Software Knowledge Graph

A tool for creating and managing comprehensive knowledge graphs for programming and software development concepts using Neo4j.

This tool leverages LLMs (like Claude or GPT) to automatically generate and populate a knowledge graph with programming concepts, their relationships, dependencies, and hierarchical structures. It uses Neo4j as the graph database to store and query the knowledge graph.

• Dynamic discovery of programming domains, subjects, topics, concepts, and techniques
• Automatic relationship creation (contains, requires, related_to)
• Hierarchical knowledge organization
• Validation and enhancement of knowledge graph consistency
• Command-line arguments for customizing the seeding process

• Node.js 22+
• Neo4j 4.4+ running locally or remotely
• LLM API access (Anthropic's Claude or OpenAI)

1. Clone this repository
2. Install dependencies:

   npm install
   

3. Copy .env.example to .env and configure:

   cp .env.example .env
   

4. Edit the .env file to add your own API keys and Neo4j configuration

npm run dev

npm run build
npm start

Available options:

• --clear: Clear existing data before seeding
• --depth=N: Maximum exploration depth (default: 3)
• --limit=N: Maximum total entities to create (default: 300)
• --skip-validation: Skip validation phase
• --retries=N: Number of retries for failed LLM parsing (default: 3)
• --no-fallbacks: Disable simplified fallback prompts
• --skip-gap-filling: Skip filling identified gaps in the graph
• --skip-description-enhancement: Skip enhancing auto-generated descriptions
• --children-per-unit=N: Number of child units to explore per parent (default: 4)
• --relationships-per-unit=N: Number of relationships per unit (default: 5)

MIT