build-graph

Builds a searchable knowledge graph from analysis files and source code using metadata.

8|Updated Dec 25, 2025
One-click install
npx skills add https://github.com/wfukatsu/refactoring-agent-for-claude-code --skill build-graph
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: build-graph
Source: https://github.com/wfukatsu/refactoring-agent-for-claude-code/tree/main/.claude/skills/build-graph
Command: npx skills add https://github.com/wfukatsu/refactoring-agent-for-claude-code --skill build-graph

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires ryugraph, and includes scripts (resource) components.

What problem does it solve?

This Skill builds a knowledge graph from system analysis results and source code, enabling semantic querying of ubiquitous language, domains, and code relationships via a RyuGraph database.

Core Features & Use Cases

  • Graph construction: Ingests /analyze-system outputs and code structure to populate a knowledge graph with nodes for terms, domains, entities, and relationships.
  • Automated metadata storage: Stores graph schema information and statistics for governance and future evolution.
  • Use Case: Given analysis results and a codebase, run this Skill to emit a knowledge graph at knowledge.ryugraph and generate reports for exploration.

Quick Start

Run the graph build against your project: /build-graph <path-to-project>

Frequently Asked Questions about build-graph

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a knowledge graph from code analysis results?

You can build a knowledge graph from analysis results by running this Skill against your project path; it parses Markdown tables of ubiquitous language and domain-code mappings into RyuGraph nodes and relationships for semantic querying.

What is a knowledge graph for software engineering and when do I need one?

A software engineering knowledge graph maps ubiquitous language, domains, and code relationships into a queryable graph. You need one to enable fast semantic querying across formal analysis outputs and source code.

Do I need Python 3.9 and RyuGraph to construct a code knowledge graph?

Yes, constructing this code knowledge graph requires Python 3.9 or higher and the ryugraph package. You also need access to formal analysis outputs like ubiquitous language tables and domain-code mappings from your source code.

Can I use Markdown tables from system analysis to populate a knowledge graph?

Yes, Markdown tables from system analysis can populate a knowledge graph. This Skill reads Markdown tables of ubiquitous language and domain-code mappings, parsing them directly into graph nodes and relationships for semantic querying.

What's the best way to query domain-code relationships in a knowledge graph?

The best way to query domain-code relationships is to build a knowledge graph using RyuGraph. This Skill stores schema and statistics alongside the graph, enabling fast semantic queries of ubiquitous language, domains, and code relationships.

Why does knowledge graph construction require formal analysis outputs?

Knowledge graph construction requires formal analysis outputs because it relies on Markdown tables containing ubiquitous language, domain-code mappings, and architecture notes to generate the nodes and relationships that structure the graph.