AWS has introduced updates for Amazon Bedrock AgentCore that fundamentally simplify and accelerate the development of AI agents.
Amazon Web Services (AWS) has introduced a series of significant updates for Amazon Bedrock AgentCore, aimed at radically simplifying and accelerating the AI agent development process. These innovations are designed to remove infrastructural barriers that previously slowed development teams at all stages — from prototyping to deployment in a production environment. The central element of the updates is the managed agent harness feature in AgentCore.
instead of days of work on configuring frameworks, storage, authentication, and deployment pipelines, now just three API calls are sufficient to define an agent, its model, tools, and instructions. This allows developers to fully focus on the agent's logic itself, without being distracted by the underlying infrastructure. AgentCore, originally designed for this purpose, continues to provide support for popular frameworks such as LangGraph and LlamaIndex. This approach significantly reduces the time from concept to a fully functional agent, from days to mere minutes. Developers gain the ability to quickly validate various hypotheses by changing only API parameters, which accelerates the iterative process and overall development. At the same time, the solution maintains a high degree of flexibility: the AgentCore harness is based on AWS's open-source Strands Agents framework.
This means that if there is a need to create specific custom orchestration logic or coordinate multiple agents, one can easily switch from configuration to code writing, using the same unified platform and isolation mechanisms. In addition, AgentCore now has the ability to persist session state in a permanent file system. This functionality makes it practical to use human-in-the-loop scenarios without the need for complex integration. Agents can effectively pause task execution and resume from the same point, which significantly simplifies the process and eliminates the need for redesign.
The set of innovations is completed by the updated AgentCore command line interface (CLI), which integrates the entire agent lifecycle into a single, optimized workflow. This gives developers the ability to create, deploy, and operate agents directly from a single terminal, ensuring unprecedented consistency and efficiency at all stages — from development to production operation.
