Xiaohongshu MCP (Xpzouying)

xiaohongshu-mcp


  ](https://hub.docker.com/r/xpzouying/xiaohongshu-mcp)

MCP for Xiaohongshu/xiaohongshu.com. Allow your AI assistant to directly access data on Xiaohongshu.

🚀 Quick Start: Choose the version that suits you best

\[!IMPORTANT\] #### 🔥 Solution A: Openclaw deep integration (recommended for developers) - Openclaw is too popular 🔥🔥🔥 ， Added Openclaw support, divided into two types, please use as needed: - xiaohongshu-mcp-skills(Applicable to users who have already deployed this project) - xiaohongshu-skills(Ready to use version)

\[!TIP\] #### ✨ Option B: x-mcp Browser Plugin Version (Recommended for non-technical students/minimalist users) - Don't want to tinker with Docker or deployment environments? try:xpzouying/x-mcp - Zero configurationInstall plugins and use them immediately, without any code, agents, or complex environment configuration required. - safe and stableDirectly run on commonly used browsers (Chrome/Edge) and local networks, without server IP risks, and can solve 90% of deployment errors.

📖 Related resources

• My blog post：haha.ai/xiaohongshu-mcp
• Contribution Guide：Contributing Guide

🛠️ difficult and complicated diseases

If you encounter problems when deploying traditional Docker versions,Be sure to check first:Various Difficult and Complicated Problems (Issues # 56).

*Tip: If the environment investigation is too time-consuming, switch to Usually a more efficient choice.*

Star History


Appreciation and support

All appreciation for this project will be used for charitable donations. All charitable donation records, please refer to DONATIONS.md.

Please indicate MCP and name when donating. If you need to correct/withdraw the signature, please open an issue or contact us via email.

Alipay (QR code not shown):

Through Alipay xpzouying@gmail.com Appreciation.

WeChat

Project Introduction

Main functions

💡 Hint Click on the function title below to expand and view the video demonstration

1. 登录和检查登录状态

The first step is necessary, Xiaohongshu needs to log in. You can check the current login status.

Login demo:

https://github.com/user-attachments/assets/8b05eb42-d437-41b7-9235-e2143f19e8b7

Check login status demonstration:

https://github.com/user-attachments/assets/bd9a9a4a-58cb-4421-b8f3-015f703ce1f9

1. 发布图文内容

Support publishing graphic and textual content to Xiaohongshu, including titles, content descriptions, and images.

Image support method:

Supports two image input methods:

1. HTTP/HTTPS image link

   ["https://example.com/image1.jpg", "https://example.com/image2.png"]

1. Absolute path of local image(Recommended)

   ["/Users/username/Pictures/image1.jpg", "/home/user/images/image2.png"]

Why is it recommended to use a local path

• ✅ Better stability, not dependent on the network
• ✅ Faster upload speed
• ✅ Prevent image link failure issues
• ✅ Support more image formats

Post a picture and text demonstration:

https://github.com/user-attachments/assets/8aee0814-eb96-40af-b871-e66e6bbb6b06

1. 发布视频内容

Support publishing video content to Xiaohongshu, including titles, content descriptions, and local video files.

Video support method:

Only supports absolute paths for local video files:

"/Users/username/Videos/video.mp4"

Features:

• ✅ Support local video file upload
• ✅ Automatically process video format conversion
• ✅ Support titles, content descriptions, and tags
• ✅ Automatically publish after waiting for video processing to be completed

Notes:

• Only supports local video files, does not support HTTP links
• The video processing time is relatively long, please be patient and wait
• Suggest that the video file size should not exceed 1GB

1. 搜索内容

Search for content on Xiaohongshu based on keywords.

Search post demonstration:

https://github.com/user-attachments/assets/03c5077d-6160-4b18-b629-2e40933a1fd3

1. 获取推荐列表

Get the recommended content list on the homepage of Xiaohongshu.

