gke-ai-troubleshooting-tpu-vbar-oom
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- Command Execution: The skill provides and uses commands for
kubectlandgcloudto inspect cluster workloads and validate logging filters. These are standard diagnostic practices for Kubernetes and Google Cloud environments. - Indirect Prompt Injection: The skill is designed to ingest and analyze log data from external sources (Google Cloud Logging). This presents a potential surface for indirect prompt injection if the logs were to contain adversarial content.
- Ingestion points: The skill uses the
query_logstool to fetch data fromserialconsole.googleapis.com%2fserial_port_1_outputandk8s_containerlogs. - Boundary markers: No explicit delimiters or instructions to ignore embedded content are used within the logging filters.
- Capability inventory: The skill utilizes
kubectlfor resource inspection andgcloud logging read(viascripts/validate_queries.sh) for log retrieval. - Sanitization: No explicit sanitization or filtering of the fetched log content is performed before it is processed by the agent.
Audit Metadata