Ai Security Audit Skills

Agent skills that audit what an attacker exploits first: Electron app surface, LLM prompt/RAG/tool exposure, invisible Unicode. Read-only; every finding cites file and line.