baoyu-post-to-wechat Skill for Claude: What It Does and How to Install It
Lets Claude turn Markdown, HTML or text into a WeChat Official Account article draft, or post an image-text set.
- Installs
- 34.6K
- on skills.sh
- Stars
- 27.2K
- on GitHub
- Files
- 28
- besides SKILL.md
- Updated
- Nov 2026
- last repo commit
baoyu-post-to-wechat sends content to a WeChat Official Account (微信公众号) in two ways. Article posting (文章) takes HTML, Markdown or plain text, applies a theme, and saves the result as a draft. Image-text posting (贴图) composes a post of up to nine images with a short caption in the WeChat editor, and with --submit saves it as a draft.
For articles you choose the WeChat API, a remote API call through your own server, or a logged-in Chrome browser. The API route needs your WeChat AppID and AppSecret. The browser route needs Chrome and a scanned login. It runs TypeScript scripts with Bun.
Good for
- Chinese-language publishers who write in Markdown and post to WeChat
- Teams managing several Official Accounts
- Developers who want article drafts created from a script
Install baoyu-post-to-wechat
Claude Code
- 1
Copy this prompt.
Install the baoyu-post-to-wechat skill in Claude Code. Run this command: npx skills add https://github.com/jimliu/baoyu-skills --skill baoyu-post-to-wechat Then check that it loaded and tell me how to use it.
- 2
Paste it into Claude Code. It shows the change and asks before running it.
Requires Node.js.
From its listing on skills.sh. Source.
What's in the skill
When it loads
Posts content to WeChat Official Account (微信公众号) via API or Chrome CDP. Supports article posting (文章) with HTML, markdown, or plain text input, and image-text posting (贴图, formerly 图文) with multiple images. Markdown article workflows default to converting ordinary external links into bottom citations for WeChat-friendly output. Use when user mentions "发布公众号", "post to wechat", "微信公众号", or "贴图/图文/文章".
Sections in SKILL.md
- User Input Tools
- Language
- Script Directory
- Preferences (EXTEND.md)
- Multi-Account Support
- Pre-flight Check (Optional)
- Image-Text Posting (图文)
- Article Posting Workflow (文章)
- Step 0: Load Preferences
- Step 1: Determine Input Type
- Step 2: Select Publishing Method and Configure
- Step 3: Resolve Theme/Color and Validate Metadata
- Step 4: Publish
- Step 5: Completion Report
- Feature Comparison
- Troubleshooting
Key facts
- Author
- jimliu (the repo owner)
- Repository
- jimliu/baoyu-skills
- Installs on skills.sh
- 34,597 as of 2026-11-02
- License
- MIT
- Last commit to the repo
- 2026-11-01
- Script dependencies
- scripts/package.json lists 5 npm packages (@jsquash/webp, baoyu-chrome-cdp, baoyu-md, jimp, socks)
- skills.sh audits
- Gen Agent Trust Hub pass, Socket warn, Snyk pass
What it includes
Creates EXTEND.md preferences before anything else
If no EXTEND.md exists, Claude asks for theme, color and publishing method and saves the file. Project, XDG and home-folder locations are checked in order.
Write ·
First-time setupArticle draft through the WeChat API
Accepts a .md or .html file, uploads images and cover, and calls draft/add on api.weixin.qq.com. Needs a cover image and AppID/AppSecret.
Try: “Post article.md to my WeChat Official Account with the grace theme.”
Runs code ·
scripts/wechat-api.tsArticle through a Chrome session
Drives Chrome with a logged-in WeChat session, using the clipboard to paste the converted content. Slower than the API route.
Runs code ·
scripts/wechat-article.tsSend the API calls through an SSH tunnel to your server
For when WeChat's IP allowlist excludes your machine. Rendering stays local, and only the HTTPS calls to api.weixin.qq.com go through an ssh SOCKS5 forward.
Runs code ·
Remote API (--remote)Image-text post with up to 9 images, saved as a draft with --submit
Takes --markdown with an images folder, or --title, --content and --image. Without --submit it only composes the post as a preview; --submit saves it as a draft, per image-text-posting.md.
Runs code ·
scripts/wechat-browser.tsMarkdown to WeChat-ready HTML
Four themes (default, grace, simple, modern) and 13 color presets or a hex color. External links become bottom citations unless you pass --no-cite.
Runs code ·
scripts/md-to-wechat.tsOptional environment check
Checks Chrome, profile isolation, Bun, macOS Accessibility, clipboard, paste keystroke, API credentials and Chrome conflicts.
