GeekNews MCP Server

The GeekNews MCP Server is a lightweight, Python‑based web scraper that turns the popular Korean tech news site GeekNews into a structured data source for AI assistants. By exposing two dedicated tools— and —the server enables Claude or other MCP‑compatible agents to fetch fresh content without developers writing custom scraping logic. This solves the common problem of integrating dynamic web content into conversational agents: instead of hard‑coding HTTP requests and parsing logic, developers can simply call a tool that returns well‑structured JSON with article titles, URLs, points, authors, timestamps, and more.

At its core, the server uses BeautifulSoup to parse GeekNews’s HTML once per day. A background scheduler triggers this parsing, storing the results in a local cache that lives under . Subsequent tool invocations read from this cache, dramatically reducing load on both the server and the GeekNews website. When the cache expires after 24 hours, the scheduler automatically refreshes it, ensuring that agents always have access to recent headlines while keeping network traffic minimal.

Key capabilities include:

• Article retrieval: Specify category (, , , ) and limit the number of items returned. Each article comes with metadata such as points, author, time elapsed, comment count, and rank.
• Weekly news extraction: Fetch a specific weekly issue or the most recent one, including title, ID, content, URL, and a list of items with their own titles, URLs, and ranks.
• Automatic caching: Daily parsing and a 24‑hour validity window keep data fresh without repeated scraping.
• Easy integration: The server is registered via Smithery, and its MCP configuration can be added to any client’s section with a single command.

Developers building AI workflows benefit from this server in several scenarios. A chatbot could answer “What’s trending on GeekNews today?” by calling with the category, while a knowledge‑base updater might periodically run to ingest new content into a database. Because the server abstracts away HTML parsing and caching, teams can focus on higher‑level conversational logic rather than maintenance of scraping pipelines. The standout advantage is the built‑in scheduler and cache, which together provide a reliable, low‑overhead data source that scales with user demand.