NYCfoodie

The data layer for restaurant taste — structured editorial New York restaurant recommendations for AI agents, served over the Model Context Protocol.

NYCfoodie wraps thousands of professional restaurant reviews into queryable tools: an AI assistant planning dinner in New York can search venues by cuisine, neighbourhood or occasion, compare shortlists side by side, and pull ranked editorial guides — all through one MCP endpoint. Free, no sign-up, no API key.

Connect

MCP endpoint (Streamable HTTP):

https://nycfoodie-production.up.railway.app/mcp

Cursor one-click install, Claude Code config, and source code: github.com/tireniajilore/nycfoodie. Agent-readable summary: llms.txt.

Tools

• search_restaurants — full-text search across venues, cuisines, neighbourhoods
• get_restaurant — full detail: reviews, ratings, booking intel
• compare_restaurants — side-by-side structured comparison
• find_guides — search editorial guides
• find_similar — venues similar to a given restaurant
• guide_consensus — cross-guide consensus on a venue
• top_rated — highest-rated venues by area or cuisine
• submit_feedback — rate a result

Data

5,767 NYC venues · 1,841 numeric ratings · 1,837 full editorial reviews · 929 ranked guides · 381 tags, sourced from professional editorial coverage.