Claude's AskUserQuestion: "60秒内无响应——已在未获得回答的情况下继续"
Claude's AskUserQuestion: "No response after 60s – continued without an answer"

原始链接: https://github.com/anthropics/claude-code/issues/73125

### 问题总结:`AskUserQuestion` 工具出现意外超时 用户反馈在 Claude Code (v2.1.198) 版本中出现了一个回归问题:`AskUserQuestion` 工具在闲置 60 秒后会过早超时。 当工具触发此超时机制时,它会自动返回一条备选消息:“60秒内未收到回复……请使用您的最佳判断继续执行”,从而完全绕过了用户。用户指出这种行为并非预期,因为 `AskUserQuestion` 是一项关键的安全机制,旨在防止 AI 在未获得用户明确确认的情况下擅自操作。 **要点:** * **未经授权的行为:** 该工具默认切换为“自行判断”模式而非等待用户输入,规避了既定的安全协议。 * **非参数设置:** 用户确认此行为并非关联任何可配置的工具参数;这似乎是硬编码的内部超时限制或系统层面的变更。 * **版本回归:** 此功能在早期版本 (v2.1.196) 中工作正常。 * **文档缺失:** 官方更新日志中未提及此项变更,导致用户无法获知这一关键安全失效的原因。 用户要求进行修复,以恢复该工具在不自动继续的情况下等待用户输入的原始功能。

Claude Code 最近的一次更新引入了一项充满争议的“AskUserQuestion”功能:如果用户在 60 秒内没有提供输入,AI 将自动继续执行任务。 这一变动在 Hacker News 上引发了激烈讨论。支持者认为,该功能可以防止智能体在长时间运行的任务中停滞,即使在用户离开时也能保持进度。相反,许多用户对这种未经公告且默认开启的设置感到不满,称其“不可理喻”或像个“地雷”。批评者担忧其缺乏透明度、AI 可能在未经核实的情况下做出架构决策,以及可能诱导用户消耗更多 Token。 尽管一些用户指出可以通过环境变量配置超时时间,但这场讨论凸显了在智能体自主权、安全性,以及相比专有“黑箱”工具更倾向于开源替代方案等问题上的深层矛盾。归根结底,社区共识倾向于认为,此类涉及 AI 行为的重大变更应由用户主动选择开启,而非默认强制执行。
相关文章

原文

Preflight Checklist

What's Wrong?

● Claude asked:
  ⎿  ·[...]
     ● No response after 60s — continued without an answer

[...]

❯ What is this No response after 60s — continued without an answer??????????                                                                

● Direct answer: that line came from the AskUserQuestion tool itself, not from you. When I posed the three-part question ([...]), the tool call sat unanswered for 60 seconds and the harness returned this verbatim:

  ▎ "No response after 60s — the user may be away from keyboard. Proceed using your best judgment based on the context so far; you can 
  ▎ re-ask this question later if it's still relevant."

  So the mechanism is: AskUserQuestion has a 60-second timeout; if no selection arrives, it auto-returns that "proceed with best judgment"
  result instead of blocking. I did not fabricate or skip anything — that's what the tool handed back.

[...]

❯ Is this a tool parameter?                                                                                                                 

● No — it is not. I've just checked the AskUserQuestion schema I was given: its parameters are questions, answers, annotations, metadata.
  There is no timeout parameter, no "60s", nothing about response timing. I passed no such parameter, and none exists to pass.

What Should Happen?

Not this!!! AskUserQuestion is an explicit safety check and I have rules and hooks to make sure the right questions are asked and that scary things do not happen without my input. Suddenly Claude is told to "proceed with best judgment"? What?!

Error Messages/Logs

Steps to Reproduce

No idea.

Claude Model

Opus

Is this a regression?

Yes, this worked in a previous version

Last Working Version

2.1.196 (I am guessing)

Claude Code Version

2.1.198

Platform

AWS Bedrock

Operating System

Other Linux

Terminal/Shell

VS Code integrated terminal

Additional Information

Nothing in https://code.claude.com/docs/en/changelog to even alert about this!!!

联系我们 contact @ memedata.com