The Agent Skills Directory

[PROMPT_INJECTION]: No prompt injection or behavior override patterns were detected. The skill focuses on structured discovery and management phases.
[DATA_EXFILTRATION]: No unauthorized data access or external network operations were found. The skill uses standard kubectl commands to interact with the local Kubernetes cluster.
[REMOTE_CODE_EXECUTION]: No remote script execution or unverifiable dependency installations were detected. The skill relies on locally installed CLI tools (kubectl, opa, jq).
[COMMAND_EXECUTION]: Command execution is limited to standard diagnostic and management tasks (kubectl get, kubectl logs, opa test, opa eval). All inputs are handled through positional shell variables in a predictable manner.
[CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or private configuration files were accessed or exposed.
[SAFE]: The skill follows security best practices for Kubernetes management, including a 'Discovery-First' approach and the use of 'dryrun' for initial policy testing.

managing-opa