gke-ai-troubleshooting-tpu-vm-repair-stuck

Pass

Audited by Gen Agent Trust Hub on Oct 20, 2026

Risk Level: SAFE
Full Analysis
  • Diagnostic Data Ingestion (Indirect Prompt Injection Surface): The skill instructs the agent to process data from external sources such as Google Cloud Audit Logs and cluster node status information. While ingesting external data is an attack surface for indirect prompt injection, the risk is minimized here because the data originates from the user's own cloud environment and the skill follows a strict 'Read-only rule' for all diagnostic operations.
  • Standard Tooling and Documentation: The skill references well-known tools like gcloud and kubectl and provides links to official documentation for their use. These resources originate from the vendor's established infrastructure, which is consistent with the skill's intended troubleshooting purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 20, 2026, 06:34 AM