How would you design an incident response system that fails securely and generates high-fidelity alerts during an attack?
