creating-secrets-using-best-practices

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it allows secret configuration data to be retrieved from remote, potentially untrusted URLs.\n
  • Ingestion points: The SOP in references/create-secrets-using-best-practices.md instructs the agent to support parameter acquisition via a URL.\n
  • Boundary markers: There are no defined delimiters or instructions provided to the agent to disregard embedded instructions within the content retrieved from the URL.\n
  • Capability inventory: The skill utilizes the call_aws tool to perform high-privilege operations, including the creation of KMS keys, management of Secrets Manager secrets, and modification of IAM policies.\n
  • Sanitization: No explicit sanitization or validation logic is defined for the remote configuration payload prior to its processing by the agent.\n- [COMMAND_EXECUTION]: The skill executes local shell commands to verify the environment's configuration.\n
  • Evidence: The SOP requires running aws sts get-caller-identity to confirm the AWS CLI is properly configured before proceeding with secret creation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 05:50 PM
Security Audit — agent-trust-hub — creating-secrets-using-best-practices