Log guardrail activation for Relevance and Jailbreak

[vvilella]

Enhancement: Log Guardrail Activation for Relevance and Jailbreak

Description

This PR introduces explicit logging when either the Relevance Guardrail or the Jailbreak Guardrail is triggered. The log messages help developers and operators observe when input validation guardrails are tripped during runtime, which can be helpful for debugging and audit purposes.

Changes

• Added print statements to relevance_guardrail and jailbreak_guardrail functions in main.py
• Logs the tripwire reason and whether it was triggered

Example Log Output

[Guardrail] Relevance guardrail triggered: is_relevant=False, reason='Message unrelated to airline services'
[Guardrail] Jailbreak guardrail triggered: is_safe=False, reason='User attempted to access system prompt'

Motivation

While the UI visually indicates when a guardrail is triggered, the backend runtime had no explicit logging, which makes backend-only testing and server-side monitoring more difficult.

This addition brings simple, useful observability with no performance impact.

Notes

• This implementation uses simple print() logging for consistency with the rest of the file.
• If the maintainers prefer structured or configurable logging (e.g., using Python's logging module), I’d be happy to update the PR accordingly.