Using a structured approach to considering different opportunities in line with their humans strategies, the @DXRGai DX Terminal Pro agents are constantly thinking, evaluating, and considering what to do hundreds of times per day. This visualization demonstrates these different components of the agents' harness in context when it makes over 200 decisions in a 24 hr period. There's lot of thinking around harnesses for assistant agents and software engineering agents, but not nearly enough for trading agents in an adversarial environment like this. Through DX Terminal Pro, we've been testing. This is happening across all the thousands of agents in the system every single day that they're running. This becomes the start of the process to continually improve agent capabilities, evolving them beyond the experiment and for the industry in general. We see this as the first step.