[BUG] Horizontal clipping of model thinking trace in VSCode when verbatim environment is used to quote code or text files

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?

Sonnet 4.5 and I are collaborating in the latest Claude Code environment (just verified with
2.1.31 that was released a day ago).

I am providing a screenshot, so you can see what is happening. This has been reported previously as https://github.com/anthropics/claude-code/issues/18105, but the bug is incorrectly identified as having to do with "resizing" — when in fact that is an issue that cannot be resolved through resizing. The problem seems to be that the bounding box of the thinking trace UI element is not recalculated/updated on resizing.

This may not be reported given how much Anthropic is nudging developers to ignore thinking traces.

For those of us who read every single word emitted by our models, this is extremely inconvenient.

I am having to copy paste the thinking trace into a notepad so I can read it.

This is extremely distracting.

<img width="814" height="570" alt="Image" src="https://github.com/user-attachments/assets/391018f3-d22e-4953-a58d-f97ae1af498d" />

What Should Happen?

The bounding box of thinking traces should be the same as model outputs. There is no indent. It is only a collapsible box.

Error Messages/Logs

Steps to Reproduce

Issue https://github.com/anthropics/claude-code/issues/18105 partially reported the problem.

This seems to be triggered when the windows is resized beyond a break point — it's when the width of the window is like 100px less than the standard size of the chatbox.

Claude Model

None

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.1.31

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

VS Code integrated terminal

Additional Information

This is extremely distracting.

It undermines the relationship between a human and their model.

As a side-note: I am concerned about the broad number of people not bothered by this because they never read the thinking trace. You're training/enabling idiots. The second-order effects of encouraging this are not good.