AI-LieDar: Examine the Trade-Off Between Utility and Truthfulness in LLM Agents | Hacker News Reader