Sources
Connect up to ten sources. Click a tile to connect it — a connected source shows a clear ✓ badge and its sync action.
PDFs, spreadsheets, Word docs, CSVs, markdown — ingested into a named collection, with date, format, and provenance. It becomes a first-class collection you can scope to and add to a project.
Load an RDF graph into your workspace — Turtle, RDF/XML, JSON-LD, or N-Triples (every format the Export button produces re-imports here, so you can restore or move a graph).
Taxonomies
Bring in a SKOS taxonomy from the SKOS editor — upload the file you export there (Turtle, RDF/XML, or JSON-LD), or paste it. Its concepts become the tags you can apply to documents.
Everything in your workspace as one RDF file — all taxonomies (crosswalk mappings included), documents, tags, metadata, and provenance.
None yet.
Every property of every concept — labels, alt labels, definitions, notes, notation, and relationships. The full governance view.
Import a taxonomy to browse its concepts.
Harvest a glossary for a topic, or match your concepts to external authorities. Every result carries the entity's real URI, added as skos:exactMatch — nothing invented.
RDF Wiki
Turn a collection of sources into a governed RDF wiki, one guided step at a time — you review and approve at every stage. Draft a schema, tag with a taxonomy, then publish to the graph.
A collection is one import of assets (a folder sync or an upload batch). A project is what you curate from them — one or more whole collections, plus any hand-picked assets — and it is the only thing a schema is applied to. Build and curate projects in Repository; everything below works on the selected project.
Claude reads the collection and proposes properties from the standard ontologies (DC · DCAT · FOAF · PROV-O · RDF · RDFS). Approve as-is, or hand-edit what to include.
None yet — generate or build one above.
Your saved SKOS taxonomies. Import more in the Taxonomies tab — the tool keeps as many as you like.
Nothing is written yet — approve concepts, drop ones that don't fit, and add free-text keywords per source.
Auto-tag to review proposals here.
Writes the approved tags (with provenance) to the graph database and finalizes the collection.
Repository
Everything you've built, in one place — your collections (folders), taxonomies, schemas, and saved projects. Open a collection to view and tag it, or save it out as RDF.
No collections yet — connect a folder in Sources.
None yet.
None yet.
None yet.
Catalog
Pick a concept scheme in the rail and confirm it — its concepts become the tags you apply to documents.
Sync a source, then refresh.
Metadata
Describe any asset in the standard ontologies. Pick an ontology, choose a property, add a value — the graph on the left highlights the asset you're describing.
The W3C / DCMI standards this asset can be described with. Select one to load its properties.
Schemas you approved in the RDF Wiki tab (AI-generated or hand-built). Select one to load its properties — required ones are marked *.
None yet — approve one in the RDF Wiki tab.
Pick an asset.
Query by meaning
Which documents cover a concept? This is the payoff of tagging.
No query run yet.
Graph
Your documents, concepts, agents, and provenance as one linked graph — drag nodes, scroll to zoom, click to inspect. You build it by tagging documents and adding metadata; it renders here as real DCAT / FOAF / PROV-O / SKOS.
Every concept in your taxonomies, with all its properties. Stays open alongside the graph for governance and meaning.
Import a taxonomy to browse its concepts.
Assistant
Build your metadata layer by talking. Ask about your content, or say things like "describe the deployment runbook with a creator and a license" or "model the partnership agreement in DCAT" — the assistant queries your graph and proposes metadata using the standard ontologies (DC · DCAT · FOAF · PROV-O · RDF · RDFS) for you to approve. (Conversational mode needs Claude connected; offline answers simple catalog questions.)