Get recommendation list demo:

https://github.com/user-attachments/assets/110fc15d-46f2-4cca-bdad-9de5b5b8cc28

1. 获取帖子详情（包括互动数据和评论）

Get complete details of Xiaohongshu posts, including:

• Post content (title, description, images, etc.)
• user information
• Interactive data (likes, favorites, shares, comments)
• Comment list and sub comments

⚠️ Important Notice:

• Need to provide post ID and xsec_token (both parameters are indispensable)
• These two parameters can be obtained from the feed list or search results
• You must log in first to use this feature

Get post details demonstration:

https://github.com/user-attachments/assets/76a26130-a216-4371-a6b3-937b8fda092a

1. 发表评论到帖子

Support automatic posting of comments on Xiaohongshu posts.

Function Description:

• Automatically locate comment input box
• Enter the comment content and publish it
• Support HTTP API and MCP tool calls

⚠️ Important Notice:

• You need to log in first to use this feature
• Need to provide post ID, xsec_token, and comment content
• These parameters can be obtained from the feed list or search results

Post a comment demonstration:

https://github.com/user-attachments/assets/cc385b6c-422c-489b-a5fc-63e92c695b80

1. 获取用户个人主页

Obtain personal homepage information of Xiaohongshu users, including basic user information and note content.

Function Description:

• Obtain basic user information (nickname, profile picture, avatar, etc.)
• Obtain statistics on the number of followers, fans, and likes received
• Retrieve the list of note content posted by users
• Support HTTP API and MCP tool calls

⚠️ Important Notice:

• You need to log in first to use this feature
• Need to provide user ID and xsec_token
• These parameters can be obtained from the feed list or search results

The returned information includes:

• User basic information: nickname, profile picture, avatar, authentication status
• Statistical data: number of followers, number of fans, number of likes, number of notes
• Note List: All public notes posted by users

1. 回复评论

Reply to specified comments under notes, supporting precise responses to specific user comments.

Function Description:

• Reply to specific comments under designated notes
• Support locating target comments through comment ID or user ID
• Need to provide feed_id, xsec_token, comment_id/user_id, and reply content

⚠️ Important Notice:

• You need to log in first to use this feature
• Provide at least one comment_id and user_id
• These parameters can be obtained from the comment list of the post details

1. 点赞/取消点赞

Like or unlike notes, intelligently detect the current status to avoid duplicate operations.

Function Description:

• Like or unlike specified notes
• Intelligent detection: Skip liking when already liked, skip canceling liking when not liked
• Need to provide feed_id and xsec_token

⚠️ Important Notice:

• You need to log in first to use this feature
• Default is like operation, setting unlike=true can cancel likes

1. 收藏/取消收藏

Collect or cancel notes, intelligently detect the current status to avoid duplicate operations.

Function Description:

• Collect specified notes or cancel collection
• Intelligent detection: Skip bookmarking when already bookmarked, skip unsubscribing when not bookmarked
• Need to provide feed_id and xsec_token

⚠️ Important Notice:

• You need to log in first to use this feature
• By default, it is a bookmarking operation. Setting 'dislike=true' can cancel bookmarking

Basic operational knowledge of Xiaohongshu

• Title: (Very Important) Xiaohongshu requires titles to not exceed 20 words
• Main text: (very important): The main text cannot exceed 1000 words
• Current support for sending text and images as well as videos: From a recommendation perspective, text and image traffic will be better than video and pure text traffic.
• (Low priority) Consider pure text support. 1. Personally, I feel that pure text will greatly increase the complexity of operations; The value of pure text in my usage scenario is relatively low.
• Tags: now supported. Adding appropriate tags can bring more traffic.
• Based on my actual experience, the daily posting volume on Xiaohongshu should be 50 articles.
• (Very important) The same account on Xiaohongshu is not allowed to log in on multiple web pagesIf you log in to the current xiaohongshu mcp, do not log in to that account on other web pages, otherwise the current MCP account will be "kicked out of login". You can use the mobile app to view the current account information.
• If the exposure is low, first check if there are any prohibited words in the content and search for many third-party free tools.
• Never engage in traffic diversion or pure transportation, as it is a key target of official crackdown.

