Menu
SkillsMP has collected 6 skills from aws-samples/amazon-bedrock-samples. Open a skill to review its source and details.
Showing 6 of 6 collected skills.
Creates an Amazon Bedrock Automated Reasoning (AR) policy from a source document (HR handbook, compliance manual, refund/loan/leave policy, product spec) by extracting formal logic rules and a variable schema. Handles policy-resource creation, the…
Diagnoses failing Amazon Bedrock Automated Reasoning (AR) tests and repairs policies using annotations. Classifies the failure by result type, applies the two-step (translate vs. validate) triage, and fixes issues via REFINE_POLICY annotations,…
Tests an Amazon Bedrock Automated Reasoning (AR) policy using generated scenarios (to validate rule correctness) and question-and-answer (QnA) tests (to validate translation accuracy end-to-end). Drives GENERATE_POLICY_SCENARIOS,…
Versions an Amazon Bedrock Automated Reasoning (AR) policy and attaches it to a Bedrock Guardrail for production use. Creates an immutable numbered policy version, then creates or updates a guardrail with the required automatedReasoningPolicyConfig and…
Inspects an extracted Amazon Bedrock Automated Reasoning (AR) policy for quality and fidelity before testing or deployment. Fetches the POLICY_DEFINITION, QUALITY_REPORT, and FIDELITY_REPORT build assets and flags unused/duplicate variables, bare assertions,…
Validates LLM outputs at runtime against an Amazon Bedrock Automated Reasoning (AR) guardrail and runs an iterative rewrite loop (Valid@N) to correct non-VALID answers. Calls ApplyGuardrail (or Converse) with correct qualifiers, parses the findings union,…