[BUG] Feedback is too easy to mis-apply

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 asks for feedback on a 1-5 scale, it uses the next input character as input. I often am using lists with claude code, and will type 1, as in the first thing it should do. This is interpreted as feedback. You are likely getting a lot of bad feedback data due to this. I suggest having the user confirm or do something to enter feedback, not just grabbing the next thing they type. The feedback prompt is extremely easy to overlook.

What Should Happen?

The "1" I type should be accepted as normal input, not feedback.

Error Messages/Logs

Steps to Reproduce

1. Use claude until the feedback prompt appears
2. Type "1" as if you are about to try and give claude a list of items
3. Your "1" is now interpreted as feedback.

Claude Model

Opus

Is this a regression?

No, this never worked

Last Working Version

_No response_

Claude Code Version

2.1.107 (Claude Code)

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Other

Additional Information

_No response_