sast-dast-analysis

SUSPICIOUS. The skill is coherent for security analysis and uses mainstream official tooling, so install/data-flow trust looks mostly proportionate. However, it grants an AI agent explicit security-scanning capability against code and live staging targets, which is intrinsically high risk for misuse and prompt-injection exposure even without clear malicious intent.