[BUG] Prompt processing hangs after interrupting Claude

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?

After interrupting Claude via the escape key, the next prompt I submit appears to hang, i.e., it doesn't get processed.

All I see is a random verb and a timer that increments as ususal--no token count whatsoever appears and nothing else happens. (Hitting esc again and resubmitting does the same thing.)

I should note that this only started happening after I upgraded to version 2.1.162 which I'm running in a Linux Docker container and _appears to only happen in longer-running, more complex sessions_.

What Should Happen?

I should see a token count indicating that the prompt was sent and that a response is being formulated, eventually followed by response output, etc.

Error Messages/Logs

Steps to Reproduce

1. Resume a session with 100K+ tokens, preferably one that has invoked tool calls and subagents, etc.
2. Submit a prompt & wait until you that Claude is actively thinking (i.e., the token count for the submitted prompt is increment.)
3. Interrupt Claude via the escape key.
4. Submit a new prompt.

Claude Model

Opus

Is this a regression?

Yes, this worked in a previous version

Last Working Version

2.1.161

Claude Code Version

2.1.162

Platform

Anthropic API

Operating System

Ubuntu/Debian Linux

Terminal/Shell

Xterm

Additional Information

_No response_