Glossary

The words the docs, the studio and the CLI use, one or two sentences each, with the page that explains them. Contributors: the names these take in the code are in CONTRIBUTING.md.

Artifact

The validated JSON a diagram source compiles to, in .diagrams/.artifacts/. publish builds its pages from artifacts, and the studio shows a .diagram.ts diagram through its artifact, so a TypeScript edit appears only after a compile (diagc studio keeps one running). A .diagram.json drawn in the studio is read from its source. See Why an artifact at all.

Badge

A count on an element; click it to open the element's note. A threat badge hangs off a box's top-left corner, or sits three-quarters of the way along an arrow, and shows how many threats are open, or a green ✓ once all are handled. A comment badge sits at the bottom-right corner, or a quarter of the way along an arrow, and shows the comment count, or ↗ when there are only links. See Comment on a diagram and How threats are shown.

Chip

A small pill-shaped control or label: the ▸/▾ fold chip and the ⤢ drill chip on a group header, the layer chips, top-bar chips such as Save positions, a plan's role chips, a legend's colour chip. See Studio.

Drill

Entering a node as its own diagram with the ⤢ chip: it becomes the whole canvas, and each node outside it that a relation reaches shows as an external stub. See Enter a node as its own diagram.

Element

A node or a relation, as something that can carry threats and comments; a node can also carry links. See Comment on a diagram.

External stub

In a drill view, a ghosted, dashed stand-in for a node outside the drilled subtree that a relation reaches, so the arrow still lands somewhere. Outside nodes that nothing reaches are left out. See Enter a node as its own diagram.

Fold

A group shown shut, as one box that hides its children. Semantic zoom decides what is folded; a pin overrides it. See Semantic zoom.

Include

A node that pulls another diagram in under itself at compile time, so one umbrella view can show several diagrams. See Compose diagrams.

Lane

In a git branching diagram, the row a branch's commits sit on. In an activity diagram, a swimlane for one actor. See Draw a git branching diagram and Draw an activity diagram.

Layer

A named, cross-cutting set of nodes and relations that a viewer switches on or off. Layers are off by default. Toggling a layer of relations never moves a box; a layer that tags nodes adds or removes those boxes, so the arrangement changes with it. See Layers.

Layout overlay

What a diagram's layout sidecar holds: saved positions, sizes, layout settings and open groups, per plane. It is kept out of the model so that meaning and coordinates change in different diffs. See LayoutOverlay.

Legend

The on-canvas key: rows derived from what the diagram uses, plus the items you add. See Add a legend.

Notation

A visual language for the whole model or one plane — c4, git-graph, threat-model, plan and others — that changes how nodes and relations are drawn and laid out. See DiagramModel.

Note

The speech bubble that opens from a badge, listing an element's threats, comments and links. In edit mode you can drag it where it reads best, and its place is saved with the layout. See How threats are shown.

Pin

A viewer's override that holds a group open or shut, whatever semantic zoom would decide. Pins are viewer state, but the groups open when you click Save positions, or while you edit, are saved in the layout overlay and open again on the next load. Elsewhere the docs also use "pin" in its everyday sense: a pinned version, style or edge end. See Semantic zoom.

Plane

An alternative containment hierarchy over the same nodes, so one model can be read as, say, an architecture view and an infrastructure view. See Planes.

Semantic zoom

The diagram rests folded, every group one box, apart from groups the layout overlay saved open. Double-click a group to unfold it; its siblings stay folded, so you read one path of detail against an overview. See Semantic zoom.

Sidecar

A file kept beside a diagram's source: <name>.layout.json holds its layout overlay, and <name>.drawings.json its freehand strokes. See Files and directories.

Starter

The small example for each diagram type that diagc init copies into a repository and diagc guide quotes. See init.

Text node

In a causal-loop diagram, a node with no type, drawn as bare text rather than a box: a variable. In other notations a typeless node is a plain box. See Draw a causal-loop diagram.

Threat register

The threats an element carries, each with a severity, a status and a mitigation. The published page also lists every register under the canvas. See Draw a threat model.

Zone

In a plan, a date bar that can nest and carries role chips. In a deployment diagram, an availability zone or another failure domain. See Draw a plan and The vocabulary.