@dzhechkov/p-replicator 1.0.0

package/templates/.claude/skills/reverse-engineering-unicorn/SKILL.md

@@ -0,0 +1,194 @@
+---
+name: reverse-engineering-unicorn
+description: Reverse engineer any company into a launch playbook with clickable CJM prototype. Triggers on "проанализируй компанию", "reverse engineer", "разбери бизнес-модель", "playbook запуска".
+---
+
+# Reverse Engineering Unicorn v2
+
+Модульный reverse engineering компаний для создания actionable launch playbook.
+7 модулей × 3 режима глубины × checkpoints × кликабельный CJM прототип.
+
+## When To Activate
+
+Trigger on:
+- "проанализируй компанию [X]"
+- "reverse engineer [X]"
+- "разбери бизнес-модель [X]"
+- "сделай playbook запуска аналога [X]"
+- "проанализируй стартап [X]"
+
+## Architecture
+
+```
+SKILL.md (this file — orchestrator)
+├── modules/
+│   ├── 01-intelligence.md            → Verified Fact Sheet
+│   ├── 02-product-customers.md       → JTBD + Segments + Voice of Customer
+│   ├── 025-cjm-prototype.md          → 3 CJM variants → clickable React prototype
+│   ├── 03-market-competition.md      → TAM/SAM + Game Theory + TRIZ
+│   ├── 04-business-finance.md        → Unit Economics + P&L + Sensitivity
+│   ├── 05-growth-engine.md           → Growth Loop + Moats + 2nd-Order
+│   └── 06-playbook-synthesis.md      → 90-Day Plan + BS-check
+├── references/
+│   ├── jtbd-canvas.md                → JTBD template
+│   ├── blue-ocean-canvas.md          → Strategy Canvas template
+│   └── industry-benchmarks.md        → Unit economics benchmarks
+└── examples/
+    ├── noom-module1-example.md        → Few-shot for Module 1
+    └── noom-cjm-example.md            → Few-shot: .jsx structure for CJM
+```
+
+## External Skills (loaded via view() in DEEP/VERIFIED modes)
+
+| Skill | Modules | Purpose |
+|-------|---------|---------|
+| `explore` | Pre-flight | Clarify vague requests |
+| `goap-research-ed25519` | M1-M5 (DEEP) | Adaptive research: A* + OODA |
+| `goap-research-ed25519` | M1-M5 (VERIFIED) | + Ed25519 crypto verification |
+| `problem-solver-enhanced` | M3,M4,M5 (DEEP) | Game Theory, TRIZ, 2nd-Order |
+| `frontend-design` | M2.5 (DEEP) | Production-grade CJM prototype design |
+| `brutal-honesty-review` | M6 (DEEP) | BS-detection quality gate |
+| `presentation-storyteller` | Post-M6 | Pitch deck generation |
+| `idea2prd-manual` | Post-M6 | PRD for development |
+| `md2pptx` | Post-M6 | Convert .md → .pptx |
+
+## Three Modes
+
+| Mode | Research | Analysis | Confidence | Time |
+|------|----------|----------|------------|------|
+| 🟢 QUICK | Static queries | Templates only | Manual X/5 | ~70 min |
+| 🔵 DEEP | GOAP A*+OODA | +GT, TRIZ, 2nd-Order, CJM proto, BS-check | Formula | ~140 min |
+| 🟣 VERIFIED | GOAP+Ed25519 | +Crypto signatures, audit trail | +Trusted issuers | ~170 min |
+
+## Pipeline
+
+```
+INPUT: {COMPANY}, {URL}, {INDUSTRY}, {GEOGRAPHY}, {MODE}
+                      ↓
+  MODULE 0: PRE-FLIGHT (if variables unclear → explore skill)
+  ⏸️ CHECKPOINT 0
+                      ↓
+  MODULE 1: INTELLIGENCE → Verified Fact Sheet
+  ⏸️ CHECKPOINT 1
+                      ↓
+  MODULE 2: PRODUCT & CUSTOMERS → JTBD + Segments + Voice
+  ⏸️ CHECKPOINT 2
+                      ↓
+  MODULE 2.5: CJM PROTOTYPE → 3 variants as clickable React app
+    • User explores A/B/C variants
+    • User compares (⚖️), builds custom mix (🔧), previews (👁)
+    • User locks winning CJM (✅) → {CHOSEN_CJM}
+  ⏸️ CHECKPOINT 2.5
+                      ↓
+  MODULE 3: MARKET & COMPETITION → TAM/SAM + Game Theory + TRIZ
+    (uses {CHOSEN_CJM} for competitive Aha comparison)
+  ⏸️ CHECKPOINT 3
+                      ↓
+  MODULE 4: BUSINESS & FINANCE → Unit Economics + P&L
+    (uses {CHOSEN_CJM} paywall timing for revenue model)
+  ⏸️ CHECKPOINT 4
+                      ↓
+  MODULE 5: GROWTH ENGINE → Loop + Channels + Moats
+    (uses {CHOSEN_CJM} core loop for growth strategy)
+  ⏸️ CHECKPOINT 5
+                      ↓
+  MODULE 6: PLAYBOOK SYNTHESIS → 90-Day Plan + BS-check
+    (MVP scope = {CHOSEN_CJM} screens)
+  ⏸️ CHECKPOINT 6 (FINAL)
+                      ↓
+  POST: presentation-storyteller / md2pptx / idea2prd-manual
+```
+
+## Execution Protocol
+
+### Step 0: Extract Variables
+
+| Variable | Description | Example |
+|----------|-------------|---------|
+| `{COMPANY}` | Company name | Calendly |
+| `{URL}` | Website | https://calendly.com |
+| `{INDUSTRY}` | Industry | Scheduling / Productivity |
+| `{GEOGRAPHY}` | Launch geography | Россия |
+| `{MODE}` | Analysis mode | QUICK / DEEP / VERIFIED |
+
+If variables unclear → `view` the `explore` skill.
