Idea → Project: Full Automated Pipeline

You are running the complete pipeline from raw idea to coding-ready project. This skill chains 3 phases automatically — the user gives an idea and walks away with everything.

FLOW RULE: Phases run automatically EXCEPT at defined "Decision Points" where you MUST pause and ask the user. Between decision points, keep working without interruption. Do NOT ask "devam edelim mi?" — only stop at the marked 🔷 Decision Points below.

Pipeline Overview

User: "Şu fikrim var: ..."
         │
         ▼
╔══════════════════════════════════════╗
║  PHASE 1: IDEA CAPTURE              ║
║  Fikri anla → sorular sor           ║
║  🔷 Kullanıcı onayı                 ║
╠══════════════════════════════════════╣
║  PHASE 2: COMPETITIVE INTEL         ║
║  20 rakip bul → otomatik analiz     ║
║  🔷 Rakip tech stack sunumu →       ║
║     Kullanıcıdan tech stack kararı  ║
║  → Kalan analizler otomatik         ║
║  → HTML + DOCX + Excel üret         ║
╠══════════════════════════════════════╣
║  PHASE 3: PROJECT BLUEPRINT         ║
║  🔷 Mimari desen + state mgmt       ║
║     seçimi (önerilerle)             ║
║  🔷 Modül haritası onayı            ║
║  → PRD + DB + API + rules otomatik  ║
╚══════════════════════════════════════╝
         │
         ▼
User: Masaüstünde tam proje klasörü hazır

Decision Points Summary

Pipeline'da tam 4 karar noktası var — sadece buralarda kullanıcıya sorulur:

#	Aşama	Ne Sorulur
🔷1	Phase 1 sonu	Fikir özeti doğru mu?
🔷2	Phase 2 ortası	Tech stack kararı (rakiplerin teknolojileri gösterilir, öneriler sunulur)
🔷3	Phase 3 başı	Mimari desen + state management + UI kit seçimi
🔷4	Phase 3 ortası	Modül haritası ve sprint planı onayı

---

PHASE 1: Idea Capture

Goal: Understand the idea quickly, produce a structured brief.

Step 1.1: First Contact

Read the user's message. Then ask ONE round of focused questions (max 5):

1. Ürün ne yapıyor? — Tek cümleyle (skip if already clear)
2. Hedef kitle kim? — B2B/B2C, meslek grubu, ülke/bölge
3. Teknoloji tercihi? — Flutter, React, Python vb. (yoksa sen öner)
4. Platform? — Web / Mobile / Desktop / Hepsi
5. Mevcut kod tabanı var mı? — Sıfırdan mı?

While waiting for answers, research:

• Search the domain for 5-10 obvious competitors
• Note market category and common features
• Identify popular tech stack for this category

Step 1.2: Expand & Confirm

Present a quick summary:

Anladığım kadarıyla:
• Ürün: [1 cümle]
• Kategori: [SaaS / Mobile / Platform]
• Hedef: [Kitle + pazar]
• Platform: [Web / Mobile]
• Tech stack: [Frontend + Backend + DB]
• Tahmini modül sayısı: X
• Bilinen rakipler: [5-10 isim]

Doğru mu? (Doğruysa tüm pipeline başlıyor — rakip analizi + proje planı otomatik çalışacak)

Step 1.3: Generate Business Brief

After confirmation, produce the structured brief internally (keep in memory for Phase 2 and 3). Also save as [ProductName]_Business_Brief.md.

Brief structure:

# Business Brief: [Ürün Adı]

## 1. Ürün Özeti
- Tek cümle, kategori, hedef pazar, dil, platform, aşama

## 2. Problem & Vizyon
- 2-3 paragraf

## 3. Özellik Haritası
- Modül → Özellik tabloları (öncelik + karmaşıklık)
- Toplam istatistikler

## 4. Teknoloji Mimarisi
- Katman → Teknoloji → Neden

## 5. Hedef Kitle
- Persona'lar

## 6. İş Modeli
- Fiyatlandırma tablosu

## 7. Rekabet Ortamı (Ön Keşif)
- Bilinen rakipler tablosu

## 8. Sonraki Adımlar

Immediately proceed to Phase 2. Do NOT wait for user input.

---

PHASE 2: Competitive Intelligence

Goal: Full competitor analysis producing HTML dashboards, DOCX reports, and Excel workbook.

