fluorite-mcp
Advanced tools
Enterprise-grade MCP server for Claude Code CLI with 6,200+ spike templates, 86 library specs, intelligent memory engine, and continuous learning capabilities.
Transform Claude Code CLI into a comprehensive development powerhouse with Fluorite MCP v0.20.1. Experience professional-grade code generation, intelligent error prevention, and accelerated prototyping through our extensive library of 6,200+ spike templates, 86 library specifications, and proven patterns.
Enterprise Solution: From rapid prototyping to production deployment - all specifications validated, maintained, and optimized for professional development workflows with continuous learning capabilities.
Quick Installation:
npm install -g fluorite-mcp
claude mcp add fluorite-mcp -- fluorite-mcp
Fluorite MCP is a comprehensive Model Context Protocol server that enhances Claude Code CLI with:
Transform your development workflow with professionally curated specifications:
Prevent issues before they occur with intelligent analysis:
Bridge the gap between idea and implementation:
Professional-grade tooling that scales with your team:
Advanced Rust-based learning system that evolves continuously:
Quick Setup (2-minute installation):
# 1. Install Fluorite MCP globally
npm install -g fluorite-mcp
# 2. Add to Claude Code CLI
claude mcp add fluorite-mcp -- fluorite-mcp
# 3. Verify installation
claude mcp list
Alternative Installation Methods:
# Using yarn
yarn global add fluorite-mcp
# Using pnpm
pnpm add -g fluorite-mcp
# From source (developers)
# Clone from your fork or local copy
cd fluorite-mcp && npm install && npm run build
Test your installation with a simple request:
"Create a modern login form with validation using react-hook-form and zod"
Expected Results:
FluoriteはClaude Codeの/fl:コマンド実行時にスパイク（テンプレート）を自動活用します。
auto-spikeにユーザー入力全文をtaskとして渡します。preview-spike→apply-spikeで差分を提示しつつ適用（サーバーはdiffのみ返却、実適用はクライアント側）。discover-spikesで候補列挙→ユーザー確認。/fl:spikeでdiscover/auto/preview/applyの運用が可能。Claude内の例:
/fl:implement "Bun Elysia の typed worker を TS で作る"
/fl:design "API Gateway + 認可 + RateLimit"
/fl:spike auto "Elysia のセキュアな plugin scaffolding"
/fl:spike discover "elysia worker typed"
セットアップ/更新:
fluorite setup
Docs CLI（ローカルドキュメントの一覧/パス表示/内容出力）:
fluorite docs
fluorite docs --show recipes.md
fluorite docs --cat one-pagers.md
Bulk generation（大規模書き出し）:
fluorite spikes synth-bulk --generated-only --filter '^strike-' --total 1000 --batch 100 --pretty
自然言語に短縮トークンを含めることで、人気スパイクをダイレクト指定できます。[alias: <token>] 形式でもOK。
elysia-worker-tselysia-plugin-secure-tsnext-route-tsnext-mw-tsfastapi-secure-pyreact-component-tsreact-hook-tsprisma-schema-tsdrizzle-schema-tsnext-service-tsreact-provider-ts, react-adapter-ts例:
/fl:implement "Next.js に typed middleware を追加 [alias: next-mw-ts]"
/fl:implement "Prisma schema を TS で生成 [alias: prisma-schema-ts]"
パラメータ（constraints）を渡すと、preview-spike の params に引き継がれます:
// 例: Prismaのモデル名を指定
/fl:implement "[alias: prisma-schema-ts]" model=User
詳しい一覧と使い方: docs/short-aliases.md 適用後のチェックリスト: docs/post-apply-checklists.md（検証例: docs/verification-examples.md） ファイル構成サンプル: docs/file-structure-samples.md 最小diffサンプル: docs/diff-samples.md 統合レシピ: docs/recipes.md（Prompt → Files → Diff → Verify） Queue/Brokerスニペット: docs/queue-snippets.md Search Index Tips: docs/search-index-tips.md One-Pagers: docs/one-pagers.md Monitoring & Alerts (MVP): docs/monitoring-alerts.md Bulk Generation Guide: docs/BULK-GENERATION.md
よく使う領域をパック化して、合成テンプレートを絞り込み・大量書き出しできます。
fluorite spikes packsfluorite spikes synth --pack search --generated-only --write --pretty --max 100詳しくは以下を参照してください。
