agentleFS
michaelalbermichaelalber/ai-toolkit/skills/react-architecture-checklist/SKILL.md

react-architecture-checklist

Skill3 starsChanged 3 months agoMITView on GitHub

Grades an existing React/TypeScript codebase. Detects React version, bundler (Vite/CRA/Next), TypeScript usage, state library, and router, then checks hooks discipline, component cohesion, effect correctness, render performance, state boundaries, accessibility, and type safety with file:line evidence. Use to review or grade a React codebase. Not for Socratic critique (architecture-review), security audits (react-security-review), or new test-first code (tdd).

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Version-gate recommendations: Do not suggest React 18/19 APIs (useTransition, Actions, use(), server components) for a React 16/17 codebase.

    line 115

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

React Architecture Checklist

"A checklist cannot fly a plane, but a pilot cannot fly safely without one." — Atul Gawande

Core Philosophy

Shared across the dotnet / python / php / rust / react architecture checklists — same values, language-specific checks.

Grounding note: the knowledge base has no React corpus. Use collection="javascript" for TS/JS idioms, collection="ui_ux" for accessibility, and collection="internal" for architecture standards; cite react.dev as the primary React authority. Never invent a react collection.

#ValueWhat it means
1Detect before judgeDetermine React version / bundler / TS / state lib before applying any item; context decides what is idiomatic.
2Evidence over opinionEvery finding cites file:line and the offending pattern. "Too many re-renders" is not a finding; "features/cart/Cart.tsx:31 recreates onAdd every render, breaking the memoized <Row>" is.
3Feature cohesionOrganized by feature, not by technical type (components/, hooks/, utils/ dumping grounds). Cross-feature imports are a violation.
4Dependencies point inwardUI components depend on hooks/services, not the reverse; data-fetching is isolated from presentation. Boundaries are explicit via barrels.
5Hooks honor the rulesHooks called unconditionally at the top level; effect dependency arrays complete; effects clean up. A lint suppression of react-hooks/exhaustive-deps is a finding.
6Config & secrets hygieneNo secrets in client bundles; only intentionally-public VITE_/NEXT_PUBLIC_ values exposed; config injected, not hardcoded.
7Version awarenessRecommendations gated to the detected React version; never suggest an API that does not exist there (e.g. use() / Actions on React 17).
8Tests gate changeUntested components/hooks are a finding; high-risk interactive components without RTL tests are prioritized.
9Graded, actionable outputA letter grade (A–F) from counted findings, plus prioritized, version-correct recommendations.

Workflow

Shared skeleton: DETECT → SCAN → REPORT → RECOMMEND.

DETECT     React version (package.json `react`), bundler (Vite/CRA/Next.js), TypeScript (tsconfig +
           strict), state library (Redux/RTK/Zustand/Context/none), router, and whether the React
           Compiler is wired (`babel-plugin-react-compiler` / Next.js `experimental.reactCompiler`).
           Record findings; version and compiler presence both change what is idiomatic.

SCAN       Run the React Checklist below section by section. Gather evidence with tooling:
             npx eslint . --max-warnings 0          # baseline — incl. eslint-plugin-react-hooks
             npx tsc --noEmit                        # type errors gate the review
             npx knip                                # dead code / unused exports / deps
             grep -rn "useEffect" src/ | wc -l       # effect surface to audit
           Every violation becomes a finding with file:line and a severity (critical/high/medium/low).

REPORT     Emit the graded report (Output Template). Grade = function of counted findings.

RECOMMEND  Prioritize: critical → quick wins → modernization. Version-gate every recommendation.

React Checklist (language-specific)

#CheckSeverity
1Hooks rules — hooks called unconditionally at top level; no hooks in loops/conditions; custom hooks prefixed useCritical
2Effect correctness — complete dependency arrays (no exhaustive-deps suppressions); cleanup returned for subscriptions/timers; no derived state that belongs in renderHigh
3Component cohesion — one responsibility per component; container/presentation separation where it earns it; no 300-line god componentsMedium
4State placement — state lives at the lowest common owner; no prop-drilling past ~3 levels (lift to context/store); server state in a query cache, not useState+useEffectHigh
5Render performance — stable list keys (never array index for dynamic lists); if the React Compiler is wired, manual memo/useMemo/useCallback are a finding unless justified (compiler bail-out, cross-boundary prop to a non-compiled dep); if not wired, use them only where a measured re-render warrants itMedium
6Type safety — tsconfig strict: true; no any without justification; props typed (no implicit any); no as casts hiding shape mismatchesHigh
7Accessibility — semantic elements over div soup; interactive elements keyboard-reachable; labels/alt/ARIA where needed; eslint-plugin-jsx-a11y cleanHigh
8Boundary & dep hygiene — no cross-feature deep imports; data-fetching isolated from presentation; error boundaries around async UI; bundle/dep weight justifiedMedium

ESLint config: eslint configuration. Full section-by-section list (with the hooks & effects audit table) + grounding query map: review checklist.

State Block

<arch-checklist-state>
language: react
mode: DETECT | SCAN | REPORT | RECOMMEND | COMPLETE
detected: [react-version | bundler | ts:strict/loose | state-lib | tests:yes/no]
issues_found: [critical:N high:N medium:N low:N]
last_action: [what was just done]
next_action: [what should happen next]
</arch-checklist-state>

Output Template

Shared across all architecture checklists.

## Architecture Checklist: [app/package] (React)
**React**: [19] | **Bundler**: [Vite/CRA/Next] | **Compiler**: [wired/absent] | **TS**: [strict/loose/none] | **State**: [RTK/Zustand/Context] | **Tests**: [yes/no]

| Section | Pass | Fail | Warn |
|---------|------|------|------|
| Hooks / Effects / Cohesion / State / Render / Types / a11y / Boundaries | … | … | … |

### Grade: [A–F]
Grading: **A** 0 crit/0 high/≤3 med · **B** 0 crit/≤2 high · **C** 0 crit, gaps in one area ·
**D** 1+ crit · **F** fundamental problems (conditional hooks, `any` everywhere, no a11y, server state in effects throughout).

| Severity | Location | Finding | Recommendation |
|----------|----------|---------|----------------|
| CRITICAL | file:line | [pattern] | [version-gated fix] |

**Hooks/effects audit**: | location | issue | dep array | cleanup | risk |
**Quick wins**: [low-effort, high-impact] · **Modernization**: [larger items with effort estimate]

AI Discipline Rules

  • Lint + types must pass first. eslint --max-warnings 0 and tsc --noEmit are the baseline; report failures before the architectural checklist.
  • Evidence or it is not a finding. Cite file:line; show the eslint/grep output. Never grade on vibes.
  • Version-gate recommendations. Do not suggest React 18/19 APIs (useTransition, Actions, use(), server components) for a React 16/17 codebase.
  • Architecture, not security. XSS, secret leakage, and dependency CVEs belong to react-security-review — note them and route there.

Integration with Other Skills

  • architecture-review — When the grade is D/F, escalate to the Socratic critic: this checklist finds what is wrong; architecture-review builds why.
  • react-security-review — Companion for the security dimension (XSS, bundle secrets, npm audit).
  • react-feature-slice — Correct-pattern reference when the checklist flags structural/cross-feature violations.
  • react-modernization-analyzer — When findings cluster around legacy patterns (class components, CRA), route to the modernization plan.
  • tdd — Methodology for adding the RTL tests the checklist flags as missing, and for driving any refactor.
  • dotnet / python / php / rust-architecture-checklist — Sibling skills sharing this exact Core Values + workflow + output.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.