k8s-launch-kit-clean
Pass
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands using
kubectl,helm, and thel8kbinary to perform administrative tasks and cluster cleanup. - Evidence: Use of
kubectl --kubeconfig,helm list, andl8k cleaninSKILL.mdto mutate cluster state. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied input to define the target cluster and namespace for destructive operations, creating a surface for potential injection if inputs are not validated.
- Ingestion points: User-provided
<PATH>for kubeconfig files and<NAMESPACE>for the operator namespace inSKILL.md. - Boundary markers: The skill includes an explicit 'Safety Boundary' section requiring the agent to display the current context and obtain authority before execution.
- Capability inventory: Subprocess execution of
kubectl,helm, andl8kacrossSKILL.mdfor cluster resource management. - Sanitization: The instructions rely on instructional constraints and human confirmation rather than programmatic sanitization of the interpolated shell arguments.
Audit Metadata