docs/ja/strike-packs.mddocs/ja/strike-recipes.mddocs/ja/pack-checklists.mddocs/ja/pack-samples.mddocs/ja/tutorials.mddocs/ja/pr-checklist.mddocs/ja/metrics-scrape.mddocs/ja/metrics-alertmanager.mddocs/ja/storage-filename-policy.mddocs/ja/storage-cdn-inline.mddocs/ja/storage-cdn-recipes.mddocs/ja/storage-cdn-ui.mddocs/ja/payments-db-design.mddocs/ja/pr-checklist.mdExamples (runnable minis): examples/pack-examples
npx tsx examples/pack-examples/meili/index.tsnpx tsx examples/pack-examples/s3/index.tsnpx tsx examples/pack-examples/pino/index.tsnpx tsx examples/pack-examples/stripe/index.tsnpx tsx examples/pack-examples/monitoring-full/index.tsnpx tsx examples/pack-examples/search-full/index.tsPORT=3003 STRIPE_WEBHOOK_SECRET=whsec_xxx npx tsx examples/pack-examples/payments-full/index.tsPORT=3004 AWS_REGION=us-east-1 npx tsx examples/pack-examples/storage-full/index.tsstrike-* を含む仮想スパイクを動的に合成し、10万件以上のテンプレートを提供します。詳細・チューニング・使い方の日本語ガイドは docs/ja/strike-scale.md を参照してください。
環境変数で自動選定の挙動を調整できます。
FLUORITE_ALIAS_ENABLE（既定 true）: 短縮エイリアス推定を有効/無効にします（0|false|no|off で無効）FLUORITE_AUTO_SPIKE_THRESHOLD（既定 0.4）: coverage がこの値未満のときは discover-spikes と確認質問を提示FLUORITE_AUTO_SPIKE_TOP（既定 5）: 詳細評価に進める候補数FLUORITE_AUTO_SPIKE_BATCH（既定 50）: メタデータバッチサイズFLUORITE_SPIKE_METADATA_MULTIPLIER（既定 2）: discovery時のスコアリング対象拡大FLUORITE_SPIKE_LIST_LIMIT: スパイク露出数の上限（パフォーマンス調整）FLUORITE_ALIAS_BOOST（既定 2.0）: 短縮エイリアス候補の優先度（0.0–5.0）Try these ready-made prompts inside Claude Code to trigger automatic spike selection and previews:
/fl:implement "Bun Elysia の typed listener を TypeScript で: /health と /metrics を実装"/fl:design "Elysia listener に CORS/Helmet/RateLimit を適用したセキュア構成"/fl:spike auto "Elysia のセキュアな plugin scaffolding (auth middleware, schema validation)"/fl:spike discover "elysia plugin secure ts"/fl:implement "Elysia の typed worker を TS で、Vitest 付きで"/fl:spike discover "elysia worker testing ts"/fl:spike auto "Elysia の基本的な migration セットアップを TS で"/fl:spike auto "Elysia migration をセキュア構成へアップグレード"/fl:spike auto "Elysia の typed seed スクリプト (TS) を作成"ヒント: auto-spike の結果で coverage_score が低い場合は discover-spikes に自動フォールバックします。preview-spike の差分を確認してから apply-spike を実行してください。
React/Next.js:
/fl:implement "Next.js で OAuth ログイン付きのダッシュボード（Rechartsでグラフ、TanStack Tableで表）」/fl:design "Next.js App Router で認可ミドルウェア＋RSC パターンのアーキテクチャ"FastAPI（Python）:
/fl:spike auto "FastAPI のJWT認証＋Refreshトークン＋/auth/* エンドポイントを実装"/fl:implement "FastAPI WebSocket チャット（Redis PubSub、pytest 付き）"Rust/Axum:
/fl:spike auto "Axum で認証ミドルウェア＋typed routing＋tower-limits を構成"/fl:design "Axum + SQLx + Tokio でAPIサーバのベストプラクティス構成"Go（Gin or Fiber）:
/fl:spike discover "gin middleware secure"/fl:implement "Fiber で REST API（validator + logger + rate limit + unit test）"Vue（Vite + Pinia）:
/fl:spike auto "Vue 3 + Pinia + Vue Router で認証付きダッシュボード"/fl:implement "Vite + Vitest 構成で UI コンポーネントとテストを追加"| Domain | Key Libraries | Specifications | Use Cases |
|---|---|---|---|
| 🎨 UI/UX | shadcn/ui, Radix UI, Material-UI, Mantine, Tremor | 15+ | Component libraries, design systems, accessibility |
| 🔄 State Management | Zustand, Jotai, TanStack Query, Redux Toolkit, Valtio | 8+ | Application state, server state, real-time data |
| 📝 Forms & Validation | react-hook-form, Zod, Valibot, Yup, Formik | 6+ | Type-safe forms, schema validation, user input |
| 🔐 Authentication | NextAuth.js, Clerk, Lucia, Supabase Auth, Auth0 | 8+ | User management, OAuth, session handling |
| 🧪 Testing & Quality | Playwright, Vitest, Cypress, Testing Library, Jest | 10+ | E2E testing, unit testing, accessibility testing |
| 📊 Data & Visualization | Recharts, D3.js, Chart.js, Tremor, TanStack Table | 8+ | Charts, dashboards, data tables, analytics |
