[Feature Request] Add option to configure security checks behavior in Claude Code CLI

**Bug Description**
I am implementing application for digital signage and trying to have it as secure as possible but the checks of security of app are constantly downgrading the model... so instead of having app as secure as possible I end up in troubles... would be nice to have option to change it

**Environment Info**
- Platform: darwin
- Terminal: WarpTerminal
- Version: 2.1.234
- Feedback ID: 6795085d-dcd6-4484-9d90-a8f91f8c7d44

**Errors**
```json
[]
```

Start by reading the Claude Code CLI's existing security-check behavior and identify where model downgrades are decided. Define the configuration surface and its allowed behavior changes, then verify that the selected option consistently affects security checks without weakening the default behavior.

Written by the indexing model from the issue text.