Xiaohongshu Automatic Image-Text Post Publishing
Feature Overview
After users provide a title and body text, automatically completes the entire process: login detection, AI-generated images, content filling, AI-generated tags, tag activation, original content declaration, and publishing. Only QR code login requires manual operation; all other steps are fully automated.
Trigger Conditions
Trigger this skill when the user mentions any of the following keywords:
- - Xiaohongshu image-text post publishing
- Publish Xiaohongshu image-text post
- Automatically publish Xiaohongshu image-text post
- Xiaohongshu image-text notes
- 小红书图文发布
- 发布小红书图文
- 自动发小红书图文
- 小红书图文笔记
Use Cases
- - Users have existing note content ready to publish on Xiaohongshu.
Workflow
CODEBLOCK0
Prerequisites
Browser Environment
- - Start the browser using host mode
- Open the Xiaohongshu Creator Platform: https://creator.xiaohongshu.com/publish/publish
Account Login
- - Only manual QR code login is required
- Automatically pauses and prompts when not logged in is detected
- Automatically continues execution after successful login
Automated Execution Steps
1. Launch Browser and Check Login Status
browser start profile=openclaw target=host
browser navigate https://creator.xiaohongshu.com/publish/publish
browser snapshot refs=aria
Login Detection Logic:
- - Check if the page contains a login form
- If not logged in, take a screenshot and prompt the user to scan the QR code
- Poll every 15 seconds automatically; continue after login
2. Switch to Image-Text Publishing Mode
CODEBLOCK2
3. AI Automatically Generates Images
CODEBLOCK3
4. Automatically Fill Title and Body Text
browser act ref=<Title Input Field> kind=type text="<User Title>"
browser act ref=<Body Input Field> kind=type text="<User Body Text>"
Character Count Check:
- - Title exceeds 20 characters → Automatically truncate and prompt
- Body text exceeds 1000 characters → Automatically truncate and prompt
5. AI Generate Tags (if user does not provide)
# Analyze title and body text, generate 5-10 relevant tags
# Tag types: Core topics, Scene tags, Emotion tags
Generation Rules:
- - Extract core keywords from the title
- Extract high-frequency words from the body text
- Supplement with popular related tags
6. Activate Topic Tags
Key Rule: Must click the first item in the pop-up recommendation list for the tag to become a searchable link
CODEBLOCK6
Activation Success Indicator:
- - Tag becomes a clickable blue link
- Format: INLINECODE0
Important Notes:
- - ⚠️ Must click the recommendation pop-up; direct input is invalid
- ⚠️ Must click the first item to ensure tag standardization
- ⚠️ Each tag needs to be activated individually
7. Check Original Content Declaration
CODEBLOCK7
8. Execute Publishing
browser act ref=<Publish Button> kind=click
browser snapshot refs=aria
Publishing Success Indicator: Page displays "Published Successfully" or redirects to the "Note Management" page
Exception Handling
| Issue | Handling Method |
|---|
| Login timeout | Take screenshot, poll automatically to detect login status |
