The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses kubectl CLI commands and Python subprocess.run to orchestrate containerized applications, which involves high-privilege operations within a cluster environment.
[EXTERNAL_DOWNLOADS]: The skill instructions include examples of pulling container images from external registries, such as nginx, which is standard for Kubernetes operations.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection. * Ingestion points: The agent reads and applies external YAML manifests (deployment.yaml, pod.yaml) which could contain malicious instructions. * Boundary markers: There are no explicit delimiters or instructions provided to distinguish between trusted commands and untrusted data within the processed files. * Capability inventory: The skill utilizes administrative cluster commands and arbitrary subprocess execution, providing a high-impact target for successful injection. * Sanitization: The documentation does not specify validation or sanitization steps for external configuration content before it is executed by the cluster management tools.

k8s