pattern-of-life-from-socials

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This skill is high-risk: it explicitly documents techniques to locate, link, and profile individuals (geolocation from images, EXIF/metadata extraction, network linkage and temporal analysis) and thus can be used for stalking, doxxing, or targeted privacy invasion.

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow described in SKILL.md ingests outsider-authored free text by reading social media posts, bios, comments/mentions, and associated metadata (e.g., “profile metadata, content analysis” and “For each post… Text verbatim”) from public platforms like Reddit/Twitter/X, Instagram/Facebook, and others, which are authored by external users.