gitops-cluster-debug

Pass

Audited by Gen Agent Trust Hub on Mar 16, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its interaction with untrusted cluster data.
  • Ingestion points: The agent reads live data from the cluster via get_kubernetes_resources (status, events, annotations) and get_kubernetes_logs (pod stdout/stderr).
  • Boundary markers: The instructions do not define clear delimiters or specific warnings to ignore instructions embedded within the data retrieved from the cluster.
  • Capability inventory: The skill allows the agent to modify cluster state using the apply_kubernetes_resource tool and change active environments via set_kubeconfig_context.
  • Sanitization: There is no mention of filtering or sanitizing the content of logs or resource metadata before processing.
  • [COMMAND_EXECUTION]: The skill provides the agent with powerful tools to manage a Kubernetes cluster. While necessary for the debugging use case, the ability to apply arbitrary YAML manifests (apply_kubernetes_resource) and switch contexts (set_kubeconfig_context) represents a high-privilege interface that could be abused if the agent's logic is subverted.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 16, 2026, 08:02 AM