Adaptive Agentic Attacks on LLM Vulnerability Detectors via Adversarial Comments | Hacker News Reader