Semantic Matter
Developer Portal

Build with THiNK

Integrate semantic documentation and catalog operations directly into your stack. Manage the complete lifecycle of source catalogs and target graphs, tracing how your data maps to computed results.

Middle Layer API Spec

API Surface

Core endpoints available for integration today.

Catalog & Documentation

Manage source (Groups, Packages, Resources) and target (Projects, Targets) models.

API/documentation/groups/{group_name}
API/documentation/packages/{package_name}
API/documentation/resources/{resource_name}/datasets/{dataset_id}
API/documentation/projects/{project_name}
API/documentation/targets/{target_name}

Composer Reads

Read computed target outputs, individual properties, and graph traversal paths.

GET/composer/{target_name}
GET/composer/{target_name}/properties/{target_property}
GET/composer/{target_name}/paths

Semantic Concepts & Files

Exchange object, entity, or ontology files and retrieve URI-based semantic concepts.

GET/resources/concepts
POST/files/upload
GET/files/download/{file_id}
Required Header: X-DataSpaces-FileType (object, entity, ontology)

Workflow Management

Backend routes to register workflows, replace definitions, and manage trigger lifecycles.

API/workflow/workflows/{workflow_name}
API/workflow/{workflow_name}/triggers
API/workflow/{workflow_name}/routes
API/routing/workflows/{workflow_name}

Performance & Discovery

Optimize your integration using built-in query parameters and traversal controls to fetch exactly what you need.

  • Use shallow listings with ?only_names=true on packages and targets to drastically reduce payload sizes during discovery.
  • Control concept neighborhood expansion by passing depth (0-5) to /resources/concepts.
  • Trace how documented source data maps to graph results via /composer/{target_name}/paths.
  • Retrieve singular target properties directly rather than pulling massive composer results.

Ecosystem Integration

THiNK acts as a semantic control layer next to your existing systems, not a replacement.

Catalogs & Data Lakes

Keep existing storage; register semantic metadata and lineage in THiNK.

Historians & Time-series

Model source resources and map outputs to targets for queryable context.

LIMS / ELN

Attach structured dataset series to resources, maintaining provenance.

Digital Twins

Use projects/targets to organize entities and read graph paths.

Enterprise Workflows

Trigger metadata refresh or mapping updates via backend workflow routes.

Integration Walkthroughs

1

Catalog Bootstrap & Model Retrieval

POST /documentation/groups
POST /documentation/packages
POST /documentation/resources
POST /documentation/projects
POST /documentation/targets

Pro tip: Use ?group={group}&only_names=true for fast discovery, then retrieve full objects with /documentation/packages/{package_name} when full metadata is needed.

2

File-Backed Semantic Enrichment

  1. Upload via /files/upload with header X-DataSpaces-FileType: ontology.
  2. Persist the returned file identifier in your resource metadata workflow.
  3. Download later via /files/download/{file_id}.
  4. Query /resources/concepts?concept_uris=...&depth=1 for direct concept context.
  5. Increase depth (up to 5) to fetch wider neighborhoods for complex validation.
3

Read Target Outputs & Workflow Triggers

  1. Fetch target results via /composer/{target_name}.
  2. Read a single field lightly via /composer/{target_name}/properties/{target_property}.
  3. Inspect provenance traversal with /composer/{target_name}/paths.
  4. Register workflows via /workflow.
  5. Add/replace triggers via /workflow/{workflow_name}/triggers.
  6. Use trigger events to re-run your read path and refresh downstream consumers.