MCP Proto-OKN Server

A Model Context Protocol (MCP) server providing seamless access to SPARQL endpoints with specialized support for the NSF-funded Proto-OKN Project (Prototype Open Knowledge Network). This server enables querying the scientific knowledge graphs hosted on the FRINK platform. In addition, third-party SPARQL endpoints can be queried.

Features

• 🔗 FRINK Integration: Automatic detection and documentation linking for FRINK-hosted knowledge graphs

• 🧬 Proto-OKN Ecosystem: Optimized support for biomedical and scientific knowledge graphs, including:

  • SPOKE - Scalable Precision Medicine Open Knowledge Engine

  • BioBricks ICE - Chemical safety and cheminformatics data

  • SAWGraph - Safe Agricultural Products and Water monitoring

  • Additional Proto-OKN knowledge graphs - Expanding ecosystem of scientific data

• ⚙️ Flexible Configuration: Support for both FRINK and custom SPARQL endpoints

• 📚 Automatic Documentation: Registry links and metadata for Proto-OKN knowledge graphs

• 🔗 Federated Query: Prompts can query multiple endpoints

Architecture

The MCP Server Proto-OKN acts as a bridge between AI assistants (like Claude) and SPARQL knowledge graphs, enabling natural language queries to be converted into structured SPARQL queries and executed against scientific databases.

Prerequisites

Before installing the MCP Server Proto-OKN, ensure you have:

• Operating System: macOS, Linux, or Windows

• Client Application: One of the following:

  • Claude Desktop with Pro or Max subscription

  • VS Code Insiders with GitHub Copilot subscription

Installation

Prerequisites

The MCP Proto-OKN server requires the uv package manager to be installed on your system. If you don't have it installed run:

Note: Once uv is installed, the uvx command in the configuration below will automatically download and run the latest version from PyPI when needed.

Claude Desktop Setup

Recommended for most users

1. Download and Install Claude Desktop

   Visit https://claude.ai/download and install Claude Desktop for your operating system.

   Requirements: Claude Pro or Max subscription is required for MCP server functionality.

2. Configure MCP Server

   Option A: Download Pre-configured File (Recommended)

   Download the pre-configured claude_desktop_config.json file with FRINK endpoints from the repository and copy it to the appropriate location:

   macOS:

   # Download the config file curl -o /tmp/claude_desktop_config.json https://raw.githubusercontent.com/sbl-sdsc/mcp-proto-okn/main/config/claude_desktop_config.json # Copy to Claude Desktop configuration directory cp /tmp/claude_desktop_config.json "$HOME/Library/Application Support/Claude/"

   Windows PowerShell:

   # Download the config file Invoke-WebRequest -Uri "https://raw.githubusercontent.com/sbl-sdsc/mcp-proto-okn/main/config/claude_desktop_config.json" -OutFile "$env:TEMP\claude_desktop_config.json" # Copy to Claude Desktop configuration directory Copy-Item "$env:TEMP\claude_desktop_config.json" "$env:APPDATA\Claude\"

   Option B: Manual Configuration

   Alternatively, you can manually edit the configuration file in Claude Desktop. Navigate to Claude->Settings->Developer->Edit Config to edit it.

   Below is an example of how to configure FRINK endpoints. Third-party SPARQL endpoints with a custom description can be added (see uniprot-sparql example below).

   { "mcpServers": { "spoke-sparql": { "command": "uvx", "args": [ "mcp-proto-okn", "--endpoint", "https://frink.apps.renci.org/spoke/sparql" ] }, "biobricks-sparql": { "command": "uvx", "args": [ "mcp-proto-okn", "--endpoint", "https://frink.apps.renci.org/biobricks-ice/sparql" ] }, "uniprot-sparql": { "command": "uvx", "args": [ "mcp-proto-okn", "--endpoint", "https://sparql.uniprot.org/sparql", "--description", "Resource for protein sequence and function information. For details: https://purl.uniprot.org/html/index-en.htm" ] } } }

   Important: If you have existing MCP server configurations, do not use Option A as it will overwrite your existing configuration. Instead, use Option B and manually merge the Proto-OKN endpoints with your existing mcpServers configuration.

3. Restart Claude Desktop

   After saving the configuration file, quit Claude Desktop completely and restart it. The application needs to restart to load the new configuration and start the MCP servers.

