The Agent Skills Directory

[Command Execution] (SAFE): The skill makes extensive use of kubectl commands to manage cluster resources. These are standard administrative tools consistent with the skill's purpose.
[Indirect Prompt Injection] (SAFE): The skill provides mechanisms to read container logs and resource descriptions which could be manipulated by a compromised workload. This is a known risk for any monitoring tool and is categorized as safe here as it is core functionality.
Ingestion points: kubectl logs, kubectl describe pod, kubectl get pod -o yaml in SKILL.md.
Boundary markers: Not present in the provided shell command templates.
Capability inventory: kubectl exec for container access, kubectl cp for file movement, and kubectl apply/delete for resource management in SKILL.md.
Sanitization: The skill does not implement custom sanitization for retrieved logs or metadata.

pod-management