Xiaohongshu (RedBook)

Clone the project, create a Python virtual environment, and activate it. Using a virtual environment is essential because the MCP client config must reference the interpreter inside it.

git clone https://github.com/chenningling/Redbook-Search-Comment-MCP2.0.git
cd Redbook-Search-Comment-MCP2.0
python3 -m venv venv
source venv/bin/activate  # macOS/Linux
# venv\Scripts\activate    # Windows

Install all Python dependencies from requirements.txt, then install the Playwright browser binaries that the server uses to drive the Xiaohongshu web interface.

pip install -r requirements.txt
pip install fastmcp
playwright install

Add the server to claude_desktop_config.json using the absolute path to the Python interpreter inside your virtual environment and the absolute path to the main script. On Windows, escape backslashes and use 'python.exe'.

{
  "mcpServers": {
    "xiaohongshu": {
      "command": "/absolute/path/to/venv/bin/python3",
      "args": ["/absolute/path/to/Redbook-Search-Comment-MCP2.0/xiaohongshu_mcp.py"]
    }
  }
}

When the server starts for the first time, call the mcp0_login tool. A browser window will open displaying a Xiaohongshu QR code. Scan it with the Xiaohongshu mobile app. The session is persisted to disk so you do not need to repeat this step on subsequent runs.

Use mcp0_search_notes to find posts by keyword and mcp0_get_note_content to read the full text, author, and timestamp of individual notes. Pass the note URL obtained from search results to the content and comment tools.

Call mcp0_post_smart_comment with a note URL and a comment_type (引流 for traffic, 点赞 for engagement, 咨询 for consultative, or 专业 for authority) to get an AI-analysed recommendation. Then call mcp0_post_comment with the final comment text to publish it.