Skip to main content
Glama
README.md
# colacloud-mcp

MCP server for [COLA Cloud](https://colacloud.us) - access US alcohol label data from AI assistants.

Query 2.5M+ Certificate of Label Approval (COLA) records from the TTB (Alcohol and Tobacco Tax and Trade Bureau) directly from Claude, ChatGPT, or other MCP-compatible AI tools.

## Installation

```bash
# Using uvx (recommended, no install required)
uvx colacloud-mcp

# Or install with pip
pip install colacloud-mcp
```

## Configuration

### Claude Desktop

Add to your Claude Desktop config (`~/Library/Application Support/Claude/claude_desktop_config.json` on macOS):

```json
{
  "mcpServers": {
    "colacloud": {
      "command": "uvx",
      "args": ["colacloud-mcp"],
      "env": {
        "COLA_API_KEY": "your-api-key-here"
      }
    }
  }
}
```

### Other MCP Clients

Set the `COLA_API_KEY` environment variable and run:

```bash
export COLA_API_KEY="your-api-key-here"
colacloud-mcp
```

## Getting an API Key

1. Create an account at [app.colacloud.us](https://app.colacloud.us)
2. Go to Dashboard > API Keys
3. Create a new API key

Free tier includes 500 requests/month.

## Available Tools

| Tool | Description |
|------|-------------|
| `search_colas` | Search and filter COLA records by brand, product type, origin, dates, ABV |
| `get_cola` | Get detailed info for a specific COLA by TTB ID |
| `search_permittees` | Search permit holders (alcohol producers/importers) |
| `get_permittee` | Get permit holder details with recent COLAs |
| `lookup_barcode` | Find COLAs by product barcode (UPC/EAN) |
| `get_api_usage` | Check your API usage and rate limits |

## Example Queries

Once configured, you can ask Claude things like:

- "Search for Tito's vodka labels"
- "Find wine labels from Napa Valley approved in 2024"
- "Look up the product with barcode 080686001010"
- "Show me recent labels from permit holder CA-I-123"
- "What craft beers have an IBU over 60?"

## Data Available

Each COLA record includes:

- **Basic info**: Brand name, product name, product type, origin
- **Dates**: Application, approval, expiration dates
- **Label images**: Front, back, neck, strip labels (1-2 per record)
- **Extracted data**: ABV, volume, barcodes (via OCR/computer vision)
- **AI enrichment**: Category, tasting notes, product descriptions

## Development

```bash
# Clone and install
git clone https://github.com/cola-cloud-us/colacloud-mcp
cd colacloud-mcp
uv sync --dev

# Run locally
export COLA_API_KEY="your-key"
uv run colacloud-mcp

# Run tests
uv run pytest

# Lint
uv run ruff check .
```

## License

MIT

TDQS

A4/5.0

Scored across 6 tools

Disambiguation5/5

Each tool has a clearly distinct purpose targeting different resources and actions within the COLA domain. get_cola retrieves specific records by TTB ID, search_colas performs filtered searches, lookup_barcode finds records by barcode, get_permittee retrieves permit holder details, search_permittees searches permit holders, and get_api_usage checks system usage - there is no functional overlap between these operations.

Naming Consistency5/5

All tools follow a consistent verb_noun naming pattern (get_api_usage, get_cola, get_permittee, lookup_barcode, search_colas, search_permittees). The verbs 'get', 'lookup', and 'search' are appropriately differentiated based on the operation type, creating a predictable and readable naming convention throughout the set.

Tool Count5/5

With 6 tools, this server is well-scoped for its COLA database domain. The count provides comprehensive coverage without bloat - each tool serves a distinct, necessary function for interacting with COLA records, permit holders, and system usage, making the set appropriately sized for the server's purpose.

Completeness4/5

The tool set provides excellent coverage for querying and retrieving COLA and permittee data, including search, lookup by various identifiers, and system monitoring. The only minor gap is the absence of write/update operations, but given the domain appears to be a regulatory database (likely read-only), the surface is appropriately complete for typical agent workflows.