Guardrail-Driven CTF Design: Level-Based Isolation + Platform Abuse Prevention

[syedDS]

This discussion proposes a guardrail-first design Module for challenges where:

-Each level defines both challenge logic and allowed attack surface
-Platform abuse is explicitly constrained without reducing realism
-Isolation boundaries evolve with difficulty

Few Questions to be discussed
1.How much realism is worth the added isolation cost as currently the platform is self deployable?
2.How do we standardize guardrail policies across different challenge types ? What telemetry is most useful for detecting platform abuse in real time?