Skip to content

MCP Guide

TomoYummy provides a public, read-only MCP server for its English editorial ramen reviews and restaurant information. No account or authentication is required.

Connect

Use this Streamable HTTP endpoint in an MCP client that supports remote servers:

https://tomoyummy.com/api/mcp

Set up your client

  1. Add a remote MCP server in your client’s settings.
  2. Select Streamable HTTP when the client asks for a transport.
  3. Enter the endpoint above and save the connection.

Clients use different configuration formats. A common shape is:

{
  "mcpServers": {
    "tomoyummy": {
      "type": "http",
      "url": "https://tomoyummy.com/api/mcp"
    }
  }
}

Use with ChatGPT

ChatGPT can add remote MCP servers through Developer mode. Download the 128 × 128 TomoYummy icon for the connector image, then follow the official ChatGPT Developer mode guide.

  1. In ChatGPT on the web, open Settings → Security and login, then turn on Developer mode.
  2. Create an app from an MCP server and enter https://tomoyummy.com/api/mcp.
  3. Use the downloaded icon when ChatGPT asks for a connector image.

Developer mode provides full MCP client access. Add MCP servers only if you understand and accept OpenAI’s security warnings; you are responsible for verifying a connector’s safety and configuration. TomoYummy’s server is read-only.

Available tools

  • search_reviews: Search published ramen reviews.
  • get_review: Read one published review in detail.
  • search_restaurants: Search TomoYummy’s restaurant information.
  • get_restaurant: Read one restaurant and its related reviews.
  • get_taxonomy: Read TomoYummy’s ramen types and rating criteria.

Scope

TomoYummy is an editorial review source. The MCP server does not provide real-time hours, operating status, reservations, or directions. When a Google Maps URL is available in TomoYummy’s published data, the server can return that link; confirm current details with Google Maps or the restaurant directly.

Successful MCP calls are counted only as non-identifying aggregate retrievals. Search terms, IP addresses, cookies, and other personal identifiers are not stored.