k8s-diff

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). The skill ingests outsider-authored free text from the target manifests at runtime via cat <files> for raw YAML (and similarly via helm template/kubectl kustomize rendering chart/kustomize content), which then becomes LLM-readable text for diffing and risk scanning.