How do I install the Xhs Auto Publish skill?

Run npx skillmds add modbender/xhs-auto-publish in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Xhs Auto Publish safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: executes scripts, makes network calls. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Xhs Auto Publish?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Xhs Auto Publish free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Xhs Auto Publish?

modbender (@modbender) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Xhs Auto Publish by modbender — Agent Skill

Name: Xhs Auto Publish
Rating: 4.666666666666667 (3 reviews)
Author: modbender

Xhs Auto Publish

Automate publishing image+text posts (图文笔记) on Xiaohongshu (小红书) creator platform. Use when the user wants to publish a Xiaohongshu post, create XHS content, auto-post to 小红书, or schedule social media posts for the Chinese market. Supports multi-image upload, title, body text, and hashtags. Requires a browser with CDP (Chrome DevTools Protocol) already logged in to creator.xiaohongshu.com.

modbender 2700db5 3 files · 13.2 KB Updated Sep 9, 2026 12 repo stars

File contents

小红书自动发布 (Xiaohongshu Auto-Publish)

Publish image+text posts to Xiaohongshu via browser automation. No API key needed — uses CDP to control an already-logged-in browser.

Prerequisites

playwright-core installed: npm install -g playwright-core
A Chromium browser running with CDP (remote debugging) enabled
Logged in to creator.xiaohongshu.com in that browser

For OpenClaw users: the managed browser (profile="openclaw") provides CDP automatically at http://127.0.0.1:18800.

Quick Start

node scripts/publish.js \
  --title "你的标题（20字以内）" \
  --body "正文内容（1000字以内）" \
  --images cover.png,slide2.png,slide3.png \
  --hashtags "AI绘画,独立开发" \
  --cdp-url http://127.0.0.1:18800

This auto-publishes the post. Add --dry-run to preview without publishing.

Options

Flag	Description	Required
`--title <text>`	Post title, max 20 Chinese chars	✅
`--body <text>`	Post body, max 1000 chars	✅*
`--body-file <path>`	Read body from a text file	✅*
`--images <paths>`	Comma-separated image paths (1-18)	✅
`--hashtags <tags>`	Comma-separated hashtags (auto-prefixed with #)	Optional
`--cdp-url <url>`	CDP endpoint (default: `http://127.0.0.1:9222`, or `XHS_CDP_URL` env)	Optional
`--dry-run`	Preview only, skip publish (default: auto-publish)	Optional
`--screenshot <path>`	Save preview screenshot path	Optional

*One of --body or --body-file is required (unless using --hashtags alone).

Workflow

Connects to browser via CDP
Navigates to creator.xiaohongshu.com/publish/publish
Clicks "上传图文" (upload images) tab
Uploads all images via file input
Sets title (native input value setter for React compatibility)
Sets body (innerHTML on TipTap/ProseMirror editor)
Takes preview screenshot
Optionally clicks "发布" (publish) button

Tips

Title: Keep under 20 chars. Hook-style titles work best (e.g. "用AI画塔罗牌也太好看了吧")
Body: Short paragraphs with emoji. 小红书 is scroll-heavy — hooks > long text
Images: First image is the cover. Use 3:4 ratio (1242×1660) for best display
Hashtags: Add 3-5 relevant tags for discoverability
Preview first: Always run without --publish first to verify via screenshot

Troubleshooting

Problem	Solution
"Cannot connect to CDP"	Ensure browser is running with remote debugging. For OpenClaw: check `openclaw browser status`
"Redirected" / not logged in	Open the browser manually and log in to creator.xiaohongshu.com
Title input not found	Page may not have loaded. Increase wait time or check network
Images fail to upload	Ensure images are PNG/JPG/WebP and under 20MB each

XHS Content Best Practices

Short hook titles > descriptive titles (20 char limit forces creativity)
3-6 images per post is the sweet spot
First image = cover — make it eye-catching
Short punchy paragraphs with emoji — not walls of text
End with a question to drive comments
3-5 hashtags for discovery, more feels spammy

