What the report is
A fictional but carefully-constructed scenario: these specific experts met at Lau Pa Sat hawker centre in Singapore after API Days, produced a joint review of INHERIT, and offered 10 prioritised recommendations. Rich wrote it as a stress-test of INHERIT’s design against the perspectives of the people most likely to critically evaluate it.
The expert panel:
- Graham Grieve — creator of HL7 FHIR
- Eve Maler — co-creator of SAML, UMA architect, DADE protocol architect (OpenID Foundation)
- Mike Kiser — OIDF Board Member, Director of Strategy & Standards at SailPoint
- Dean Saxe — DADE co-chair
- Heather Flanagan — identity standards expert, NIST mDL contributor
- Juan Cruz Viotti — JSON Schema architect, Sourcemeta toolchain, co-author of Unifying Business, Data, and Code
- Cassian Smith — author of JSON Mastery for API Development (2026)
- Jon Scheele — API Days APAC producer (host)
Headline verdict
The report endorses the current INHERIT JSON Schema + TemporalRule + OpenAPI 3.1 architecture. It recommends 10 enhancements, not paradigm change.
Key endorsements:
- Juan: “this is someone who read chapter 4 of my book and actually understood it” (on JSON Schema usage)
- Graham: “structurally what I’d expect from someone who studied FHIR carefully… temporal rules genuinely clever”
- Cassian: “OpenAPI 3.1.0 maps cleanly to JSON Schema 2020-12 dialect… field naming consistent camelCase”
- Eve: “The data layer is solid — what’s missing is the protocol layer on top. But he knows that. He’s left a clean seam for us.”
- Dean: “We don’t need to build a data layer. We need to build a bridge to this one.”
Explicit “do not change” items (Part 3 of the report):
- Do NOT add XML support (Graham, FHIR lesson)
- Do NOT move to HL7 governance (Mike)
- Do NOT weaken
unevaluatedProperties: false(Juan) - Do NOT adopt FHIR’s resource naming (Graham — INHERIT’s naming is better)
- Do NOT conflate data and access (Eve — keep two-layer)
The 10 recommendations (prioritised in Part 4)
| # | Recommendation | Priority | Effort | Lead |
|---|---|---|---|---|
| 1 | JSON Profile Document (parser-boundary behaviour) | HIGH | Low | Cassian |
| 2 | Schema Diff + Contract Testing in CI | HIGH | Medium | Juan |
| 3 | Data Product framing + Concept Map | HIGH | Medium | Juan |
| 4 | Two-Layer Architecture INHERIT + DADE formalisation | HIGH | Low | Eve + Dean |
| 5 | Non-TypeScript SDK generation (Python, Go, C#, Java) | MEDIUM | Medium | Cassian |
| 6 | Canonical JSON + signing readiness | MEDIUM | Low | Cassian + Eve |
| 7 | Legal-professional maturity signalling (self-verified / peer-reviewed / institutionally endorsed) | MEDIUM | Low | Graham + Heather |
| 8 | RFC 9457 Error Reporting | MEDIUM | Low | Cassian |
| 9 | Minimal Viable Estate Levels (0–3) | MEDIUM | Low | Juan |
| 10 | Conformance Kit as adoption flywheel (FHIR Connectathon model) | HIGH | Medium | Graham + Jon |
Already absorbed
Recs 7 and 9 are already in Scoping Framework v1.1+:
- Review-status vocabulary (self_verified / peer_reviewed / institutionally_endorsed) = Rec 7
- MVE levels L0–L3 = Rec 9
Rec 3 “Data Product framing” is already referenced in Scoping Framework v1.1+ (Viotti & Itelman).
Implications for foundation-architecture decision
The report is a material piece of evidence that:
- The current JSON Schema foundation is endorsed, not challenged
- DADE’s integration is planned around JSON-format INHERIT entities
- FHIR-lessons-learned argue AGAINST multi-serialisation burdens (explicit anti-XML advice extends to RDF/OWL/SHACL as a paradigm shift)
- OpenAPI 3.1 + RFC 9457 + canonical JSON direction is supported
- The TemporalRule mechanism is praised (explicitly “better than FHIR did it”)
- Juan’s JSON Schema praise implies Sourcemeta relationship has technical legitimacy
The v2.4 foundation-architecture thesis (RDF/OWL/SHACL/SPARQL/LegalRuleML/Akoma Ntoso/OWL-Time replacing JSON Schema) is at odds with this report’s endorsements. After honest self-correction (Friday 17 April 2026), the thesis was narrowed to selective additions that do NOT displace JSON Schema — option C from the three-option choice presented to Rich.
What the narrowed thesis preserves
- Akoma Ntoso URIs for statute references (additive)
- JSON-LD as concentric alternative serialisation (additive, not replacing)
- LegalRuleML as optional formal-rules format alongside Rego (additive)
- FHIR-inspired resource modelling (already visible, make explicit)
- OASIS TC hosting pathway to ISO/IEC JTC 1 (works for JSON-based specs; UBL precedent)
- Agent-ready layer: OpenAPI 3.1 + Arazzo + RFC 9457 + Jentic AI-Readiness scoring
What the narrowed thesis drops
- SHACL replacing JSON Schema
- RDF/OWL as primary ontology
- SPARQL as primary query
- OWL-Time replacing TemporalRule
References
- Source document:
standard/orgs/dade/singapore-bar-report.mdv1.0 (15 April 2026) - Companion:
standard/orgs/dade/complementary-positioning.md(TT’s DADE positioning) - INHERIT-FHIR thesis:
standard/orgs/dade/fhir-of-us-probate-thesis.md - Foundation-architecture handoff prompt evolution:
- v2.4 (broad paradigm-shift thesis) — superseded
- v3.0 (narrowed thesis + Lau Pa Sat as primary evidence)