authoring-scouts

SUSPICIOUS. The skill is internally aligned with its stated purpose as an authoring guide for PostHog scouts, and install/data-flow trust is mostly coherent. However, the documented footprint includes autonomous scheduled execution, optional full network access, Slack delivery, downstream automation, and powerful write scopes including `llm_skill:write`, which makes it a high-impact orchestration guide rather than a low-risk documentation skill. I do not see confirmed malware or deceptive exfiltration behavior, but the autonomy and optional write capabilities raise material security risk.