What problem does it solve? Understanding an unfamiliar or large codebase requires reading hundreds of files manually. This Skill automates that by scanning the project, analyzing files in semantic batches, and producing a knowledge-graph.json that powers an interactive dashboard for exploring architecture, components, and their relationships. ## Core Features & Use Cases - Full and incremental analysis: Runs a complete 7-phase pipeline on first use, then performs deterministic incremental updates on subsequent commits, re-analyzing only structurally changed files. - Semantic batching with Louvain community detection: Groups files by import-graph modularity so related modules are analyzed together, with cross-batch neighbor maps preserving edge confidence. - Multi-language import resolution: Deterministically resolves imports for TypeScript/JavaScript (including tsconfig path aliases and NodeNext ESM), Go modules, PHP autoloads, Swift packages, Java, Kotlin, Scala, and C#. - Use Case: Point it at a newly inherited monorepo and get an interactive graph showing every module, its exported symbols, dependencies, and test coverage edges, with multilingual summaries generated in your preferred language. ## Quick Start Ask the AI to run /understand on the current project directory to build the architecture knowledge graph.
