The Guardian MCP. Query decades of news archives directly with your AI agent.

The Guardian MCP Server gives your AI agent deep access to decades of reporting across The Guardian’s entire content archive. You're not just browsing; you're running full-text queries against a massive, structured database of journalism. This server lets your agent pull everything—from the raw body text and byline details for any specific article to mapping out every contributor and topic area they’ve ever covered.

You can start by figuring out the scope. If you want to know what content exists, you'll first use list_sections to grab a list of all major editorial categories—things like 'Technology,' 'Sport,' or 'Politics.' You'd then run list_tags to pull every single keyword or contributor tag they’ve used over the years.

To map out the entire scope, you can also call list_editions, which lists regional versions of The Guardian, so you know if you need US data, UK data, or Australian feeds.

When you're ready to search, you get powerful filtering options. If you only care about one area, calling search_by_section lets your agent browse content strictly filtered by a chosen category, and it handles pagination for you. You can narrow the focus even more using search_by_tag, which filters results based on specific keywords or named contributors.

For ultimate control, you've got search_content. This tool runs across every possible criteria at once: section name, keyword tag, start and end date range, ordering preference, and pagination. You don't have to run three separate searches; this one query does it all.

If your search is time-bound, you can use search_by_date_range to pull articles published between specific dates you provide. Or, if you know the exact section and the date range, you've got multiple ways to combine those filters for super precise results.

Once a query returns a list of promising article IDs, getting the full details is easy. You use get_item to fetch the complete body text, the byline, and all the necessary metadata for one single article ID. If you just want a snapshot of what’s hot right now, calling get_latest_content returns a list ordered from most recent straight back.

To find out about featured content within any given category, you can check get_section_details, which pulls the editorial highlights and most-viewed pieces for that specific section. It’s all structured so your AI agent gets clean data it can work with immediately.