mcp -oxigraph
MCP server for connecting to Oxigraph
claude mcp add --transport stdio identimoji-mcp-server-oxigraph uv run mcp-server-oxigraph
How to use
This MCP server provides a Claude-friendly wrapper around Oxigraph via PyOxigraph, exposing RDF storage and SPARQL capabilities through the Model Context Protocol. It enables your Claude workspace to manage one or more Oxigraph stores, run SPARQL queries and updates, and serialize/import RDF data in a variety of formats. The server supports creating and opening multiple file-based stores, querying with SPARQL, and handling common RDF operations such as adding quads, retrieving results, and exporting stored graphs. You can interact with the server using the Oxigraph-specific tool commands exposed in Claude, enabling quick experimentation with RDF data models and semantic queries in a local, persistent store.
To use it, install the MCP server, configure Claude to launch the server via uv with the package mcp-server-oxigraph, and (optionally) set environment variables like OXIGRAPH_DEFAULT_STORE to point to a preferred store. The README provides example Claude configurations and demonstrates how to create, query, and manage stores from within Claude-friendly calls such as oxigraph_create_named_node, oxigraph_query, and oxigraph_parse, among others. The server also exposes rich RDF serialization options and a full SPARQL implementation to work with graphs locally.
How to install
Prerequisites:
- Python 3.11+ (or a compatible Python runtime)
- uv (the MCP runner) installed and accessible in your PATH
Step 1: Install the Oxigraph MCP Server package via uv
uv pip install mcp-server-oxigraph
Step 2: Verify installation
uv list
You should see mcp-server-oxigraph available for launch.
Step 3: Run the MCP server (example)
uv run mcp-server-oxigraph
Step 4: Configure Claude to load the MCP server (example minimal configuration)
"oxigraph": {
"command": "uv",
"args": ["run", "mcp-server-oxigraph"],
"env": {}
}
Optional: Configure a default store by setting environment variables (see Additional Notes) before starting the server.
Additional notes
Notes and tips:
- The server initializes both a system default store and, if provided, a user default store. Operations without an explicit store_path use the appropriate default store.
- To set a custom store location, define OXIGRAPH_DEFAULT_STORE in the environment, for example: OXIGRAPH_DEFAULT_STORE=~/.claude/my_knowledge_graph.oxigraph
- Supported RDF formats include Turtle, N-Triples, N-Quads, TriG, RDF/XML, and N3. Use oxigraph_get_supported_formats() to fetch capabilities at runtime.
- The server exposes a comprehensive API for store management (create/open/close/backup/restore/optimize/list) and RDF data operations (add/remove/query/serialize). You can mix default and custom stores depending on your workflow.
- If you encounter permission or path resolution issues, ensure the store paths exist and are writable by the user running the MCP server. For Windows users, adjust path styles accordingly.
- When debugging, check that the uv invocation is correctly pointing to mcp-server-oxigraph and that Python dependencies for PyOxigraph are installed in your environment.
Related MCP Servers
mcp-vegalite
MCP server from isaacwasserman/mcp-vegalite-server
github-chat
A Model Context Protocol (MCP) for analyzing and querying GitHub repositories using the GitHub Chat API.
nautex
MCP server for guiding Coding Agents via end-to-end requirements to implementation plan pipeline
pagerduty
PagerDuty's official local MCP (Model Context Protocol) server which provides tools to interact with your PagerDuty account directly from your MCP-enabled client.
futu-stock
mcp server for futuniuniu stock
mcp -boilerplate
Boilerplate using one of the 'better' ways to build MCP Servers. Written using FastMCP