Worldbox Business Intelligence: MCP Server

The Model Context Protocol (MCP) in Worldbox Business Intelligence on AWS Marketplace enables AI agents to interact with company data. Here's how to use it:

1. Connecting your AI Agent to Worldbox's MCP Server:

• Get your Worldbox API Key: Subscribe to the service through the AWS Marketplace. Create a new Worldbox account or login to your existing one.
• Connect your AI agent (MCP Client): Configure your AI agent using a Worldbox API Key for authentication.

2. Core MCP Tools and Their Usage:

The MCP Server provides specific tools for AI agents to interact with the Worldbox companies’ database:

• worldbox_search_companies:
  • Purpose: Search for companies using natural language queries. Accessible at no cost.
  • Inputs: query (company name) and country (ISO country code).
  • Output: Basic company information, including Worldbox Identifier Number (WIN).
  • Example: {"name": "worldbox_search_companies", "arguments": {"query": "Nestle", "country": "CH"}}
• worldbox_search_products:
  • Purpose: List available data products for a specific company. Accessible at no cost.
  • Input: win (Worldbox Identifier Number).
  • Output: List of available products, including productcode, name, description, units (cost).
  • Example: {"name": "worldbox_search_products", "arguments": {"win": "CH0000163067"}}
• worldbox_purchase_product:
  • Purpose: Purchase specific data products. Sample datasets can be purchased at no cost (0 units) for evaluation and integration purposes - check section 5 below for more information.
  • Inputs: win, productcode, units (0 for samples, 1-1000 for full), and billing_reference (optional - this is to be defined by you for billing/ tracking purposes).
  • Output: Confirms the purchase with ordernumber, status, and product details.
  • Example: {"name": "worldbox_purchase_product", "arguments": {"win": "CH0000163067", "productcode": "CMP001", "units": 1, "billing_reference": "Nestle-BasicInfo-07.2025"}}
• worldbox_retrieve_purchase:
  • Purpose: Retrieve the content of a previously purchased dataset or report.
  • Inputs: ordernumber (from purchase).
  • Output: The actual data content.
  • Example: {"name": "worldbox_retrieve_purchase", "arguments": {"ordernumber": "5966220803"}}

3. Workflow and Phased Implementation:

• Phase 1: Search and Discovery: Subscribe to the service, connect your AI agent (MCP Client), use worldbox_search_companies and worldbox_search_products to find companies and explore their data products. Get a first hand experience of the service by consuming sample datasets at no cost (0 units) to validate integration.
• Phase 2: Scale Intelligence Operations: Implement intelligent purchasing using worldbox_purchase_product based on your product requirements. Execute automated data products retrieval data using worldbox_retrieve_purchase.
• Phase 3: Billing Monitoring: Monitor consumption and usage through AWS Marketplace billing.

4. Key Considerations for Using MCP:

• Authentication: Requires a Worldbox API Key.
• Rate Limits: Be aware of rate limits for both no cost (0 units) and paid tiers (>0 units).
• Billing: All purchases are automatically billed through AWS Marketplace based on unit price with monthly billing cycles.
• Data Freshness: Varies by product type. For more information, contact us.

5. Sample Datasets (Reports) Details:

You can use the following sample prompts within your preferred AI Agent to search for companies and products:

• "Find all reports available for 'Example Watch AG' in Switzerland"
• "Find all reports available for 'Example Watch Asia Ltd.' in Thailand"
• "Find all reports available for 'Example Watch North America INC.' in USA"

6. MCP Client Configuration:

For more information, check out the API Usage Instructions Tab.