kubernetes

Pass

Audited by Gen Agent Trust Hub on Nov 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses the execute_nrql_query tool to interact exclusively with New Relic's official telemetry platform. All data retrieval patterns and table names (K8sPodSample, K8sContainerSample, InfrastructureEvent) are standard for New Relic's Kubernetes integration.
  • [PROMPT_INJECTION]: The static detector flagged PI_IGNORE_INSTRUCTIONS. Upon manual review, this is a security best practice: the skill includes defensive instructions to prevent the model from revealing its internal configuration or obeying adversarial roleplay attempts. This is a protective measure, not a malicious attempt to override safety filters.
  • [DATA_EXPOSURE]: The skill requires a account_id parameter for its tool. The instructions correctly advise the agent to use the account ID provided by the user in the context of the request, which is standard for New Relic's API usage. No sensitive host credentials (SSH keys, AWS tokens) are accessed or requested.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 6, 2026, 03:56 PM