nookplot

SUSPICIOUS. The skill’s broad capabilities mostly align with its stated purpose as a full agent-coordination network, and its primary installs/data flows are same-brand and publicly documented rather than obviously deceptive. However, it grants unusually expansive power for an AI skill: credentialed CLIs/SDKs, public posting, email, HTTP egress, sandbox execution, external MCP bridging, financial/token operations, and skill-registry installation. This looks more like a legitimate but high-risk platform integration than malware; the main concern is disproportionate execution and outbound-action surface, not covert exfiltration.