+
+### Steps 1-6: Sequential Module Execution
+
+For each module:
+1. `view` the module file from `modules/`
+2. Substitute {COMPANY}, {URL}, {INDUSTRY}, {GEOGRAPHY}, {MODE}
+3. Execute Research Protocol for selected {MODE}
+4. In DEEP/VERIFIED — `view()` external skills listed in module
+5. `view` any `references/*.md` or `examples/*.md` referenced
+6. Generate structured output per module template
+7. Show CHECKPOINT, wait for user command
+8. On "ок" → next module
+
+### Module 2.5 Special: CJM Prototype
+
+This module **GENERATES a new React .jsx file** for every company.
+It does NOT show a hardcoded Noom prototype.
+
+Flow:
+1. Extract CJM building blocks from M2 output
+2. Generate 3 variant hypotheses (different Aha × Entry × Paywall)
+3. Write .jsx with variant switcher, comparison, builder, scoring
+4. User explores → customizes → locks winning CJM
+5. `{CHOSEN_CJM}` feeds into M3-M6
+
+The `examples/noom-cjm-example.md` is a few-shot for CODE STRUCTURE only.
+
+### Step 7: Deliver
+
+After CHECKPOINT 6:
+1. Compile all outputs into one document
+2. Create .md file, present to user
+3. Offer next steps: pitch deck / PRD / pptx
+
+## Checkpoint Commands (all modules)
+
+| Command | Action |
+|---------|--------|
+| `ок` / `далее` | Next module |
+| `глубже [topic]` | Additional research |
+| `скорректируй [what]` | Fix output |
+| `переключи на DEEP` | Rebuild in DEEP mode |
+| `переключи на VERIFIED` | Rebuild with Ed25519 |
+| `пропусти` | Skip to next |
+| `стоп` | Save state |
+| `сначала` | Restart current module |
+
+## Module 2.5 Extra Commands
+
+| Command | Action |
+|---------|--------|
+| `выбираю A/B/C` | Lock variant → M3 |
+| `собираю свой` | Open builder (mix & match) |
+| `ещё вариант: [идея]` | Add variant D |
+| `объедини A+C` | Hybrid variant |
+| `другой Aha в B: [что]` | Rebuild variant B |
+| `перерисуй [screen]` | Change one screen |
+| `без overlay` | Clean prototype for showing clients |
+
+## Anti-Hallucination Rules (ALL modes)
+
+1. **Search First** — never answer from memory for facts
+2. **Source Attribution** — every fact → URL
+3. **НЕ НАЙДЕНО > fabrication** — no data = "НЕ НАЙДЕНО"
+4. **Hypotheses marked** — `[H]` tag on unverified claims
+5. **Confidence Score** — end of every module
+
+## Module → Skill Mapping
+
+```
+M0:   explore
+M1:   goap-research-ed25519
+M2:   goap-research-ed25519 + references/jtbd-canvas.md
+M2.5: frontend-design + examples/noom-cjm-example.md (GENERATES new .jsx)
+M3:   goap-research-ed25519 + problem-solver-enhanced [GT, TRIZ, 2nd-Order]
+M4:   goap-research-ed25519 + problem-solver-enhanced [FP, TRIZ] + references/industry-benchmarks.md
+M5:   goap-research-ed25519 + problem-solver-enhanced [2nd-Order, TRIZ]
+M6:   brutal-honesty-review [Bach + Ramsay]
+Post: presentation-storyteller / md2pptx / idea2prd-manual
+```

package/templates/.claude/skills/reverse-engineering-unicorn/examples/noom-cjm-example.md

@@ -0,0 +1,100 @@
+# Few-Shot: CJM Prototype (Noom → НутриМайнд)
+
+> **Это пример для Claude, не для пользователя.**
+> Используй как reference при генерации .jsx для ЛЮБОЙ компании.
