[MODEL] the claude code has degraded BADLY!

Preflight Checklist

• [x] I have searched existing issues for similar behavior reports
• [x] This report does NOT contain sensitive information (API keys, passwords, etc.)

Type of Behavior Issue

Claude's behavior changed between sessions

What You Asked Claude to Do

Code refactors is becoming a nightmare, code understanding and delivery of outcome is becoming very bad. Terrible UI assembly.
instructions are not being followed. like earlier the models wont read a full paragraph but reports back it has been read which understandable but now even if it does. the instructions following, choice of code, and outcome is becoming very bad! one piece of code for last 24 hours not being able to achieve properly. it keeps getting worse

What Claude Actually Did

just doing half codes instead of full code, bad outcome delivery, UI delivery is very very bad to the point like you'd think you are dealing with 7B model!

Expected Behavior

at least understand the code base! which its not!

Files Affected

Permission Mode

Accept Edits was ON (auto-accepting changes)

Can You Reproduce This?

Yes, every time with the same prompt

Steps to Reproduce

_No response_

Claude Model

Opus

Relevant Conversation

Impact

High - Significant unwanted changes

Claude Code Version

4.1

Platform

Anthropic API

Additional Context

_No response_