| 🌐 API & Backend | tRPC, Prisma, Drizzle, GraphQL, REST patterns | 12+ | Type-safe APIs, database integration, real-time |
| 🚀 Performance | React Query, SWR, Optimization patterns, Bundling | 6+ | Caching, performance monitoring, optimization |
| Language | Frameworks & Tools | Specifications | Primary Use Cases |
|---|---|---|---|
| 🟦 TypeScript/JavaScript | Next.js, React, Vue, Svelte, Astro | 25+ | Full-stack web applications, SPAs, SSR |
| 🐍 Python | FastAPI, Django, Flask, Pydantic, SQLAlchemy | 12+ | APIs, data science, machine learning, automation |
| 🦀 Rust | Tauri, Axum, Actix, Serde, Tokio | 8+ | Systems programming, desktop apps, WebAssembly |
| 🐹 Go | Gin, Fiber, Echo, GORM, Chi | 6+ | High-performance backends, microservices, CLI tools |
| 💎 Ruby | Rails, Sinatra, Sidekiq, ActiveRecord | 4+ | Full-stack web applications, rapid prototyping |
| ☕ Java | Spring Boot, Hibernate, Maven, Gradle | 6+ | Enterprise applications, microservices, Android |
| ⚡ C#/.NET | ASP.NET Core, Entity Framework, Blazor | 5+ | Enterprise software, web APIs, desktop applications |
| 🎯 Dart/Flutter | Flutter, Dart packages, State management | 4+ | Cross-platform mobile apps, desktop applications |
| 🧊 Elixir | Phoenix, LiveView, Ecto, OTP | 3+ | Real-time applications, distributed systems |
| ⚙️ Zig | Zig standard library, Build system | 2+ | Systems programming, high-performance applications |
| 🔧 C/C++ | Modern C++ patterns, Build systems | 3+ | Systems programming, game development, embedded |
| 📱 Swift | SwiftUI, Combine, Foundation | 3+ | iOS/macOS development, Apple ecosystem apps |
Fluorite MCP bridges the gap between natural language requirements and production-ready code through a sophisticated multi-stage process:
| Your Natural Language Request | Fluorite MCP Intelligence | Production-Ready Output |
|---|---|---|
| "Create a responsive data table with sorting, filtering, and pagination" | Analyzes: Project framework, existing UI patterns, performance requirements Selects: TanStack Table + your UI library Generates: Feature-complete component | Fully typed data table with: • Responsive design • Accessible sorting & filtering • Virtual scrolling for performance • Comprehensive test coverage |
| "Set up user authentication with social login" | Analyzes: Next.js project, security requirements, OAuth providers Selects: NextAuth.js + appropriate providers Generates: Complete auth system | Production auth flow with: • Secure session management • OAuth integration (Google, GitHub, etc.) • Protected routes & middleware • CSRF & XSS protection |
| "Build a real-time chat component" | Analyzes: Framework capabilities, real-time requirements, scalability needs Selects: Socket.IO or WebSockets + state management Generates: Full chat implementation | Enterprise-grade chat with: • Real-time messaging • Typing indicators & presence • Message persistence • Accessibility features |
| "Create a dashboard with charts and metrics" | Analyzes: Data visualization needs, performance constraints, design system Selects: Recharts/D3.js + responsive layout Generates: Interactive dashboard | Professional dashboard with: • Interactive charts & graphs • Real-time data updates • Mobile-responsive design • Export & filtering capabilities |
Transform ideas into working prototypes in minutes with our comprehensive template ecosystem:
| Category | Templates | Key Features | Time Savings |
|---|---|---|---|
| 🌐 Frontend Frameworks | 180+ | Next.js (SSR, API routes, middleware), React (hooks, contexts), Vue (Composition API) | 80% faster setup |
| 🚀 Backend APIs | 120+ | FastAPI (auth, WebSockets, databases), Express, Go/Rust servers | 70% faster API development |
