@eventcatalog/visualiser@4.0.0-beta.0

Major Changes

• 728ca67: Add support for a new systems collection. Systems are a versioned resource that can be defined in any folder (including inside domains), rendered as documentation pages, and listed in the sidebar with their own icon and color. Systems can group services, flows, entities, data stores (containers), and diagrams, have their own architecture overview page (/architecture/systems/[id]) listing their entities, services, external integrations, data stores, and flows, be visualised as an architecture map, be browsed and filtered on a dedicated discover page (/discover/systems), and be targeted by architecture decision records (appliesTo: [{ type: 'system' }]). Domains can reference one or more systems, which are listed in the domain sidebar, rendered as expandable sections on the domain's architecture page (listing the system's services, like subdomains, with external services shown under "External Integrations"), and merged into the domain's architecture map (each referenced system's services are grouped within the domain graph). Teams and users can own systems, which are surfaced on their profile pages.

  Systems can also declare relationships to other systems (each with an optional version and label), which power a new Context Diagram visualiser (/visualiser/systems/[id]/[version]/context, linked from the system sidebar under Architecture). Starting from a system, the diagram walks its relationships outward to build the reachable neighbourhood of systems, rendering each as a node (showing its service, entity, and data store counts) connected by labelled edges. Clicking a system node opens that system's architecture map.

  Systems support a scope of internal (default) or external. External systems represent third-party/SaaS systems you integrate with (e.g. "Resend", "Stripe") and are shaded and badged as "External System" in the Context Diagram.

  Systems can also declare actors — people or roles that interact with the system (e.g. a Customer or Support Agent). Each actor has an id (used to de-duplicate the same actor across systems), an optional name/label, and a direction (inbound = actor → system, e.g. "logs into"; outbound = system → actor, e.g. "sends notifications to"). Actors are rendered as nodes on the Context Diagram connected to their system by a labelled, directional edge.