Risk Disclosure

1. This project is open sourced based on another project of our own. The original project has been running steadily for over a year without any account suspension, except for the need to log in again after cookies expire.
2. I used Claude Code to connect, and after several weeks of stable automated operation, I verified that there were no issues before opening it up.
3. If the account does not have real name authentication, especially for new accounts, it usually triggers Real-name authentication The message reminder (see figure below). ⚠️ This is not a ban, even without MCP, real name authentication will be required. After real name authentication, the account will function normally. It is recommended to use your real name before using this project.

This project is based on the purpose of learning and prohibits all illegal activities.

Practical results

On the first day, the number of likes/favorites reached 999+,

About a week's worth of results

1. Usage tutorial

1.1. Quick Start (Recommended)

Method 1: Download precompiled binary files

directly from Download the binary file for the corresponding platform:

Main program (MCP service):

• macOS Apple Silicon: xiaohongshu-mcp-darwin-arm64
• macOS Intel: xiaohongshu-mcp-darwin-amd64
• Windows x64: xiaohongshu-mcp-windows-amd64.exe
• Linux x64: xiaohongshu-mcp-linux-amd64

Login tool:

• macOS Apple Silicon: xiaohongshu-login-darwin-arm64
• macOS Intel: xiaohongshu-login-darwin-amd64
• Windows x64: xiaohongshu-login-windows-amd64.exe
• Linux x64: xiaohongshu-login-linux-amd64

Usage steps:

# 1. 首先运行登录工具
chmod +x xiaohongshu-login-darwin-arm64
./xiaohongshu-login-darwin-arm64

# 2. 然后启动 MCP 服务
chmod +x xiaohongshu-mcp-darwin-arm64
./xiaohongshu-mcp-darwin-arm64

⚠️ Important NoticeThe headless browser (approximately 150MB) will be automatically downloaded during the first run, please ensure that the network connection is normal. Subsequent runs do not require repeated downloads.

Method 2: Source Code Compilation

源码编译安装详情

Dependent on Golang environment, please refer to the installation method .

Set up a proxy for Go domestic sources,

# 配置 GOPROXY 环境变量，以下三选一

# 1. 七牛 CDN
go env -w  GOPROXY=https://goproxy.cn,direct

# 2. 阿里云
go env -w GOPROXY=https://mirrors.aliyun.com/goproxy/,direct

# 3. 官方
go env -w  GOPROXY=https://goproxy.io,direct

Method 3: Use Docker containers (simplest)

Docker 部署详情

Deploying with Docker is the simplest way, without the need to install any development environment.

1. Pull images from Docker Hub (recommended)

We provide pre built Docker images that can be directly pulled from Docker Hub for use:

# 拉取最新镜像
docker pull xpzouying/xiaohongshu-mcp

Docker Hub Address:

2. Start using Docker Compose (recommended)

We provide pre configured docker-compose.yml The file can be directly used:

# 下载 docker-compose.yml
wget https://raw.githubusercontent.com/xpzouying/xiaohongshu-mcp/main/docker/docker-compose.yml

# 或者如果已经克隆了项目，进入 docker 目录
cd docker

# 启动服务
docker compose up -d

# 查看日志
docker compose logs -f

# 停止服务
docker compose stop

3. Build your own image (optional)

# 在项目根目录运行
docker build -t xpzouying/xiaohongshu-mcp .

4. Configuration Description

The Docker version will automatically:

• Configure Chrome browser and Chinese fonts
• mount ./data Used for storing cookies
• mount ./images Used for storing published images
• Expose port 18060 for MCP connection

For detailed usage instructions, please refer to:

When encountering problems with Windows, first look here:Windows Installation Guide

1.2. login

The first time you need to manually log in, you need to save the login status of Xiaohongshu.

Using binary files：

# 运行对应平台的登录工具
./xiaohongshu-login-darwin-arm64

Using source code：

go run cmd/login/main.go

1.3. Start MCP service

Start the xiaohongshu mcp service.

