LLM Output Manipulation for Malicious Code Generation via Adversarial Prompt Chaining | AI Breaking Wire