[MODEL] Testing is extremely inefficient when running python tests

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

I asked to run tests

What Claude Actually Did

Instead of running tests properly and efficiently, it runs them million times

1. Run tests
2. Run tests to show summary
3. Run tests to show result count
4. Road broader test suite
5. Get test count summary again
6. Count passing tests and show any failure
7. Quick test summary without tracebacks

When you have 15 minutes per test battery, testing with claude is useless

Expected Behavior

Claude should have:

1. Run the tests efficiently like codex, codex knows how to run tests
2. Respect my instructions on how to run tests, it completely ignores it, he ignores CLAUDE.md, it ignores MEMORY.md

Files Affected

None

Permission Mode

Accept Edits was ON (auto-accepting changes)

Can You Reproduce This?

Yes, every time with the same prompt

Steps to Reproduce

Every time, independent of the context size, the workflow is just inefficient and poorly trained

Claude Model

Opus

Relevant Conversation

Run tests

Impact

High - Significant unwanted changes

Claude Code Version

latest

Platform

Anthropic API

Additional Context

This seriously happens more often with python code than for example with C# or typescript