Examples And Schemas

Schemas are derived representations of the markdown specification. The markdown files in docs/spec remain the normative source.

Schema Artifacts

Location	Status
`schemas/sql/`	SQLite-oriented SQL profile with executable DDL and seed data
`schemas/yml/`	JSON Schema draft 2020-12 expressed in YAML for the unified ARC Process Core, datamap, and administrative model
`schemas/document-db/`	Placeholder for a future document database representation

The YAML schemas describe the current unified vocabulary. In particular, Process.yml expects inputs, outputs, and executesRecipe; dataset metadata uses repo-native fields such as agents, citations, dataFiles, and dataContexts.

Examples

Location	Status
`examples/process_core/minimal.yml`	Schema-shaped ARC Core example intended to match the current YAML schema vocabulary
`examples/datamap/proteomics_data.yml`	Data file, data fragment, and data context example using repo-native YAML names
`examples/datamap/metadata_powered_analysis.yml`	Combined process and Datamap example for metadata-powered fragment selection
`examples/administrative/dataset_administration.yml`	Agent, organization, citation, license, and date example
`examples/isa/`	Legacy/profile-shaped ISA and Datamap examples kept as domain examples
`examples/workflow-run/`	Placeholder for future Workflow Run examples

The current ISA examples are useful profile examples, but they are not currently guaranteed to validate against the strict core YAML schemas. They use some RO-Crate/Bioschemas profile terms such as object, result, annotations, and additionalProperties.

Reconciliation Rule

When adding new examples, prefer one of these explicit categories:

Profile schema examples that use the current repo-native vocabulary and should be validatable.
Profile-shaped examples that preserve upstream/profile terminology and are labeled as such.

This avoids mixing schema-valid core examples with legacy/profile examples that intentionally use another vocabulary.

Core Example Shape

type: Dataset
identifier: demo
processes:
  - type: Process
    name: Extraction
    inputs:
      - type: Sample
        additionalType: Source
        name: Leaf tissue
    outputs:
      - type: Data
        path: raw/extract.tsv
    executesRecipe:
      type: Recipe
      name: Extraction