The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes Python scripts to wrap and execute kubectl CLI commands, allowing the agent to interact directly with a Kubernetes cluster.
Evidence: scripts/debug_pod.py, scripts/get_resources.py, and scripts/cluster_health.py all use subprocess.run() to execute kubectl with various arguments. The implementation uses list-based arguments which mitigates shell injection risks, and the commands are central to the skill's stated purpose.
[INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially attacker-controlled data from within the Kubernetes cluster, creating a surface for indirect prompt injection.
Ingestion points: scripts/debug_pod.py (via kubectl logs) and scripts/cluster_health.py (via kubectl get pods and events) ingest raw logs, pod statuses, and event messages into the agent's context.
Boundary markers: Absent; the scripts print the raw output from Kubernetes directly to the console without delimiters or instructions to the agent to ignore embedded commands.
Capability inventory: The skill enables high-privilege operations including kubectl exec (shell access), kubectl delete (resource removal), and kubectl apply (configuration changes), as documented in SKILL.md.
Sanitization: Absent; the external content from logs and events is not filtered or escaped before being presented to the agent.

kubernetes-operations