Graph Snapshots and AI Tutorials

Aphelion graph documents use versioned JSON with the aphelion-graph format. They are safe data files: they contain nodes, properties, positions, connections, groups, and annotations, but no executable code.

CLI

aphelion nodes list
aphelion nodes describe "Floor Tracker" --json
aphelion nodes export-schema > nodes.json
aphelion graph schema > graph-schema.json
aphelion graph validate floor-tracking.apgraph --json
aphelion graph render floor-tracking.apgraph --output floor.png --scale 2 --theme dark
aphelion graph inspect floor-tracking.apgraph
aphelion graph format floor-tracking.apgraph --write
aphelion graph export project.aph --output main.apgraph
aphelion graph import main.apgraph --output imported.aph
aphelion tutorial validate examples/floor-tracking.aptutorial --json
aphelion tutorial render examples/floor-tracking.aptutorial --output floor-tutorial
aphelion tutorial schema > tutorial-schema.json

Graph rendering is headless and uses model coordinates, so it does not depend on the current graph viewport or zoom. --no-watermark, --padding, --max-width, and --max-height control export safety and presentation.

AI Workflow

1. Query aphelion nodes export-schema.
2. Generate an .apgraph or .aptutorial file using actual node and port names.
3. Run validation with --json and repair stable error codes such as UNKNOWN_INPUT or INVALID_CONNECTION_TYPE.
4. Render the validated graph or tutorial into PNG and Markdown.

The sample examples/floor-tracking.aptutorial demonstrates a multi-step, operation-based tutorial that builds a floor tracker and corner-pin setup.