Using binary files：

# 默认：无头模式，没有浏览器界面
./xiaohongshu-mcp-darwin-arm64

# 非无头模式，有浏览器界面
./xiaohongshu-mcp-darwin-arm64 -headless=false

Using source code：

# 默认：无头模式，没有浏览器界面
go run .

# 非无头模式，有浏览器界面
go run . -headless=false

Configure proxy (optional)：

If proxy access is required, it can be set XHS_PROXY Environment variables:

# 设置代理后启动
XHS_PROXY=http://user:pass@proxy:port ./xiaohongshu-mcp-darwin-arm64

# 或使用源码
XHS_PROXY=http://proxy:port go run .

Supports HTTP/HTTPS/SOCKS5 proxy, and the authentication information of the proxy will be automatically hidden in the log.

1.4. Verify MCP

npx @modelcontextprotocol/inspector

运行 Inspector

After running, open the link marked in red, configure MCP inspector, and enter http://localhost:18060/mcp , click Connect button

Note: Is the option in the left border correct.

After configuring MCP inspector as above, click List Tools Button to view all tools.

1.5. Publish using MCP

Check login status

检查登录状态

Publish images and text

In the example, it is from https://unsplash.com/ I randomly selected an image for testing.

发布图文

Search Content

Use the search function to search for content on Xiaohongshu based on keywords:

搜索内容

2. MCP client access

This service supports the standard Model Context Protocol (MCP) and can connect to various AI clients that support MCP.

2.1. Quick Start

Start MCP service

# 启动服务（默认无头模式）
go run .

# 或者有界面模式
go run . -headless=false

The service will run on:http://localhost:18060/mcp

Verify service status

# 测试 MCP 连接
curl -X POST http://localhost:18060/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","method":"initialize","params":{},"id":1}'

Claude Code CLI access

# 添加 HTTP MCP 服务器
claude mcp add --transport http xiaohongshu-mcp http://localhost:18060/mcp

# 检查 MCP 是否添加成功（确保 MCP 已经启动的前提下，运行下面命令）
claude mcp list

2.2. Supported clients

Claude Code CLI

The official command-line tool has been shown in the quick start section above:

# 添加 HTTP MCP 服务器
claude mcp add --transport http xiaohongshu-mcp http://localhost:18060/mcp

# 检查 MCP 是否添加成功（确保 MCP 已经启动的前提下，运行下面命令）
claude mcp list

Open Code CLI

Add MCP Server using interactive commands:

opencode mcp add

To add xiaohongshu-mcp For example:

┌  Add MCP server
│
◇  Enter MCP server name
│  xiaohongshu-mcp
│
◇  Select MCP server type
│  Remote
│
◇  Enter MCP server URL
│  http://localhost:18060/mcp
│
◇  Does this server require OAuth authentication?
│  No
│
◆  MCP server "xiaohongshu-mcp" added to C:\Users\admin\.config\opencode\opencode.json
│
└  MCP server added successfully

Verify if the addition was successful (ensuring MCP is started):

opencode mcp list

┌  MCP Servers
│
●  ✓ xiaohongshu-mcp connected

Cursor

How to configure files

Create or edit MCP configuration file:

Project level configuration(Recommended): Create in the root directory of the project .cursor/mcp.json：

{
  "mcpServers": {
    "xiaohongshu-mcp": {
      "url": "http://localhost:18060/mcp",
      "description": "小红书内容发布服务 - MCP Streamable HTTP"
    }
  }
}

Global Configuration： Create in user directory ~/.cursor/mcp.json (Same content).

Usage steps

1. Ensure that the MCP service on Xiaohongshu is running
2. After saving the configuration file, restart cursor
3. In cursor chat, tools should be automatically available
4. You can view the connected MCP tools through the "Available Tools" in the chat interface

Demo

Plugin MCP connection:

cursor_mcp_settings

Call MCP tool: (Taking checking login status as an example)

cursor_mcp_check_login

VSCode

Method 1: Configure using the command panel

1. press Ctrl/Cmd + Shift + P Open Command Panel
2. run MCP: Add Server command
3. choice HTTP Method.
4. Enter address: http://localhost:18060/mcpOr modify it to the corresponding server address.
5. Enter MCP name: xiaohongshu-mcp.

