Summary
The IDD spec format currently evolves through documentation, examples, and validator behavior, but it does not yet have a single normative schema that defines which artifact forms are legal and how they compose.
This creates ambiguity when introducing richer representations such as:
- annotated journey references
- lifecycle/state-machine shorthand
- aggregate-root model documents
- catalog/registry documents
- shared value-object bundles
Problem
Today, format evaluation is operational rather than fully formal:
docs/idd/front-matter-spec.md defines part of the shape
- validators enforce parts of the format
- examples demonstrate accepted forms
- certification verifies traceability and evidence
That is useful, but it means the canonical language is partly implicit in tool behavior. As the spec surface expands, that increases the risk of drift between documentation, examples, and validators.
Desired outcome
Define a formal language-theoretic schema for IDD specifications so that:
- each artifact kind has a normative grammar or meta-schema
- allowed document families are explicit rather than inferred
- validator behavior is derived from the documented schema
- examples act as acceptance fixtures rather than de facto standards
- new spec forms can be evaluated against a published extension model
Scope
This issue should establish the schema approach, not necessarily complete every validator rewrite in one pass.
In scope:
- decide the formalism to use for each artifact class
- document canonical forms for narrative, model, contract, and traceability metadata
- define how extensions are introduced without breaking compatibility
- identify which current formats are normative, legacy-compatible, or experimental
- define conformance levels for docs, examples, validators, and certification
Out of scope:
- redesigning the entire IDD methodology
- changing downstream implementation skills unless needed for schema conformance
Acceptance criteria
Why this matters
Without a formal schema, tool behavior becomes the effective standard. That slows evolution, makes validator bugs harder to classify, and makes it unclear whether a new representation is invalid, unsupported, or simply undocumented.
A formal schema would let IDD evolve intentionally instead of extemporaneously.
Summary
The IDD spec format currently evolves through documentation, examples, and validator behavior, but it does not yet have a single normative schema that defines which artifact forms are legal and how they compose.
This creates ambiguity when introducing richer representations such as:
Problem
Today, format evaluation is operational rather than fully formal:
docs/idd/front-matter-spec.mddefines part of the shapeThat is useful, but it means the canonical language is partly implicit in tool behavior. As the spec surface expands, that increases the risk of drift between documentation, examples, and validators.
Desired outcome
Define a formal language-theoretic schema for IDD specifications so that:
Scope
This issue should establish the schema approach, not necessarily complete every validator rewrite in one pass.
In scope:
Out of scope:
Acceptance criteria
Why this matters
Without a formal schema, tool behavior becomes the effective standard. That slows evolution, makes validator bugs harder to classify, and makes it unclear whether a new representation is invalid, unsupported, or simply undocumented.
A formal schema would let IDD evolve intentionally instead of extemporaneously.