A VS Code extension for architects and senior developers —
see your codebase as a neon graph, not a wall of text.

OverviewProduct

VisualVS drops directly into VS Code and lets you understand huge codebases by looking at a topology map rather than reading thousands of lines.

Combining LLM, LangChain, and Memgraph, it extracts not just structural relationships but semantic ones — and it scales the same trick up to Kubernetes-distributed systems.

Key FeaturesCapabilities

① Instant Topology Maps

See one diagram instead of reading 1,000 lines. Auto mode re-analyses whenever you switch files, so the neon map of functions, modules, and dependencies is always one keystroke away.

② AI-driven Semantic Analysis

The LLM layer reads the meaning of the code, not just its syntax — producing logical visualisations that show why something is called, not merely what.

③ Plugin Architecture

AST parsing, AI generation, graph storage, and retrieval each live in their own pipeline stage. Drop in new languages or analysis strategies without touching the rest of the system.

④ Kubernetes-Native Topology

Built-in adaptation for distributed system analysis — extract service-to-service dependencies directly from running clusters and overlay them on the same neon graph.

⑤ Cypher Console

The AI-generated Cypher queries stay visible and editable. Power users can tune the query, drill deeper, or extract custom views without leaving the editor.

Tech StackEngineering

Format

VS Code extension (TypeScript)

AI layer

LLM (NVIDIA NIM / OpenAI / Ollama), LangChain

Graph store

Memgraph (Cypher v1 subset)

Parsing

AST parsers, modular plugin pipeline

Deploy modes

Local · Kubernetes auto-deploy · local Ollama

License

Apache License 2.0

Who It's ForAudience

Architects inheriting a legacy codebase. Senior engineers planning a refactor. SREs mapping microservice dependency surfaces.

"Where is this method called from?" "What does this service depend on?" — VisualVS answers in one diagram.

Open SourceCommunity

VisualVS is released under the Apache License 2.0.

Get in Touch

For deployment help or custom integration, drop us a line.