An offline-first VS Code extension that uses local parsing & Gemma 4 to transform raw code into interactive system architecture diagrams. Onboard onto unfamiliar codebases, safely analyze dependencies, and understand AI-generated features instantly.
Hey Product Hunt! 👋
I’m Kennedy, the creator of CodeArchy.
As developers, we spend nearly 60% of our time reading and trying to understand code rather than writing it. With the explosion of AI coding agents generating thousands of files in minutes, and legacy codebases piling up tech debt, the question is always: "Where do I even start?"
CodeArchy was built to solve this, entirely locally and privately. It acts like a senior architect who instantly draws your codebase on a whiteboard and walks you through it.
How it helps you ship better code:
- See How Code is Connected: It parses your local code via Tree-sitter, structures it into dependency graphs, and renders clean logical subsystems in VS Code. No more traversing folder lists to guess how services talk.
- Demystify AI-Generated Code: AI agents are fast, but their code is often architecturally opaque. CodeArchy maps features instantly, giving your team the visual clarity needed.
- Understand and Contribute to Open Source Safely: Clone any repository and get an instant visualization into how feature codes are connected, and get interactive walk-through.
- 100% Offline & Private: Code never leaves your machine. Powered by Ollama (Gemma 4) and Kokoro neural TTS, everything from parsing to interactive visual chat and spoken explanations, runs on your local CPU/GPU.
CodeArchy is fully open-source, offline, and ready to make architecture education and codebase onboarding simple. I’d love to hear your feedback, feature ideas, and what codebases you analyze first!
Let’s chat in the comments! 👇
Report
No reviews yetBe the first to leave a review for CodeArchy