Xiaohongshu MCP Server
521w/xiaohongshu-mcp
Templates e estilos estruturados para conteúdo Xiaohongshu via MCP
Ver no GitHubServidor MCP que fornece templates estruturados, estilos de escrita e sugestões de hashtags para conteúdo no Xiaohongshu (小红书). Sem dependência de APIs externas, funciona localmente para gerar drafts consistentes de posts em redes sociais chinesas com fluxos de trabalho repetíveis.
Instalação
npx convez-ai 521w/xiaohongshu-mcp1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Geração de posts de recomendação de produtos (种草文)
- Roteiros para vídeos curtos e storytelling diário (Vlog 文案)
- Templates de tutoriais e guias passo-a-passo (教程文)
- Workflows automatizados de produção de conteúdo para marcas
- Assistentes de IA para criadores de conteúdo em plataformas chinesas
Veredito

Ferramenta especializada e bem-estruturada para criadores que trabalham com Xiaohongshu e redes sociais chinesas. Abordagem local e sem APIs torna-a prática, mas o escopo é bastante nicho — útil principalmente para agências de criação de conteúdo ou marcas com foco em mercado chinês.
README
A tradução para o português deste README ainda está na fila.
Xiaohongshu MCP Server
An MCP server that gives AI agents structured Xiaohongshu-style content templates, writing styles, and hashtag suggestions.
It is designed for creators, operators, and AI assistants that need repeatable Chinese social-media content workflows.
What It Does
- Lists supported Xiaohongshu content styles
- Returns structured prompt templates for each style
- Provides recommended hashtags
- Helps agents generate more consistent Xiaohongshu drafts
- Works without external API keys
Tools
| Tool | Description |
|---|---|
get_xiaohongshu_styles | List available content styles with descriptions |
get_xiaohongshu_prompt | Get a structured prompt template for a specific style |
get_xiaohongshu_tags | Get recommended hashtags for a specific style |
Supported Styles
| Style | Use Case |
|---|---|
| 种草文 | Product recommendation and lifestyle sharing |
| 测评文 | Comparison review and pros/cons analysis |
| 教程文 | Step-by-step guide or how-to post |
| 开箱文 | Unboxing and first-impression content |
| Vlog 文案 | Short video script and daily-life storytelling |
Good For
- Xiaohongshu content drafting
- AI creator assistants
- Product review templates
- Social media operation workflows
- Repeatable brand or personal-account content production
Installation
pip install git+https://github.com/521w/xiaohongshu-mcp.git
Or install from source:
git clone https://github.com/521w/xiaohongshu-mcp.git
cd xiaohongshu-mcp
pip install -e .
MCP Configuration
{
"mcpServers": {
"xiaohongshu": {
"command": "python3",
"args": ["-m", "xiaohongshu_mcp.server"]
}
}
}
Example Agent Flow
User asks for a Xiaohongshu product review.
Agent calls get_xiaohongshu_styles.
Agent calls get_xiaohongshu_prompt with the chosen style.
Agent calls get_xiaohongshu_tags for matching hashtags.
Agent drafts the final post using the returned structure.
Example Agent Tasks
- "Write a Xiaohongshu product recommendation post for wireless earbuds."
- "Turn this travel experience into a Xiaohongshu Vlog script."
- "Give me hashtags for a skincare comparison review."
- "Create a reusable Xiaohongshu content workflow for a small brand."
Why This Server
- No API key: fully local prompt/template tool
- Agent-native: exposes reusable writing assets through MCP
- Consistent output: reduces random style drift across drafts
- Productizable: useful as a content-ops component for creator services or small businesses
Requirements
- Python >= 3.10
mcp >= 1.0.0
Notes
This server provides writing structures and prompt assets. It does not publish content, scrape Xiaohongshu, or bypass platform controls.
Verified Status
Verified on Termux/Android with Python 3.13:
python -m venv .venv
. .venv/bin/activate
pip install -e .
xiaohongshu-mcp
MCP initialization, tool listing, and get_xiaohongshu_styles tool calls were verified successfully.
License
MIT