webtoon-harness

SUSPICIOUS: the skill's purpose is coherent, but its footprint is elevated by transitive trust, multi-repo installation, shell-based scaffolding, and untrusted web-content ingestion with execution/write capabilities. No confirmed credential theft or overtly malicious behavior is shown, but the installation and delegation model make it medium-high risk.

No direct malware behavior is evident in the script logic itself (it only clones, copies, and optionally runs npx to install a skill). The main security concern is supply-chain trust: it fetches remote repository contents based on environment-controlled URL/ref and copies them into the consuming project’s `.claude/` directory without pinning or integrity verification; it also optionally performs an unpinned global npx install from an external repo with reduced output visibility. This warrants scrutiny of upstream sources and installation-time environment integrity.