Actor

An Actor is whoever or whatever issues a Command. It can be a human – a customer, an administrator, a clerk – or it can be an automated piece of the system itself.

Actors are how the model captures agency. Every Command in ESDM has at least one issuer, and every issuer is named, scoped, and typed. That makes the question who can do this? a model-level question rather than a deployment-time one.

Human and Non-Human Actors

ESDM distinguishes Actor types. A human Actor represents a real person or a role a real person takes on. A non-human Actor – such as a scheduler, a watchdog, or a system service – represents an automated trigger. The distinction matters because the two come with different concerns: human Actors usually have permissions and interfaces, non-human Actors usually have schedules and runtimes.

A human Actor cannot also be backed by a system component – the schema enforces that distinction so the model doesn't quietly conflate a person with a process. When a system component issues Commands on behalf of a human, model both: the human as the Actor of intent, the system component as the External System that delivers the Command.

Actor and Permissions

Actors are also where authorization can live. A Command that should only be issued by an administrator carries that constraint by listing only the administrative Actor as a permitted issuer. The linter doesn't enforce business-level access policies – that's a runtime concern – but it does verify that every Command has a permitted issuer at all, which is a useful baseline.