[Feature Request] Allow model selection override for security/game development use cases

**Bug Description**
Compromises the whole point of a model as powerful as fable when I can't use it to help secure our own title... I sort of understand the scenario and it could be used to jailbreak but is there any way to gain an exception or anything because I'm certainly not trusting Opus' output when compared to Fable's capabilities on this for a cyber security assessment on our own Unity source code.

**Environment Info**
- Platform: win32
- Terminal: windows-terminal
- Version: 2.1.245
- Feedback ID: f67d786f-2886-422b-895d-fcf13a5dc6b5

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

I'm assuming session context is available on your own end.

No files, tests, or entry points are identified in the issue. Start by clarifying how model selection overrides should work for the stated Unity security-assessment use case, including the security and jailbreak concerns; done would require an agreed implementation scope and acceptance criteria.

Written by the indexing model from the issue text.