| 🧪 Testing & Quality | 150+ | Playwright (E2E, accessibility), Vitest, Cypress, performance testing | 60% faster test setup |
| 🏗️ Infrastructure | 200+ | Docker, Kubernetes, CI/CD pipelines, monitoring, deployment | 90% faster DevOps setup |
| 🎨 UI Components | 100+ | Component libraries, design systems, interactive elements | 85% faster UI development |
Your Request: "I need JWT authentication for my FastAPI project"
Fluorite MCP Process:
1. 🔍 Detects FastAPI project context
2. 🎯 Selects "fastapi-jwt-auth" template
3. ⚡ Generates complete auth system
4. 🔧 Adapts to your existing database setup
5. ✅ Includes security best practices
Generated Output:
• JWT token generation & validation
• User registration & login endpoints
• Password hashing with bcrypt
• Refresh token rotation
• Rate limiting & brute force protection
• Comprehensive test suite
• API documentation with examples
| Resource | Time Investment | What You'll Learn | Best For |
|---|---|---|---|
| ⚡ Quick Start Guide | 5 minutes | Installation, first commands, basic usage | New users getting started |
| 🔧 Installation Guide | 10 minutes | Detailed setup, troubleshooting, environment configuration | System administrators, DevOps |
| 📖 Command Reference | 15 minutes | Complete command syntax, advanced features | Power users, automation |
| Resource | Audience | Content | Value |
|---|---|---|---|
| 🔍 API Documentation | Developers, integrators | Complete MCP protocol reference, 200+ documented functions with TSDoc | Build custom integrations and understand internal APIs |
| 👩💻 Developer Guide | Contributors, advanced users | Architecture, comprehensive function documentation, contributing guidelines | Extend and customize with full API understanding |
| 🏭 Template Creation Guide | Template authors | Create custom Spike templates and library specifications | Add new frameworks/libraries |
| 📊 Static Analysis Guide | Quality engineers | 50+ validation rules, error prediction, React/Vue/Next.js analyzers, dependency analysis | Improve code quality with comprehensive tooling |
| Topic | Documentation | Target Use Case |
|---|---|---|
| 🧪 Rapid Prototyping | Spike Templates Guide | Learn template system, create prototypes |
| 🔗 Integration Patterns | Claude Code Integration | Understand MCP integration, customize workflow |
| 🏗️ Architecture Deep-Dive | Architecture Documentation | System design, deployment, scalability |
| 📈 Performance & Optimization | Performance Guide | Optimization strategies, monitoring |
| Support Channel | Response Time | Best For | Access |
|---|---|---|---|
| 🔧 Troubleshooting Guide | Immediate | Common issues, error resolution | Self-service |
| 🌐 Online Documentation | Immediate | Searchable docs, latest updates | Comprehensive reference |
Explore our comprehensive library knowledge base:
| Domain | Specification Count | Coverage | Access |
|---|---|---|---|
| 🎨 UI/UX Libraries | 15+ specifications | Component libraries, design systems, accessibility | Browse UI Specs |
| 🔐 Authentication | 8+ specifications | OAuth, JWT, session management, security | Browse Auth Specs |
| 🌐 Backend & APIs | 12+ specifications | REST, GraphQL, databases, real-time | Browse Backend Specs |
| 🧪 Testing & Quality | 10+ specifications | E2E testing, unit testing, performance | Browse Testing Specs |
| 🚀 Performance & State | 8+ specifications | Caching, state management, optimization | Browse Performance Specs |
# Clone and setup
git clone [local repository]
cd fluorite-mcp && npm install
# Development workflow
npm run dev # Start development server with hot reload
