Agent-Native Restaurant Directory

Restaurant intelligence
for AI agents

Structured menus, per-dish dietary flags, allergen data, and reservation links for Minneapolis‑St. Paul. One MCP connection. No API key required.

Connect via MCP See it work

Restaurants

2,143

Menu items

91%

Avg completeness

Capabilities

What your agent gets

☰

Structured menus

Every dish with name, description, price, and course. Not scraped HTML fragments — clean, normalized JSON.

⚖

Dietary & allergen flags

Per-dish flags: gluten_free, vegan, vegetarian, dairy_free. Plus allergens: nuts, shellfish, wheat, soy. Inferred and explicit.

📅

Reservation links

Platform, booking URL, walk-in policy. Resy, Tock, OpenTable — one call to check_availability returns the link.

Example

One query, structured results

An agent calls search_restaurants and gets back structured profiles — not web pages to parse.

MCP Tool Call

1{
2  "cuisine_type": "Italian",
3  "neighborhood": "North Loop",
4  "price_range": "$$$"
5}

Response

1[{
2  "name": "Bar La Grassa",
3  "cuisine": "Italian",
4  "price_range": "$$$",
5  "completeness": 1.0,
6  "reservation": "Resy"
7}]

API Surface

Four MCP tools

01 search_restaurants

Faceted search by cuisine, price, neighborhood, or natural language query

02 get_restaurant

Full structured profile — cuisine, experience, hours, reservation info

03 get_menu

Menu items with prices, dietary flags, allergens, ingredients. Filter by course or diet.

04 check_availability

Reservation platform, booking URL, walk-in policy

Get Started

Connect in 30 seconds

No API key. No sign-up. Point your MCP client at the endpoint.

MCP Endpoint (Streamable HTTP)

$ https://misen.ai/mcp

claude_desktop_config.json

1{
2  "mcpServers": {
3    "misen": {
4      "url": "https://misen.ai/mcp"
5    }
6  }
7}

REST API

$ curl https://misen.ai/api/v1/search?cuisine_type=Italian

Restaurant intelligence for AI agents