Developers · MCP

Connect QuvtyQR to your AI assistant

QuvtyQR runs a public Model Context Protocol (MCP) server. Add one URL to your assistant and it can read any published QuvtyQR menu: dishes, prices, allergens, nutrition and languages.

Server URL

https://quvtyqr.com/mcp

Read-only
No sign-in or API key
Streamable HTTP, stateless

What it is

MCP is an open standard that lets AI assistants call tools on external services. The QuvtyQR MCP server exposes two tools. There is nothing to install and no account needed: your assistant sends requests to the server URL and gets structured data back.

What your assistant can do with it

Read a restaurant's published menu by its slug and summarise it for you.
Answer questions about dishes: ingredients, vegan or vegetarian options, calories and macros.
Check prices and current campaign discounts in the menu's currency.
Filter by allergens, e.g. find dishes without gluten or nuts.
Read the menu in another language the business offers.
Explain what QuvtyQR is and how the Starter and Business plans differ.

Good to know

It only reads published menus. It cannot create, edit or delete anything: no orders, no menu changes, no account access.
Some data stays hidden on purpose: if a business hides prices or allergens, the server leaves them out too. Menus that can only be opened by scanning the QR code on-site are not available.
Requests are rate-limited (60 per minute per IP).
Managing your own menu from an AI assistant is coming later.

Tools

`get_menu`

Returns a business's public QR menu: categories, products, prices (campaign discounts already applied), allergen and content labels, nutrition values, active campaigns and business info such as address, opening hours and the menu link.

Arguments

slug string: Required. The business's slug (the subdomain of its menu address).
language string?: Optional. Menu language code (e.g. en, tr, de). If the menu doesn't offer that language, its default language is used.

`get_service_info`

Returns what QuvtyQR is, its plans (Starter and Business) and links to the site's machine-readable resources.

Arguments

No arguments.

Finding a slug

Every QuvtyQR menu lives at https://<slug>.quvtyqr.com. The slug is the first part of that address; for https://cafe-luna.quvtyqr.com it is cafe-luna. You can also just give your assistant the menu link.

Connect your assistant

Pick your app. Every client uses the same server URL; no authentication is required.

Open claude.ai or the Claude desktop app and go to Settings → Connectors.
Choose "Add custom connector".
Name it QuvtyQR, paste the server URL below and add it. Leave the authentication fields empty.
Server URL
```
https://quvtyqr.com/mcp
```
In a chat, enable the QuvtyQR connector from the tools menu and ask about a menu.

Run this command in your terminal:

Terminal

claude mcp add --transport http quvtyqr https://quvtyqr.com/mcp

Check it with claude mcp list, or type /mcp inside Claude Code.

Create .cursor/mcp.json in your project (or ~/.cursor/mcp.json for all projects) with:
.cursor/mcp.json
```
{
  "mcpServers": {
    "quvtyqr": {
      "url": "https://quvtyqr.com/mcp"
    }
  }
}
```
Open Cursor Settings → MCP and make sure quvtyqr is enabled.

Create .vscode/mcp.json in your workspace with:

.vscode/mcp.json

{
  "servers": {
    "quvtyqr": {
      "type": "http",
      "url": "https://quvtyqr.com/mcp"
    }
  }
}

Start the server from the file (or via the "MCP: List Servers" command), then use it in Copilot Chat's agent mode.

In ChatGPT settings, enable developer mode for apps/connectors (available on supported plans).
Create a new custom connector and paste the server URL below. Choose no authentication.
Server URL
```
https://quvtyqr.com/mcp
```
Enable it in a chat and ask about a menu. Menu names in ChatGPT change often, so look for "connectors" or "apps".

Any MCP client that supports the Streamable HTTP transport can connect. Point it at the server URL:
Server URL
```
https://quvtyqr.com/mcp
```

Many clients accept a JSON config like the one below. Clients that only support local (stdio) servers can use a bridge such as mcp-remote.

mcp.json

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

stdio · mcp-remote

{
  "mcpServers": {
    "quvtyqr": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://quvtyqr.com/mcp"
      ]
    }
  }
}

Try these prompts

Replace cafe-luna with the slug of a real QuvtyQR menu.

What vegan dishes does cafe-luna have on QuvtyQR?
Show me the cafe-luna menu in English and list the three cheapest mains.
Which desserts at cafe-luna are free of gluten and nuts?
Is cafe-luna open now, and what's on their current campaigns?
Which drinks at cafe-luna have fewer than 100 calories?

Raw JSON-RPC (for developers)

The server speaks JSON-RPC 2.0 over HTTP POST. Each request is independent (no session ID is needed), and every response is a single JSON body.

1. Initialize

curl -s https://quvtyqr.com/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-06-18","capabilities":{},"clientInfo":{"name":"curl","version":"1.0"}}}'

2. List tools

curl -s https://quvtyqr.com/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":2,"method":"tools/list"}'

3. Call get_menu

curl -s https://quvtyqr.com/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"get_menu","arguments":{"slug":"cafe-luna","language":"en"}}}'

Supported protocol versions: 2025-06-18 (latest), 2025-03-26 and 2024-11-05. GET requests return 405 because the server doesn't open SSE streams.

Resources

Your menu, readable by AI

Publish your menu on QuvtyQR and assistants connected to this server can answer guests' questions about it.

Start free