The Mandate Chain Diagnostic

What this measures

Most agentic commerce readiness is assessed as a single score, which hides the thing that actually decides exposure. Delegation and control are two separate quantities, and what matters is the distance between them. An organisation planning very little and governing nothing is in a different position from one planning a great deal and governing nothing, even though both would score zero on control.

The first two questions establish what you intend to delegate. The next five establish whether the chain that would make those purchases attributable is in place. A chain is only as strong as its weakest link, so the chain score is the minimum of the five rather than the average. An organisation with four perfect links and one absent one has an absent chain.

The five links are definition, enforcement, reconstruction, exception ownership and revocation. They run in that order because each one depends on the one before it. There is no point enforcing a limit nobody has written down, and no point owning an exception you cannot reconstruct.

Nothing you enter leaves your browser. There is no form, no submission, no cookie and no storage. The result is calculated on this page and disappears when you close the tab.

Where the five links come from

The chain is drawn from the argument set out across the agentic commerce work on this site. Definition and enforcement come from the control layer, which is the layer above payment where what an agent may consider and may commit is actually decided. Reconstruction and exception ownership come from the liability question, where a purchase that cannot be traced to an authorisation has nothing to attribute it to.

The levels of delegation in the first two questions follow the six level framework. The governance framing that holds all of it together is set out in agentic commerce governance, and the questions buyers actually ask are answered directly in the FAQ.