weibo

This MCP server provides access to Weibo hot search data, including the hot search list, detailed topic information, and comments. It exposes endpoints under the MCP framework to fetch the current hot search rankings, view details for individual topics (such as category, description, link, claim info, and basic statistics), and retrieve comments associated with the top post or a post linked from a hot search topic. After starting the MCP server with the configured environment variable for Weibo cookies, you can query the server for the hot list, topic details, and comments using the MCP tooling provided by your environment (the MCP client will route requests to the weibo_mcp_server implementation). The server relies on Weibo cookies to access data that is otherwise public via the site, and the cookie should be kept secure in your MCP configuration.

Tools and capabilities:

• Get Weibo Hot Search List: rankings, keywords, and ranking indices.
• Get Hot Search Details: topic category, description, link, claim information, and statistics (reads, discussions, interactions, original posts).
• Get Hot Search Comments: fetch comments for the first Weibo post or the post linked from a hot search, with configurable max comment count and display of content and like counts.

To use, ensure the MCP runner (uv) is pointed at the weibo_mcp_server module and supply the weibo_COOKIE in the environment. Then call the MCP endpoints to retrieve the data you need for display, analytics, or aggregation in your application.

Prerequisites:

• Python >= 3.10
• internet access to install dependencies

Installation from sources:

1. Clone the repository

git clone https://github.com/Yooki-K/weibo-mcp-server.git

2. Install dependencies

pip install -r requirements.txt

Note: If you are running via uv, dependencies will be installed automatically when you start the server, so you may skip a separate pip install in that scenario.

Installation from PyPI:

pip install weibo-mcp-server

Run locally using MCP integration (example with uv):

uv --directory /path/to/weibo-mcp-server run weibo_mcp_server

Environment:

• Set weibo_COOKIE in your MCP config as shown below (example):

{
  "mcpServers": {
    "weibo": {
      "command": "uv",
      "args": [
        "--directory",
        "path/to/weibo-mcp-server",
        "run",
        "weibo_mcp_server"
      ],
      "env": {
        "weibo_COOKIE": "YOUR_WEIBO_COOKIE"
      }
    }
  }
}

Tips and caveats:

• The Weibo cookie is required to access certain data endpoints. Keep the cookie secure and do not expose it publicly.
• If you see authentication errors, verify the cookie value and ensure it has not expired.
• When running via MCP, the directory path must point to the weibo-mcp-server package/module containing the MCP implementation.
• If you upgrade dependencies, re-check compatibility with Python 3.10+ and the MCP framework version (mcp>=1.0.0).
• For production deployments, consider containerizing the service and mounting the cookie as a secret/environment variable rather than embedding in config files.
• The hot search data is time-sensitive; implement caching strategies in your application if you need to minimize repeated requests.
• Review Weibo’s terms of service when programmatically scraping or accessing data to ensure compliant usage.