Making Amazon Quick enterprise-ready: Automated, auditable cross-account resource promotion

Amazon Quick is a work-focused agentic AI companion that allows users to build agents that reason over data and call action connectors to complete multi-step tasks. Promoting these resources from a development to a production AWS account is a manual and error-prone process. To address this, Amazon has developed an automated, auditable Model Context Protocol (MCP) server on Amazon Bedrock AgentCore.

This server promotes resources in an idempotent and auditable manner, making it easy for enterprises to transition from development to production.

The building blocks of an agentic solution include agents, action connectors, knowledge bases, and spaces. Teams assemble and iterate on these components in a development account before promoting them to production. However, there is no native, one-click way to promote these resources, leading to slow, difficult-to-audit, and error-prone processes.

Teams often rebuild each resource manually, recreating agents with the same instructions and starter prompts, re-attaching action connectors, granting resource permissions, and provisioning Amazon S3 buckets and data sources for knowledge bases.

Amazon Quick resources can be programmatically managed through the Amazon Quick API, which provides the full resource lifecycle, including create, read, update, delete, and list operations. This programmable surface allows for governed promotion of resources. Instead of manually recreating each resource, the MCP server reads the resource and its permissions from the source account and re-applies them in the target account, ensuring permissions are faithfully copied.

The migrator in this post composes these operations into a repeatable workflow, ensuring that a single tool call can promote resources in a single run.

The Quick Resource Migrator is a sample MCP server hosted on Amazon Bedrock AgentCore. It is resource-driven, allowing users to pick a resource type (agent, connector, knowledge base, flow, or space) and select by id, name, or all. The migrator is idempotent, ensuring that each resource is created or updated safely. Every update is protected by a versioned backup written to Amazon S3, providing a full history for review and rollback.

A read-only preview reports what the migration would create or update before committing, and the migration itself runs on Amazon Bedrock AgentCore, making it easy to drive from Amazon Quick or any MCP-compatible client.