aws-fis-experiment-prepare

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The content contains an intentional persistence pattern for Kubernetes RBAC (fixed ServiceAccount/Role/RoleBinding names created by an inline Lambda and explicitly not deleted on stack removal), which creates a persistent privileged access vector that can be abused as a backdoor across deployments.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires calling aws___read_documentation / aws___search_documentation at runtime to fetch AWS docs (e.g. https://docs.aws.amazon.com/en_us/fis/latest/userguide/az-availability-scenario.html and https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/aws-resource-fis-experimenttemplate.html) whose content is used to extract the authoritative JSON experiment templates and thus directly controls the generated CFN templates — a required runtime dependency.