Content filtering and topic-restriction policy for foundation models.
ai-ml
category11
settings1
inputs1
outputs
| Setting | Type | Required | Default |
|---|
| Guardrail name | Text | Yes | — |
| Description | Text | — | — |
| Blocked input message | Text | — | I cannot answer that. |
| Blocked output message | Text | — | I cannot answer that. |
| KMS key ARN | Text | — | — |
| Topic policy (JSON) | Text | — | — |
| Content filter policy (JSON) | Text | — | — |
| Word policy (JSON) | Text | — | — |
| PII policy (JSON) | Text | — | — |
| Contextual grounding (JSON) | Text | — | — |
| Tags | Key–value | — | — |
| Socket | Direction | Accepts | Terraform arg |
|---|
| KMS key | Input | aws.kms-key | kms_key_arn |
| Bedrock invocations | Output | any | — |