LegacyLens started from a frustration I kept running into while exploring large repositories. Most code analysis tools give you metrics, static reports, or endless tables, but they rarely help you *see* where architectural risk actually lives. I wanted something more visual and explainable. So I started experimenting with the idea of turning repositories into an interactive “Code City”, where: - folders become districts - files become buildings - height reflects size + complexity - color reflects risk Under the hood, the project combines AST-based analysis, weighted risk scoring, aggregated folder analytics, and optional AI-generated insights. One of the hardest parts wasn’t parsing code — it was making the system understandable. Balancing metrics, visualization, UX, and explainability together turned out to be much more difficult than I expected. Another interesting challenge was scaling: large repositories quickly become visual noise, so I had to introduce adaptive aggregation and logical grouping strategies for monorepos and high-volume projects. The project is still evolving, and I’d genuinely love feedback, especially from engineers working with large or legacy codebases.

LegacyLens

Visualize GitHub repos as a Code City. Find tech debt.

Visualize GitHub repos as a Code City. Find tech debt.