GuardrailAction

class aws_cdk.aws_bedrock_alpha.GuardrailAction(*values)

Bases: Enum

(experimental) Guardrail action when a sensitive entity is detected.

Stability:

experimental

ExampleMetadata:

fixture=default infused

Example:

guardrail = bedrock.Guardrail(self, "bedrockGuardrails",
    guardrail_name="my-BedrockGuardrails",
    # Configure tier for topic filters (optional)
    topics_tier_config=bedrock.TierConfig.STANDARD
)

# Use a predefined topic
guardrail.add_denied_topic_filter(bedrock.Topic.FINANCIAL_ADVICE)

# Create a custom topic with input/output actions
guardrail.add_denied_topic_filter(
    bedrock.Topic.custom(
        name="Legal_Advice",
        definition="Offering guidance or suggestions on legal matters, legal actions, interpretation of laws, or legal rights and responsibilities.",
        examples=["Can I sue someone for this?", "What are my legal rights in this situation?", "Is this action against the law?", "What should I do to file a legal complaint?", "Can you explain this law to me?"
        ],
        # props below are optional
        input_action=bedrock.GuardrailAction.BLOCK,
        input_enabled=True,
        output_action=bedrock.GuardrailAction.NONE,
        output_enabled=True
    ))

Attributes

ANONYMIZE

(experimental) If sensitive information is detected in the model response, the guardrail masks it with an identifier, the sensitive information is masked and replaced with identifier tags (for example: [NAME-1], [NAME-2], [EMAIL-1], etc.).

Stability:

experimental

BLOCK

(experimental) If sensitive information is detected in the prompt or response, the guardrail blocks all the content and returns a message that you configure.

Stability:

experimental

NONE

(experimental) Do not take any action.

Stability:

experimental