k8s-platform-tenancy

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: An indirect prompt injection surface is identified where user-provided data is interpolated into Kubernetes YAML templates.
  • Ingestion points: SKILL.md defines templates for Namespaces and ResourceQuotas that use variables such as ${TENANT_NAME}, ${ENV}, and ${OWNER_EMAIL}.
  • Boundary markers: No specific delimiters or boundary markers are suggested to prevent malicious input from escaping the intended YAML structure.
  • Capability inventory: The skill utilizes high-impact capabilities through mcp__flux-operator-mcp__apply_kubernetes_manifest to modify the cluster configuration.
  • Sanitization: The instructions do not specify validation or sanitization requirements for the inputs used to populate the resource templates.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 08:58 AM