本文へ移動
cccskills
無料GitHub で公開

project-analysis-nextjs

Use for deep Next.js analysis: server vs client boundaries, routing, data fetching, caching, rendering modes, and hydration/runtime issues.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md3.7 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

project-analysis-nextjs

When to use

Use this skill when:

  • The project uses Next.js
  • There are SSR, SSG, ISR, or hydration issues
  • Data fetching or caching behavior is unclear
  • Client vs server behavior is inconsistent
  • universal-project-analysis routes here after detecting Next.js

Do NOT use when:

  • The task is a simple UI change
  • The issue is purely React-local and not Next-specific
  • The project is not Next.js

Core principles

  • Server and client are different execution environments
  • Rendering mode determines behavior
  • Data fetching defines lifecycle
  • Caching changes reality
  • Hydration must match server output
  • Edge and Node runtimes behave differently

Procedure

1. Identify Next.js setup

Check: Next.js version, App Router vs Pages Router, deployment target (Node vs Edge), TypeScript or JavaScript, config files. Validate: routing system is clear, runtime environment is known.

2. Determine rendering model

Identify: SSR, SSG, ISR, client-side rendering. Check: which pages/components use which mode, unexpected rendering mode switches, static vs dynamic boundaries.

3. Analyze server vs client boundaries

Inspect: server components, client components ("use client"), shared components.

Check:

  • server-only code in client components
  • client-only logic in server components
  • unexpected serialization issues

4. Trace data fetching

Inspect: fetch, getServerSideProps/getStaticProps (Pages Router), server actions, API routes.

Check:

  • where data is fetched
  • duplication of requests
  • incorrect caching
  • missing revalidation

5. Analyze caching behavior

Check: fetch caching options, revalidation settings, route segment config, browser vs server cache differences. Validate: stale data issues, unexpected cache hits, missing invalidation logic.

6. Analyze routing and navigation

Inspect: route structure, dynamic routes, layouts, navigation patterns. Check: incorrect routing assumptions, layout reuse issues, navigation inconsistencies.

7. Detect hydration and runtime issues

Check: server vs client output mismatch, conditional rendering differences, browser-only APIs used on server, timezone/randomness differences. Validate: hydration warnings, inconsistent UI between server and client.

8. Detect common Next.js anti-patterns

  • mixing server and client concerns incorrectly
  • overusing client components
  • fetching data in the wrong layer
  • ignoring caching behavior
  • duplicating fetch logic across layers

9. Validate Next.js analysis quality

Check:

  • rendering mode is explicit
  • server/client boundary is clear
  • data fetching flow is mapped
  • caching behavior is understood
  • hydration issues are explained

Output format

  1. Next.js setup summary
  2. Rendering model overview
  3. Server vs client boundary analysis
  4. Data fetching findings
  5. Caching behavior findings
  6. Routing findings
  7. Hydration/runtime issues
  8. Key risks and fixes

Gotcha

  • Many bugs come from misunderstanding server vs client boundaries.
  • Caching can make correct code appear broken.
  • Hydration errors often indicate mismatched assumptions, not simple bugs.

Do NOT

  • Do NOT assume everything runs on the client
  • Do NOT ignore caching behavior
  • Do NOT mix server and client logic blindly
  • Do NOT debug hydration issues without checking rendering differences
  • Do NOT treat Next.js as "just React"

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Use when reviewing UI for accessibility — WCAG 2.2 AA, keyboard nav, focus, ARIA, contrast, screen-reader semantics — even on 'is this a11y-OK?' or 'mach das barrierefrei'.

日本語の概要は準備中です。原文の説明を表示しています。

event4u-app/agent-config112026年10月12日 更新

Use when defining or auditing the activation event — aha-moment selection, retention correlation, falsifiable definition. Triggers on 'what is our aha moment', 'redefine activation'.

日本語の概要は準備中です。原文の説明を表示しています。

event4u-app/agent-config112026年10月12日 更新

Use when capturing an architectural decision — file naming, next ADR number, Status / Context / Decision / Consequences, index regen; fires even without saying 'ADR'.

日本語の概要は準備中です。原文の説明を表示しています。

event4u-app/agent-config112026年10月12日 更新

Adversarial critique — devil's advocate, stress-test, honest teardown ('poke holes', 'be brutal', 'was hältst du davon'); explicit request only. Routine code or design review → code-review.

日本語の概要は準備中です。原文の説明を表示しています。

event4u-app/agent-config112026年10月12日 更新

Use when reading, creating, or updating agent documentation, module docs, roadmaps, or AGENTS.md. Understands the full .augment/, agents/, and copilot-instructions structure.

日本語の概要は準備中です。原文の説明を表示しています。

event4u-app/agent-config112026年10月12日 更新

Use for an adversarial red-team / blue-team / auditor review of an AI agent's CONFIG + behaviour (rules, skills, MCP, hooks, permissions) — attack-chain → defensive-gap list, not a code audit.

日本語の概要は準備中です。原文の説明を表示しています。

event4u-app/agent-config112026年10月12日 更新

event4u-app のスキルをすべて見る

このスキルの問題を報告する