Forma v1 · Final

Canonical UI representation for AI systems

Forma v1 is the final first version of a production-ready intermediate representation for user interfaces. It defines how systems like Tropos emit normalized, semantic, token-aware UI data that AI agents and third-party tools can consume consistently.

Forma is intended to make design input interoperable, machine-readable, versioned, and predictable across producers and consumers.

Read the specification Download JSON Schema

At a glance

Production-readyNormative requirements for producers and consumers.

InteroperableStable document model, conformance rules, extensions, and versioning.

DeterministicNormalization guidance for consistent Tropos output.

Machine-checkableJSON Schema and examples included.

Overview

What Forma solves

Raw design APIs are not a dependable standard for AI. They are vendor-specific, structurally inconsistent, and too low-level for reliable downstream reasoning. Forma defines the missing intermediate layer between source design data and consumers such as AI agents, code generators, validation systems, and design infrastructure.

Specification map

Core specificationDocument model, node model, required concepts, versioning base Schema referenceField-by-field object definitions, enums, required fields Normalization rulesHow Tropos transforms source data into Forma ConformanceProducer and consumer obligations, validation, compatibility Examples corpusReference documents and patterns Extensions and versioningEvolution policy, vendor fields, migration rules Errors and ambiguityMissing tokens, unknown roles, low confidence, unsupported features Working GroupGovernance and release process

Implementation baseline

What is included in Forma v1

1. Schema reference

Formal object definitions, required fields, enums, and JSON Schema.

2. Normalization rules

Deterministic guidance for transforming raw source structures into Forma.

3. Conformance

Rules for producers, consumers, validation, and compatibility.

4. Example corpus

Sample documents for common UI patterns and interoperability testing.

5. Versioning

Rules for major and minor evolution, migration, and extensions.

6. Error model

Standardized representation of ambiguity, unsupported features, and fallback behavior.

JSON Schema

Machine-checkable schema package for validators and toolchains.

Working Group

Release governance for specification stability and interoperability.