Symbol Ontology

IMPORTANT: This software is dual-licensed under Mozilla Public License 2.0 (for non-commercial use) and a Commercial License (for business/for-profit use). Commercial use requires a paid license. Learn more

A symbolic reasoning engine built in Rust that serves as an MCP-compliant server for symbolic ontology. This server enables applications to query and reason about symbols, their meanings, and their relationships.

Project Structure

The Symbol Ontology project is structured as a Rust workspace with multiple crates:

symbol-ontology/
├── ontology-core/           # Core domain models and database logic
├── ontology-api-server/     # API server implementation
├── symbol-mcp-client/       # Public MCP client binary
├── tests/                   # Integration tests
└── docs/                    # Documentation

Features

MCP Client Features

• Direct database connectivity without requiring the API server
• Full MCP protocol compliance with SSE transport for Claude integration
• Available MCP methods:
  • get_symbols - List all symbols with optional filtering
  • search_symbols - Search symbols by text query
  • filter_by_category - Get symbols filtered by category
  • get_categories - Get all available symbol categories
  • get_symbol_sets - List all symbol sets
  • search_symbol_sets - Search symbol sets by name or description

API Server Features

• REST API endpoints for symbol management
• JSON response format
• Database-backed persistence

Core Library Features

• Domain models for symbols and symbol sets
• Database repository interfaces and implementations
• Shared utilities for both client and server

Getting Started

Installation

Install from GitHub

# Install the symbol-mcp-client package directly from GitHub
cargo install --git https://github.com/yogimathius/symbolic-ontology-mcp symbol-mcp-client

# Verify installation
symbol-mcp --help

Build from Source

# Clone the repository
git clone https://github.com/yogimathius/symbolic-ontology-mcp.git
cd symbolic-ontology-mcp

# Build the project
cargo build --release

# Run the MCP client
cargo run -p symbol-mcp-client

Database Setup

Before running the MCP client, you need a PostgreSQL database:

# Set up database connection string
export DATABASE_URL=postgres://username:password@localhost:5432/symbol_ontology

# Create database (if it doesn't exist)
createdb symbol_ontology

Required Environment Variables

The Symbol Ontology MCP client requires the following environment variables:

Docker Deployment

You can easily run the Symbol Ontology MCP client using Docker:

# Build the Docker image
docker build -t symbol-ontology-mcp .

# Run the container
docker run -p 3002:3002 \
  -e DATABASE_URL=postgres://user:pass@host:5432/symbol_ontology \
  -e RUST_LOG=info \
  -e MCP_PORT=3002 \
  symbol-ontology-mcp

Using docker-compose

# Create a .env file with your DATABASE_URL
echo "DATABASE_URL=postgres://user:pass@host:5432/symbol_ontology" > .env

# Start the service
docker-compose up -d

Deploying to Fly.io

This project includes configuration for deploying to Fly.io:

1. Install the Fly CLI:

   curl -L https://fly.io/install.sh | sh
   

2. Login to Fly.io:

   fly auth login
   

3. Set your database secret:

   fly secrets set DATABASE_URL=postgres://postgres:password@your-project.supabase.co:5432/postgres
   

4. Deploy the application:

   fly deploy
   

Running the Client

Start the MCP client to serve symbolic queries:

# Run with default settings
symbol-mcp

# Run with custom database URL
symbol-mcp --database-url postgres://username:password@localhost:5432/symbol_ontology

# Run with debug logging
symbol-mcp --verbose

Integration with Cursor

To use Symbol Ontology with Cursor AI:

1. Start the MCP server locally:

   symbol-mcp
   

   The server will start at http://localhost:3000 by default.

2. In Cursor, open the settings:

   • Navigate to the "MCP" section

3. Add a new MCP provider:

   • Name: Symbol Ontology
   • URL: http://localhost:3000/sse
   • Authentication: None (or as required)

4. Save your settings

5. The symbol-ontology tools will now be available in Cursor AI.

Running the API Server (Optional)

The API server provides REST endpoints if needed:

# Run from the project directory
cargo run -p ontology-api-server

API Reference

MCP Methods

REST API Endpoints

Upcoming Features

The following features are planned for future releases:

• Database seeding tool for easy initialization
• Improved SSE integration for the MCP client
• Authentication for the API server
• Rate limiting for the API server
• License validation
• Centralized logging setup
• Improved configuration system
• More comprehensive documentation

License

This project is dual-licensed under:

• Mozilla Public License 2.0 for non-commercial use
• Commercial License for business/for-profit use

Please see the LICENSE files for details.

For commercial licensing inquiries, please contact [info@yogimathius.dev].