Agent credentials: the token that says "yes, actually let this agent buy."
Before an AI agent can act on anyone's behalf, something has to vouch for it. An agent credential is the short-lived, scoped proof a business checks before it trusts a request enough to act on it.
5 min read
The problem a credential solves
A credential is small and boring by design, and it does almost all of the actual trust-checking in this hub, which is exactly why it's worth understanding properly.
Nothing stops a system from sending a request that claims "I'm an agent, acting for a customer, please proceed." Without something to check, a business has no way to tell that claim apart from a script written to abuse the same endpoint. A credential is the thing a business checks instead of taking the claim at face value.
What a real credential contains
A useful credential is issued by a party the business already trusts, scoped to a specific task rather than standing access, time-limited, and cryptographically signed so it can't be forged or reused past its window. It answers a narrow question well: is this specific request, right now, coming from an agent this business has reason to trust.
How this differs from a customer's own login
A customer's account login authorises a standing relationship: this person, indefinitely, until they log out. An agent credential authorises one bounded task: this agent, this request, this window. Confusing the two is how a business ends up either trusting far more than it should, or building account-style friction into something that was only ever meant to be checked once.
Where a business checks this
The credential check happens before anything else in the sequence runs, identity first, everything else after. Selfe verifies it automatically at that first step for every request it forwards, so a business's own system only ever sees requests that have already cleared it.
Who issues an agent credential?
A trusted issuer the business and the agent platform both recognise, not the agent itself.
What happens if a credential is missing or expired?
The request is rejected before it reaches anything further down the chain, the same way an expired card would be.
The plumbing behind every agent purchase.
Every purchase an AI agent completes, whatever it's buying, passes through the same four jobs: proving who the agent is, proving what it's allowed to spend, connecting to the business's own systems, and completing a transaction the business can settle. The protocols in this hub each do one of those jobs. None of them does all four.
A2A, explained: how one agent proves what it's asking another to do.
A shopping AI agent rarely finishes the whole job itself. It hands specific tasks to other, more specialised agents along the way. Agent2Agent protocol is how one agent describes a task to another clearly enough that the second agent can trust it and act on it.
AP2 and agent payments: how an agent proves it's allowed to spend on someone's behalf.
A credential proves an AI agent is who it says it is. AP2 (Agent Payments Protocol) proves something narrower and more specific: that a real person authorised this exact spend, up to this exact limit, and a business can verify that before it charges anything.