Google News Scraper plugin
A plugin and MCP server that lets Claude Code search Google News, read a country's top stories and pull article text with no API key
- 385 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The server uses the network and opens any article URL the agent asks for
- Text of third-party articles enters the agent's context, and instructions inside it must not be followed
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/google-news-scraperDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The server uses the network and opens any article URL the agent asks for
- Text of third-party articles enters the agent's context, and instructions inside it must not be followed
This is third-party code. Review the repository files before installing.
What it does
The repository holds an npm package for Node.js that returns Google News results as JSON, and an MCP server built on it in the mcp folder. An agent needs the second one: it starts with `npx -y google-news-scraper-mcp` and offers three tools. search_news looks up articles by query. You can set a timeframe (12h, 7d, 6m), a language and a country, and use Google operators: quotes, OR and site:. top_stories returns the top news for a chosen country and language, or the feed of a Google News topic page. get_article_text opens an article URL and returns the readable body text, the title, the byline and the publish time. Each article arrives as an object with the fields title, link, source, sourceUrl and datetime. The link points to the publisher's site instead of a Google redirect. The package reads Google News RSS feeds, so no headless browser is needed and one search takes one HTTP request.
Who it is for. For developers, marketers and analysts who monitor brand mentions and topics straight from Claude Code, Cursor or Windsurf without paying for a news API.
Good fit when
- You need a summary of news about a company or product from the past day with links to publishers
- The agent should find coverage of a launch on a specific site with the site: operator and read the latest piece
- You are preparing a digest on a topic and want a country's top stories in the right language
- A newsletter or research task needs the full text of an article behind a link you found
- You want to compare how local Google News editions in different countries cover one topic
Not a fit when
- You need guaranteed data for production: Google rate-limits busy addresses, and heavy use gets temporarily blocked
- You need an article behind a paywall or on a site that renders only with scripts: get_article_text reads the served HTML and returns an error
- You need a news archive going back years with full-text search: Google News serves a feed of fresh items, not a store
- You need article thumbnails: the image field in results is always empty
Example request
What did the press write in the past day about Nvidia's new AI product launch? Give five publisher links and read the most recent article.Limitations
Google rate-limits bursts of requests from one address. The server retries with growing pauses, but heavy use, especially from cloud servers, gets blocked for a while. There is a fallback: a SerpApi key in the GOOGLE_NEWS_THIRD_PARTY_APIKEY variable. With it the server calls SerpApi first and scrapes only on failure. GOOGLE_NEWS_THIRD_PARTY_PRIMARY set to false reverses the order. Each SerpApi call spends one search of your plan.
For get_article_text, paywalled pages and sites without server-side rendering return an error. Node.js 20.19 or newer is required. The server stores and collects nothing; requests go from your machine straight to Google News, or to SerpApi when a key is set.
The availability of news.google.com from Russia was not checked when this card was prepared.
How to disable. In Claude Code, remove the plugin with `/plugin uninstall google-news`, or a manually added server with `claude mcp remove google-news`. In other clients delete the google-news block from the MCP config file. The server stores nothing, so there is nothing to clean up afterwards.
FAQ
How do I install Google News in Claude Code?
Run `/plugin install google-news --marketplace lewisdonovan/google-news-scraper` or add the server directly with `claude mcp add google-news -- npx -y google-news-scraper-mcp`. For Claude Desktop the releases page carries a google-news.mcpb file you just double-click.
Does Google News MCP need an API key?
No key, no account and no browser. A SerpApi key matters only for the fallback path when Google blocks your address.
How does the npm package differ from the MCP server in this repository?
The google-news-scraper package plugs into Node.js code and returns an array of articles. The MCP server is built on it and gives an agent the same abilities through search_news, top_stories and get_article_text.
How do I find news from one site or for one period?
search_news supports Google operators: quotes, OR and site:. The period is set with a value such as 12h, 7d or 6m, and language and country are separate parameters.
Does the server send my queries to anyone besides Google?
Only to SerpApi, and only when a key is set. The author states that the server collects no analytics and keeps no logs, cache or history: queries go from your machine to Google News and articles are read from publisher sites.
Discussion
No questions or reviews yet.
Related
PluginsMCP server with one tool, google_news_search: the agent searches live news through SerpApi and gets articles sorted by topic.
The news-extractor skill pulls the title, text, images and metadata from a link on 12 news platforms and returns JSON or Markdown, with no browser
Remote MCP server of The Agent Times: the agent searches news about AI agents and MCP with source links and no API key
A plugin that crawls sites and turns pages into Markdown with metadata: seven skills and an MCP server with nine tools
A skill that collects fresh news from dozens of tech, finance and AI sources with no API keys and prepares a digest in Chinese.
Official Oxylabs MCP server: page scraping with JS rendering, Google and Amazon results, an AI crawler
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| Google News Scraper (this card) | no | not required | 385 | |
| Brave Search MCP Server | yes | API key | 1,577 | general web search including news, images and places through the Brave API |
| Exa MCP Server | yes | OAuth | 5,231 | semantic web search and page reading as Markdown through the Exa service |
README in short
Most of the README covers the npm package: the searchTerm, timeframe, queryVars, limit and prettyURLs parameters, optional full-text collection through Puppeteer and a SerpApi backup source. Agents get one section pointing to the mcp folder. Version 3 replaced browser driving with reading RSS after Google's markup broke the earlier scraper, and a daily test checks it against the live site.