generative-ai-cdk-constructs

@cdklabs/generative-ai-cdk-constructs

---

@cdklabs/generative-ai-cdk-constructs / bedrock / GuardrailProps

Interface: GuardrailProps

Properties for creating a Guardrail.

Properties

blockedInputMessaging?

readonly optional blockedInputMessaging: string

The message to return when the guardrail blocks a prompt.

Default

"Sorry, your query violates our usage policy."

---

blockedOutputsMessaging?

readonly optional blockedOutputsMessaging: string

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

Default

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

---

contentFilters?

readonly optional contentFilters: ContentFilter[]

The content filters to apply to the guardrail. Note, if one of

---

contextualGroundingFilters?

readonly optional contextualGroundingFilters: ContextualGroundingFilter[]

The contextual grounding filters to apply to the guardrail.

---

deniedTopics?

readonly optional deniedTopics: Topic[]

Up to 30 denied topics to block user inputs or model responses associated with the topic.

---

description?

readonly optional description: string

The description of the guardrail.

---

kmsKey?

readonly optional kmsKey: IKey

A custom KMS key to use for encrypting data.

Default

"Your data is encrypted by default with a key that AWS owns and manages for you."

---

managedWordListFilters?

readonly optional managedWordListFilters: ManagedWordFilter[]

The managed word filters to apply to the guardrail.

---

name

readonly name: string

The name of the guardrail.

---

piiFilters?

readonly optional piiFilters: PIIFilter[]

The PII filters to apply to the guardrail.

---

regexFilters?

readonly optional regexFilters: RegexFilter[]

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

---

wordFilters?

readonly optional wordFilters: WordFilter[]

The word filters to apply to the guardrail.

This site is open source. Improve this page.