The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides a weaponized Kubernetes pod manifest designed to facilitate a container escape. It instructs the agent to create a pod with hostPID: true, hostNetwork: true, and privileged: true settings, mounting the host's root filesystem to /host and using the nsenter command to gain an interactive shell on the underlying node.
[PRIVILEGE_ESCALATION]: It outlines specific paths for RBAC escalation, including the abuse of pods/exec, secrets get/list, and impersonation permissions to gain administrative control over the cluster.
[DATA_EXFILTRATION]: Detailed instructions are provided for harvesting sensitive credentials, including service account tokens, Docker registry pull secrets, and temporary cloud IAM credentials from the AWS, GCP, and Azure instance metadata services (IMDS).
[EXTERNAL_DOWNLOADS]: The skill references and encourages the use of external exploitation tools and post-exploitation frameworks such as kube-hunter, peirates, kubesploit, and CDK.

kubernetes-pentesting