testing-mobile-applications

SUSPICIOUS. The skill is internally aligned with mobile pentesting, but it gives an AI agent offensive security capabilities: runtime bypasses, device shell access, credential/keychain extraction, and traffic interception. There is no clear exfiltration to attacker infrastructure or deceptive install chain, so this is high-risk offensive enablement rather than confirmed malware.