npm test # Run comprehensive test suite
npm run build # Build for production
npm run docs:dev # Start documentation server
# Quality assurance
npm run lint # Code linting and formatting
npm run typecheck # TypeScript type checking
npm run test:coverage # Test coverage analysis
Join our mission to democratize professional development patterns:
| Contribution Type | Impact | Getting Started | Recognition |
|---|---|---|---|
| 📚 Library Specifications | Help developers discover and use libraries effectively | Specification Guide | Listed in contributors, library attribution |
| 🧪 Spike Templates | Accelerate prototyping across the ecosystem | Template Creation Guide | Template attribution, featured examples |
| 🌍 Language Ecosystems | Expand multi-language support | Ecosystem Guide | Ecosystem maintainer recognition |
| 🔍 Static Analysis Rules | Improve code quality and error prevention | Analysis Rules Guide | Quality improvement impact metrics |
| 📖 Documentation | Enhance developer experience and adoption | Documentation Guide | Documentation contributor badge |
Fluorite MCPは、Claude Code CLIを強力なエンタープライズ開発ツールに変換する包括的なMCPサーバーです。プロフェッショナルグレードのコード生成、インテリジェントなエラー予防、高速プロトタイピングを通じて開発ワークフローを革新します。
プロフェッショナルに厳選された仕様で開発ワークフローを変革：
問題が発生する前にインテリジェントな解析で予防：
アイデアと実装の間のギャップを橋渡し：
チームと共にスケールするプロフェッショナルグレードツール：
# 1. Fluorite MCPをグローバルインストール
npm install -g fluorite-mcp
# 2. Claude Code CLIに追加
claude mcp add fluorite-mcp -- fluorite-mcp
# 3. インストール確認
claude mcp list
| リソース | 時間投資 | 学習内容 | 対象者 |
|---|---|---|---|
| ⚡ クイックスタートガイド | 5分 | インストール、基本コマンド、基本使用法 | 新規ユーザー |
| 🔧 インストールガイド | 15分 | 詳細セットアップ、トラブルシューティング、環境設定 | システム管理者、DevOps |
| 📖 コマンドリファレンス | 30分 | 完全なコマンド構文、高度な機能 | パワーユーザー、自動化 |
| リソース | 対象者 | 内容 | 価値 |
|---|---|---|---|
| 🔍 APIドキュメント | 開発者、統合担当者 | 完全なMCPプロトコルリファレンス、200以上の文書化された関数とTSDoc | カスタム統合の構築と内部APIの理解 |
| 👩💻 開発者ガイド | 貢献者、上級ユーザー | アーキテクチャ、包括的関数ドキュメント、貢献ガイドライン | 完全なAPI理解での拡張とカスタマイズ |
| 🏭 テンプレート作成ガイド | テンプレート作成者 | カスタムスパイクテンプレートとライブラリ仕様の作成 | 新しいフレームワーク/ライブラリの追加 |
| 📊 静的解析ガイド | 品質エンジニア | 50以上の検証ルール、エラー予測、React/Vue/Next.js解析器、依存関係解析 | 包括的ツールでコード品質を向上 |
| トピック | ドキュメント | 対象使用例 |
|---|---|---|
| 🧪 高速プロトタイピング | スパイクテンプレートガイド | テンプレートシステムを学び、プロトタイプを作成 |
| 🔗 統合パターン | Claude Code統合 | MCP統合を理解し、ワークフローをカスタマイズ |
| 🏗️ アーキテクチャ詳細 | アーキテクチャドキュメント | システム設計、デプロイ、スケーラビリティ |
| 📈 パフォーマンス・最適化 | パフォーマンスガイド | 最適化戦略、監視 |
要求: "react-hook-formとzodを使った検証付きのモダンなログインフォームを作成"
結果:
✅ 自動ライブラリ検出: react-hook-formとzodの仕様にアクセス
✅ 本番対応コード: TypeScript型を含む完全なコンポーネント
✅ アクセシビリティ準拠: ARIAラベル、キーボードナビゲーション、スクリーンリーダーサポート
✅ セキュリティベストプラクティス: 入力検証、XSS防止、安全なフォーム処理
✅ パフォーマンス最適化: 最小バンドル影響と最適化されたレンダリング
✅ テスト対応: テスト例とアクセシビリティテストパターンを含む
要求: "JWT認証とPydantic検証を含むFastAPIエンドポイントを作成"
結果:
✅ フレームワークネイティブパターン: FastAPIの慣例と慣用的パターンに従う
✅ 型安全性: 完全なTypeScript定義と適切な型推論とエラー予防
✅ セキュリティベストプラクティス: 一般的な脆弱性（XSS、CSRF、インジェクション攻撃）に対する組み込み保護
✅ パフォーマンス解析: バンドル影響評価と最適化推奨事項
要求: "チャートとメトリクスを含むレスポンシブダッシュボードを作成"
結果:
✅ プロフェッショナルダッシュボード: インタラクティブチャートとグラフ
✅ リアルタイムデータ更新: ライブデータフィードとリアルタイム更新
✅ モバイル対応設計: レスポンシブレイアウトとモバイル最適化
✅ エクスポート・フィルタリング機能: データエクスポートとフィルタリング機能
| ドメイン | 主要ライブラリ | 仕様数 | 使用例 |
|---|---|---|---|
| 🎨 UI/UX | shadcn/ui、Radix UI、Material-UI、Mantine、Tremor | 15以上 | コンポーネントライブラリ、デザインシステム、アクセシビリティ |
| 🔄 状態管理 | Zustand、Jotai、TanStack Query、Redux Toolkit、Valtio | 8以上 | アプリケーション状態、サーバー状態、リアルタイムデータ |
| 📝 フォーム・検証 | react-hook-form、Zod、Valibot、Yup、Formik | 6以上 | 型安全フォーム、スキーマ検証、ユーザー入力 |
| 🔐 認証 | NextAuth.js、Clerk、Lucia、Supabase Auth、Auth0 | 8以上 | ユーザー管理、OAuth、セッション処理 |
| 🧪 テスト・品質 | Playwright、Vitest、Cypress、Testing Library、Jest | 10以上 | E2Eテスト、ユニットテスト、アクセシビリティテスト |
| 📊 データ・可視化 | Recharts、D3.js、Chart.js、Tremor、TanStack Table | 8以上 | チャート、ダッシュボード、データテーブル、アナリティクス |
