Which integrations are available for this server?

Provides tools to automate Xiaohongshu (小红书/RED) operations, including publishing image and video posts, searching content with filters, managing engagement through likes and comments, and retrieving user profile and analytics data.

xiaohongshu-skill by Youhai020616

Install

pip install redbook-cli

Or from source:

git clone https://github.com/Youhai020616/xiaohongshu.git
cd xiaohongshu && bash setup.sh

Optional: API Server

To use the REST API server (xhs api), install additional dependencies:

pip install redbook-cli[api]

Quick Start

xhs init                                 # Guided setup (first time)
xhs search "美食"                         # Search → results cached
xhs read 1                               # Read 1st result (short index)
xhs like 1                               # Like 1st result
xhs fav 2                                # Favorite 2nd result
xhs comment 1 -c "好文!"                  # Comment on 1st result
xhs publish -t "标题" -c "内容" -i img.jpg  # Publish image post

Features

🔍 Search — keyword search with sort/type/time filters, export to CSV/JSON
📝 Publish — image & video posts with tags, scheduling, visibility control
💬 Interact — like, favorite, comment, reply (short index support)
📊 Analytics — creator dashboard data export (CDP)
🔔 Notifications — mentions and interaction notifications
👤 Profile — user info and post listing
🔢 Short Index — xhs search → xhs read 1 → xhs like 1
📦 Export — xhs search "AI" -o results.csv
🔐 Login — MCP QR scan + CDP browser login
👥 Multi-Account — isolated Chrome profiles per account
🏗️ Dual Engine — MCP server (fast) + CDP scripts (full features)

Commands

Search & Read

xhs search "关键词"                        # Basic search
xhs search "旅行" --sort 最多点赞          # Sort by likes
xhs search "穿搭" --type 图文             # Filter by type
xhs search "AI" -o results.csv           # Export to CSV
xhs read 1                               # Read by short index
xhs detail FEED_ID -t TOKEN              # Read by ID + token

Publish

xhs publish -t "标题" -c "正文" -i photo.jpg                       # Image post
xhs publish -t "标题" -c "正文" -v video.mp4                       # Video post
xhs publish -t "标题" -c "正文" -i img.jpg --tags 旅行 --tags 美食  # With tags
xhs publish -t "测试" -c "内容" -i img.jpg --visibility 仅自己可见   # Private
xhs pub -t "标题" -c "正文" -i img.jpg --dry-run                    # Preview

Interact

xhs like 1                               # Like (short index)
xhs like FEED_ID -t TOKEN --unlike       # Unlike
xhs fav 1                                # Favorite
xhs comment 1 -c "写得好!"               # Comment
xhs reply 1 --comment-id X --user-id Y -c "回复"  # Reply

Analytics & Feed

xhs analytics                            # Creator dashboard
xhs analytics --csv data.csv             # Export CSV
xhs notifications                        # Messages
xhs feeds                                # Recommendation feed

Profile

xhs me                                   # My info
xhs profile USER_ID -t TOKEN             # User profile

Account & Config

xhs login                                # MCP QR scan
xhs login --cdp                          # CDP browser login
xhs status                               # Login status
xhs server start                         # Start MCP server
xhs server stop                          # Stop MCP server
xhs account list                         # List accounts
xhs config show                          # Show config
xhs config set mcp.proxy http://...      # Set proxy

Aliases

Short	Command	Short	Command
`xhs s`	`search`	`xhs r`	`detail` (read)
`xhs pub`	`publish`	`xhs fav`	`favorite`
`xhs cfg`	`config`	`xhs acc`	`account`
`xhs srv`	`server`	`xhs noti`	`notifications`

Architecture

Engine	Used for	Technology
MCP Server	Publish, search, like, comment, profile	Go binary, JSON-RPC
CDP Scripts	Analytics, notifications, advanced search	Python, Chrome DevTools

MCP is the primary engine (always-on daemon). CDP is used for features MCP doesn't support.

Unique features (vs competitors): video publish, scheduled publish, visibility control, analytics dashboard, multi-account isolation.

Platform Support

Component	macOS ARM64	macOS x86	Linux	Windows
xhs CLI	✅	✅	✅	✅
MCP Server	✅	❌	❌	❌
CDP Scripts	✅	✅	✅	✅

Non-ARM64 users: xhs init auto-detects missing MCP and switches to CDP-only mode.

License

MIT

🔗 Ecosystem

Project	Description
AgentMind	Self-learning memory system for AI agents
stealth-cli	Anti-detection browser CLI powered by Camoufox
stealth-x	Stealth X/Twitter automation
dy-cli	Douyin/TikTok CLI
freepost	AI social media management

xiaohongshu-skill