Adopting SemType

SemType is designed to be adopted a layer at a time. You can consume the open standard directly, drop in ready-made editing components, lean on the wider open-source stack built around it, or adopt the full HASH platform end-to-end.

As an open standard

The full specification, an explanation of the different kinds of types, and the design rationale all live at semtype.org:

• Specification — the canonical definition of the type system and data model.
• Type Registry — publicly-available types made discoverable by SemType-supporting applications.

SemType is dually licensed under the MIT and Apache 2.0 licenses (at your option), so you're free to implement it in your own tools and ecosystems.

As part of HASH

The most direct way to use SemType is as part of HASH, where the type system powers every entity, link, and workflow. Users with a hash.ai account can create and browse types directly within the app.

As embeddable components

Two React components let you bring SemType-typed data and authoring into your own application without reimplementing the spec:

• SemType entity editor — view and edit semantically typed entities.
• SemType type editor — author entity types compatible with SemType.

As part of the hgres stack

SemType is the foundation of a small stack of open-source projects, each extending the same discipline — semantic types that travel with the data — into a different part of the stack:

• hGraph — a multi-temporal property graph that implements the SemType data model directly.
• HashQL — a typed query language for the SemType ecosystem.
• HaRPC — a typed, versioned transport that lets SemType definitions travel with the data across service boundaries.