GuardrailProps

class aws_cdk.aws_bedrock_alpha.GuardrailProps(*, guardrail_name, blocked_input_messaging=None, blocked_outputs_messaging=None, content_filters=None, content_filters_tier_config=None, contextual_grounding_filters=None, cross_region_config=None, denied_topics=None, description=None, kms_key=None, managed_word_list_filters=None, pii_filters=None, regex_filters=None, topics_tier_config=None, word_filters=None)

Bases: object

(experimental) Properties for creating a Guardrail.

Parameters:

  • guardrail_name (str) – (experimental) The name of the guardrail. This will be used as the physical name of the guardrail.

  • blocked_input_messaging (Optional[str]) – (experimental) The message to return when the guardrail blocks a prompt. Must be between 1 and 500 characters. Default: “Sorry, your query violates our usage policy.”

  • blocked_outputs_messaging (Optional[str]) – (experimental) The message to return when the guardrail blocks a model response. Must be between 1 and 500 characters. Default: “Sorry, I am unable to answer your question because of our usage policy.”

  • content_filters (Optional[Sequence[Union[ContentFilter, Dict[str, Any]]]]) – (experimental) The content filters to apply to the guardrail. Default: []

  • content_filters_tier_config (Optional[TierConfig]) – (experimental) The tier configuration to apply to the guardrail. Default: filters.TierConfig.CLASSIC

  • contextual_grounding_filters (Optional[Sequence[Union[ContextualGroundingFilter, Dict[str, Any]]]]) – (experimental) The contextual grounding filters to apply to the guardrail. Default: []

  • cross_region_config (Union[GuardrailCrossRegionConfigProperty, Dict[str, Any], None]) – (experimental) The cross-region configuration for the guardrail. This is optional and when provided, it should be of type GuardrailCrossRegionConfigProperty. Default: - No cross-region configuration

  • denied_topics (Optional[Sequence[Topic]]) – (experimental) A list of policies related to topics that the guardrail should deny. Default: []

  • description (Optional[str]) – (experimental) The description of the guardrail. Default: - No description

  • kms_key (Optional[IKey]) – (experimental) A custom KMS key to use for encrypting data. Default: - Data is encrypted by default with a key that AWS owns and manages for you

  • managed_word_list_filters (Optional[Sequence[Union[ManagedWordFilter, Dict[str, Any]]]]) – (experimental) The managed word filters to apply to the guardrail. Default: []

  • pii_filters (Optional[Sequence[Union[PIIFilter, Dict[str, Any]]]]) – (experimental) The PII filters to apply to the guardrail. Default: []

  • regex_filters (Optional[Sequence[Union[RegexFilter, Dict[str, Any]]]]) – (experimental) The regular expression (regex) filters to apply to the guardrail. Default: []

  • topics_tier_config (Optional[TierConfig]) – (experimental) The tier configuration to apply to the guardrail. Default: filters.TierConfig.CLASSIC

  • word_filters (Optional[Sequence[Union[WordFilter, Dict[str, Any]]]]) – (experimental) The word filters to apply to the guardrail. Default: []

Stability:

experimental

ExampleMetadata:

fixture=default infused

Example:

guardrail = bedrock.Guardrail(self, "bedrockGuardrails",
    guardrail_name="my-BedrockGuardrails"
)
# Add regex filter with input/output actions
guardrail.add_regex_filter(
    name="TestRegexFilter",
    pattern="test-pattern",
    action=bedrock.GuardrailAction.ANONYMIZE,
    # below props are optional
    description="This is a test regex filter",
    input_action=bedrock.GuardrailAction.BLOCK,
    input_enabled=True,
    output_action=bedrock.GuardrailAction.ANONYMIZE,
    output_enabled=True
)

Attributes

blocked_input_messaging

(experimental) The message to return when the guardrail blocks a prompt.

Must be between 1 and 500 characters.

Default:

“Sorry, your query violates our usage policy.”

Stability:

experimental

blocked_outputs_messaging

(experimental) The message to return when the guardrail blocks a model response.

Must be between 1 and 500 characters.

Default:

“Sorry, I am unable to answer your question because of our usage policy.”

Stability:

experimental

content_filters

(experimental) The content filters to apply to the guardrail.

Default:

[]

Stability:

experimental

content_filters_tier_config

(experimental) The tier configuration to apply to the guardrail.

Default:

filters.TierConfig.CLASSIC

Stability:

experimental

contextual_grounding_filters

(experimental) The contextual grounding filters to apply to the guardrail.

Default:

[]

Stability:

experimental

cross_region_config

(experimental) The cross-region configuration for the guardrail.

This is optional and when provided, it should be of type GuardrailCrossRegionConfigProperty.

Default:

  • No cross-region configuration

Stability:

experimental

denied_topics

(experimental) A list of policies related to topics that the guardrail should deny.

Default:

[]

Stability:

experimental

description

(experimental) The description of the guardrail.

Default:

  • No description

Stability:

experimental

guardrail_name

(experimental) The name of the guardrail.

This will be used as the physical name of the guardrail.

Stability:

experimental

kms_key

(experimental) A custom KMS key to use for encrypting data.

Default:

  • Data is encrypted by default with a key that AWS owns and manages for you

Stability:

experimental

managed_word_list_filters

(experimental) The managed word filters to apply to the guardrail.

Default:

[]

Stability:

experimental

pii_filters

(experimental) The PII filters to apply to the guardrail.

Default:

[]

Stability:

experimental

regex_filters

(experimental) The regular expression (regex) filters to apply to the guardrail.

Default:

[]

Stability:

experimental

topics_tier_config

(experimental) The tier configuration to apply to the guardrail.

Default:

filters.TierConfig.CLASSIC

Stability:

experimental

word_filters

(experimental) The word filters to apply to the guardrail.

Default:

[]

Stability:

experimental