k8s-kubevirt
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from the Kubernetes cluster, such as VM status and events, which can be manipulated by external entities to influence agent behavior.
- Ingestion points: Reads cluster resource status and logs via
kubevirt_vm_get,kubevirt_vmis_list, andkubectl_get(SKILL.md). - Boundary markers: No explicit delimiters or instructions are provided to the agent to treat cluster data as untrusted.
- Capability inventory: The skill has extensive write capabilities, including
kubectl_applyand various VM lifecycle management tools (SKILL.md). - Sanitization: There is no evidence of data validation or sanitization before cluster information is interpolated into the agent's context.
- [CREDENTIALS_UNSAFE]: The skill contains a hardcoded password within a provided example manifest.
- Evidence: The example for creating a VM includes a cloud-init configuration with
password: fedora(SKILL.md). While appropriate for a demo image, hardcoding credentials in templates is a risk if adopted into production workflows.
Audit Metadata