Redbook-Search-Comment-MCP2.0
Free基于 Playwright 开发，支持自动登录、关键词搜索、笔记获取及 AI 评论发布 (适配 Claude 桌面版)。
FreeFree tier
About Redbook-Search-Comment-MCP2.0
Redbook-Search-Comment-MCP2.0 is an open-source MCP (Model Context Protocol) server built with Playwright that automates Xiaohongshu (Little Red Book) search and commenting. It integrates with MCP clients like Claude for Desktop to handle login, keyword search, note retrieval, and AI-powered comment generation. The tool features persistent login, modular design (note analysis, comment generation, comment publishing), four comment types (drainage, like, consultation, professional), and structured data output. Version 2.0 enhances content acquisition with four methods, improves AI comment generation via client-side models, and adds robust error handling.
Key Features
Persistent browser login with QR code support
Smart keyword search with configurable result count
Multi-dimensional note content acquisition (4 methods)
AI comment generation leveraging MCP client models
Support for 4 comment types: drainage, like, consultation, professional
Modular design: note analysis, comment generation, comment publishing
Structured JSON output for notes and comments
Automatic login state detection and management
Docker and virtual environment setup support
Pros & Cons
Pros
- Deep integration with AI models via MCP for natural comments
- Persistent login eliminates repeated QR scanning
- Modular design improves maintainability and extensibility
- Multiple content acquisition methods ensure reliable note retrieval
- Supports various comment strategies for different use cases
- Free and open-source with active community and documentation
Cons
- Requires manual QR code login on first use
- Dependent on Playwright browser automation, which may break with Xiaohongshu website changes
- Configuration requires technical knowledge of MCP and paths
- Comments are limited by AI model quality and client capabilities
- No built-in scheduling or batch processing
Best For
Automating Xiaohongshu search and comment workflowsIntegrating with Claude for Desktop for AI-powered social media interactionBuilding marketing or research tools for Xiaohongshu contentTesting MCP server capabilities for Chinese social platformsEducational projects demonstrating Playwright and MCP integration
FAQ
How do I log in to Xiaohongshu?
Use the mcp0_login() function in the MCP client. On first run, a browser window opens for manual QR code scan. Subsequent uses are automatic due to persistent browser context.
What Python version is required?
Python 3.8 or higher. It is recommended to use a virtual environment.
How do I configure this tool with Claude for Desktop?
Add a JSON entry in the MCP client config file, specifying the full absolute path to the Python interpreter in your virtual environment and the xiaohongshu_mcp.py script with the --stdio argument.
What comment types are supported?
Four types: drainage (引导型), like (点赞型), consultation (咨询型), and professional (专业型). The type is chosen by the MCP client AI based on the note analysis results.
Is the tool free?
Yes, the project is open-source and free to use. There are no usage limits, though users need their own Xiaohongshu account and MCP client.