MCP App Store
by KeenEthics
Media

Futrumi

by Fingerlicky, s.r.o.

Overview

Futrumi is a Czech food-tip platform: named food critics, chefs, sommeliers, and gastronomy journalists share where they actually like to eat and drink, with the exact dish or drink they recommend. This MCP server exposes the published, curated data to AI assistants so users can ask things like "najdi mi vínárnu v Karlíně" or "co doporučuje Michal Daněk" and get real expert opinions, not generic web results. Each result includes the expert's quote in the original Czech, the business profile, and a deep link to the Futrumi mobile app.

Tools

find_recommendations_near

Claude
Geo-only lookup: return expert-recommended businesses near a point, sorted by distance. Use this when the user just wants to know "what good places are around X" without a specific cuisine/category. For category- or vibe-specific queries use search_recommendations instead. Returns markdown.

get_business

Claude
Return full details for a single business: address, opening hours, web/menu/phone/social links, featured quotes, photos (venue, per-dish, and other expert photos), and every expert recommendation about it. Use when the user is zeroing in on one place or wants to see its photos.

get_expert

Claude
Return an expert's bio plus a list of their recommendations (business name, address, quote, meals). Use when the user asks about a specific expert.

get_recommendation

Claude
Return the full text of a single expert recommendation including the strong quote, the long description, meals with their descriptions and photos, any other photos, publish date, and business basics. Use after search_recommendations when the user wants more depth on one specific tip.

list_experts

Claude
Paginated directory of all Futrumi experts with their bio and recommendation count. Use this when the user asks who's behind the recommendations, or wants to browse by expert.

search_recommendations

Claude
Search and rank expert-recommended restaurants/cafes/bars/etc around a location. Use this when the user asks for a specific cuisine, meal, occasion, vibe, or category ("good ramen in Vinohrady", "brunch in Brno", "vinný bar na rande"). Pass the user intent in query; the server ranks candidates by semantic-ish food/occasion relevance before returning markdown. The data is Czech-only and includes only PUBLISHED recommendations.

App Stats

6

Tools

Claude

Platforms

Works with

Claude

Data refreshed daily