The Agent Skills Directory

Administrative Command Execution: The skill utilizes gcloud and kubectl commands to perform security audits and apply configurations. This is a standard practice for managing GKE environments and involves localized execution within the user's terminal.
Vendor Image Reference: The skill references an official Google Cloud SDK container image hosted on gcr.io. As this is a well-known service provided by the vendor, it is considered a safe and standard dependency for testing workload identity.
Configuration Metadata Access: The audit script retrieves cluster configuration details using gcloud container clusters describe. This allows the tool to verify security settings like Shielded Nodes and Binary Authorization, providing a summary of the cluster's security posture.
Indirect Prompt Injection Surface: The audit_cluster.sh script accepts cluster-specific parameters (name, region, project ID) which are then used in shell commands.
Ingestion points: Positional arguments in scripts/audit_cluster.sh.
Boundary markers: None present for command-line arguments.
Capability inventory: Shell command execution via gcloud and jq.
Sanitization: The script uses double-quoting for variables (e.g., "${CLUSTER_NAME}") as a standard practice for handling shell arguments.

gke-workload-security