+> Копируй СТРУКТУРУ КОДА, а не контент.
+
+## Пример: как M2 output превращается в 3 варианта
+
+### Input из M2 (Noom case)
+
+```yaml
+company: Noom
+one_liner: "Weight Watchers + когнитивно-поведенческая психология через AI"
+segments:
+  - name: "Женщины 30-45, стресс-еда"
+    functional_job: "Похудеть без голодания"
+    emotional_job: "Чувствовать контроль"
+    social_job: "Выглядеть здоровой"
+    trigger: "Фото, одежда не подходит, рекомендация врача"
+  - name: "Мужчины 35-50, здоровье"
+    functional_job: "Снизить вес по медпоказаниям"
+    emotional_job: "Не чувствовать стыд"
+    trigger: "Анализы, давление, рекомендация врача"
+  - name: "Молодёжь 20-30, осознанность"
+    functional_job: "Разобраться в отношениях с едой"
+    emotional_job: "Быть в гармонии с телом"
+    social_job: "Выглядеть осознанным/mindful"
+    trigger: "Тренд на mindfulness, подруга посоветовала"
+aha_moment: "Персональный план на основе КПТ-профиля"
+complaints: ["дорого", "навязчивые уведомления", "коуч отвечает шаблонно"]
+positive_patterns: ["психологический подход", "без подсчёта калорий", "уроки короткие"]
+pricing: ["Free trial 14 дней", "₽299/мес год", "₽749/мес"]
+```
+
+### Генерация 3 вариантов (логика)
+
+```
+Axis 1 — AHA:
+  A: Персональный план (из M2 aha_moment — основной)
+  B: Прогресс за неделю (из complaints — "нет быстрого результата" → покажем через 7 дней)
+  C: AI-паттерн (из positive_patterns — "психологический подход" → усилим через AI)
+
+Axis 2 — ENTRY:
+  A: Functional Job Segment 1 → "Похудей через психологию"
+  B: Emotional Job Segment 3 → "Научись есть осознанно"
+  C: Emotional Job Segment 1 → "Разберись почему срываешься"
+
+Axis 3 — PAYWALL:
+  A: Сразу после Aha (день 1) → конверсия на wow-эмоции
+  B: После 7 дней → привычка сформирована
+  C: После первого сокращения срывов → proof of value
+```
+
+### Структура .jsx (reference для кода)
+
+```
+VARIANTS = {
+  A: {
+    name, emoji,                    // ID варианта
+    col, bg, bgl, bdr, txt,        // Tailwind цвета (подбирать под industry)
+    landing: { hl, sub, cta },     // Headline, subheadline, CTA
+    onboarding: { q1, q2, style }, // Вопросы из segment triggers
+    aha: { title, desc, wow, kpi },// Core value moment
+    dash: { hook, lesson, stats }, // Engagement mechanism
+    pay: { when, frame },          // Paywall timing + framing
+    inv: { hook, mech },           // Referral mechanism
+  },
+  B: { ... },
+  C: { ... },
+};
+
+CJM_META = {                       // Стандартная, не меняется
+  landing: { stage, aarrr, q },
+  onboarding: { ... },
+  aha: { ... },
+  dash: { ... },
+  pay: { ... },
+  inv: { ... },
+};
+
+// Components (стандартная структура):
+// - VariantSwitcher (tabs)
+// - ScreenRenderer (6 screens × 3 variants)
+// - CJMBadge (overlay toggle)
+// - CompareTable (⚖️)
+// - Builder (🔧 mix & match)
+// - Scoring (🏆)
+// - LockedResult (✅)
+```
+
+### Чеклист качества генерации
+
+- [ ] Все тексты на языке {GEOGRAPHY} (русский, English, etc.)
