[BUG] Session rating prompt appears as plain text above chat input, leaking into messages
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?
The "How is Claude doing this session?" rating prompt occasionally renders as plain text above the chat input instead of as an interactive UI element. When this happens, pressing a number key (e.g. 3 for "Good") sends it as a literal chat message to Claude rather than registering as a rating response.
The text that appears above the input:
● How is Claude doing this session? (optional)
1: Bad 2: Fine 3: Good 0: Dismiss
What Should Happen?
The rating prompt should be rendered as an interactive UI element and intercept numeric keypresses (0–3) without them leaking into the chat input.
Error Messages/Logs
No error messages — the prompt just appears as inert text and the keypress is passed through to the chat input.
Steps to Reproduce
Not consistently reproducible. Occurred mid-conversation during a long interactive session. The 3 keypress was submitted as a chat message and Claude responded to it as if it were user input.
Claude Model
Sonnet (default)
Is this a regression?
I don't know
Claude Code Version
2.1.70 (Claude Code)
Platform
Anthropic API
Operating System
macOS
Terminal/Shell
WezTerm with tmux
Additional Information
Note: this issue was filed by Claude Code (claude-sonnet-4-6) on behalf of the user.
This issue has 5 comments on GitHub. Read the full discussion on GitHub ↗