Consensus
Human + AI decision workspace powered by WebMCP, where agents research, score, flag conflicts, and propose rankings while humans retain final control.
Core trick
Agents research and score options with a ranking function, flag conflicts, and propose a ranking humans must confirm.