code-dissect

Generate structured analysis reports for AI/ML codebases.

Updated Apr 26, 2026
One-click install
npx skills add https://github.com/skywalkjian/skywalkjian-skills --skill code-dissect
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-dissect
Source: https://github.com/skywalkjian/skywalkjian-skills/tree/main/ai-code-dissect
Command: npx skills add https://github.com/skywalkjian/skywalkjian-skills --skill code-dissect

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

本技能提供对 AI/ML 代码库的结构化、全面分析,帮助研究人员在不逐行阅读代码的情况下快速理解项目架构、数据流和训练/推理流程。

Core Features & Use Cases

  • 完整的代码库结构分析与可视化,快速把握模块边界
  • 自动定位入口点、数据流路径和关键依赖,清晰呈现训练和推理流程
  • 生成可分享的分析报告,便于团队沟通与评审

Quick Start

将一个代码仓库输入给模型,即可生成完整的项目分析报告。

Frequently Asked Questions about code-dissect

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I analyze an unfamiliar AI repo structure without reading code line by line?

To analyze an unfamiliar AI repo structure without reading code line by line, you can generate a comprehensive analysis report that maps module boundaries, entry points, and inter-module relationships. This provides a complete overview of the project architecture.

What is the best way to trace data flow and training pipelines in an ML codebase?

Tracing data flow and training pipelines in an ML codebase is best achieved by automating the extraction of data-to-output flow and entry points. This approach clearly visualizes the complete training and inference pipelines without manual code tracing.

Can I automatically generate a visual report of inter-module relationships for an AI project?

Yes, you can automatically generate a visual report of inter-module relationships for an AI project. The analysis interprets repository metadata and module connections to outline a complete data-to-output flow, which can be shared for team communication and review.

Does this codebase analysis approach work for complex training and inference pipelines?

Yes, this codebase analysis approach works specifically for complex training and inference pipelines. It automatically locates key dependencies and data flow paths within AI/ML repositories to outline the complete pipeline structure.

How do I quickly grasp project structure and module boundaries in a new repository?

To quickly grasp project structure and module boundaries in a new repository, generate a structured analysis report. This report dissects repository metadata and inter-module relationships to provide a complete visualization of the project architecture.