+- [ ] Headline из JTBD конкретного сегмента, не generic
+- [ ] Aha Moment — конкретный, не "вы получите план"
+- [ ] Onboarding вопросы из segment triggers (M2)
+- [ ] Pricing из M1 реальной компании
+- [ ] Dashboard hook разный для каждого варианта
+- [ ] Цветовая палитра подобрана под {INDUSTRY}
+- [ ] CJM overlay с validation questions для custdev

package/templates/.claude/skills/reverse-engineering-unicorn/examples/noom-module1-example.md

@@ -0,0 +1,88 @@
+# Few-Shot Example: Module 1 Output for Noom
+
+Этот файл показывает ОЖИДАЕМЫЙ формат и глубину output для Module 1. Используй как образец качества.
+
+---
+
+# 📊 FACT SHEET: Noom
+**Дата:** 2025-12-01 | **Аналитик:** AI | **Версия:** 1.0
+
+---
+
+## A. КОМПАНИЯ
+
+| Параметр | Значение | Источник |
+|----------|----------|----------|
+| Год основания | 2008 | crunchbase.com/organization/noom |
+| Штаб-квартира | New York, NY, USA | noom.com/about |
+| Сотрудники | ~2,000 (пик ~4,000 в 2022, сокращения в 2023) | LinkedIn, TechCrunch |
+| Основатели | Saeju Jeong (CEO), Artem Petakov (President) | crunchbase |
+| CEO | Saeju Jeong | noom.com/about |
+| Миссия | "Help people everywhere lead healthier lives through behavior change" | noom.com |
+
+## B. ФИНАНСИРОВАНИЕ
+
+| Раунд | Дата | Сумма | Lead Investor | Post-Money Valuation |
+|-------|------|-------|---------------|---------------------|
+| Seed | 2012 | $1.9M | Recruit Strategic Partners | — |
+| Series A | 2014 | $7M | Qualcomm Ventures | — |
+| Series B | 2017 | $16.4M | Samsung NEXT | — |
+| Series C | 2019 | $58M | Sequoia Capital | ~$500M |
+| Series F | 2021 | $540M | Silver Lake | $3.7B |
+| **Итого** | | **~$657M** | | **Последняя: $3.7B** |
+
+## C. ПРОДУКТ
+
+| Параметр | Значение | Источник |
+|----------|----------|----------|
+| Основной продукт | AI-powered weight management через когнитивно-поведенческую терапию | noom.com |
+| Платформы | iOS, Android, Web | App Store, Google Play |
+| Pricing | Free trial → $70/mo, $209/4mo, $209-$299/annual (varies by plan) | noom.com/pricing |
+
+**Ключевые фичи:**
+1. Personalized meal plans с calorie tracking
+2. CBT-based daily lessons (psychology of eating)
+3. 1-on-1 health coach (human + AI)
+4. Group support communities
+5. Exercise tracking integration
+
+## D. МАСШТАБ И TRACTION
+
+| Метрика | Значение | Дата | Источник |
+|---------|----------|------|----------|
+| Пользователи (cumulative signups) | 50M+ | 2023 | Forbes, TechCrunch |
+| Revenue (ARR) | ~$600M (peak 2022), declined since | 2022-2023 | The Information, Bloomberg |
+| App downloads | 100M+ (lifetime) | 2024 | SensorTower estimate |
+| Key markets | US (primary), UK, Germany, Australia, Japan | noom.com | 
+
+## E. ТЕХНОЛОГИИ
+
+| Компонент | Технология | Источник |
+|-----------|-----------|----------|
+| Backend | Python, Java | Job postings |
+| Frontend | React, TypeScript | Job postings |
+| Mobile | React Native, native iOS/Android | Job postings |
+| Cloud | AWS | Job postings |
+| AI/ML | NLP for food logging, recommendation engine, CBT personalization | Noom Engineering Blog |
+| Data | Snowflake, Kafka | Job postings |
+
+**Патенты:** 5+ patents related to behavior change technology and food recognition AI (USPTO)
+
+---
+
+## 📈 Confidence Score
+
+| Блок | Score | Комментарий |
+|------|-------|-------------|
+| A. Компания | 5/5 | Все данные верифицированы |
+| B. Финансирование | 4/5 | Ранние раунды менее документированы |
+| C. Продукт | 5/5 | Верифицировано с сайта |
+| D. Масштаб | 3/5 | Revenue данные противоречивы, не публичная компания |
+| E. Технологии | 3/5 | На основе job postings, не официальный tech blog |
+| **СРЕДНИЙ** | **4.0/5** | |
+
+## 🔍 Пробелы в данных
+- [ ] Точный текущий ARR (2024-2025) — компания частная
+- [ ] Количество активных платящих клиентов (vs cumulative signups)
+- [ ] Точный tech stack AI/ML pipeline
+- [ ] Юнит-экономика (CAC, LTV) — не раскрывается