Runs code ·
scripts/check-permissions.tsaccounts: block in EXTEND.md
With two or more accounts, Claude asks which to use, or you pass --account <alias>.
Write ·
Multi-account
From the official docs
Permissions
Medium risk
This skill runs scripts that use your WeChat AppSecret, can drive your Chrome session and clipboard, can open an SSH tunnel to a server you name, and upload images to your Official Account. Its output is a draft in your account, not a public post.
What it can do
- Read your data
- Create and change things
- Delete things
- Read-only mode
Watch out for
Output is a draft, but images are uploaded to your account
SKILL.md ends the flow at "Draft saved to WeChat Official Account" and image-text posting only saves a draft with --submit. The API route calls uploadimg, add_material and draft/add, so images and the cover are sent to your account before you review the draft. Publishing to readers is a step you take in the WeChat console. github.com
Your AppSecret sits in a .env file
The guided setup appends WECHAT_APP_ID and WECHAT_APP_SECRET to .baoyu-skills/.env. Keep that file out of git and prefer the user-level file. github.com
The browser route controls Chrome and the clipboard
It needs a logged-in WeChat session and, on macOS, Accessibility permission for your terminal. The SKILL.md offers a pre-flight check script for this. github.com
Remote API uses an SSH tunnel
It starts ssh with a SOCKS5 forward to the host you configure. Use a key you trust and a host you control. github.com
It writes files in your project
Plain-text input is saved as post-to-wechat/YYYY-MM-DD/<slug>.md, and converted HTML is written next to the source. github.com
The browser route keeps a logged-in profile
The pre-flight table says the browser route uses a shared Chrome profile at baoyu-skills/chrome-profile, and the first run asks you to scan a QR code to log in. Anyone using that profile is signed in to your account. github.com
Scripts pull npm packages
scripts/package.json lists 5 npm dependencies, and the skill runs Bun, or npx -y bun if Bun is missing. github.com
Technical details
| Turning on read-only | No read-only switch is documented in SKILL.md. |
|---|---|
| Talks to | api.weixin.qq.com (token, image upload, draft/add). Your SSH host when --remote is used. Telegram, only if TELEGRAM_BOT_TOKEN and TELEGRAM_CHAT_ID are set, to send the login QR |
| Sign-in and keys | The API credentials are saved in plain text in a .env file you choose. The README says not to commit it. For remote publishing the SKILL.md says AppSecret never leaves the local process. |
| Scopes |
|
Questions
What does baoyu-post-to-wechat do?
It lets Claude publish to a WeChat Official Account. Articles are converted from HTML, Markdown or plain text and saved as a draft, and image-text posts carry up to nine images.
How do I install baoyu-post-to-wechat?
Use the install box at the top of this page. The README warns the repo has 20+ skills and says to install only the ones you need.
Which publishing method should I pick?
The SKILL.md marks the API as recommended and fast, but it needs credentials and an allowlisted IP. The browser route is slow and needs Chrome with a logged-in session. Remote API is for when only a server's IP is allowlisted.
Where are my WeChat credentials stored?
The guided setup asks where to save WECHAT_APP_ID and WECHAT_APP_SECRET: a .baoyu-skills/.env file in the project, or ~/.baoyu-skills/.env for all projects. The README says not to commit project-level .env files.
Is baoyu-post-to-wechat safe to use?
Treat it as a tool that acts on your WeChat account. It runs scripts that call the WeChat API with your AppSecret, can drive your Chrome with the clipboard and keystrokes, and can open an SSH tunnel to a server you name. The SKILL.md and image-text-posting.md describe every route as saving a draft, which you then review in the WeChat console, but the scripts also upload images to your account first. Try it on a test account.
Does it need the other baoyu skills?
No. The README says Markdown to HTML is already included. It lists baoyu-cover-image and baoyu-article-illustrator as the usual extras for a WeChat article workflow, and baoyu-format-markdown for turning raw text into structured Markdown.
Information
updated Nov 2, 2026
- Repository
- jimliu/baoyu-skills
- Folder
skills/baoyu-post-to-wechat
Sources
Docs, issue trackers and discussions this page is built from.
Show all 5Hide
- baoyu-post-to-wechat SKILL.md · github.com
- api-setup.md · github.com
- image-text-posting.md · github.com
- jimliu/baoyu-skills README · github.com
- scripts/package.json · github.com