[Feature] Implement a guard layer for malicious code

[Gauravpadam]

The problem

The environments in which code is executed are isolated. However, it would be nice to have a precaution layer which handles code inspection before the evaluation starts.

Proposed solution

Upto the user who picks it up, It would be nice to see something deterministic first

Alternatives considered

What could be an alternative to security? 😄

Additional context

This would need a layer of parsing before the code reaches execution layers