The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The implementation examples use standard DevOps tools (kubectl, helm) in a predictable and secure manner. Shell commands are used for deployment logic without incorporating unvalidated external inputs.
[CREDENTIALS_UNSAFE] (SAFE): The skill correctly uses placeholders for sensitive information via GitHub Actions secrets syntax (e.g., ${{ secrets.KUBE_CONFIG }}), avoiding hardcoded credentials.
[EXTERNAL_DOWNLOADS] (SAFE): External dependencies are restricted to trusted sources such as GitHub-managed actions (actions/checkout) and Azure's official kubectl setup action (azure/setup-kubectl).
[DATA_EXFILTRATION] (SAFE): No unauthorized network operations or exfiltration patterns were detected. Network activity is limited to standard deployment operations (helm repo update, helm upgrade) against defined endpoints.
[PRIVILEGE_ESCALATION] (SAFE): The use of chmod 600 on the kubeconfig file is a security best practice to restrict file permissions, and no attempts to use sudo or bypass system restrictions were found.

deployment-automation