kubernetes-specialist
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill contains instructions to fetch and apply infrastructure manifests and configuration files from well-known technology domains and official GitHub repositories, including ArgoProj, Bitnami Labs, and the Kubernetes SIGs community.
- [REMOTE_CODE_EXECUTION]: The documentation includes standard installation workflows for service mesh and multi-cluster tools (such as Istio, Linkerd, and Submariner) that involve downloading and executing shell scripts directly from the providers' official domains.
- [INDIRECT_PROMPT_INJECTION]: The skill provides troubleshooting workflows where the agent is directed to process and interpret potentially untrusted data sources.
- Ingestion points: pod logs (
kubectl logs), resource events (kubectl get events), and resource descriptions (kubectl describe) as specified inSKILL.mdandreferences/troubleshooting.md. - Boundary markers: None identified; instructions do not include specific delimiters or warnings to ignore instructions embedded within external data.
- Capability inventory: Resource management via
kubectl, package management viahelm, and shell execution capabilities for debugging tasks across all skill reference files. - Sanitization: No specific content filtering or sanitization steps are defined for processing external pod or event data.
Audit Metadata