create-test-scenarios

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.65). The skill’s required workflow calls MCP tools like mcp__content-agent__get_rendered_rule(product, rule_id) to ingest rendered OVAL/remediation code (which is derived from rule content, i.e., not user-authored), and that rendered prose/code is then used to generate test scenarios—creating an indirect prompt-injection exposure path from outsider-authored rule text into the agent’s LLM context.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill explicitly instructs creating and editing system files and configs (e.g., /etc/ssh/sshd_config, /etc/sysctl.conf, grub, systemctl actions), manipulating services, partitions, and installing packages — all actions that modify the machine state and typically require root privileges.