How it works

Agents plan experiments, control lab hardware, and read results in a closed loop. Each cycle feeds back into the next hypothesis without human input.