[BUG] Messages typed during "Thinking/Deliberating/Manifesting" state aren't delivered until user hits Escape
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?
When Claude is mid-turn (showing "Thinking", "Deliberating", "Manifesting", etc.), new user messages typed into the input box are not delivered to Claude until the user manually interrupts with Escape. This happens even when the new message is short and Claude's current turn has no background tasks. The only reliable way to get Claude to see a new message mid-turn is to interrupt the current turn.
What Should Happen?
New messages typed mid-turn should be queued and processed as the next turn automatically, without requiring the user to manually interrupt. The UI should ideally acknowledge the queued message visibly.
Error Messages/Logs
Steps to Reproduce
Send a message to Claude that will trigger a multi-step response (tool calls, chained thinking).
While Claude is still generating (UI shows "Thinking…" / "Deliberating…" / "Manifesting…"), type and submit a new message.
Observe: Claude continues its existing turn and does not acknowledge the new message.
Hit Escape to interrupt. Claude then receives the message (and the session injects a <system-reminder> saying "The user sent a new message while you were working").
Claude Model
Not sure / Multiple models
Is this a regression?
Yes, this worked in a previous version
Last Working Version
_No response_
Claude Code Version
2.1.112 (Claude Code)
Platform
Anthropic API
Operating System
Windows
Terminal/Shell
VS Code integrated terminal
Additional Information
_No response_
This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