| 🌐 API・バックエンド | tRPC、Prisma、Drizzle、GraphQL、RESTパターン | 12以上 | 型安全API、データベース統合、リアルタイム |
| 🚀 パフォーマンス | React Query、SWR、最適化パターン、バンドリング | 6以上 | キャッシング、パフォーマンス監視、最適化 |
| 🎵 音声処理 | X-SAMPA/IPA変換、ToBI韻律注釈、音声合成・解析 | 3以上 | 音韻解析、韻律分析、多言語音声処理 |
| 言語 | フレームワーク・ツール | 仕様数 | 主要使用例 |
|---|---|---|---|
| 🟦 TypeScript/JavaScript | Next.js、React、Vue、Svelte、Astro | 25以上 | フルスタックWebアプリ、SPA、SSR |
| 🐍 Python | FastAPI、Django、Flask、Pydantic、SQLAlchemy | 12以上 | API、データサイエンス、機械学習、自動化 |
| 🦀 Rust | Tauri、Axum、Actix、Serde、Tokio | 8以上 | システムプログラミング、デスクトップアプリ、WebAssembly |
| 🐹 Go | Gin、Fiber、Echo、GORM、Chi | 6以上 | 高性能バックエンド、マイクロサービス、CLIツール |
| 💎 Ruby | Rails、Sinatra、Sidekiq、ActiveRecord | 4以上 | フルスタックWebアプリ、高速プロトタイピング |
| サポートチャンネル | 応答時間 | 最適な用途 | アクセス |
|---|---|---|---|
| 🔧 トラブルシューティングガイド | 即時 | 一般的な問題、エラー解決 | セルフサービス |
| 🌐 オンラインドキュメント | 即時 | 検索可能ドキュメント、最新更新 | 包括的リファレンス |
技術スタックを具体的に:
✅ "react-hook-form、Zod検証、shadcn/uiコンポーネントでフォームを作成"
❌ "フォームを作成"
フレームワークコンテキストに言及:
✅ "tRPCとPrismaでNext.js APIルートを構築"
❌ "APIを構築"
要件を含める:
✅ "ソート、フィルタリング、ページネーション、TypeScriptでデータテーブルを作成"
❌ "テーブルを作成"
ベストプラクティスを求める:
✅ "NextAuth.js実装のセキュリティベストプラクティスは何ですか？"
✅ "このReactコンポーネントのパフォーマンス最適化を教えて"
| Metric | Current Status | Quality Standard | Industry Benchmark |
|---|---|---|---|
| Library Coverage | 86+ specifications | 100% production-tested | Leading in MCP ecosystem |
| Language Ecosystems | 12+ comprehensive | Full-stack coverage | Most comprehensive available |
| Static Analysis Rules | 50+ framework-specific | Zero false positives | Advanced error prediction |
| Template Ecosystem | 6,200+ production templates | Enterprise-grade quality | Largest collection |
| Documentation Quality | 100+ pages, searchable | Professional technical writing | Award-winning clarity |
# 5-minute team setup
npm install -g fluorite-mcp
claude mcp add fluorite-mcp -- fluorite-mcp
fluorite-mcp --self-test
# Instant project analysis
fluorite-mcp --static-analysis --project-path . --framework auto-detect
# Start continuous evolution (optional)
~/fluorite-evolve-config.sh # Configure and start learning system
# Comprehensive project analysis with error prediction
fluorite-mcp --static-analysis --project-path /my/project --framework nextjs --predict-errors
# Quick validation during development
fluorite-mcp --quick-validate --file src/components/Button.tsx --framework react
# Security-focused analysis
fluorite-mcp --static-analysis --focus security --compliance-check
# Performance analysis
fluorite-mcp --static-analysis --focus performance --analyze-bundle-size
# Discover templates for your needs
fluorite-mcp --discover-spikes --query "React form validation"
# Auto-select and apply template based on natural language
fluorite-mcp --auto-spike --task "Create authentication system" --framework nextjs
# Apply specific template with parameters
fluorite-mcp --apply-spike nextjs-auth-nextauth-credentials --params '{"app_name":"MyApp"}'
# Preview template before applying
fluorite-mcp --preview-spike fastapi-jwt-auth --params '{"project_name":"MyAPI"}'
Note: In addition to thousands of file-based spikes under src/spikes, Fluorite MCP synthesizes 10k+ virtual spikes at runtime. These generated templates have IDs prefixed with gen- (e.g., gen-express-route-basic-ts) and also support Strike-branded IDs prefixed with strike- (e.g., strike-express-route-basic-ts). Use the environment variable FLUORITE_GENERATED_SPIKES_LIMIT to cap generation for testing, and FLUORITE_SPIKE_LIST_LIMIT to cap overall listing size.
To inspect current spike coverage and duplicates, use the new spike-stats tool (exposed via MCP): it reports total spikes, counts by source (file vs generated), duplicates, and a small sample of IDs.
