xhs-mcp
Enables AI agents to automatically publish photo notes to Xiaohongshu (RedNote) using Playwright browser automation. It supports title, body, and image uploads while maintaining persistent login sessions for seamless automated posting.
README
xhs-mcp 🍠
An MCP server that lets your AI agent publish notes to Xiaohongshu (RedNote / 小红书) automatically using Playwright browser automation.
Features
- 📸 Publish photo notes with title, body, and image
- 🤖 MCP-compatible — works with Claude, Cursor, and any MCP-supported agent
- 🔐 Session persistence — log in once, post forever
- 🧠 Human-like typing to avoid bot detection
Requirements
- Python 3.10+
- macOS (tested on macOS with Apple Silicon)
Installation
1. Clone the repo
git clone https://github.com/FrancoSbaffi/xhs-mcp.git
cd xhs-mcp
2. Install dependencies
pip install -r requirements.txt
3. Install Chromium
python -m playwright install chromium
4. Set up your session (first time only)
Run the login script and log in manually in the browser that opens:
python post.py
Once logged in, press ENTER in the terminal. Your session will be saved to session/auth.json.
Usage
Option A — Run standalone
Edit content/texto.txt with your post content (first line = title, rest = body).
Add your image as content/imagen.jpg. Then run:
python post.py
Option B — Use as MCP server with Claude
Add this to your Claude MCP config (claude_desktop_config.json):
{
"mcpServers": {
"xiaohongshu": {
"command": "python3.11",
"args": ["/absolute/path/to/xhs-mcp/server.py"]
}
}
}
Then in Claude, you can say:
"Post this to my Xiaohongshu: title 'Hello World', body 'This is my first automated post', image at /path/to/image.jpg"
MCP Tool Reference
post_note
| Parameter | Type | Description |
|---|---|---|
title |
string | Post title (max 20 characters) |
body |
string | Post body text |
image_path |
string | Absolute path to image (jpg/png) |
Project Structure
xhs-mcp/
├── post.py # Standalone posting script
├── server.py # MCP server
├── requirements.txt # Dependencies
├── content/ # Default content folder
│ ├── texto.txt # Post text (title on first line)
│ └── imagen.jpg # Post image
└── session/ # Saved login session (git ignored)
└── auth.json
Important Notes
- ⚠️ This project uses browser automation and is not affiliated with Xiaohongshu
- ⚠️ Use responsibly — avoid posting at high frequency
- ⚠️ Session may expire and require re-login
- 🔒 Never commit
session/auth.jsonto GitHub — it contains your login credentials
.gitignore
Make sure to add this before pushing:
session/
.env
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。