Method 2: Directly edit the configuration file

Workplace configuration(Recommended): Create in the root directory of the project .vscode/mcp.json：

{
  "servers": {
    "xiaohongshu-mcp": {
      "url": "http://localhost:18060/mcp",
      "type": "http"
    }
  },
  "inputs": []
}

View configuration：

vscode_config

1. Confirm the operational status.
2. view tools Is it correctly detected.

Demo

Taking searching for post content as an example:

vscode_mcp_search

Google Gemini CLI

in ~/.gemini/settings.json Or project directory .gemini/settings.json Medium configuration:

{
  "mcpServers": {
    "xiaohongshu": {
      "httpUrl": "http://localhost:18060/mcp",
      "timeout": 30000
    }
  }
}

For more information, please refer to Gemini CLI MCP documentation

MCP Inspector

Debugging tool for testing MCP connections:

# 启动 MCP Inspector
npx @modelcontextprotocol/inspector

# 在浏览器中连接到：http://localhost:18060/mcp

Usage steps:

• Test the connection using MCP Inspector
• Test Ping Server functionality to verify connection
• Check if List Tools returns 13 tools

Cline

Cline is a powerful AI programming assistant that supports MCP protocol integration.

configuration method

Add the following configuration in Cline's MCP settings:

{
  "xiaohongshu-mcp": {
    "url": "http://localhost:18060/mcp",
    "type": "streamableHttp",
    "autoApprove": [],
    "disabled": false
  }
}

Usage steps

1. Ensure that the MCP service on Xiaohongshu is running（http://localhost:18060/mcp）
2. Open MCP settings in Cline
3. Add the above configuration to the MCP server list
4. Save configuration and restart Cline
5. You can directly use the functions related to Xiaohongshu in the conversation

Configuration Instructions

• urlMCP service address
• type: Use streamableHttp Type for better performance
• autoApproveList of configurable tools for automatic approval (leaving blank indicates manual approval)
• disabled: Set as false Enable this MCP service

Usage example

After configuration, you can directly use natural language to operate Xiaohongshu in Cline:

帮我检查小红书登录状态

帮我发布一篇关于春天的图文到小红书，使用这张图片：/path/to/spring.jpg

搜索小红书上关于"美食"的内容

OpenClaw（通过 MCPorter）

Please ensure that xiaohongshu mcp has completed local deployment before use.not recommendedDirectly throw the GitHub link to OpenClaw for deployment.

Due to OpenClaw's current lack of native support for MCP, the official recommendation is to use it MCPorter Call MCP service.

💡 Hint MCPorter is not the best solution for calling MCP, and there may be some compatibility issues during use. Please be aware.

Installation and configuration steps

Directly throw the following three lines of commands to OpenClaw at once (which can be Control UI, Telegram, Feishu, etc.), and Openclaw will deploy MCPorter on its behalf.

npm i -g mcporter
npx mcporter config add xiaohongshu-mcp http://localhost:18060/mcp
npx mcporter list xiaohongshu-mcp

After completing the above steps, all functions of xiaohongshu mcp can be called through natural language in OpenClaw.

其他支持 HTTP MCP 的客户端

Any client that supports the HTTP MCP protocol can connect to:http://localhost:18060/mcp

Basic configuration template:

{
  "name": "xiaohongshu-mcp",
  "url": "http://localhost:18060/mcp",
  "type": "http"
}

2.3. MCP tools available

After successful connection, the following MCP tools can be used:

• check_login_status -Check the login status of Xiaohongshu (no parameters)
• get_login_qrcode -Obtain login QR code, return Base64 image and timeout (no parameters)
• delete_cookies -Delete cookie files, reset login status, and need to log in again after deletion (no parameters)
• publish_content -Publish graphic and textual content to Xiaohongshu (required: title, content, images)

