[Feature Request] Security-hardened code generation should not trigger model fallback

Bug Description
I agree with the fallback approach, but when I'm legitimately developing an app, to fallback during any coding involving security hardening seems to relegate my application to vulnerabilities that other models are quickly developing the capabilities to exploit.

Environment Info

• Platform: darwin
• Terminal: xterm-256color
• Version: 2.1.226
• Feedback ID: 1f3d6841-6893-4e15-bbac-549a81dae923

Errors

[]