Uni-app Vue Project Analyzer

A static health check for uni-app and Vue projects

When taking over a uni-app or Vue repository, engineers often need to determine the project version, directory structure, page routes, dependencies, and overall code health. Reading only package.json or opening a few random files can miss platform-specific files such as manifest.json, pages.json, and uni.scss, and it is hard to turn Vue 2/3 migration issues, large files, low comments, and complex functions into actionable findings. This skill turns those concerns into a structured project report, useful before code review, refactoring, or legacy project onboarding.

From project identification to report output

It first identifies project type and root directory from marker files: uni-app projects look for manifest.json, pages.json, and uni.scss, while Vue projects inspect package.json, vue.config.js, vite.config.js, and related config. It then moves through configuration analysis, code quality assessment, and file statistics. Core capabilities include:

• Technical debt scoring: quantifies code health from 0 to 100, deducting points for var, long functions, oversized files, hard-coded values, console.log, and TODO items.
• Vue migration checks: flags Vue 2 patterns such as destroyed, Vue.use(), v-model, and /deep/, while avoiding false positives on uni-app lifecycles like onShow and onHide.
• Architecture recognition: identifies common patterns such as singleton, factory, observer, module, strategy, and Mixin.
• Statistics and reporting: counts .vue, .js, .ts, style, and config files, then generates a Markdown report; JSON, HTML, or bilingual output can be produced when requested.

Boundaries

It relies on existing repository configuration and source code, and it does not replace type checking, tests, or runtime validation. Files larger than 1 MB are usually reduced to basic statistics. The analysis excludes generated directories such as node_modules/, unpackage/, uni_modules/, and dist/, so it is best suited for project-level triage rather than fine-grained single-file defect analysis.