- imagesList of image paths (at least 1 image), supports HTTP links or local absolute paths, recommends using local paths - tags: Topic tag list (optional), such as ["美食", "旅行", "生活"] - schedule_at: Scheduled release time (optional), ISO8601 format, supports 1 hour to 14 days - is_original: Whether to declare originality (optional), default not to declare - visibilityVisible range (optional), supported 公开可见(default)仅自己可见、仅互关好友可见 - productsList of product keywords (optional), used to bind products for sale. Fill in the product name or product ID, and the system will automatically search and select the first matching result. The account must have activated the product function. Example: \[facial mask, sunscreen SPF50\]

• publish_with_video -Publish video content to Xiaohongshu (required: title, content, video)

- videoAbsolute path of local video file (only supports single video file) - tags: Topic tag list (optional), such as ["美食", "旅行", "生活"] - schedule_at: Scheduled release time (optional), ISO8601 format, supports 1 hour to 14 days - visibilityVisible range (optional), supported 公开可见(default)仅自己可见、仅互关好友可见 - productsList of product keywords (optional), used to bind products for sale. Fill in the product name or product ID, and the system will automatically search and select the first matching result. The account must have activated the product function. Example: \[facial mask, sunscreen SPF50\]

• list_feeds -Get the recommended list of Xiaohongshu homepage (no parameters)
• search_feeds -Search for content on Xiaohongshu (required: keyword)

- filters: Filter options (optional) - sort_bySorting criteria- 综合(Default)| 最新 | 最多点赞 | 最多评论 | 最多收藏 - note_type: Note type- 不限(Default)| 视频 | 图文 - publish_time: Release time- 不限(Default)| 一天内 | 一周内 | 半年内 - search_scope: Search scope- 不限(Default)| 已看过 | 未看过 | 已关注 - location: Location distance- 不限(Default)| 同城 | 附近

• get_feed_detail -Get post details, including interaction data and comments (required: feed_id, xsec_token)

- load_all_comments: Whether to load all comments (optional), default false returns only the first 10 first level comments - limitLimit the number of first level comments loaded (optional), only effective when load_all_commends=true, default is 20 - click_more_repliesWhether to expand the second level reply (optional), only takes effect when load_all_commends=true, defaults to false - reply_limitSkip comments with too many replies (optional), only takes effect when click_mare-replies=true, default 10 - scroll_speedRolling speed (optional),slow | normal | fast, only takes effect when load_all_comments=true

• post_comment_to_feed -Leave a comment on a post on Xiaohongshu (required: feed_id, xsec_token, content)
• reply_comment_in_feed -Reply to the designated comment under the note (required: feed_id, xsec_token, content, and at least one comment_id or user_id)
• like_feed -Like/Cancel Like (Required: feed_id, xsec_token)

- unlike: Cancel like (optional), true means cancel like, default is like

• favorite_feed -Collect/Cancel Collection (Required: feed_id, xsec_token)

- unfavorite: Cancel collection (optional), true means cancel collection, default is collection

• user_profile -Obtain user's personal homepage information (required: user_id, xsec_token)

2.4. Usage example

Publish content to Xiaohongshu using Claude Code:

Example 1: Using HTTP image links

帮我写一篇帖子发布到小红书上，
配图为：https://cn.bing.com/th?id=OHR.MaoriRock_EN-US6499689741_UHD.jpg&w=3840
图片是："纽西兰陶波湖的Ngātoroirangi矿湾毛利岩雕（© Joppi/Getty Images）"

使用 xiaohongshu-mcp 进行发布。

Example 2: Use local image path (recommended)

帮我写一篇关于春天的帖子发布到小红书上，
使用这些本地图片：
- /Users/username/Pictures/spring_flowers.jpg
- /Users/username/Pictures/cherry_blossom.jpg

使用 xiaohongshu-mcp 进行发布。

Example 3: Posting video content

帮我写一篇关于美食制作的视频发布到小红书上，
使用这个本地视频文件：
- /Users/username/Videos/cooking_tutorial.mp4

使用 xiaohongshu-mcp 的视频发布功能。

claude-cli 进行发布

Release results:

