vulnerability-verification

SUSPICIOUS: the skill is internally aligned as a penetration-testing verification workflow, but it gives an AI agent high-risk offensive capabilities, includes stealth/WAF-evasion guidance, routes proof traffic to attacker-controlled endpoints, and trusts unseen local scripts plus a transitive sub-skill. Not confirmed malware, but dangerous and inappropriate outside a tightly controlled authorized security environment.