agent-autopilot

SUSPICIOUS: the skill's purpose matches its capabilities, and there is no clear credential theft or external exfiltration. Risk comes from autonomous action/reporting and the transitive trust dependency on another community skill with unpinned local script execution.