2.5. 💬 Frequently Asked Questions about MCP Usage

______________________________________________________________________

⚠️ The following are known risks when using OpenClaw+MCPorter, please fully understand before use:

• The AI automatic deployment behavior of OpenClaw is not within the maintenance scope of this project, and the deployment results cannot be guaranteed
• MCPorter, as an intermediate layer, may introduce additional compatibility issues that are unrelated to xiaohongshu mcp itself
• If you encounter problems such as connection failure or abnormal tool calls, please first check the configuration of MCPorter itself, rather than submitting an issue
• Before asking a question to a community or group, please confirm if the question can be addressed inNot using OpenClaw Reproduce under the circumstances

If you do not have a strong demand for OpenClaw usage, it is strongly recommended to use it instead Claude Code CLITheCursor or Cline Waiting for clients that natively support HTTP MCP will result in a more stable experience.

______________________________________________________________________

Q: Why is checking the login username displayed xiaghgngshu-mcp？ A: The username is written dead.

______________________________________________________________________

Q: After successfully publishing, but not actually displaying? A: The troubleshooting steps are as follows:

1. use Non Headless Mode Publish again.
2. replace different content Republish.
3. Log in to the web version of Xiaohongshu and check if your account has been compromised Risk control restrictions on web version release.
4. check image size Is it too large.
5. confirm There are no Chinese characters in the image path.
6. If using a network image address, please confirm The image link can be accessed normally.

______________________________________________________________________

Q: How to solve the problem of MCP program crashing when running on the device? A:

1. suggestion Install from source code.
2. or use Docker installation of xiaohongshu mcp, Tutorial reference:

- - X-MCP Project Page

______________________________________________________________________

Q: use http://localhost:18060/mcp Unable to connect when performing MCP verification? A:

• in Docker environment Below, please use

👉

• in Non Docker environment Below, please use Local IPv4 address visit.

______________________________________________________________________

three 🌟 Community Showcases: Practical Case Studies

💡 Strongly recommend viewingThese are real use cases of community contributors, including detailed configuration steps and practical experience!

📚 Complete tutorial list

1. N8n Complete Integration Tutorial -Integration of workflow automation platform
2. Cherry Studio Complete Configuration Tutorial -Perfect integration of AI client
3. Claude Code+Kimi K2 Integration Tutorial -Claude Code threshold is too high, so let's connect it to Kimi's domestic large model
4. AnythingLLM Complete Guide -AnythingLLM is an all-in-one multimodal AI client that supports workflow definition, multiple large models, and plugin extensions.

🎯 hintClick on the link above to view detailed graphic tutorials and quickly get started with various integration solutions! 📢 Welcome to contributeIf you have any new integration cases, please feel free to submit a PR to share with the community!

4. Xiaohongshu MCP Mutual Aid Group

Important: Before asking questions in the group, please carefully read the README document and review the Issues.

WeChat group

Feishu Group

Note: 1. The QR code in the WeChat group has a time limit and sometimes I forget to update it. Please wait for the update or submit an issue to urge me to update. 1. Feishu group, if some groups are full, you can try scanning another group, there are always pitfalls.

🙏 Thanks to contributors ✨

Thank you to all the friends who have contributed to this project! (Ranked in no particular order)

zy 💻 🤔 📖 🎨 🚧 🚇 👀

clearwater 💻

Zhongpeng 💻

Duong Tran 💻

Angiin 💻

Henan Mu 💻

Journey 💻

Eve Yu 💻

CooperGuo 💻

Banghao Chi 💻

varz1 💻

Melo Y Guan 💻

lmxdawn 💻

haikow 💻

Carlo 💻

hrz 💻

Ctrlz 💻

flippancy 💻

Yuhang Lu 💻

Bryan Thompson 💻

tan jun 💻

coldmountain 💻

mamage 💻 📖

Runyang YOU 💻 📖

e0_7 💻 📖

prehisle 💻 📖

Xinhao Chen 💻 📖

✨ Special thanks

@wanpengxie

@tanxxjun321

@Angiin

This project follows all-contributors standard. Welcome any form of contribution!