| Title too long |
Automatically truncate to 20 characters, prompt user to confirm |
| Publishing failed | Take screenshot record, prompt user to retry or escalate to manual |
Key Considerations
Account Security
- - Only QR code login requires manual operation
- Recommended publishing interval ≥ 30 minutes
- Automatically check content compliance before publishing
Content Compliance
- - Title ≤ 20 characters, Body text ≤ 1000 characters
- Avoid absolute terms, medical efficacy claims, price inducement language
Recommended Publishing Frequency
- - New accounts: Daily or every other day
- Established accounts: Adjust based on follower activity times
Quick Command Reference
CODEBLOCK9
Example
CODEBLOCK10
小红书自动图文发布
功能概述
用户提供标题和正文后,自动完成全流程:登录检测、AI生成图片、内容填充、AI生成标签、标签激活、声明原创、发布。仅需手动扫码登录,其余步骤全自动完成。
触发条件
当用户提及以下任一关键词时触发该技能:
- - 小红书图文发布
- 发布小红书图文
- 自动发布小红书图文
- 小红书图文笔记
- 小红书图文发布
- 发布小红书图文
- 自动发小红书图文
- 小红书图文笔记
使用场景
工作流程
你提供内容 → 启动浏览器 → 检查登录 → AI生成图片 → 自动填充 → AI生成标签 → 激活标签 → 声明原创 → 发布笔记
↑
仅扫码登录需手动操作
前置条件
浏览器环境
- - 使用主机模式启动浏览器
- 打开小红书创作者平台:https://creator.xiaohongshu.com/publish/publish
账号登录
- - 仅需手动扫码登录
- 检测到未登录时自动暂停并提示
- 登录成功后自动继续执行
自动化执行步骤
1. 启动浏览器并检查登录状态
bash
browser start profile=openclaw target=host
browser navigate https://creator.xiaohongshu.com/publish/publish
browser snapshot refs=aria
登录检测逻辑:
- - 检查页面是否包含登录表单
- 若未登录,截图并提示用户扫码
- 每15秒自动轮询,登录后继续
2. 切换到图文发布模式
bash
browser act ref=<上传图文按钮> kind=click
3. AI自动生成图片
bash
点击文生图按钮
browser act ref=<文生图按钮> kind=click
输入图片描述
browser act ref=<文本框> kind=type text=<用户标题>
生成图片并选择风格(默认基础)
browser act ref=<生成按钮> kind=click
browser act ref=<风格按钮> kind=click
browser act ref=<下一步按钮> kind=click
4. 自动填充标题和正文
bash
browser act ref=<标题输入框> kind=type text=<用户标题>
browser act ref=<正文输入框> kind=type text=<用户正文>
字数检查:
- - 标题超过20字 → 自动截断并提示
- 正文超过1000字 → 自动截断并提示
5. AI生成标签(若用户未提供)
bash
分析标题和正文,生成5-10个相关标签
标签类型:核心话题、场景标签、情感标签
生成规则:
- - 提取标题中的核心关键词
- 提取正文中的高频词汇
- 补充相关热门标签
6. 激活话题标签
关键规则:必须点击弹出推荐列表中的第一项,标签才能成为可搜索链接
bash
对每个标签执行以下流程
for each tag in tag list:
# 步骤1:点击话题按钮打开输入框
browser act ref=<话题按钮> kind=click
# 步骤2:输入标签名称(包含#符号)
browser act ref=<话题输入框> kind=type text=#标签名
# 步骤3:等待推荐弹窗出现(约1秒)
browser snapshot refs=aria
# 步骤4:点击推荐列表中的第一项(必须!)
browser act ref=<推荐列表第一项> kind=click
# 步骤5:确认标签已激活(显示为蓝色链接)
# 继续下一个标签
激活成功标志:
- - 标签变为可点击的蓝色链接
- 格式:[话题]#标签名
重要提示:
- - ⚠️ 必须点击推荐弹窗,直接输入无效
- ⚠️ 必须点击第一项以确保标签规范化
- ⚠️ 每个标签需单独激活
7. 检查原创声明
bash
browser act ref=<原创声明复选框> kind=click
browser act ref=<同意复选框> kind=click
browser act ref=<声明原创按钮> kind=click
8. 执行发布
bash
browser act ref=<发布按钮> kind=click
browser snapshot refs=aria
发布成功标志:页面显示发布成功或跳转到笔记管理页面
异常处理
| 问题 | 处理方式 |
|---|
| 登录超时 | 截图,自动轮询检测登录状态 |
| 标题过长 |
自动截断至20字,提示用户确认 |
| 发布失败 | 截图记录,提示用户重试或转人工处理 |
关键注意事项
账号安全
- - 仅扫码登录需手动操作
- 建议发布间隔 ≥ 30分钟
- 发布前自动检查内容合规性
内容合规
- - 标题 ≤ 20字,正文 ≤ 1000字
- 避免绝对化用语、医疗功效、价格诱导
建议发布频率
- - 新账号:每日或隔日
- 老账号:根据粉丝活跃时间调整
快速命令参考
bash
browser navigate
# 打开页面
browser snapshot refs=aria # 获取页面元素
browser act ref=<元素> kind=click # 点击元素
browser act ref=<元素> kind=type text=内容 # 输入文本
browser screenshot # 截图
示例
使用 xiaohongshu-image-auto 技能。汇总今日国内AI领域重要发布。