OpenClaw XHS (Xiaohongshu) Setup Guide¶

Source: zhjiang22/openclaw-xhs (113 ⭐) Category: Social Media / Growth

OpenClaw XHS integrates Xiaohongshu (小红书, "Little Red Book") — China's 300M+ user lifestyle and shopping platform — into your AI agent workflow. Built as an MCP server, it enables your Hermes or OpenClaw agent to read, monitor, and analyze Xiaohongshu content, track hot topics, and export personal memory banks.

Quick Install¶

# Via skills.sh
npx skills add zhjiang22/openclaw-xhs@xiaohongshu

# Direct from GitHub
cd ~/.hermes/skills/
git clone https://github.com/zhjiang22/openclaw-xhs.git
cd openclaw-xhs
npm install

Prerequisites¶

Requirement	Details
Node.js	18+
Hermes Agent or OpenClaw	v0.20.0+ (Hermes) or latest OpenClaw
Xiaohongshu account	Required for authenticated API access
MCP support	Both Hermes and OpenClaw support MCP natively

Key Capabilities¶

1. Hot Topic Tracking (热点跟踪)¶

Monitor trending topics on Xiaohongshu in real-time:

# Start the MCP server
cd ~/openclaw-xhs && npm start

# Agent commands (via MCP tools):
# - get_trending_topics(category="tech", limit=20)
# - search_notes(keyword="AI agent", sort="hot")
# - get_note_detail(note_id="...")

2. Content Reading & Analysis¶

Your agent can read and analyze Xiaohongshu notes (posts):

Tool	Description
`search_notes`	Search notes by keyword, category, or creator
`get_note_detail`	Full note content including images, comments, engagement
`get_creator_profile`	Creator stats, follower count, content history
`get_comments`	Comment thread with sentiment analysis

3. Personal Memory Bank Export (个人记忆库导出)¶

Export your Xiaohongshu browsing history, saved notes, and interactions as structured data for your agent's memory:

# Export personal data
node export-memory.js --format json --output ~/xhs-memory.json

# Import into GBrain
gbrain import ~/xhs-memory.json --type xhs-notes

MCP Server Configuration¶

Register the XHS MCP server with Hermes:

# ~/.hermes/profiles/corpusiq/config.yaml
mcp_servers:
  xiaohongshu:
    transport: stdio
    command: node
    args:
      - ~/.hermes/skills/openclaw-xhs/server.js
    env:
      XHS_COOKIE: ${XHS_COOKIE}  # Your Xiaohongshu session cookie

Authentication¶

Xiaohongshu requires browser cookies for API access:

Log into xiaohongshu.com in your browser
Export cookies (use browser dev tools or a cookie manager extension)
Set XHS_COOKIE environment variable with the cookie string
Cookies typically last 24-48 hours before needing refresh

Use Cases¶

Use Case	How It Works
Competitive monitoring	Track competitor brand mentions and sentiment on XHS
Content research	Analyze top-performing content in your niche
Trend detection	Identify emerging trends before they hit Western platforms
Creator discovery	Find and evaluate KOLs/KOCs for partnerships
Market intelligence	Understand Chinese consumer preferences and language

Why This Matters¶

Xiaohongshu has 300M+ monthly active users, predominantly young, affluent Chinese consumers. It's the #1 platform for:

Product discovery and reviews
Lifestyle trends and shopping decisions
Brand perception in the Chinese market

For AI companies and SaaS products expanding into APAC, Xiaohongshu intelligence is critical — and this MCP integration makes it accessible directly from your agent workflow.

Production Notes¶

113 GitHub stars, actively maintained (updated July 2026)
Chinese-language README with English code comments
Built as MCP server — compatible with any MCP client (Hermes, OpenClaw, Claude Desktop)
Includes hot topic tracking, memory export, and content analysis tools
Cookie-based auth (no API key required — uses standard web session)

Limitations¶

Cookie expiry: Session cookies expire every 24-48h and must be manually refreshed
Rate limiting: Xiaohongshu may rate-limit aggressive scraping
Chinese-language content: Most content is in Chinese — pair with translation tools for non-Chinese speakers
No posting: Read-only at this time (posting requires additional anti-bot measures)

← Memory Merger Setup | Skills Catalog → Powered by CorpusIQ