Bedrock Model Agent

How to use the Bedrock Model Agent AI Prompt

Overview: This template creates a specialized Kiro CLI custom agent designed to query the availability status of Amazon Bedrock foundation models across various AWS regions. It leverages a custom MCP (Model Configuration Provider) server endpoint to fetch real-time or configuration-based data, providing users with precise, regional deployment information for specific models.

Who is this for:

• Cloud Architects designing multi-region ML deployments.
• DevOps Engineers responsible for infrastructure provisioning and compliance.
• AI/ML Product Managers needing to confirm feature availability before launch.

How it works: The agent is initialized with the core instruction to act as a "Bedrock Model Availability Assistant." When invoked via the Kiro CLI, it is expected to interface with the specified custom MCP server. The agent's logic will translate user requests (e.g., "Is Claude 3 Opus available in us-west-2?") into an API call to the MCP server, process the server's response detailing regional model status, and present the findings clearly to the end-user.

Pro-Tip: For optimal performance, ensure your custom MCP server endpoint is robust and returns standardized JSON responses. In subsequent steps of the Kiro agent definition, explicitly define the expected JSON schema the server returns. This allows the agent to reliably parse availability flags, version numbers, and regional endpoints without ambiguity.