The minimum confidence level required for Automated Reasoning policy violations to trigger guardrail actions. Values range from 0.0 to 1.0.
The list of Automated Reasoning policy ARNs that should be applied as part of this guardrail configuration.