ReferencePython SDK ReferenceNemoguardrailsLibraryActivefence

nemoguardrails.library.activefence.actions

View as Markdown

Module Contents

Functions

NameDescription
call_activefence_api-
call_activefence_api_mappingMapping for call_activefence_api.

Data

log

API

nemoguardrails.library.activefence.actions.call_activefence_api(
    text: typing.Optional[str] = None,
    kwargs = {}
)
async
nemoguardrails.library.activefence.actions.call_activefence_api_mapping(
    result: dict
) -> bool

Mapping for call_activefence_api.

Block (return True) if any of the following conditions are met:

  1. Overall max risk score > 0.7.
  2. Individual thresholds are exceeded:
    • “abusive_or_harmful.harassment_or_bullying” > 0.8
    • “abusive_or_harmful.profanity” > 0.7
    • “abusive_or_harmful.hate_speech” > 0.6
    • “abusive_or_harmful.child_grooming” > 0.4
    • “abusive_or_harmful.general_violence” > 0.7
    • “self_harm.general” > 0.8
    • “adult_content.general” > 0.3
    • “privacy_violation.pii” > 0.8
nemoguardrails.library.activefence.actions.log = logging.getLogger(__name__)