Examples of generated spike IDs and their specialized files:
gen-nextjs-route-basic-ts: adds app/api/health/route.ts (Next.js Route Handler)gen-prisma-crud-basic-ts: adds src/prisma.ts and prisma/schema.prismagen-graphql-service-basic-ts: adds schema.graphql and src/graphql/resolvers.tsgen-next-auth-config-basic-ts: adds app/api/auth/[...nextauth]/route.tsgen-github-actions-config-basic-ts: adds .github/workflows/ci.yml# Pre-commit hook integration
git add -A && fluorite-mcp --quick-validate --staged-files
# CI/CD pipeline integration
fluorite-mcp --static-analysis --framework auto-detect --max-issues 0
# Development workflow with npm scripts
npm run fluorite:analyze && npm run build
# Real-time validation
fluorite-mcp --realtime-validation --file src/components/Form.tsx --watch-mode
# Configure and start the evolution system
~/fluorite-evolve-config.sh
# Evolution modes:
# - Normal Mode: Learn every hour (production)
# - Fast Mode: Learn every 5 minutes (testing)
# - Overnight Mode: Learn every 30 minutes (sleep time)
# Monitor evolution progress
tail -f ~/tmp/fluorite-evolution/logs/evolution-$(date +%Y%m%d).log
# View performance improvements
cat /Users/sware/Projects/github.com/kotsutsumi/fluorite-mcp/METRICS.md
# Next.js app with authentication
fluorite-mcp --auto-spike --task "Create Next.js app with NextAuth and Prisma"
# Analyze Next.js specific issues
fluorite-mcp --static-analysis --framework nextjs --enabled-rules nextjs-hydration,nextjs-server-components
# React component with form validation
fluorite-mcp --auto-spike --task "Create React form with React Hook Form and Zod validation"
# React performance analysis
fluorite-mcp --static-analysis --framework react --focus performance --enabled-rules react-hooks-deps
# FastAPI with JWT auth and database
fluorite-mcp --auto-spike --task "Create FastAPI with JWT authentication and PostgreSQL"
# Python code analysis
fluorite-mcp --static-analysis --framework fastapi --language python --focus security
// .vscode/tasks.json - VS Code integration
{
"version": "2.0.0",
"tasks": [
{
"label": "Fluorite: Analyze Current File",
"type": "shell",
"command": "fluorite-mcp",
"args": ["--quick-validate", "--file", "${file}", "--framework", "auto-detect"]
},
{
"label": "Fluorite: Discover Templates",
"type": "shell",
"command": "fluorite-mcp",
"args": ["--discover-spikes", "--query", "${input:spikeQuery}"]
}
]
}
// package.json - npm scripts integration
{
"scripts": {
"fluorite:analyze": "fluorite-mcp --static-analysis --project-path . --framework auto-detect",
"fluorite:validate": "fluorite-mcp --quick-validate --framework auto-detect",
"fluorite:spikes": "fluorite-mcp --discover-spikes",
"precommit": "fluorite-mcp --quick-validate --staged-files",
"prebuild": "npm run fluorite:validate"
}
}
📚 More Examples: For comprehensive real-world scenarios including:
- Team development workflows with startup, enterprise, and multi-vendor setups
- Advanced CI/CD integration patterns with blue-green deployments
- IDE and editor integrations (VS Code, Cursor, etc.)
- Large-scale monorepo management strategies
- Production troubleshooting and performance monitoring
See our detailed Use Cases & Examples Guide
graph TB
A[Claude Code CLI] --> B[MCP Protocol Layer]
B --> C[Fluorite MCP Server]
C --> D[Library Specification Engine]
C --> E[Static Analysis Engine]
C --> F[Spike Template System]
C --> G[Intelligent Code Generation]
D --> H[86+ YAML Specifications]
E --> I[50+ Validation Rules]
F --> J[2500+ Production Templates]
G --> K[Framework-Native Output]
H --> L[TypeScript/JavaScript]
H --> M[Python/FastAPI]
H --> N[Rust/Go/Other Languages]
style A fill:#e1f5fe
style C fill:#f3e5f5
style K fill:#e8f5e8
fluorite-mcp/ # Enterprise MCP server
├── 📁 src/catalog/ # Library specifications (86+ YAML files)
├── 📁 src/core/ # Core MCP functionality & analysis engine
├── 📁 src/spikes/ # Spike template ecosystem (6,200+ templates)
├── 📁 memory-engine/ # Rust-based intelligent memory system
│ ├── 📁 fluorite-memory/ # Core chunk storage implementation
│ ├── 📁 fluorite-ml/ # ML integration with PyO3
│ ├── 📁 fluorite-bridge/ # Node.js N-API bridge
│ └── 📁 fluorite-learner/ # Learning pipeline
├── 📁 performance-reports/ # Auto-generated performance metrics
├── 📁 docs/ # VitePress documentation site
│ ├── 📁 specs/ # Specification documentation
│ ├── 📁 architecture/ # Technical architecture docs
│ └── 📁 guides/ # User and developer guides
├── 📁 dist/ # Compiled production artifacts
├── 📄 API.md # Complete API reference
├── 📄 METRICS.md # Real-time performance metrics
└── 📄 package.json # Enterprise package configuration
Fluorite MCP features extensively documented APIs with professional-grade TSDoc coverage:
| Component | Functions | Documentation Coverage | Key Features |
|---|---|---|---|
| Static Analysis Handlers | 5 core functions | Complete TSDoc with examples | Comprehensive project analysis, quick validation, real-time validation |
| React Analyzer | 10+ validation rule categories | Full class and method documentation | Hooks validation, performance optimization, security checks |
| Vue Analyzer | 8+ validation rule categories | Complete API documentation | Composition API validation, template syntax, reactivity system |
| Dependency Analyzer | 6+ core analysis functions | Comprehensive documentation | Version conflicts, security vulnerabilities, circular dependencies |
| Component | Functions | Documentation Coverage | Key Features |
|---|---|---|---|
| Setup Command | 7 documented functions | Complete implementation guide | Auto-detection, MCP registration, command mapping, verification |
| Command Generation | 17+ enhanced commands | Full parameter documentation | Tool permission mapping, natural language processing |
import { handleStaticAnalysisTool } from './src/core/static-analysis-handlers.js';
// Comprehensive project analysis with framework detection
const result = await handleStaticAnalysisTool({
projectPath: './my-nextjs-app',
framework: 'nextjs',
predictErrors: true,
analyzeDependencies: true,
autoFix: true
});
console.log(result.metadata.issuesFound); // Number of issues detected
console.log(result.metadata.autoFixes); // Generated auto-fixes
import { ReactAnalyzer } from './src/core/react-analyzer.js';
import { VueAnalyzer } from './src/core/vue-analyzer.js';
// React component validation
const reactAnalyzer = new ReactAnalyzer();
const reactRules = reactAnalyzer.getRules();
// Returns 10+ rule categories: hooks, state, performance, security, etc.
