Afternoon Primus Ai agents are learning to work, transact and make decisions independently. But theres one problem we shouldn’t ignore How do we know the information behind those decisions is real? That’s where @primus_labs gets interesting. Check here it s.kaito.ai/j2BXKsq Through zktls primus exploring a way to verify data from websites without exposing the entire underlying information. Think about an ai trading agent fetching a market price. Instead of simply trusting the api response the agent could verify evidence of where that data came from and whether it was altered. Now connect that idea with cluster Protocol Cluster is building infrastructure for ai services, data inference and agent payments. Primus brings a complementary layer verifiable evidence for the information those systems rely on. Identity tells us who an agent is. Payments let it transact. Verification helps establish whether its data can be trusted. That distinction matters when agents start making economic decisions without constant human supervision. Of course proving data provenance doesn't automatically prove that an ai answer is correct. The verification process and its security assumptions still matter. But making information independently verifiable is an important step toward a more accountable ai economy. For me primus is interesting because trust shouldn't depend entirely on who makes a claim. It should depend on what can actually be verified. As ai agents become economic actors could verification become just as essential as compute and payments?