mcp -weibo

This MCP server exposes a Weibo data access API powered by the Weibo MCP implementation. It provides capabilities to search for users, fetch user profiles, retrieve user posts (feeds), and obtain trending topics. Typical usage flows include searching for a user by keyword, getting a user profile by UID, listing a user’s recent posts with rich metadata, or pulling the latest hot topics. The server is designed to be consumed by AI applications that need structured Weibo data such as user metadata, post content with text and media, engagement counts, and trend information, all in JSON formats suitable for ingestion and analysis.

Prerequisites:\n- Node.js environment with npx or uv installed (as configured by MCP provider).\n- Access to run commands with network access to fetch dependencies.\n\nStep 1: Install prerequisites and clone the server (via UVX from GitHub)\n- Ensure you have uvx installed.\n- Run the following to install directly from source:\n\nbash\n# From source installation (preferred in README)\nuvx --from git+https://github.com/qinyuanpei/mcp-server-weibo.git mcp-server-weibo\n``\n\nStep 2: Run the MCP server locally (example)\n- Start the server using the provided command in the MCP config:\n\nbash\nuvx --from git+https://github.com/qinyuanpei/mcp-server-weibo.git mcp-server-weibo\n\n\nStep 3: Optional containerized installation (Docker)\n- Build the image and run:\n```bash\ndocker build -t mcp-server-weibo .\ndocker run -d --name mcp-server-weibo -p 4200:4200 mcp-server-weibo\n\n\nStep 4: Validate configuration\n- Sample MCP config (already provided in this documentation) should point to the server URL like:\njson\n{\n "mcpServers": {\n "weibo": {\n "url": "http://localhost:4200/mcp"\n }\n }\n}\n

Notes and tips:\n- The Weibo MCP server supports searching for users (search_users), retrieving full profiles (get_profile), fetching user feeds (get_feeds), and obtaining trendings (get_trendings).\n- Ensure network access to the Weibo endpoints or mock data when testing locally.\n- If you run behind a proxy, configure your environment variables accordingly (e.g., HTTP_PROXY, HTTPS_PROXY).\n- The provided example responses in the README show the expected JSON shapes for each function. Adapt your integration to handle fields like id, screen_name, profile_image_url, follow_count, followers_count, text, created_at, comments_count, attitudes_count, etc.\n- When deploying, set the server URL in the mcp config so that clients know where to query (the sample url is http://localhost:4200/mcp).