4. Verify Installation

   1. Launch Claude Desktop

   2. Navigate to Claude->Settings->Connectors

   3. Verify that the configured Proto-OKN endpoints appear in the connector list

   4. You can configure each service to always ask for permission or to run it unsupervised (recommended)

VS Code Setup

For advanced users and developers

1. Install VS Code Insiders

   Download and install VS Code Insiders from https://code.visualstudio.com/insiders/

   Note: VS Code Insiders is required as it includes the latest MCP (Model Context Protocol) features.

2. Install GitHub Copilot Extension

   • Open VS Code Insiders

   • Sign in with your GitHub account

   • Install the GitHub Copilot extension

   Requirements: GitHub Copilot subscription is required for MCP integration.

3. Configure Workspace

   Create or edit .vscode/mcp.json in your workspace:

   { "servers": { "mcp-spoke-sparql": { "command": "uvx", "args": [ "mcp-proto-okn", "--endpoint", "https://frink.apps.renci.org/spoke/sparql" ] } } }

   Note: For the VS Code Insiders configuration, change "mcpServers" to "servers".

4. Use the MCP Server

   1. Open a new chat window in VS Code

   2. Select Agent mode

   3. Choose Claude Sonnet 4 or later model for optimal performance

   4. The MCP servers will automatically connect and provide knowledge graph access

Quick Start

Once configured, you can immediately start querying knowledge graphs through natural language prompts in Claude Desktop or VS Code chat interface.

Example Queries

1. Knowledge Graph Overview

   Provide a concise overview of the SPOKE knowledge graph, including its main purpose, data sources, and key features.

2. Multi-Entity Analysis

   Antibiotic contamination can contribute to antimicrobial resistance. Find locations with antibiotic contamination.

3. Cross-Knowledge Graph Comparison

   What type of data is available for perfluorooctanoic acid in SPOKE, BioBricks, and SAWGraph?

The AI assistant will automatically convert your natural language queries into appropriate SPARQL queries, execute them against the configured endpoints, and return structured, interpretable results.

Development

Installing from Source

If you want to run a development version:

Building and Publishing (maintainers only)

API Reference

Available Tools

query

Executes SPARQL queries against the configured endpoint.

Parameters:

• query_string (string, required): A valid SPARQL query

Returns:

• JSON object containing query results

get_description

Retrieves endpoint metadata and documentation.

Parameters:

• None

Returns:

• String containing endpoint description, PI information, funding details, and related documentation links

Command Line Interface

Required Parameters:

• --endpoint : SPARQL endpoint URL (e.g., https://frink.apps.renci.org/spoke/sparql)

Optional Parameters:

• --description : Custom description for the SPARQL endpoint (auto-generated for FRINK endpoints)

Example Usage:

Troubleshooting

Common Issues

MCP server not appearing in Claude Desktop:

• Ensure you've completely quit and restarted Claude Desktop (not just closed the window)

• Check that your JSON configuration is valid (use a JSON validator)

• Verify that uvx is installed and accessible in your PATH

Connection errors:

• Check your internet connection

• Verify the SPARQL endpoint URL is correct and accessible

• Some endpoints may have rate limits or temporary downtime

Performance issues:

• Complex SPARQL queries may take time to execute

• Consider breaking down complex queries into smaller parts

• Check the endpoint's documentation for query best practices

License

This project is licensed under the BSD 3-Clause License. See the LICENSE file for details.

Citation

If you use MCP Server Proto-OKN in your research, please cite the following works:

Related Publications

• Nelson, C.A., Rose, P.W., Soman, K., Sanders, L.M., Gebre, S.G., Costes, S.V., Baranzini, S.E. (2025). "Nasa Genelab-Knowledge Graph Fabric Enables Deep Biomedical Analysis of Multi-Omics Datasets." NASA Technical Reports, 20250000723. Link

• Sanders, L., Costes, S., Soman, K., Rose, P., Nelson, C., Sawyer, A., Gebre, S., Baranzini, S. (2024). "Biomedical Knowledge Graph Capability for Space Biology Knowledge Gain." 45th COSPAR Scientific Assembly, July 13-21, 2024. Link

Acknowledgments

Funding

This work is supported by:

• National Science Foundation Award #2333819: "Proto-OKN Theme 1: Connecting Biomedical information on Earth and in Space via the SPOKE knowledge graph"

Related Projects

• Proto-OKN Project - Prototype Open Knowledge Network initiative

• FRINK Platform - Knowledge graph hosting infrastructure

• Knowledge Graph Registry - Catalog of available knowledge graphs

• Model Context Protocol - AI assistant integration standard

• Original MCP Server SPARQL - Base implementation reference

For questions, issues, or contributions, please visit our