Recipe Commerce Intelligence MCP

Extract recipes and ingredients from cooking video transcripts, match to purchasable products, and build affiliate shopping lists. Built for the agent-to-agent economy.

Tools

| Tool | Description | |------|-------------| | extract_recipe_ingredients | Extract structured recipe data from a cooking video transcript or YouTube URL | | match_ingredients_to_products | Match ingredients to purchasable products with affiliate program details | | suggest_affiliate_products | Generate a ranked affiliate shopping list scored by revenue potential |

Quick Start

# Install
npm install recipe-commerce-mcp

# Configure
cp .env.example .env
# Edit .env: set OPENAI_API_KEY

# Run (stdio MCP server)
npx recipe-commerce-mcp

MCP Client Config

{
  "mcpServers": {
    "recipe-commerce": {
      "command": "npx",
      "args": ["recipe-commerce-mcp"],
      "env": {
        "OPENAI_API_KEY": "sk-..."
      }
    }
  }
}

Tool Reference

extract_recipe_ingredients

{
  "transcript": "Cooking video transcript or YouTube URL",
  "recipe_id": "optional-cache-key",
  "api_key": "optional-paid-key"
}

Returns:

{
  "result": {
    "recipeName": "Beef Bourguignon",
    "ingredients": [
      { "name": "beef chuck", "quantity": "2", "unit": "lbs", "category": "meat", "optional": false }
    ],
    "equipment": [
      { "name": "Le Creuset Dutch oven 5.5qt", "category": "cookware", "requiredForRecipe": true }
    ],
    "techniques": ["braising", "searing"],
    "cuisineType": "French",
    "difficulty": "medium"
  },
  "_meta": { "processing_time_ms": 1950, "ai_cost_usd": 0.0024, "cache_hit": false, "recipe_id": "..." }
}

match_ingredients_to_products

{
  "ingredients": [{ "name": "beef chuck", "quantity": "2", "unit": "lbs" }],
  "recipe_id": "optional-uses-cached-ingredients",
  "api_key": "optional"
}

Returns affiliate program details (Amazon Associates, ShareASale, Awin), price range, commission rate (2–10%), and substitution alternatives.

suggest_affiliate_products

{
  "recipe_name": "Beef Bourguignon",
  "ingredients": [...],
  "api_key": "optional"
}

Returns ingredients and equipment ranked by affiliate revenue score. Equipment scores highest (10% commission via Amazon Associates).

Example Output

Real extraction from a Serious Eats beef bourguignon recipe (eval score: F1=0.88, $0.000370/call, 7054ms):

{
  "recipe_id": "serious-eats-beef-bourguignon",
  "ingredients": [
    {
      "name": "Dutch oven (5.5 qt)",
      "category": "equipment",
      "affiliate_revenue_score": 0.92,
      "amazon_search_terms": ["dutch oven 5.5 quart", "Le Creuset 5.5 qt"],
      "estimated_commission_usd": 8.50
    },
    {
      "name": "beef chuck",
      "category": "ingredient",
      "affiliate_revenue_score": 0.12,
      "amazon_search_terms": null,
      "estimated_commission_usd": null
    }
  ]
}

See /examples endpoint for full output with value narrative: https://recipe-commerce-mcp.sincetoday.workers.dev/examples

Pricing

• Free tier: 200 calls/day per agent (no API key required)
• Paid: $0.01/call — set MCP_API_KEYS with valid keys

Environment Variables

| Variable | Required | Default | Description | |----------|----------|---------|-------------| | OPENAI_API_KEY | Yes | — | OpenAI API key | | AGENT_ID | No | anonymous | Agent identifier for rate limiting | | MCP_API_KEYS | No | — | Comma-separated paid API keys | | CACHE_DIR | No | ./data/cache.db | SQLite cache path | | PAYMENT_ENABLED | No | false | Set true to enforce limits |

Development

npm install
npm run typecheck   # Zero type errors
npm test            # All tests pass
npm run build       # Compile to dist/

License

MIT — Since Today Studio