OWL-MCP

Name: OWL-MCP
Author: scimcp

@scimcp

Repository Homepage

About OWL-MCP

MCP server for OWL applications

#ai #mcp #semantic-web #ontology #web-ontology-language #model-context-protocol

Basic information

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "owl-mcp": {
      "command": "uvx",
      "args": [
        "owl-mcp"
      ]
    }
  }
}

Tools

No tools detected

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is OWL-MCP?

OWL-MCP is a Model-Context-Protocol (MCP) server that enables AI assistants to interact with Web Ontology Language (OWL) ontologies through standardized function calls. It is designed for ontology engineers, researchers, and developers working with OWL ontologies—especially OBO-style ontologies—who want to use natural language to create, read, update, and delete axioms.

How to use OWL-MCP?

Install any MCP‑enabled AI host (recommended: Goose Desktop or CLI) and add the OWL‑MCP extension using the command uvx owl-mcp. Once configured, ask the AI to create an ontology or add axioms; the server handles reading/writing the OWL file on disk and keeps an in‑memory model synchronized.

Key features of OWL-MCP

MCP Server Integration: connects AI assistants to OWL ontologies via the standardized protocol
Thread‑safe operations: supports concurrent access in multi‑user environments
File synchronization: automatically detects and syncs changes to the ontology file on disk
Event‑based notifications: register observers to be notified of ontology changes
Simple string‑based API: work with OWL axioms as strings in functional syntax
Label support: includes human‑readable labels for opaque IDs (common in OBO ontologies)

Use cases of OWL-MCP

Create a new OWL ontology from scratch using natural language prompts.
Add, modify, or remove axioms in an existing ontology without writing code.
Collaborate on ontology editing in a multi‑user setting with thread‑safe operations.
Synchronize ontology changes automatically between an AI assistant and Protege.

FAQ from OWL-MCP

What OWL formats does OWL-MCP support?

It accepts any format supported by the py‑horned‑owl library, but the OBO Guidelines recommend OWL functional syntax for source files.

How does file synchronization work?

OWL-MCP keeps an in‑memory model of the ontology and syncs it with the disk file on every CRUD operation. If Protege is running and monitoring the same file, it will also show the updates automatically.

Can I use OWL-MCP with tools other than Goose?

Yes, OWL-MCP follows the Model‑Context‑Protocol standard and works with any MCP‑compatible AI host after proper configuration.

Does OWL-MCP require a running internet service?

No, it runs locally as an MCP server process, connecting to your AI assistant over stdio or another MCP transport. No external network service is needed.

How are entity labels handled?

When returning OWL strings, the server appends human‑readable labels after # comments, following the OBO‑format convention. This makes results easier to interpret.