Anthropic Enhances Claude Managed Agents with New Features
Anthropic has introduced three new features to its Claude Managed Agents, aimed at improving the functionality and efficiency of AI agents. The first feature, 'dreaming,' allows agents to review past sessions to identify patterns and enhance memory, thereby improving over time. The second feature, 'outcomes,' enables users to define success criteria for tasks, with a separate grader evaluating the agent's performance. The third feature, 'multiagent orchestration,' allows a lead agent to delegate tasks to specialized subagents, enhancing collaborative problem-solving. These updates are designed to streamline the deployment and operation of AI agents, making them more adaptable and efficient. 
