ontosphere

Overview

Ontosphere is a browser-based RDF/ontology knowledge graph editor that loads data from files, URLs, or SPARQL endpoints. It provides a visual canvas for creating and editing nodes and edges, runs OWL-RL reasoning, and applies multiple layout algorithms. The MCP server integration exposes tools for AI agents to interact with the knowledge graph, including loading data, querying the graph, updating nodes and links, and running reasoning operations. All computation happens client-side in the browser using Web Workers.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• loadRdf

  Load RDF data from files, URLs, or SPARQL endpoints

  Try:Load RDF data from this URL: https://example.org/data.ttlTry:Load data from the SPARQL endpoint at https://example.org/sparql
• loadOntology

  Load a specific ontology into the knowledge graph

  Try:Load the BFO ontologyTry:Load the DCAT ontology and automatically import dependencies
• queryGraph

  Query the loaded knowledge graph with SPARQL

  Try:Find all instances of the class 'Person'Try:Show me all relationships between organizations and people
• exportGraph

  Export the current graph in various RDF formats

  Try:Export the graph as TurtleTry:Save the current graph as RDF/XML
• addNode

  Add a new node to the knowledge graph

  Try:Add a new node representing 'Company X'Try:Create a node for the organization 'Research Institute'
• removeNode

  Remove a node from the knowledge graph

  Try:Remove the node labeled 'Obsolete Department'Try:Delete the node with ID 'node123'
• addLink

  Add a relationship between two nodes

  Try:Connect 'Alice' to 'Bob' with a 'knows' relationshipTry:Add a 'worksFor' link between 'John' and 'Tech Corp'
• runReasoning

  Run OWL-RL reasoning on the knowledge graph

  Try:Apply reasoning to infer new relationshipsTry:Run OWL-RL to find implicit class hierarchies
• runLayout

  Apply layout algorithms to organize the graph visualization

  Try:Layout the graph using the Dagre algorithmTry:Apply the ELK force-directed layout
• clusterNodes

  Automatically cluster nodes in the graph based on algorithms

  Try:Cluster nodes using label propagationTry:Group nodes with the K-Means algorithm
• getNodeDetails

  Retrieve detailed information about a specific node

  Try:Show details for the node 'Organization'Try:Get all properties of the node with ID 'node456'
• findPath

  Find a path between two nodes in the knowledge graph

  Try:Find a path between 'Research Institute' and 'Government Agency'Try:Show connections between 'Person A' and 'Company B'

Comparable tools

Installation

"mcpServers": {
  "ontosphere": {
    "command": "npx",
    "args": ["@thhanke/ontosphere-mcp"]
  }
}

FAQ

How do I connect an AI agent to Ontosphere?

Ontosphere exposes MCP tools via the browser's navigator.modelContext API. You can connect using Claude Code with Playwright for full automation or use the AI Relay Bridge for ChatGPT, Gemini, and Claude.ai.

What reasoning capabilities does Ontosphere support?

It runs OWL-RL inference in the browser using the N3.js BGP-only Reasoner. It supports rdfs:subClassOf transitivity, owl:equivalentClass, owl:someValuesFrom restrictions, owl:inverseOf, and other OWL constructs, but not full OWL-RL features requiring EYE reasoner.

Compare ontosphere with