Setup

npm install docx --prefix /sessions/laughing-keen-darwin/.npm-global 2>/dev/null
pip install openpyxl --break-system-packages 2>/dev/null

Step 2.1: Competitor Discovery

Using the brief from Phase 1, web search for competitors:

• Direct competitors — Same market, similar features
• Partial competitors — Overlapping features, different focus
• Adjacent competitors — Related market, could expand
• Local competitors — Region-specific alternatives

Rules:

• Maximum 20 competitors (pick top 20 by market presence)
• For each: name, website, category, founded, users, revenue, pricing

Step 2.2: Feature Matrix

Create a comprehensive feature matrix comparing all competitors:

• List ALL features from the brief's feature map
• For each competitor: research which features they have
• Values: "Tam" (full), "Kısmi" (partial), "Yok" (missing)
• Calculate completeness scores

Step 2.3: Gap Analysis & AI Opportunities

• Identify features competitors have that user's product is missing
• Identify features NO competitor has (blue ocean opportunities)
• Map AI/automation opportunities for each module

Step 2.4: Target Market Analysis

Using competitor data + brief's personas:

• Refine market segments
• Estimate market sizes
• Identify underserved segments

Step 2.5: Pricing Strategy

• Research all competitor pricing pages
• Build pricing comparison table
• Recommend package structure and pricing
• Revenue projections

Step 2.6: Competitor Complaints

• Search: "[competitor] complaints", "switched from [competitor]", Reddit, G2, Trustpilot, Şikayetvar
• Categorize complaints by theme
• Identify strategic opportunities from pain points

Step 2.7: Brand & Name Suggestions

• 30 name ideas inspired by competitor gap analysis
• Color palette suggestions based on competitor brand positioning
• Tagline/slogan ideas

Step 2.8: Tech Stack & Library Analysis — Competitor Detection

First, detect competitor tech stacks:

• Search StackShare, BuiltWith, job postings, engineering blogs, GitHub orgs
• Research competitor UI kits and design systems (Storybook, Figma, component libraries)

🔷 Decision Point 2: Tech Stack Kararı

After competitor tech detection, STOP and present to user:

📊 Rakiplerin Teknoloji Haritası:

| Rakip | Frontend | Mobile | Backend | Database | Cloud | AI | UI Kit |
|-------|----------|--------|---------|----------|-------|----|--------|
| Comp1 | React    | RN     | Node.js | Postgres | AWS   | OpenAI | MUI |
| Comp2 | Vue      | Flutter| Django  | MongoDB  | GCP   | Custom | Vuetify |
| Comp3 | Angular  | Native | Go      | Postgres | AWS   | - | Custom |
| ...   | ...      | ...    | ...     | ...      | ...   | ... | ... |

📌 Senin mevcut tercihin: [Phase 1'den gelen tech stack]

💡 Önerilerim:
• Frontend/Mobile: [Öneri + neden — popülerlik, ekosistem, rakip trendleri]
• Backend: [Öneri + neden]
• Database: [Öneri + neden]
• AI/ML: [Öneri + neden]
• UI Kit: [Öneri + neden — rakiplerin kullandığı vs alternatifler]

Seçeneklerin:
1. Mevcut tercihlerinle devam et ✅
2. Önerilerimi kabul et
3. Karışık seçim yap (hangilerini değiştirmek istediğini söyle)

Wait for user's tech stack decision. After they confirm:

• Finalize the tech stack
• Recommend libraries/packages for the chosen stack (17 categories)
• Recommend UI kits for the chosen framework (component count, theming, dark mode, accessibility)
• Then continue automatically to 2.9

Step 2.9: Generate All CI Deliverables

Read the generation script templates from this skill's references/ and produce:

1. [ProductName]_Tam_Rapor.html — Master interactive dashboard (dark theme)

   • All sections combined: competitors, features, gaps, AI, pricing, complaints, brand, tech
   • Embedded data as JS variables, no external dependencies
   • Interactive tabs, filters, search

2. [ProductName]_Strateji_Raporu.docx — Strategy report

   • Feature matrix, gap analysis, AI opportunities, target segments

3. [ProductName]_Fiyatlandirma_Raporu.docx — Pricing report

   • Competitor pricing, packages, revenue projections

4. [ProductName]_Analiz.xlsx — Multi-sheet Excel (11 sheets)

   • Feature matrix, AI, segments, gaps, pricing, packages, local pricing, revenue, tech stack, libraries, UI kits

5. [ProductName]_Sikayetler.html — Complaints analysis

6. [ProductName]_Marka_Onerileri.html — Brand suggestions

7. [ProductName]_Tech_Stack.html — Tech stack report

Immediately proceed to Phase 3. Do NOT wait for user input.

---

PHASE 3: Project Blueprint

Goal: Using the brief (Phase 1) and CI results (Phase 2), generate a complete project scaffold.

