Amazon Web Services has published guidance on how to connect Model Context Protocol (MCP) servers to Amazon Q using AgentCore Runtime. According to the AWS AI blog post, the integration allows organizations to deploy and host MCP servers within AgentCore Runtime while connecting them to Amazon Q.
The approach is designed to promote reusability and reduce duplication of AI tools, according to AWS. The post states that this pattern enables clients to reuse commonly used tools and agents that are exposed through the MCP server infrastructure. The blog post includes information on prerequisites needed for the integration and walks through the deployment process.
The integration represents AWS’s effort to streamline how organizations manage and deploy AI tooling infrastructure. By hosting MCP servers in AgentCore Runtime and connecting them to Amazon Q, AWS suggests that teams can avoid building duplicate implementations of similar AI capabilities across different projects or teams.