trees, configurations, records, and schemas as file. the native format for machine-readable knowledge in the cybergraph
source format: JSON, TOML — any hierarchical key-value structure
rendering
json/toml source → parse → tree layout → collapsible glyph tree → GPU render
nodes expand and collapse. keys and values have distinct styling. depth encoded visually. the robot renders any struct file as an interactive tree regardless of nesting depth or key count
in the cybergraph
struct is how machines describe their own state — and how the graph describes complex objects with named parts
types of struct files: smart contract ABIs, network configurations, protocol parameters, API schemas, scientific metadata, genomic annotations, experimental conditions, machine learning model configs, governance proposals, identity documents
a struct file is often the metadata companion to another file: the pixels file of a satellite image may have a struct file linked that contains coordinates, timestamp, sensor calibration, and resolution
properties
- machine-readable natively — parseable by any conformant JSON or TOML parser without transformation
- schema-flexible — struct does not require a fixed schema. the cybergraph discovers schema by topology: files that share struct shapes cluster via motifs
- queryable by datalog — any key path in a struct file is accessible as a datalog term
- composable — struct files embed in component files as data sources for tables and forms
relation to other languages
struct is the configuration language of the cybergraph. text carries argument; struct carries specification. a component reads a struct and renders it as interactive form fields. datalog queries struct fields directly
see json for the primary source format. see table for 2D structured data. see component for interactive composition
discover all concepts