k8s-multicluster

Pass

Audited by Gen Agent Trust Hub on Nov 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external Kubernetes clusters, creating a surface for indirect prompt injection where malicious data in the cluster could influence agent behavior.\n
  • Ingestion points: Data enters the agent context via kubectl_get, capi_get_cluster, helm_list, gitops_apps_list, cilium_list_nodes, and istio_proxy_status (found in SKILL.md and CONTEXT-SWITCHING.md).\n
  • Boundary markers: The instructions do not specify any delimiters or warnings to ignore potentially malicious instructions embedded within the retrieved Kubernetes resource data or logs.\n
  • Capability inventory: The skill possesses significant operational capabilities including kubectl_apply (resource creation/modification), install_helm_chart (software deployment), capi_scale_machinedeployment (infrastructure scaling), and gitops_app_sync (synchronization triggers).\n
  • Sanitization: There are no visible mechanisms for sanitizing, validating, or filtering the content retrieved from the Kubernetes API before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 9, 2026, 06:52 PM