Copilot is too lazy sometimes

### Describe the bug

Since a few versions i noticed with the claude-opus 4.6 (high) model that copilot has become what I would call "lazy"... If I ask it to implement a new feature there is about a 50% chance that it starts exploring the code, then creates a plan for the implementation but do nothing but waiting for me to say "start the implementation" instead of implement it right away.

Given that each interaction requires 3x premium request this effectively often doubles the cost before some code is actually created (not counting possible refinements) beside its being annoying as I already _have_ described it and usually copilot after making a plan then starts to explore things again to know how to implement its own plan :-\

### Affected version

_No response_

### Steps to reproduce the behavior

_No response_

### Expected behavior

_No response_

### Additional context

_No response_