Show HN: AutoThink – Boosts local LLM performance with adaptive reasoning May 28, 2025 by kamal Comments