Skills
Navigate
skills/rohitg00/kubectl-mcp-server/k8s-networking/Gen Agent Trust Hub

k8s-networking

Pass

Audited by Gen Agent Trust Hub on Nov 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes live Kubernetes resource data (e.g., service definitions, endpoint lists) which can be manipulated by malicious actors within the cluster to influence the agent's logic.
  • Ingestion points: Data retrieved from the cluster using kubectl_get and kubectl_describe (SKILL.md).
  • Boundary markers: No delimiters or warnings are present to distinguish between trusted instructions and untrusted data from the cluster.
  • Capability inventory: The agent can apply manifests via kubectl_apply and execute shell commands inside containers using exec_in_pod (SKILL.md).
  • Sanitization: No input validation or sanitization mechanisms are implemented for data ingested from the cluster.
  • [COMMAND_EXECUTION]: The skill enables the agent to execute arbitrary shell commands within Kubernetes pods, which provides a significant capability for environment manipulation.
  • Evidence: The skill provides examples for using exec_in_pod to perform network troubleshooting commands like nslookup (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 9, 2026, 06:52 PM