enterprise-vpn-attack

SUSPICIOUS: the skill is internally consistent as an offensive VPN attack playbook, but that coherence does not make it safe. It materially increases an AI agent's ability to conduct external exploitation, credential testing, and attack chaining against real systems; no obvious third-party exfiltration is present, but the offensive scope is inherently high risk.