Definition
A graph runtime is an execution environment in which addressable entities and typed relationships participate directly in contextual computation. Nodes are not only records to retrieve. They can carry identity, state and capabilities; relationships can constrain or enable how events propagate.
Scope
The concept concerns the minimum structures required for a graph to execute meaningful operations: stable identity, typed relations, context, events, permissions and observable state transitions.
Electronic Artefacts distinguishes a graph runtime from a static knowledge graph. A knowledge graph may describe facts and connections. A graph runtime uses graph structure as part of the mechanism that decides what can execute, under which context and with which consequences.
Electronic Artefacts position
Graph execution is useful only when it remains legible and governable. Context should not be hidden inside application code when it determines authority, visibility or behavior. Identity and relation types therefore belong to the runtime model rather than being treated as decorative metadata.
Applications
VASTE is the primary program through which Electronic Artefacts develops this model. Vestiges applies the runtime thesis to cultural and craft knowledge, where people, techniques, materials, institutions, places and documents need stable identities and explicit historical relationships.
Runtime Theory supplies the research question beneath the implementation: what is the minimum necessary for a universe of entities to execute events coherently?
Limitations
A graph runtime introduces modeling and governance costs. Poorly defined relations create ambiguity, high-degree graphs can become difficult to reason about, and unrestricted context propagation can produce security and performance failures. Not every workflow requires a graph runtime.
References
This definition is grounded in the VASTE program, Runtime Theory research field, Vestiges project architecture and the preserved foundational lineage of Electronic Artefacts.