ok so this whole thing started because i kept asking chatgpt stuff and realising it just agreed with whatever i already wanted to do. like id ask should i do X or Y and it would basically back my lean every time. felt like a yes-man, not a second opinion. so i built the thing i actually wanted — war table. you ask a question and 5 models (claude, gpt-5, gemini, grok, qwen) actually debate it across a few rounds, take opposite sides and poke holes in each others reasoning. then a hidden chairma...