[MODEL] CLAUDE.md is systematically violated

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 ignored my instructions or configuration

What You Asked Claude to Do

Systematically violates constraints in CLAUDE.md. After I catch its violation: it answers like Ah yes, you said me not to do it but I did. Sorry, let's proceed.
It happening systematically!!!!
I can't trust you.
If you don't fix this behaviuor I need to find better AI solution! AI agent should do what I need without surprises!

What Claude Actually Did

Constant violation of the constraints in CLAUDE.md of the project. e.g., I forbade git, but it uses it without asking permission. I required it to check all the answers and qualify them by rules: VERIFIED, INFERENCE, UNKNOWN. It doesn't care about my rules. And may-many others!!!!!! I can't trust it - I can't predict the result!!! I need to check each step to be sure that it didn't lie to me

Expected Behavior

As GPT does: it checks all the restrictions in AGENTS.md and sticks to them. If it really needs to break my restrictions, it asks me EACH TIME if it can do it!

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

Claude Code: 2.1.22

Platform

Anthropic API

Additional Context

_No response_