates a draft.
Boundaries And Caveats
Conversion-only use has no extra prerequisites; WeChat push requires ~/.md_push_wechat/config.yaml with appid and secret, and the flow must not auto-create or search for it. Juejin push depends on juejin.cookie, which is valid for about 30 days. Image upload is best-effort, and failed uploads leave the original reference intact. The skill does not verify facts, handle copyright, or satisfy WeChat review rules; it focuses on formatting, digest preparation, and draft delivery.
Use Cases
- Technical writers convert a finished Markdown post into WeChat-compatible HTML and push it to a WeChat draft.
- Ops editors format an AI weekly report into sectioned news-mode HTML for a WeChat daily digest.
- Engineers upload local images to Juejin CDN and create a Juejin draft with category and tags.
- Documentation authors split long posts over 20,000 characters by H2 and push multiple image-text parts.
Best For
- Tech WeChat authors who regularly turn Markdown long-form posts into WeChat HTML and drafts.
- AI weekly report editors who need sectioned news-mode styling and controlled digest length.
- Juejin technical authors who need category, tags, cover images, and local image CDN replacement.
- Documentation engineers who want long posts split by H2 when limits are hit while keeping WeChat rendering.
Related Skills
A Chinese financial writing editor that removes AI-ish research-report tone, template phrases, and vague claims, rewriting text into natural, concrete prose and exporting Word files on demand.
Generates a Markdown user scenario document from model designs and PRDs, covering overview, MVP, user stories, and non-functional requirements.
Continues user-provided story fragments while preserving worldbuilding, characters, and style, with options for genre, length, foreshadowing, and endings.
Generate ready-to-use Weibo hot-topic copy, reply plans, checklists, and risk notes for brand moments and public topics.