kube-bench

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the kube-bench CLI to audit Kubernetes configurations against CIS benchmarks. The commands provided, such as run --targets master and run --targets node, are standard and legitimate usage for this security utility.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to capture and analyze output from the kube-bench tool. This creates a surface where external data (benchmark results) is ingested into the agent's context.
  • Ingestion points: Tool output from kube-bench commands (SKILL.md).
  • Boundary markers: None explicitly defined in the instructions.
  • Capability inventory: Execution of shell commands (kube-bench) and file redirection (> contoso-kube-bench.json) (SKILL.md).
  • Sanitization: The procedure instructs the agent to triage results for material gaps rather than treating every difference as a finding, which incorporates a manual review logic to mitigate accidental obedience to data-embedded instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 02:38 AM
Security Audit — agent-trust-hub — kube-bench