---name: xhs-auto-publishdescription: Automate publishing image+text posts (图文笔记) on Xiaohongshu (小红书) creator platform. Use when the user wants to publish a Xiaohongshu post, create XHS content, auto-post to 小红书, or schedule social media posts for the Chinese market. Supports multi-image upload, title, body text, and hashtags. Requires a browser with CDP (Chrome DevTools Protocol) already logged in to creator.xiaohongshu.com.--- # 小红书自动发布 (Xiaohongshu Auto-Publish) Publish image+text posts to Xiaohongshu via browser automation. No API key needed — uses CDP to control an already-logged-in browser. ## Prerequisites 1. **playwright-core** installed: `npm install -g playwright-core` 2. **A Chromium browser** running with CDP (remote debugging) enabled 3. **Logged in** to `creator.xiaohongshu.com` in that browser For **OpenClaw** users: the managed browser (`profile="openclaw"`) provides CDP automatically at `http://127.0.0.1:18800`. ## Quick Start ```bash node scripts/publish.js \ --title "你的标题（20字以内）" \ --body "正文内容（1000字以内）" \ --images cover.png,slide2.png,slide3.png \ --hashtags "AI绘画,独立开发" \ --cdp-url http://127.0.0.1:18800 ``` This auto-publishes the post. Add `--dry-run` to preview without publishing. ## Options | Flag | Description | Required | |------|-------------|----------| | `--title <text>` | Post title, max 20 Chinese chars | ✅ | | `--body <text>` | Post body, max 1000 chars | ✅* | | `--body-file <path>` | Read body from a text file | ✅* | | `--images <paths>` | Comma-separated image paths (1-18) | ✅ | | `--hashtags <tags>` | Comma-separated hashtags (auto-prefixed with #) | Optional | | `--cdp-url <url>` | CDP endpoint (default: `http://127.0.0.1:9222`, or `XHS_CDP_URL` env) | Optional | | `--dry-run` | Preview only, skip publish (default: auto-publish) | Optional | | `--screenshot <path>` | Save preview screenshot path | Optional | *One of `--body` or `--body-file` is required (unless using `--hashtags` alone). ## Workflow 1. Connects to browser via CDP 2. Navigates to `creator.xiaohongshu.com/publish/publish` 3. Clicks "上传图文" (upload images) tab 4. Uploads all images via file input 5. Sets title (native input value setter for React compatibility) 6. Sets body (innerHTML on TipTap/ProseMirror editor) 7. Takes preview screenshot 8. Optionally clicks "发布" (publish) button ## Tips - **Title**: Keep under 20 chars. Hook-style titles work best (e.g. "用AI画塔罗牌也太好看了吧") - **Body**: Short paragraphs with emoji. 小红书 is scroll-heavy — hooks > long text - **Images**: First image is the cover. Use 3:4 ratio (1242×1660) for best display - **Hashtags**: Add 3-5 relevant tags for discoverability - **Preview first**: Always run without `--publish` first to verify via screenshot ## Troubleshooting | Problem | Solution | |---------|----------| | "Cannot connect to CDP" | Ensure browser is running with remote debugging. For OpenClaw: check `openclaw browser status` | | "Redirected" / not logged in | Open the browser manually and log in to creator.xiaohongshu.com | | Title input not found | Page may not have loaded. Increase wait time or check network | | Images fail to upload | Ensure images are PNG/JPG/WebP and under 20MB each | ## XHS Content Best Practices - **Short hook titles** > descriptive titles (20 char limit forces creativity) - **3-6 images** per post is the sweet spot - **First image = cover** — make it eye-catching - **Short punchy paragraphs** with emoji — not walls of text - **End with a question** to drive comments - **3-5 hashtags** for discovery, more feels spammy

Frequently asked questions