jQAssistant Graph RAG FAQs

Question 1

What is jQAssistant Graph RAG?

Accepted Answer

jQAssistant Graph RAG transforms standard jQAssistant-generated Neo4j graphs into a highly enriched, semantically intelligent knowledge graph. It integrates source code with bytecode structure, enabling deep, AI-driven analysis of Java/Kotlin projects.

Question 2

How does it enhance jQAssistant graphs for AI?

Accepted Answer

It performs a two-stage process: first, graph enrichment by adding source code structure and class hierarchy overlays, normalizing properties. Second, RAG generation by using LLMs to create hierarchical, dependency-aware summaries for all codebase components, from methods to the entire project.

Question 3

What kind of insights can I gain from the enriched graph?

Accepted Answer

You can uncover deep insights into software architecture, behavior, and design. Examples include understanding package responsibilities, identifying key components in a directory, tracing workflows, and getting refactoring suggestions, all through natural language queries with AI agents.

Question 4

Can it handle large and complex codebases?

Accepted Answer

Yes, jQAssistant Graph RAG is designed for scalability. It intelligently handles large source files and complex components by chunking context to fit LLM token limits and supports incremental, resumable processing with caching to save time and LLM costs.

Question 5

What are the primary prerequisites to use this tool?

Accepted Answer

You need a pre-existing jQAssistant-generated Neo4j graph of your Java/Kotlin project (including both source and compiled files) and a Python 3.13+ environment with the required dependencies installed. An LLM API is needed for summary generation, though a 'fake' API is available for testing.

jQAssistant Graph RAG

jQAssistant Graph RAG

概要

主な機能

ユースケース

概要

主な機能

ユースケース