The surface mirrors the underlying API rather than layering over it — the same by_city, by_state and by_type filters, the same page and per_page, and an id lookup for the record behind a search hit. brewery/random is not a query at all: it takes size and hands back that many at random, which is what makes this as useful for filling a fixture as for answering a question. There is no coordinate or radius parameter anywhere here, so a question shaped like "breweries near me" has to be turned into a city or a state before it can be asked.
A client for the Open Brewery DB dataset with three tools: a paged search across name, city, state and type, a detail lookup by brewery id, and a random sampler.
- brewery/search takes a free-text query or the by_city, by_state and by_type filters, paging through page and per_page.
- brewery/by-id returns the detail for a single brewery from its id.
- brewery/random returns random breweries, with size deciding how many come back.
Nothing — no account, no key.
One command — npm install -g @jimbuck/open-brewery-db-mcp