skill-auditor（技能审核员）

Skill Auditor

Analyze your repository and generate a customized SKILL.md file so AI understands your project's architecture, tech stack, and conventions.

When to Use This Skill

Use this skill when the user:

• Says "audit my repo" or "audit repo"
• Says "analyze project" or "analyze my project"
• Says "generate skill file" or "create skill file"
• Asks "/skill-audit"
• Wants to understand project structure
• After major refactoring or dependency updates

Workflow

Step 1: Scan Repository

Check for these files to understand the project:

Dependencies:

• package.json → Node.js/TypeScript project
• go.mod → Go project
• requirements.txt or pyproject.toml → Python project

Extract: project name, description, language version, framework, key dependencies

Folder Structure:

Check src/, lib/, app/, or pkg/ for these patterns:

Infrastructure:

• Dockerfile → Docker deployment
• docker-compose.yml → Check for postgres, mongo, redis
• .gitlab-ci.yml or .github/workflows/ → CI/CD platform

Code Style:

• .eslintrc.*, .prettierrc.* → Formatting rules
• tsconfig.json → TypeScript settings

Step 2: Generate Rules

Based on detected patterns, create rules:

Clean Architecture:

• ✅ Keep business logic in domain layer
• ❌ Don't import infrastructure into domain
• ✅ Use dependency injection

Payload CMS:

• ✅ Define collections in payload.config.ts
• ❌ Don't bypass Payload's access control
• ✅ Use Payload hooks for business logic

Go:

• ✅ Use standard library where possible
• ❌ Don't panic, return errors
• ✅ Follow Go naming conventions

Step 3: Create Skill File

Generate .kiro/skills/[project-name]/SKILL.md:

---
name: [project-name]
description: [description]
---

# [Project Name] - Custom Skill

> 🤖 Auto-generated by Skill Auditor

## Tech Stack
- Language: [detected]
- Framework: [detected]
- Key dependencies: [top 5-8]

## Architecture Pattern
**[Detected Pattern]**

Folder structure:
- src/domain/ - Business entities
- src/application/ - Use cases
- src/infrastructure/ - External integrations

Rules:
- ✅ [Do's]
- ❌ [Don'ts]

## Development Workflow
- dev: pnpm dev
- test: pnpm test
- build: pnpm build

## Infrastructure
- Deployment: Docker
- Database: PostgreSQL
- CI/CD: GitLab CI

## Project-Specific Rules
- [Generated rules based on architecture]
- [Framework-specific best practices]

Step 4: Confirm

Show summary and confirm the skill file is active.

Example

User: /skill-audit

Agent:

🔍 Analyzing repository...

Found:
- Language: Go 1.22
- Architecture: Clean Architecture
- Deployment: Docker + Kubernetes
- Database: PostgreSQL

✅ Generated: .kiro/skills/payment-service/SKILL.md

AI now understands your project structure.

Configuration (Optional)

Add .skill-auditor.config.json for custom rules:

{
  "customRules": [
    "All API responses must include 'success' field",
    "Use Zod for validation"
  ]
}

Tips

• Run after major changes to keep skill file current
• Commit skill file to share with team
• Hooks will prompt you when dependencies change
• Use /skill-audit --force to regenerate