What should a node card help readers understand? #688

I've been looking at the card that opens when you click a node (the Semantic Passport). For me, the source references are currently the only part that adds much value: they let me inspect where the diagram's claims came from. The other fields—type, tags, node ID, relationship counts, and upstream/downstream reachability—often feel less useful in the card, even though some of those interactions may be useful elsewhere.

I haven't settled on what the card should become. I'd like to discuss what people actually want to learn after clicking a node before choosing more fields or a new layout.

For example, a node called “Describe the system” currently shows “Natural language / Mermaid,” its type and ID, a source link, and its connections. After opening it, I still want to know something the diagram didn't already tell me. Adding “This node receives a system description” would just repeat the label.

Some possibilities are a concrete explanation of the node's role, important conditions or exceptions, an input/output example, or a link to a more detailed view. These are possibilities to test, not a proposed checklist for every node. A compact source card may be enough when there is nothing useful to add.

A few questions:

• When you click a node, what question are you trying to answer? Could you share a specific node and the information you wanted but couldn't find?
• Which existing parts of the card do you actually use, and which could be less prominent or omitted?
• What additional content would be worth opening the card for? Should it be visible immediately or available on demand?

PR #398 explores author-provided explanations. It is a related candidate, but the broader content and interaction direction is still open. Concrete examples would help us judge it and other approaches.