Diagrams
In this section
| Page | What it covers | Code | Cowork | Desktop chat | VS Code |
|---|---|---|---|---|---|
| ea-diagram-advisor | Decide which diagram type and viewpoint to draw for what you want to understand, before anything is drawn. | Yes | No | No | Yes |
| ea-diagram-composition | Draw a diagram from model content using your notation's conventions, then check and correct it until it reads as deliberate. | Yes | No | No | Yes |
Where it runs: Yes = runs fully · With you = runs, and you do the steps in EA’s window · Partly = one part needs a capability the client lacks · No = not available · TBC = to be confirmed
Two skills turn model content into diagrams. They split the work along the line between a decision and a computation:
| Skill | Question it answers | Logic |
|---|---|---|
ea-diagram-advisor |
Which diagram type and viewpoint should this be? | AI + AI Power Tools checks (a rule-based module ranks the candidates) |
ea-diagram-composition |
Where does everything go, and does the result read well? | AI + AI Power Tools checks (deterministic geometry, rule-based checks, AI judgment) |
The advisor never draws; the composition skill never chooses the diagram type. The advisor's recommendation is what you hand to the composition skill. Behind both is an engine of small, tested modules that carry no modeling-language knowledge of their own: the diagram advisor engine (Diagram advisor engine), the layout grammars (Layout grammars), the composition modules (Composition modules), the notation bindings that supply each language's conventions (Notation bindings), and the verification modules (Verification modules).
Skills for hierarchies of diagrams with drill-down links (ea-navigation-diagrams) and for the operations that tidy and style a diagram that already exists (ea_diagram) sit alongside these two.