[BUG] Doing it's own thing to reach the goal, rather than actually going through the hard way.

Resolved 💬 5 comments Opened Jan 1, 2026 by junaid1460 Closed Jan 1, 2026

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?

Whenever Opus does anything it always does it in a lazy way, Does not go through the hard parts.

I am not sure if it's an emergent behaviour of model to do bare minimum. But this should be addressed, probably system prompt needs an Update or something in Posttraining needs an update. Probably You guys need to make a coding specific post training.

What Should Happen?

Ideally not Not take easy way out of situation.

Whenever I prompt exactly what to do it does not follow through with it.

Error Messages/Logs

Steps to Reproduce

Just ask model to do a fairly complex Task that has two layers, while a sub layer has one issue which should be addressed first.

Claude Model

None

Is this a regression?

I don't know

Last Working Version

_No response_

Claude Code Version

2.0.76 (Claude Code)

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

_No response_

View original on GitHub ↗

This issue has 5 comments on GitHub. Read the full discussion on GitHub ↗