Create HC AWS

Pass

Audited by Gen Agent Trust Hub on Jun 6, 2026

Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill references and utilizes local sensitive configuration files, including AWS credentials ($AWS_CREDENTIALS), OpenShift pull secrets ($PULL_SECRET), and management cluster kubeconfigs ($MGMT_KUBECONFIG). This access is a requirement for the skill's primary function of provisioning and managing cloud resources and container clusters.
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute a local binary (./bin/hypershift) and the kubectl CLI to interact with the AWS and Kubernetes APIs. These operations are standard for the intended administrative tasks.
  • [PROMPT_INJECTION]: An indirect prompt injection surface exists as the skill interpolates user-provided placeholders, such as <CLUSTER_NAME>, into shell commands.
  • Ingestion points: User-provided cluster names and custom image tags in SKILL.md.
  • Boundary markers: Absent; placeholders are directly interpolated into bash commands.
  • Capability inventory: Subprocess execution of ./bin/hypershift and kubectl in SKILL.md.
  • Sanitization: No explicit sanitization or validation of the cluster name is described in the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 6, 2026, 04:59 PM
Security Audit — agent-trust-hub — Create HC AWS