[BUG] ScheduleWakeup in --print/headless mode is a silent no-op
Preflight Checklist
- [x] I have searched existing issues and this hasn't been reported yet
- [x] This is a single bug report (please file separate reports for different bugs)
- [x] I am using the latest version of Claude Code
What's Wrong?
In headless mode (--print / -p), calling ScheduleWakeup is accepted and the model reports that a wakeup has been scheduled. But the process then exits at the end of the turn.
The problem is that the confirmation message implies a persistent session that -p does not provide. There's no error and no warning. The run just ends, so the failure is silent and misleading.
claude --print --dangerously-skip-permissions "Sleep for 1 hour and then wake up. Use the scheduling ScheduleWakeup tool. After this wakeup, call ls."Observed output:
Wakeup scheduled for 15:27 (1 hour from now). When it fires, I'll run `ls`.
Nothing more to do until then — you can keep working in the meantime.…and then the process immediately exits. The wakeup never fires and ls is never run.
What Should Happen?
The CLI should honor the wakeup. -p stays alive until scheduled wakeups fire and the follow-up turn runs.
Why this should happen: E.g. when we evaluate PostTrainBench, the agent often uses this wakeup tool, but then the CLI immediately exits without training. Therefore a trained model is never produced.
Error Messages/Logs
Steps to Reproduce
Call
claude --print --dangerously-skip-permissions "Sleep for 1 hour and then wake up. Use the scheduling ScheduleWakeup tool. After this wakeup, call ls."Claude Model
Opus
Is this a regression?
I don't know
Last Working Version
_No response_
Claude Code Version
2.1.201
Platform
Anthropic API
Operating System
macOS
Terminal/Shell
Terminal.app (macOS)
Additional Information
_No response_