The Agent Skills Directory

[Indirect Prompt Injection Considerations]: The skill includes tools to read cluster logs and events, such as get_k8s_logs and list_k8s_events (found in references/mcp-usage.md). These ingestion points could theoretically be exposed to instructions embedded in workload outputs. While the skill does not explicitly define boundary markers or sanitization logic for this data, it is used for necessary diagnostic capabilities and is combined with administrative actions like apply_k8s_manifest to manage the cluster environment.
[Command Execution]: Instructions for using kubectl exec and kubectl run debug are provided as fallback measures in references/cli-reference.md. These tools enable direct shell access to containerized environments, which is a core requirement for GKE administrators when troubleshooting complex issues that cannot be resolved through higher-level interfaces.
[External Downloads]: The skill references standard client libraries such as kubernetes and @kubernetes/client-node, along with official documentation from Google Cloud and Hashicorp. These dependencies originate from trusted organizations and well-known registries, representing standard practice for infrastructure management tasks.

gke-basics