landing-page

SUSPICIOUS. The skill’s overall purpose is coherent, and most capabilities match landing-page research/generation. However, risk is elevated by broad ingestion of untrusted external content, optional reading of sensitive CV data, transitive skill installation, and reliance on an externally installed CLI delivered via curl|sh. Data flows are mostly proportionate to the task, but the trust chain and prompt-injection surface make this a medium-risk skill rather than benign.