[BUG] Inline thinking output with MAX_THINKING_TOKENS environment variable set no longer appears automatically

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?

A while back I added MAX_THINKING_TOKENS: 1024 in my environment variables section of ~/.claude/settings.json, and prior to today's update Claude would show the thinking content output in the VSCode terminal as it reasoned through whatever prompt I gave it. This update got rid of that, I have to hit ctrl+o to see the thinking output but that prevents me from seeing its actual output or from typing into the terminal, which really hinders the workflow.

What Should Happen?

The thinking tokens functionality as described above should return to how it worked before this update.

Error Messages/Logs

Steps to Reproduce

In ~/.claude/settings.json on Windows, make sure the file includes:
{
"env": {
"MAX_THINKING_TOKENS": "1024"
}
}

And then open Claude in the VSCode terminal, and type any command, even as simple as "hi." Claude would go through the thought process of describing what type of interaction it was and how it was meant to address me by my name, and then give whatever output, something like "Hi, how can I help you today?" but now it will just show you the output, and will make you hit ctrl+o to see that thought process.

Claude Model

Sonnet (default)

Is this a regression?

Yes, this worked in a previous version

Last Working Version

1.0.126

Claude Code Version

2.0.0 (Claude Code)

Platform

AWS Bedrock

Operating System

Windows

Terminal/Shell

VS Code integrated terminal

Additional Information

_No response_