Most AI trading agents are still half-blind. They connect to one account, see the balance and positions inside it, then try to apply your risk rules on top. The problem is obvious once you think about it: that single account is rarely your full exposure. If you’re long somewhere else, short somewhere else, or have size in prediction markets, the agent has no idea. So the position caps and stops it’s enforcing are protecting a fraction of the real risk. Alphio is now integrating with Verona to fix exactly that. Verona lets you verify a fact once (like your total exposure across accounts) and own the proof. Any agent you authorize can reuse that proof without ever seeing the underlying data. When Alphio runs on top of it, the agent finally knows the real picture it’s supposed to be managing. The caps actually constrain your whole position, and if you revoke access, the proof goes with it. It’s a small shift in how the data is shared, but it changes the difference between an agent you have to constantly watch and one you can actually trust with rules.