Project Structure Analysis

Problem

When taking over an unfamiliar codebase, the directory tree, config files, build scripts, and core entry points are often buried under noise such as node_modules, dist, and .git. Reading source file by file is slow, and it is hard to quickly determine the project type, stack, layered responsibilities, and startup path.

How It Works

The skill turns project context into a structured analysis instead of simply summarizing source code. Inputs can be a directory list, source snippets, or a directory screenshot. It performs the following tasks:
- Identify project type: classify frontend, backend, full-stack, mini program, desktop, and infer stacks such as Vue/React, Java/Spring, Node.js, Python.
- Generate a concise tree: filter out caches, build artifacts, and VCS directories like node_modules, dist, build, and .git, keeping core directories relevant to architecture.
- Map layers and key files: describe layer responsibilities and locate entry points, routing, configuration, and business modules.
- Suggest startup steps: provide minimal run steps, structural trade-offs, and targeted optimization directions.

Boundaries

It is best for a first-pass understanding of a project skeleton, useful before code review, onboarding to legacy code, stack assessment, or drafting project documentation. Output quality depends on input completeness: full directory lists and core snippets improve confidence, while vague screenshots may lead to inferred key files and startup steps. It does not replace reading source code and cannot guarantee coverage of every private build convention.