skills/mukul975/anthropic-cybersecurity-skills/conducting-mobile-application-penetration-test/Socket
conducting-mobile-application-penetration-test
Warn
Audited by Socket on Mar 15, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS: the skill is internally coherent and uses mostly legitimate tooling, but it equips an AI agent with offensive mobile exploitation capabilities, including SSL pinning bypass, root/jailbreak bypass, keychain/keystore dumping, and API tampering. Risk is high due to the nature of the capability, not because of deceptive install sources or hidden exfiltration.
Confidence: 92%Severity: 88%
Audit Metadata