Taxonomy catalogue
Open-reference taxonomies and Autotwin methodologies. The locked typed shapes every twin reasons against.
A taxonomy is a typed-object schema every twin can reason against — names, fields, relationships, invariants. It is the locked shape that survives across upstream systems, downstream consumers, and model rotations.
Senses and taxonomies are separate but tightly coupled. A taxonomy can be derived at sense adoption (the upstream schema becomes a typed contract the moment a sense is wired in) or authored as an open-reference standard or methodology that any twin can adopt without an upstream. See /docs/architecture/senses for the sense side of the pair — the wiring, the upstream catalogue, and the derivation act.
STANDARDSOpen-reference taxonomies
Authored shapes lifted from public standards bodies. Adopted by any twin without an upstream sense — the taxonomy is the entire contract.
- FDA medical devices. 7,066 device types from openFDA. Used by CSP for medical-device sourcing.
- Odoo (ERP shape). Sales, purchase, inventory, accounting. Lifted as a shape — no Odoo code ships.
- EDIFACT. UN/EDIFACT B2B message types (ORDERS, INVOIC, DESADV). The wire format of mature supply chains.
METHODSAutotwin methodologies
Best-practice frameworks codified as typed objects. The agent populates, scores, and cites the same shape every twin reasons against.
- HBSAANY DD. Harvard due-diligence framework. Typed assessment the agent can populate, score, cite.
- Balanced Scorecard. Kaplan & Norton four-perspective strategy measurement. Codified as typed objects.
- MIT Sloan DT. MIT Sloan digital-transformation maturity model. Twins assess themselves against it.
Where to go next
- /docs/architecture/senses — the sense side of the pair, including the full 145-entry Windsor connector catalogue (each connector derives its own taxonomy at adoption).
- /contributors — who contributes, how attribution works.
- /contributors/join — the CLA flow and how to publish a new taxonomy.