🔷 Decision Point 3: Mimari & Teknoloji Seçimleri

Before generating anything, STOP and present architecture options:

Based on the chosen tech stack (from Decision Point 2) and competitor analysis, present:

🏗️ Mimari Kararlar

1. MİMARİ DESEN
   Rakiplerin yaklaşımları:
   • [Comp1]: Clean Architecture + BLoC
   • [Comp2]: MVVM + Provider
   • [Comp3]: Feature-first + GetX

   Seçeneklerin:
   a) Clean Architecture (Önerilen ✅) — Katmanlı, test edilebilir, büyük projeler için ideal
   b) MVVM — Daha basit, orta ölçek projeler
   c) Hexagonal — Domain-heavy, iş kuralları ağır projeler

2. STATE MANAGEMENT
   Rakipler:
   • [Comp1]: Redux / Zustand
   • [Comp2]: BLoC
   • [Comp3]: Riverpod

   Seçeneklerin:
   a) Riverpod (Önerilen ✅) — Modern, compile-safe, test kolay
   b) BLoC — Battle-tested, event-driven
   c) GetX — Kolay ama büyük projelerde karmaşık

3. UI KİT / DESIGN SYSTEM
   Rakiplerin kullandığı UI kitler:
   • [Comp1]: Custom design system
   • [Comp2]: Material + shadcn/ui
   • [Comp3]: Ant Design

   Seçeneklerin:
   a) [Framework-specific top recommendation] (Önerilen ✅)
   b) [Alternative 1]
   c) [Alternative 2]
   d) Custom design system (sıfırdan)

4. BACKEND YAKLAŞIMI
   a) Supabase (BaaS) — Hızlı, RLS, real-time built-in
   b) Custom API — Tam kontrol, daha fazla iş
   c) Hibrit — Supabase + Edge Functions for complex logic

Hangilerini tercih ediyorsun? (Numara + harf ile cevapla, örn: "1a, 2a, 3a, 4a")

Wait for user's architecture decisions. Then proceed.

Step 3.1: Create Blueprint Folder

Create [ProductName]_Blueprint/ with all files below.

Step 3.2: Generate Overview & PRDs

00_OVERVIEW.md — Reading guide, file map, tech summary

01_HIGH_LEVEL_PRD.md — Business-facing:

• Product vision, personas, user stories (US-MODULE-NUM format)
• Module map, success KPIs, competitive advantage, scope exclusions

02_TECHNICAL_PRD.md — Engineering-facing:

• System architecture, tech decisions with rationale
• Module technical requirements, non-functional requirements
• Integration points, data flow diagrams

Step 3.3: Architecture & Database

03_ARCHITECTURE.md (use the user's choices from Decision Point 3):

• Chosen architecture pattern with WHY
• Layer diagram, dependency rules
• Chosen state management with patterns & examples
• Chosen UI kit integration guide
• Error handling, caching, offline strategy

04_DATABASE_SCHEMA.md:

• Every table: columns, types, nullable, defaults, descriptions
• Indexes with rationale
• Foreign keys, relationships
• RLS policies (for Supabase)
• ER diagram (Mermaid syntax)
• Migration strategy, seed data

Read references/database_patterns.md for schema design patterns.

🔷 Decision Point 4: Modül Haritası & Sprint Planı Onayı

Before generating API contracts and module specs, STOP and present the module map:

📦 Modül Haritası — [ProductName]

| # | Modül | Özellik Sayısı | Ekran | DB Tablo | Sprint | Karmaşıklık |
|---|-------|---------------|-------|----------|--------|-------------|
| 1 | auth | 5 | 3 | 2 | Sprint 0 | Orta |
| 2 | dashboard | 8 | 2 | 1 | Sprint 1 | Orta |
| 3 | [module] | X | X | X | Sprint X | High/Med/Low |
| ... | ... | ... | ... | ... | ... | ... |

Toplam: X modül, Y özellik, Z ekran, W tablo

📅 Sprint Planı:
• Sprint 0 (1 hafta): Altyapı + Auth
• Sprint 1 (2 hafta): [Modules]
• Sprint 2 (2 hafta): [Modules]
• Sprint 3 (2 hafta): [Modules]
• Sprint 4 (1 hafta): Polish + Launch

Tahmini toplam süre: ~X hafta

Bu modül haritası ve sprint planı uygun mu?
Eklemek/çıkarmak/birleştirmek istediğin modüller var mı?

Wait for user confirmation. Adjust if needed, then continue generating all remaining files automatically.

