WeChat-Ready HTML from Markdown: An Agent Skill

Where this fits

You keep a draft in Markdown and ask your agent to "format this for WeChat." A generic md-to-HTML converter outputs <style> tags and classes—then the WeChat editor strips them, leaving plain text. This skill solves that exact failure: it runs Markdown through doocs/md and produces HTML with every style inline, ready to paste. It sits naturally between writing (wechat-writer) and publishing (wechat-publisher). For a quick overview of metadata and source, see its OpenAgentSkill page.

Why agents benefit

• Your agent can reliably produce WeChat-compatible HTML in one script call, instead of hand-inlining CSS or using a converter that loses styles.
• It can pick from eight doocs/md themes based on article mood—orange for personal growth, blue for technical reviews, etc.—so formatting choices feel intentional.
• It keeps the publishing pipeline modular: writer → formatter → publisher. The formatter only transforms markup, so you can swap themes or inputs without touching other steps.
• It gives agents a clear trigger vocabulary (“公众号排版”, “md 转 html”, “微信编辑器不识别我的样式”) so no manual mode switch is needed.
• Output is copy-paste-ready: agents can hand the HTML file directly to a human or pass it to an automation step.

Practical scenarios

From draft to WeChat-ready HTML

A user finishes drafts/launch.md and says "排版一下". The agent checks the article's tone, selects blue, and runs the formatter, producing launch.html with all styles inlined. The user pastes it into the WeChat editor and sees the same headings, quotes, and code blocks.

Themed batch conversion for a blog migration

A developer has 50 Markdown posts from a personal blog and wants to republish selected pieces on WeChat. The agent iterates over files, classifies each article as technical, story, or aesthetic, applies the matching theme, and generates a folder of HTML files ready for review.

Automatic formatting inside a larger publishing workflow

An agent orchestration has wechat-writer produce the article and wechat-publisher handle upload. The formatter is the middle step: whenever the writer finishes, the agent invokes the skill, verifies the HTML contains no <style> tag, then hands it to the publisher.

Add it to your agent workflow

Install the skill with:

npx skills add YiShu5/claude-skills --skill wechat-formatter

Then use it from a draft directory:

python3 scripts/markdown_to_wechat_doocs.py \
  --input drafts/article.md \
  --output drafts/article.html \
  --theme green

Your agent should treat the script as the definitive answer whenever the user mentions WeChat formatting.

Compare before adopting

Before adding it, compare against three things: output quality (paste the generated HTML into WeChat and check styling), maintenance freshness (last commit, issue responses, doocs/md version), and workflow fit (do you already depend on another md-to-HTML path or prefer a different theme set). Alternatives include calling doocs/md directly or maintaining your own inline-CSS template, but that means taking on that maintenance yourself.

Why it is worth tracking

This skill is worth tracking because it solves a concrete, recurring problem—WeChat's editor ignoring standard HTML styling—and it does so by wrapping a proven open-source tool rather than inventing a fragile format. With 84 GitHub stars and a narrow, well-documented trigger scope, it is a good signal that the author understands agent workflows. Evaluate it when you build publishing agents for Chinese content platforms or any pipeline where Markdown must become editor-safe HTML.