Turns codebases, knowledge bases, and documentation into an interactive, searchable knowledge graph dashboard. It uses a hybrid Tree-sitter and multi-agent pipeline to map structural code relationships and business domains.
Project overview
Separates deterministic structural parsing from semantic summaries, and allows viewing the resulting dashboard entirely read-only from local disk without requiring an LLM or API key.
Project type
AI Coding · RAG
Use cases
Knowledge Q&A
Deployment
Refer to project documentation
License
MIT
Best for
Developers and AI engineers needing to visualize structural and business logic relationships in unfamiliar codebases or knowledge bases.
Users who require a read-only dashboard viewer that operates from local disk without an LLM or API key.
Key capabilities
Generates an interactive web dashboard that visualizes the codebase as a knowledge graph, searchable and clickable, color-coded by architectural layer.
Analyzes a project with a multi-agent pipeline to build a knowledge graph of every file, function, class, and dependency saved to a local JSON file.
Switches to a domain view to see how code maps to real business processes — domains, flows, and steps laid out as a horizontal graph.
Extracts wikilinks and categories from an LLM wiki index.md to generate a force-directed knowledge graph with community clustering.
Allows viewing the generated graph without an LLM or API key by serving everything read-only from local disk using Node.js.
Limitations and risks
Initial whole codebase analysis can consume a significant number of tokens.
Getting started
Installation difficulty is rated easy via a simple marketplace add or curl script. The viewer runs via npx. First success path includes running marketplace add, plugin install, and the understand command.