Capabilities — the building blocks

Why. Capabilities are the major things your business does — its "blocks of work". They are found first because everything else in the model hangs off them.

How. Click Generate with LLM. A map of boxes appears, joined by "depends on" lines. Click any box to read or edit it in the side panel; use + Capability to add one. Enrich suggests blocks a typical business in your industry usually has that yours is missing — you keep only the ones that fit.

Wiring a dependency. The lines on the map are a projection — you don't draw them there. To make one capability depend on another, open it and, under Depends on in the side panel, pick the other capability from the list. This is also how you clear an "isolated (no relationships)" hint: open the flagged capability and pick what it depends on. The picker only offers capabilities that exist, so it can never leave a broken link.

The side-panel fields are pickers. Most fields relate the capability to something else in the model, so instead of free text they let you pick from what exists: Depends on picks a capability, Actors picks a modelled role, and Produces / Consumes pick an entity (Produces/Consumes also let you type a loose object that isn't a stored entity yet). When a field's values live on another screen, the picker ends with a "Manage in …" link — e.g. Actors → Roles, Produces → Entities — so you can jump there, author the missing value in its home screen, and come back without losing your place. Only Outcomes stays free text.

Example. For a solar installer you get boxes like Lead Management, Solution Design, Procurement, Installation, and Billing.

Capabilities is the one review-only layer: because everything downstream is derived from it, regenerating it would reset the layers below. You refine capabilities by editing them directly.