Can I generate visual diagrams of my code structure?

Yes, the skill includes a diagram command that generates Mermaid-compatible output for module dependency graphs, domain concept maps, and specific entity neighborhoods.

What is a project ontology in the context of Claude Code?

A project ontology is a structured map of entities, relations, and domain concepts within your codebase. It provides a high-level representation of your system's architecture, helping both humans and AI understand how components interact.

How does the auto-scan feature prevent incorrect data?

The auto-scan feature includes a safety check mode. Users should run the scan with the --check flag first to review discovery reports before merging them into the permanent ontology with the --write flag.

Does this skill help with project documentation?

Absolutely. The gen-kb command allows you to automatically regenerate knowledge base summaries (kb-summary.md) based on the current state of the project ontology, keeping documentation in sync with the code.

Project Ontology & Knowledge Mapping

Name: Project Ontology & Knowledge Mapping
Author: seokan-jeong

byseokan-jeong

•

Learning & Documentation

Manages, queries, and visualizes project-level ontologies to help developers and AI agents navigate complex codebases.

The Ontology skill provides a sophisticated framework for mapping the structural and conceptual landscape of a software project. It allows users to perform automated scans of codebases to identify entities and relationships, utilize LLM-powered discovery for domain concepts, and generate visual Mermaid diagrams of architecture and dependencies. By maintaining a centralized ontology, this skill ensures that project knowledge is searchable, visualizable, and consistently synchronized with the codebase, facilitating better architectural oversight and more context-aware AI interactions.

Key Features

01Automatic synchronization between the ontology and project knowledge base files

02Multi-dimensional querying by entity name, type, or relationship

037 GitHub stars

04Dynamic Mermaid diagram generation for module and domain visualization

05Automated codebase scanning for entity and relation extraction

06LLM-powered auto-scan for intelligent discovery of domain concepts and patterns

Use Cases

01Visualizing complex module dependencies and domain maps to understand architectural bottlenecks

02Maintaining an AI-readable source of truth for high-level project decisions and design patterns

03Onboarding developers to large codebases by providing a searchable map of project entities

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add seokan-jeong/team-shinchan ontology

For use in Claude.ai and ChatGPT

Download Skill