// Vue.js component validation
const vueAnalyzer = new VueAnalyzer();
const vueRules = vueAnalyzer.getRules();
// Returns 8+ rule categories: composition API, template, lifecycle, etc.
import { setupCommand } from './src/cli/commands/setup.js';
// Programmatic setup with configuration
await setupCommand.action({
force: true, // Override existing configuration
dryRun: false // Actually perform setup operations
});
// Automatically configures Claude Code CLI integration
MIT License - Open source with commercial-friendly terms
Copyright: 2024 Fluorite MCP Contributors
Full License: LICENSE file contains complete terms
All library specifications include proper attribution and licensing information from original maintainers.
| Channel | Description | Purpose | Update Frequency |
|---|---|---|---|
| 📦 NPM Registry | npmjs.com/package/fluorite-mcp | Primary distribution | Every release |
| 🐙 GitHub Repository | Source code repository | Source code & issues | Continuous |
| 📖 Documentation Site | Complete documentation | Documentation reference | Weekly |
| Platform | Description | Best For | Response Time |
|---|---|---|---|
| 📈 Roadmap | GitHub Projects | Development planning | Monthly updates |
node dist/cli/index.js spikes synth --filter '^strike-' --max 50node dist/cli/index.js spikes synth --filter '^strike-' --max 12000 --write --prettyFLUORITE_GENERATED_SPIKES_LIMIT: Cap number of virtual generated spikesFLUORITE_SPIKE_LIST_LIMIT: Cap total returned IDs in listingNote: synth は仮想生成テンプレート（gen-/strike-）を列挙し、必要に応じて src/spikes/ に JSON を書き出します（--write 指定時のみ）。
list-generated-spikes ツールで strike/gen の仮想IDを高速列挙できます。
list-generated-spikes { libs: ["nextjs","fastapi"], patterns: ["route"], styles: ["typed","secure"], langs: ["ts","py"], limit: 50, prefix: "strike" }npm run materialize:spikes -- --libs nextjs,fastapi --patterns route --styles typed,secure --langs ts,py --limit 500 --prefix strikenode dist/cli/index.js spikes synth --filter '^strike-' --max 1000--overwrite を明示--merge-metadata を利用Materialized: written=..., overwritten=..., skipped=...node dist/cli/index.js spikes generate src/cli/data/strike-batch-XX.json --prefix strike --out-dir src/spikes --merge-metadata --dry-runnode dist/cli/index.js spikes generate src/cli/data/strike-batch-XX.json --prefix strike --out-dir src/spikes --merge-metadataFluorite MCP: Where natural language meets production-ready code generation
Featuring 200+ documented functions with comprehensive TSDoc coverage
Developed with ❤️ by the open source community • Enterprise support available
Get Started in 2 Minutes | View Documentation | API Reference
FAQs
Enterprise-grade MCP server for Claude Code CLI with 6,200+ spike templates, 86 library specs, intelligent memory engine, and continuous learning capabilities.
The npm package fluorite-mcp receives a total of 568 weekly downloads. As such, fluorite-mcp popularity was classified as not popular.
We found that fluorite-mcp demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 1 open source maintainer collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Product
Socket’s new Tier 1 Reachability filters out up to 80% of irrelevant CVEs, so security teams can focus on the vulnerabilities that matter.
Research
/Security News
Ongoing npm supply chain attack spreads to DuckDB: multiple packages compromised with the same wallet-drainer malware.
Security News
The MCP Steering Committee has launched the official MCP Registry in preview, a central hub for discovering and publishing MCP servers.