opus 4.6 and opus 4.7 the quality is not the same. This is now the second time i had to use a second grade model deepseek v4 so fix opus mistakes.

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 modified files I didn't ask it to modify

What You Asked Claude to Do

You models are not up to standard anymore, this is the second time i had to ask a second grade model like deepseek v4 to fix opus errors and issues. Please i am not sure if it is the app or you models please investigate. Also that 5 hour limit is ok but why only 100k tokens make it more please. How do you expect me to get my work done. I only use like 5 % of the weekly usage but 100 % on the 5 hour the reason being 100k token lmao really.

What Claude Actually Did

Claude make alot of mistakes and it was never like this. it is stubborn and full of stories and does not want to do the job, it argues with you and tell you what it wants to do and not what you want. Before i could get my app working 5hour limit i took the same issue to Deepseek it took 5 min and boom problem solved with extras that it picked up.

Expected Behavior

Should do what i want and not what it wants to do.

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

Sonnet

Relevant Conversation

Impact

Critical - Data loss or corrupted project

Claude Code Version

opus 4.7

Platform

Anthropic API

Additional Context

_No response_