GraphDB MCP Server

Name: GraphDB MCP Server
Author: ianormy

@ianormy

リポジトリ

GraphDB MCP Server について

MCP server tools for GraphDB

#graphdb #mcp-server #database-tools

基本情報

カテゴリ

その他

ライセンス

MIT license

ランタイム

python

トランスポート

stdio

公開者

ianormy

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "mcp-graphdb-ianormy": {
      "command": "uv",
      "args": [
        "venv"
      ]
    }
  }
}

ツール

Execute SPARQL read queries to read data from a repository

Get a list of the RDF triples that define the ontology for the GraphDB database

概要

What is GraphDB MCP Server?

An MCP server that enables AI assistants to interact with GraphDB by executing SPARQL queries and retrieving database schemas. It is built for developers and knowledge workers who want to explore graph data through natural language interfaces.

How to use GraphDB MCP Server?

Install uv, clone the repository, set up a virtual environment, and run uvx with --db-url, --schema-file, --username, and --password arguments. Add the server to your Claude Desktop configuration using the mcp-graphdb key.

Key features of GraphDB MCP Server

Execute SPARQL read queries against a GraphDB repository.
Retrieve the ontology schema as RDF triples.
Uses the Model Context Protocol for integration.
Supports local GraphDB instances via HTTP.

Use cases of GraphDB MCP Server

Querying graph databases through conversational AI assistants.
Exploring ontology structure without writing low-level SPARQL.
Integrating GraphDB data into AI-powered documentation or dashboards.

FAQ from GraphDB MCP Server

What dependencies does GraphDB MCP Server require?

Python and uv (Universal Virtualenv) are required for development and execution.

What transport protocol does GraphDB MCP Server use?

The README configures the server as a stdio-based MCP server (via command in Claude Desktop config). No explicit transport details are given.

How is authentication handled?

Authentication is passed via --username and --password command-line arguments. The example uses "ignore" for both.

Where does the data live?

Data resides in a GraphDB repository specified by the --db-url parameter. The example uses a Star Wars ontology at http://localhost:7200/repositories/starwars.