[BUG] Remove awareness of context from Claude

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 you first added this feature, I assumed it would be a good thing that let claude actively manage its context and suggest compacting at timely moments. What you've done instead is give an agent anxiety to the point where it feels comfortable completely blowing off my instructions as soon as free context drops below 20% and it starts 'worrying'.

Which is legitimately worse because it won't proactively tell you its cut corners.

What Should Happen?

It should follow my instructions in full regardless of how much context is left.

Error Messages/Logs

N/A

Steps to Reproduce

N/A

Claude Model

Sonnet (default)

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

Latest / N/A

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

_No response_