Token导航 LogoToken导航TokenDH.com
前端设计执行命令github未标认证来源可访问许可证需确认审计通过

architecture-audit架构审计

Agent Skill

用于辅助安全审计、权限检查、凭据风险、认证流程和常见漏洞排查。它适合让 Agent 梳理敏感配置、检查依赖风险、分析鉴权逻辑或生成安全复核清单。使用时不能把工具输出直接当最终结论,涉及密钥、令牌、用户数据或生产系统时,应先确认最小权限、脱敏方式和操作边界。

总安装

563

周安装

23

GitHub Stars

7

下载量

180
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:architecture-audit(架构审计)
来源仓库:https://github.com/helderberto/skills
仓库路径:skills/architecture-audit
安装命令:
npx skills add https://github.com/helderberto/skills --skill architecture-audit
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/helderberto/skills --skill architecture-audit

简介

用于深度模块结构分析,识别接口复杂度和实现耦合问题。

  • 提供可测试性评估和纯函数提取等改进建议。
  • 适用于代码重构和技术债务治理场景。architecture-audit 属于前端设计类 Skill,可作为该场景下的辅助能力补充。
  • 安装需指定 GitHub 仓库,使用前应确认代码库访问权限。
  • 涉及生产环境时应先进行脱敏和最小权限验证。

SKILL.md

Architecture Audit

A deep module has a simple interface hiding a large implementation. Deep modules are more testable, more AI-navigable, and let you test at the boundary instead of inside. See references/deep-modules.md for examples and anti-patterns.

Workflow

1. Explore

Use the Agent tool with subagent_type=Explore to navigate the codebase organically. Note where you experience friction:

  • Understanding one concept requires bouncing between many small files
  • A module's interface is nearly as complex as its implementation
  • Pure functions extracted just for testability, but real bugs hide in how they're called
  • Tightly-coupled modules create integration risk at the seams between them
  • Areas that are untested or hard to test

The friction you encounter IS the signal.

2. Present candidates

Show a numbered list. For each candidate:

  • Cluster: which modules/concepts are involved
  • Why they're coupled: shared types, call patterns, co-ownership of a concept
  • Dependency category: see references/dependency-categories.md
  • Test impact: what existing tests would be replaced by boundary tests

Do NOT propose interfaces yet. Ask: "Which candidate would you like to explore?" — list each candidate as an option with its cluster name as label and coupling summary as description. Use AskUserQuestion when available; otherwise present as a numbered list.

3. User picks a candidate

4. Frame the problem space

Write a user-facing explanation of the chosen candidate:

  • Constraints any new interface must satisfy
  • Dependencies it would need to rely on
  • A rough illustrative code sketch to ground the constraints (not a proposal)

Show this to the user, then immediately proceed to Step 5. User reads while sub-agents work.

5. Design multiple interfaces

Spawn 3+ sub-agents in parallel using the Agent tool. Each produces a radically different interface. Give each a separate technical brief (file paths, coupling details, dependency category). Assign distinct constraints:

  • Agent 1: "Minimize — aim for 1–3 entry points max"
  • Agent 2: "Maximize flexibility — support many use cases and extension"
  • Agent 3: "Optimize for the most common caller — make the default case trivial"
  • Agent 4 (if applicable): "Design around ports & adapters for cross-boundary dependencies"

Each sub-agent outputs:

  1. Interface signature
  2. Usage example
  3. What complexity it hides internally
  4. Dependency strategy (see references/dependency-categories.md)
  5. Trade-offs

Present designs sequentially, compare in prose, then give your own recommendation. Be opinionated — if elements from multiple designs combine well, propose a hybrid.

6. User picks an interface

Ask: "Which interface design should we use?" — list each design as an option with preview showing the interface signature. Add "Your recommendation" as first option (Recommended) with the hybrid/recommended design in preview. Use AskUserQuestion when available; otherwise present as a numbered list.

7. Write improvement PRD

Save a markdown file named architecture-<cluster-name>.md using the template in references/improvement-template.md. If .tracerkit/prds/ exists, save there; otherwise, save in prds/.

Fill with concrete details: file paths, function names, migration steps. Share the file path with the user when done.

Rules

  • Never propose an interface before the user picks a candidate (Step 3)
  • Old unit tests on shallow modules are waste once boundary tests exist — note them for deletion
  • Classify every candidate's dependency type before designing interfaces
  • Show the problem space framing (Step 4) before sub-agents finish — don't wait
  • PRD must reference concrete files and functions, not abstract concepts

Error Handling

  • If exploration reveals a candidate scope much larger than expected → surface this and ask user to re-scope before proceeding

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

33.45%
按下载量换算60

Claude

31.11%
按下载量换算56

Cursor

18%
按下载量换算32

Gemini CLI

8.75%
按下载量换算16

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

执行命令

安装流程涉及命令执行,可能通过 npx skills add https://github.com/helderberto/skills --skill architecture-audit 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills