Ontologies, seen your way.
Arspec is a viewer and editor for RDF ontologies, vocabularies, taxonomies and knowledge graphs. Instead of one fixed screen for every file, you choose how a file is presented: which things to list, what to show about each of them, and how they are grouped. Edit the data right in that view, and keep the view with the file so others see it the same way.
Opens Turtle, N3, TriG, N-Quads, RDF/XML, OWL/XML and JSON-LD — and plain JSON, XML, CSV, YAML and TOML — from your disk or from a URL. Nothing to install on a server. Windows, macOS, Linux, iOS, Android and the web.
@prefix : <http://ex/team#> .
:ada a :Person ;
:name "Ada" ;
:knows :grace .
:grace a :Person ;
:name "Grace" ;
:role "Engineer" .
Facts, written down as statements.
RDF is the W3C standard for recording facts as simple statements: Ada — knows — Grace, Grace — has role — Engineer. Each statement links a thing, a relation and a value, and together the statements form a graph. Ontologies, vocabularies, taxonomies, metadata schemes and knowledge graphs are commonly stored this way, in text files such as Turtle. The files are precise, but reading them as text means rebuilding the picture in your head. Arspec draws the picture for you, and lets you decide what the picture should be.
Views you choose, not views you are given.
Most tools decide in advance how an ontology looks: a class tree, a page per subject, or one long list of statements. Arspec comes with a few sensible views built in, and lets you make your own. A perspective is a saved view: which items to list, what to show about each one, and what is nested under what. Make one for the task in front of you, keep it with the file, and anyone who opens that file gets your view along with the data.
List every person in the file, once each.
Under each person, show everything the file says about them: the relation and its value.
Let me edit what I see, and add new people from the list.
Press Configure — the sliders at the top right — to see how the view on screen is put together, or create a new one. Each list has a When for its conditions, a Display for what it shows, and Subordinates for the lists nested under it.
Change something and the view updates as you go. Turn on the help hints and every part of the form explains itself.
p:perspec a ars:Perspective ;
ars:title "People" ;
ars:rootDataSource p:people .
p:people a ars:DataSource ;
ars:subjectVisible true ;
ars:distinctBy rdf:subject ;
ars:filter [
a ars:FilterCompare ;
ars:left [ a ars:TermValue ;
ars:role rdf:subject ] ;
ars:op ars:typeIs ;
ars:value [ a ars:IriValue ;
ars:nsUri "http://ex/team#" ;
ars:localName "Person" ] ] ;
ars:child p:props .
p:props a ars:DataSource ;
ars:order 0 ;
ars:predicateVisible true ;
ars:objectVisible true ;
ars:filter [
a ars:FilterCompare ;
ars:left [ a ars:TermValue ;
ars:role rdf:subject ] ;
ars:op ars:eq ;
ars:value [ a ars:TermValue ;
ars:role rdf:subject ;
ars:bindingDistance 1 ] ] .You never have to type this. But because a perspective is text in the same language as your data, it can be kept in version control, sent to a colleague, and drafted by an AI assistant.
Made for editing, not only for looking.
Edit right in the view
Rename a term, change a value, add an item, delete a statement, all where it is shown. Namespaces and value types are picked from a menu, so you do not have to remember the spelling. Every change can be undone with Ctrl+Z.
Look into any item
Click a name to open it in place and see what the file says about it, or open it on a page of its own and follow references from there. A view that describes one item is called a lens, and you can write your own.
Views that travel with the data
A topic is a small .arspec file that records which files belong together and which views apply to them. Hand it to a colleague, or put it in version control next to the files it describes.
An assistant that edits with you
Ask questions about a file, have a view drafted, or ask for a change. Nothing is changed until you approve it in Arspec's own window, and every approved change can be undone. Use the built-in assistant, or connect the AI chat you already have.
RDF in seven formats, plain data in five more.
Arspec recognises a file by what is inside it rather than by its name, so a URL with no extension opens fine. Whatever the format, you work with the same views. TriG, N-Quads and JSON-LD may hold named graphs, and each graph opens as a row of its own under its file. Plain JSON, XML, YAML, TOML and tables are read as items and their properties — every object, element, mapping or row an item, every field, attribute, key or column a property — and written back in the shape they came, comments included. OWL/XML can be opened but not saved back: its structure is flattened on the way in and cannot be rebuilt. Save such a file as Turtle or RDF/XML instead.
- Turtle
- N3
- RDF/XML
- JSON-LD
- TriG
- N-Quads
- JSON
- YAML
- TOML
- XML
- CSV / TSV
- OWL/XML
↔ read and written
Paste a URL, pick files with the system's file dialog, or drop them on the window. Several files can be opened together as one topic, and files a source imports from beside it are loaded along with it.
No database and no server. The file on your disk is what you are editing, and on the desktop your edits are saved back automatically.
Your own chat. On a desktop, Claude Desktop, Claude Code, Cursor and other chat apps that support MCP can connect to Arspec and work on the file you have open, using the account you already have.
The built-in assistant. A chat pane inside Arspec, talking to Anthropic, OpenAI, xAI, Mistral, a local Ollama, or any compatible service on your own key. This is how it works on phones and in the browser.
Bring the assistant you already use.
Either way the assistant gets the same abilities: an overview of the file, search, the description of any term, the views that apply, and one way to propose a change. A proposed change opens in a confirmation dialog in Arspec, shown as a difference against the file, before anything is touched. You approve it where you can see the data, not in the chat.
Same app, every platform.
The browser version is the complete editor with nothing to install. A web page cannot write to your disk, so there you save by downloading a copy. The desktop apps save changes back to the file automatically and can connect to your own AI chat.
Downloads, with what each platform needs →
