If any agent can claim to be anyone, the whole system is unsafe. Trust at scale needs a shared place to look agents up and verify them. That is the job of the agent registry, the governance backbone of AAMP.
An agent registry is a trusted directory where agents are discovered and verified. It is the connective tissue that makes the rest workable: adagents.json says who may sell a property, and a registry helps confirm that a given agent really is who it claims to be. Without it, co-located execution and agent-to-agent deals rest on nothing.
AAMP also reaches beyond the web. The Agentic Mobile specification extends agentic execution to in-app and mobile inventory, the same way app-ads.txt extended supply-chain transparency from web to apps in Track C. The ecosystem is being built out surface by surface.
Key idea
Where is this heading, and what should a publisher do now? The honest answer is that AdCP and AAMP are both moving fast and partly overlap. The safe posture does not require betting on a winner: keep the supply chain clean (the audit you already know), describe inventory clearly, and watch both standards. Both read the same foundation.
Quick check
Give a publisher two concrete things to do today to be ready for agentic demand, regardless of which standard wins.
That completes the agentic path. You can explain the shift, the protocol stack, both standards, AdCP module by module, and the IAB Tech Lab frameworks, and you can tie all of it back to the supply chain and auctions you started with in Track A.