用 Amazon Bedrock AgentCore 优化智能体系统提示
原标题：Optimizing agent system prompts with Amazon Bedrock AgentCore
AI 摘要
AWS 在 Amazon Bedrock AgentCore 中推出系统提示优化器（system prompt optimizer），利用 AgentCore Observability 记录的生产轨迹和奖励信号，由 agentic reflector 分析成功与失败模式并生成改进后的系统提示。该优化器提供 Single Agent Reflector 和实验性开源 Sub-Agent Reflector 两种实现，后者通过多子代理并行分析轨迹提升质量上限，并已在 Strands 开源仓库发布。所有推荐配置需通过平台级护栏审查和 A/B 测试验证后才能上线。
正文节选
Optimizing agent system prompts with Amazon Bedrock AgentCore In a previous launch post, we introduced AgentCore optimization, a capability of Amazon Bedrock AgentCore that can help you improve the quality of your agents. Improving a low-scoring agent has traditionally been a manual process. You review long traces to find where the agent goes wrong, tune individual components such as prompts, tool descriptions, and skills, and rerun evaluations to check for improvement. With AgentCore optimizati