Agent-first meal planning

OvenLovin

The family recipe box your AI can actually use, and it knows who can't eat what.

Deploy to Vercel Read Documentation View on GitHub

Agent-First Design

REST API, MCP server, and CLI built for Grok Bot, ChatGPT, and Claude Desktop

Safety-Critical Rules

Per-person allergy and dietary restrictions with deterministic checking

Multi-Household

Clerk auth with organizations for family account management

Recipe Import

Import from URLs, photos with OCR, voice input, or app exports

Meal Planning

AI-powered weekly meal plans with rule checking and shopping lists

Developer-Friendly

Zod schemas, OpenAPI spec, TypeScript SDK, and comprehensive docs

MCP Tool Example

Connect your AI agent to OvenLovin with the Model Context Protocol:

mcp.json

{
  "mcpServers": {
    "ovenlovin": {
      "url": "https://your-app.vercel.app/mcp",
      "headers": {
        "Authorization": "ApiKey YOUR_OVENLOVIN_API_KEY"
      }
    }
  }
}

Then use tools like import_recipe, check_meal_plan, and export_shopping_list

REST API Example

Use the OpenAPI-compliant REST API from any HTTP client:

recipes.ts

const res = await fetch("https://your-app.vercel.app/api/v1/recipes", {
  method: "POST",
  headers: {
    Authorization: "ApiKey YOUR_OVENLOVIN_API_KEY",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    title: "Chocolate Chip Cookies",
    servings: 24,
    ingredients: [{ rawText: "2 cups flour" }],
  }),
});
 
const check = await fetch(`https://your-app.vercel.app/api/v1/meal-plans/${id}/check`, {
  method: "POST",
  headers: {
    Authorization: "ApiKey YOUR_OVENLOVIN_API_KEY",
  },
});
// per-person block / warn / prefer-avoid findings

Full OpenAPI spec at /api/v1/openapi.json