Browse the schemas at https://schemas.sourcemeta.com/sourcemeta/std
Building professional OpenAPI specifications demands precision, but mastering JSON Schema is hard. This library eliminates that burden by providing production-ready schemas that encode industry standards and best practices. Reference them directly in your API specifications and focus on what matters: designing great APIs.
Led and maintained by a member of the JSON Schema Technical Steering Committee
Use this library to:
- Build production-grade OpenAPI specifications without writing schemas from scratch
- Skip the JSON Schema learning curve while maintaining expert-level quality
- Meet compliance and regulatory requirements with standards-based validation
- Establish a solid foundation for your organisation's API governance program
Warning
This project is in its early stages with much more to come. We need your feedback to shape its future. Please share your thoughts and suggestions on what you would like to see.
Note
All schemas target JSON Schema 2020-12, the dialect used by OpenAPI v3.1 and later. Earlier JSON Schema dialects will be supported in the future.
The library is designed to evolve without ever breaking its consumers. The key principles are:
- Paths mirror the authority. Each collection is versioned the way its
governing body publishes: edition years for ISO, IEEE, BIPM, and W3C
standards, RFC numbers for the IETF, dated releases for dated registries,
and in-place evolution for living registries. Self-published specifications
live under
other/. - Every schema is individually versioned. Each concept lives at
<concept>/v1.jsonand there are no unversioned aliases. - Versions never change their validation behavior. A published version may gain annotations, linked data mappings, or clearer wording, but it accepts and rejects exactly the same instances forever. Any change that is visible to validation mints a sibling version instead.
- Compliance bugs are fixed in place. When a schema demonstrably disagrees with its governing document, the fix lands on the same version file, so consumers get important corrections right away. The never-changing rule protects design decisions, not defects.
- Living registries drift in place by contract. Collections backed by an evolving registry, such as currency codes or the XBRL UTR, track their upstream within the same version, since their value space is defined by the registry rather than by a frozen document.
- Supersession is explicit. Replaced versions are marked with
deprecated: trueand are never deleted. - References pin exact versions, so the blast radius of any new version is always visible.
- Linked data carries identity, not metadata. Schemas mint dereferenceable IRIs from the governing authorities, and consumers obtain names, symbols, and classifications by following them. Every IRI is verified against its authority before it ships, and none are ever fabricated.
- Generation follows upstream data. Schemas are either generated from vendored authority data or written by hand, never generated from hand-maintained intermediates.
If you use this library in your research or project, please cite it using the DOI provided above. You can find citation information in various formats (BibTeX, APA, etc.) by clicking the "Cite as" button on the Zenodo record.
While the project is publicly available on GitHub, it operates under a source-available license rather than a traditional open-source model. You may freely use these schemas for non-commercial purposes, but commercial use requires a paid commercial license.
We are happy to discuss OEM and white-label distribution options for incorporating these schemas into commercial products.
We welcome contributions! By sending a pull request, you agree to our contributing guidelines. Meaningful contributions to this repository can be taken into consideration towards a discounted (or even free) commercial license.
Tip
Do you want to level up your JSON Schema skills? Check out learnjsonschema.com, our growing JSON Schema documentation website, our JSON Schema for OpenAPI video course, and our O'Reilly book Unifying Business, Data, and Code: Designing Data Products with JSON Schema.
If you have any questions or comments, don't hesitate in opening a ticket on GitHub Discussions or writing to us at hello@sourcemeta.com.
