AWS Cuts AI Agent Setup To 3 API Calls In AgentCore Update

AWS previewed AgentCore, a managed AI agent harness in Amazon Bedrock that can launch autonomous agents with three API calls.

Each agent session runs in an isolated microVM with a persistent filesystem, reducing the infrastructure work needed for secure, stateful execution.

AgentCore supports multiple frameworks and models, adds a new CLI, and includes prebuilt coding-assistant skills for tools like Kiro, Claude Code, and Codex.

The update targets a major enterprise blocker: the time and engineering effort required to provision, secure, and persist agent infrastructure before workflows can run.