performing-dark-web-monitoring-for-threats

SUSPICIOUS. The skill is mostly purpose-aligned and uses largely official endpoints, with no clear credential theft or malicious exfiltration. However, it grants an AI agent high-risk security-monitoring capability and processes untrusted dark web content, creating substantial misuse and indirect prompt-injection risk; the GitHub raw ransomware feed is a minor trust concern.