# nutrients-mcp

Give your AI assistant the ability to analyze food images and look up nutrition data for any food — calories, macros, vitamins, minerals, allergens, and dietary flags in under 1 second — powered by [TastyAPI](https://tastyapi.com). Much faster and more accurate than Gemini and GPT-4 Vision (4-10 seconds).

## Requirements

- Node.js 18+
- A TastyAPI API key — [get one at tastyapi.com/pricing](https://tastyapi.com/pricing)

## Quick Start

### Claude Code

```bash
claude mcp add nutrients -- npx -y nutrients-mcp
```

Then set your API key in the server's environment config:

```
TASTYAPI_KEY=YOUR_API_KEY
```

### Claude Desktop

Add to `~/Library/Application Support/Claude/claude_desktop_config.json`:

```json
{
  "mcpServers": {
    "nutrients": {
      "command": "npx",
      "args": ["-y", "nutrients-mcp"],
      "env": { "TASTYAPI_KEY": "YOUR_API_KEY" }
    }
  }
}
```

### Cursor / Windsurf

Add to your `mcp.json`:

```json
{
  "mcpServers": {
    "nutrients": {
      "command": "npx",
      "args": ["-y", "nutrients-mcp"],
      "env": { "TASTYAPI_KEY": "YOUR_API_KEY" }
    }
  }
}
```

## Tools

### `analyze_food_image`

Fetches a food image from a public URL and returns a full nutritional breakdown.

| Parameter   | Type   | Required | Description                           |
|-------------|--------|----------|---------------------------------------|
| `image_url` | string | yes      | Public URL of a food image to analyze |

**Example prompt:**
> "What are the macros in this meal? https://example.com/pasta.jpg"

**Example output:**
```json
{
  "food_name": "Pasta Bolognese",
  "calories": 520,
  "macros": {
    "protein_g": 24,
    "carbohydrates_g": 68,
    "fat_g": 16,
    "fiber_g": 4
  },
  "vitamins": { "vitamin_c_mg": 8, "vitamin_b12_mcg": 1.4 },
  "minerals": { "iron_mg": 3.2, "calcium_mg": 60 },
  "allergens": ["gluten", "dairy"],
  "dietary": { "vegan": false, "vegetarian": false, "gluten_free": false }
}
```

---

### `analyze_food`

Returns a full nutritional breakdown for a food item by name or description. No image required.

| Parameter   | Type   | Required | Description                                                   |
|-------------|--------|----------|---------------------------------------------------------------|
| `food_name` | string | yes      | Name or description of the food (e.g. `"oatmeal with banana, 1 cup"`) |

**Example prompt:**
> "How many calories are in a large avocado?"

**Example output:**
```json
{
  "food_name": "Large Avocado",
  "calories": 322,
  "macros": {
    "protein_g": 4,
    "carbohydrates_g": 17,
    "fat_g": 29,
    "fiber_g": 13
  },
  "vitamins": { "vitamin_k_mcg": 42, "folate_mcg": 163 },
  "minerals": { "potassium_mg": 975, "magnesium_mg": 58 },
  "allergens": [],
  "dietary": { "vegan": true, "vegetarian": true, "gluten_free": true }
}
```

## Pricing

- **$0.015 per request** + $5/month base fee (usage-based)
- Or choose a flat-rate plan: Monthly ($29/mo · 5k calls) or Professional ($69/mo · 15k calls)

See [tastyapi.com/pricing](https://tastyapi.com/pricing) for full details.

## License

MIT — © EB Tech LLC