ETF Agent Maps the Agent Client Interaction Lifecycle
An ETF Agent engineering article presents an interaction framework for research-oriented Agent clients, illustrated with a fictional fund-comparison scenario. It covers scope confirmation, progressive answers, concurrent questions, source verification, failure handling, result delivery and resuming work after leaving the page. The article emphasizes visible context, user intervention, honest gaps and continuity across conversations. Its example wording and layouts are design proposals rather than evidence that every feature has been implemented, and it provides no benchmark results for the cited engineering practices.
The article gives product teams a concrete framework for evaluating research clients across long-running tasks, citations, delivery, reopening and follow-up questions. It also argues that interface features should be added only where users encounter real workflow breaks, rather than exposing every backend capability.
` 一份报告出现在页面上，并不意味着一次使用已经完成。
用户还可能不知道：系统有没有读懂自己的限制，等待时能不能问别的问题，刚才补充的要求是否被采用，结论中的数字从哪里来，以及关闭页面以后去哪里继续看。
Agent 客户端要组织的，是一段可理解、可干预、可续接的协作过程。 好的体验不仅让结果出现，也让用户在过程中始终知道自己交代了什么、现在能做什么，以及下一步应该怎样判断。
下面用一个教学情境贯穿全文：
“比较我选的两只基金，重点看持仓重合和费用，先给我关键差异。我可能会离开一会儿，回来继续看。”
这里不使用真实基金数据，也不推导投资结论。ETF Agent 已有的并发对话、渐进回答、引用阅读和异步交付实践提供工程背景；本文进一步讨论交互取舍。示例文案与布局是设计表达，不代表每个细节都已在产品中实现。