wulaosiji

One-click security hardening system for AI agents and skills. Deploys input filters, command interceptors, model locks, persistent guards, and audit logging to protect against prompt injection, social engineering, dangerous command execution, and information leakage. Use when: "安全加固", "security har…