While Sierra and Meta's Personal Agent Protocol has been drawing attention this month, a separate but related effort surfaced from Decagon on October 6: the Personal Agent Consent & Trust Protocol, or PACT, co-developed with Instinct and released as open source. PACT tackles a narrower problem than the broader Sierra/Meta effort. Built on the existing Agent2Agent protocol and OAuth 2.0, it lets businesses verify whom a personal agent represents and what that customer has authorized it to do. The timing isn't coincidental — Decagon is also joining the Personal Agent Protocol working group that Sierra and Meta announced, so PACT looks like an attempt to shape the consent layer of that emerging standard rather than compete with it outright. The problem PACT targets is real and growing. Personal agents like Muse, Instinct, and dots are starting to book travel, manage purchases, and resolve issues on their users' behalf, and as they do, they increasingly run into AI agents acting on behalf of businesses rather than humans. Right now there's no standard way for the business-side agent to know whether the personal agent actually has permission to, say, cancel a subscription or authorize a refund — it just has to trust whatever the personal agent claims. What PACT actually adds is a way to make that permission explicit and checkable. Rather than a business agent inferring consent from context or taking a personal agent's word for it, PACT is designed to give these interactions explicit, verifiable customer permissions — a structured claim that can be inspected rather than assumed. This is a smaller, more specific piece than a full identity or discovery layer. It doesn't tell a business agent who the personal agent's vendor is, or how to find it, or how to route a task to it. It just answers one question: did this user actually authorize this specific action? That's a scope narrow enough to ship quickly, which may be the point — publishing it ahead of the Personal Agent Protocol's own v0.1 spec gives Decagon and Instinct a concrete reference implementation to point to while the bigger standard is still being drafted. Whether PACT gets folded into the Personal Agent Protocol itself or persists as a parallel consent layer is an open question. Standards efforts in this space have a pattern of starting as separate proposals and either merging or fragmenting depending on who ships working code first. For now, PACT is notable mainly for being specific: it's not trying to solve agent identity or agent discovery, just the one gap — verifiable consent — that sits underneath both.