Step 3.4: API & Microservices

05_API_CONTRACTS.md:

• Every endpoint: method, path, description, auth, rate limit
• Request/response JSON schemas
• Error codes

06_MICROSERVICE_MAP.md (based on confirmed module map):

• Each feature → isolated micro-module
• Responsibility, dependencies, tables, endpoints, screens, complexity
• Internal folder structure per module

Step 3.5: Frontend & Backend Structure

07_FRONTEND_STRUCTURE.md:

• Complete folder tree with explanations
• Routing, theme system, responsive design, shared components

08_BACKEND_STRUCTURE.md:

• Folder tree (Supabase or custom backend)
• Edge functions, auth strategy, background jobs, file storage

Read references/architecture_patterns.md for framework-specific structures.

Step 3.6: Rules & Testing

09_CODING_RULES.md:

• General principles (DRY, SOLID, KISS, YAGNI)
• Naming conventions table
• File organization rules
• Git commit rules (conventional commits)
• Code review checklist
• Linting config

10_TESTING_STRATEGY.md:

• Test pyramid: Unit 70% / Integration 20% / E2E 10%
• Test naming convention
• Module test matrix
• Mock strategy, coverage goals

Step 3.7: Multi-Agent & Roadmap

11_MULTI_AGENT_ARCHITECTURE.md:

• Agent roles: Architect → Module Agents (parallel) → Integration → QA → DevOps
• Agent isolation rules
• Communication protocol
• Workflow diagram

12_IMPLEMENTATION_ROADMAP.md:

• Sprint 0: Infrastructure
• Sprint 1-N: Feature development (ordered by priority)
• Each sprint: checklist of deliverables

13_DEVOPS_PIPELINE.md:

• Environments (dev/staging/prod)
• CI pipeline, branch strategy
• Environment variables list
• Monitoring setup

Step 3.8: CLAUDE.md & README

CLAUDE.md — Master AI agent guide:

• Project summary, architecture, folder structure
• Top coding rules, module structure
• How to add new modules
• Common mistakes, commands

README.md — Standard project README

Step 3.9: Per-Module Specs

For EACH module from the feature map, create modules/[module_name]/:

SPEC.md — Module description, user stories, screens, state management, API endpoints, dependencies

CLAUDE.md — AI agent context for this module: what it does, key files, data flow, do's/don'ts, implementation steps

SCHEMA.md — Tables owned by this module, relationships, migration SQL, RLS, seed data

---

Output Delivery

Folder Structure

Save everything under the workspace folder:

[ProductName]/
├── analysis/                        # Phase 2 outputs
│   ├── [ProductName]_Tam_Rapor.html
│   ├── [ProductName]_Strateji_Raporu.docx
│   ├── [ProductName]_Fiyatlandirma_Raporu.docx
│   ├── [ProductName]_Analiz.xlsx
│   ├── [ProductName]_Sikayetler.html
│   ├── [ProductName]_Marka_Onerileri.html
│   └── [ProductName]_Tech_Stack.html
├── blueprint/                       # Phase 3 outputs
│   ├── 00_OVERVIEW.md
│   ├── 01_HIGH_LEVEL_PRD.md
│   ├── ... (all 14 files)
│   ├── CLAUDE.md
│   ├── README.md
│   └── modules/
│       ├── auth/
│       ├── dashboard/
│       └── [each module]/
└── [ProductName]_Business_Brief.md  # Phase 1 output

Final Message

After everything is generated, present the complete output to the user:

✅ Pipeline tamamlandı! İşte [ProductName] için hazırlanan dosyalar:

📋 Business Brief — Fikrin detaylı açıklaması
📊 Rakip Analizi — X rakip, Y özellik karşılaştırması
   • Interaktif HTML Dashboard
   • Strateji Raporu (DOCX)
   • Fiyatlandırma Raporu (DOCX)
   • Detaylı Excel (11 sheet)
   • Şikayet Analizi + Marka Önerileri + Tech Stack
🏗️ Proje Blueprint — Kodlamaya hazır mimari
   • 14 ana dosya (PRD, DB şema, API, rules, roadmap...)
   • X modül için ayrı spec dosyaları
   • Multi-agent geliştirme planı
   • CLAUDE.md + README.md

Tüm dosyalar: [klasör linki]

---

Reference Files

• references/database_patterns.md — DB schema design patterns, Supabase RLS, common schemas
• references/architecture_patterns.md — Flutter Clean Arch, Next.js App Router, multi-agent patterns

---

Rules

1. Language: Use the user's language throughout (Turkish → Turkish, English → English)
2. No placeholders: Every section must have real, researched data
3. No stopping: Run all 3 phases automatically after Phase 1 confirmation
4. Quality: Every file should be production-grade — a developer should be able to start coding immediately
5. Consistency: Tech stack, module names, and table names must be consistent across ALL files