claude-skills/

Anthropic公式スキル・プラグインの日本語ディレクトリ

last sync 22h ago
スキルKnowledge Work

🤖ai-firstify

プラグイン
AI-Firstify
ソース
GitHub で見る ↗

説明

AI-First Design原則に合わせてプロジェクトを分析、再設計、またはブートストラップします。 プロジェクトのレビュー、監査、改善、「AI-First化」、または新規プロジェクトの開始を依頼された場合に使用します。 7つのディメンション全体にわたる深い分析を実施し、既存プロジェクトを積極的に再構築するか、発見的な質問を通じて新規プロジェクトのセットアップをガイドします。TechWolf AI-First Bootcampの9つの設計原則と7つの設計パターンに基づいています。

原文を表示

Analyze, re-engineer, or bootstrap projects to align with AI-first design principles. Use when asked to review, audit, improve, 'ai-firstify', or start a new project. Performs deep analysis across 7 dimensions, actively restructures existing projects, or guides new project setup through discovery questions. Based on the 9 design principles and 7 design patterns from the TechWolf AI-First Bootcamp.

ユースケース

  • ✓プロジェクトのレビューを依頼されたとき
  • ✓既存プロジェクトを監査・改善したいとき
  • ✓プロジェクトをAI-First化したいとき
  • ✓新規プロジェクトを開始するとき

本文

AI-Firstify

Analyze and re-engineer projects to align with AI-first design principles.

Trigger

Use this skill when asked to:

  • Audit mode (read-only): "review", "audit", "analyze", "check", "assess"
  • Re-Engineer mode (active changes): "ai-firstify", "fix", "improve", "re-engineer", "transform"
  • Bootstrap mode (new project): "start", "new project", "bootstrap", "set up", "build from scratch"

Mode 1: Audit (Read-Only Analysis)

Perform a comprehensive read-only analysis across 7 dimensions. Output a scored report with recommendations. Do NOT modify any files.

Read references/mode-audit.md for the full audit procedure and dimension checklist.

Mode 2: Re-Engineer (Active Transformation)

First, perform the full audit (Mode 1). Then actively fix issues in 7 phases: foundation, de-agentification, skill extraction, complexity reduction, context hygiene, safety hardening, workflow optimization.

Read references/mode-reengineer.md for the full re-engineering procedure.

Mode 3: Bootstrap (New Project Setup)

Guide the user through setting up a new AI-first project from scratch. Interactive: ask discovery questions, recommend architecture, scaffold, and test.

Read references/mode-bootstrap.md for the full bootstrap procedure.

Reference Files

Domain knowledge (load on demand per dimension):

  • references/principles.md: All 9 AI-first design principles in depth
  • references/patterns.md: All 7 design patterns with implementation guidance
  • references/anti-patterns.md: Common mistakes with detection patterns and fixes
  • references/skill-architecture.md: How to structure skills, sub-agents, and workflows
  • references/project-structure.md: Ideal project layouts, CLAUDE.md templates, .gitignore
  • references/assessment-rubric.md: Scoring criteria and report template

Load the relevant reference file for the dimension you are currently analyzing. Do not load all references at once. Use progressive disclosure.

Tools

  • scripts/validate-report.sh: Validates that a generated assessment report has all required sections, dimensions, and scores

原文・著作権は Anthropic および各プラグイン作者に帰属します。日本語訳は Claude API による自動翻訳です。