gcp-gke-monitoring-observability

Warn

Audited by Gen Agent Trust Hub on Feb 24, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill references a sensitive file path used for service account authentication.\n
  • Evidence: Reference to /var/secrets/google/key.json in the Spring Cloud GCP configuration example within references/advanced-configuration.md.\n- [COMMAND_EXECUTION]: The skill utilizes administrative commands to modify cluster state and interact with running workloads.\n
  • Evidence: Extensive use of gcloud container clusters update to modify logging and monitoring configurations in SKILL.md and examples/examples.md.\n
  • Evidence: Use of kubectl exec to perform health checks and metric queries within running pods in examples/examples.md.\n- [PROMPT_INJECTION]: The skill processes untrusted external data from Kubernetes logs, creating a potential surface for indirect prompt injection.\n
  • Ingestion points: Data is ingested from container logs using gcloud logging read and kubectl logs in SKILL.md and examples/examples.md.\n
  • Boundary markers: The skill does not implement delimiters or safety instructions when handling ingested log data.\n
  • Capability inventory: The skill possesses the ability to update cluster settings, deploy resources, and execute commands within pods.\n
  • Sanitization: Log data is read and processed without validation or sanitization of its content.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 24, 2026, 05:23 PM