codymaster 7.0.1 → 7.0.3

package/skills/cm-growth-hacking/trigger-system.md

@@ -0,0 +1,342 @@
+# Trigger System
+
+> Khi nào hiện bottom sheet? Trigger system quyết định.
+> Hỗ trợ: scroll, time, exit-intent, click, return-visitor, interaction.
+
+---
+
+## 🎯 Trigger Types
+
+| Type | Description | Best For |
+|------|-------------|----------|
+| `scroll` | User scrolls past X% of page | Service pages, blog content |
+| `time` | X seconds after page load | Homepage, landing pages |
+| `exit` | Mouse moves toward browser chrome (desktop) | Lead capture, promo |
+| `click` | User clicks specific element | Manual CTA buttons |
+| `return` | Returning visitor (cookie-based) | Re-engagement, loyalty |
+| `interaction` | User completes N interactions | Surveys after using tool |
+| `dual` | Time + Scroll combined (both must fire) | Audio CRO, booking nudge |
+
+---
+
+## ⚙️ TriggerManager Architecture
+
+```javascript
+class TriggerManager {
+  constructor(config) {
+    this.config = config;
+    this.triggered = false;
+    this.listeners = [];
+  }
+  
+  init() {
+    if (this.isDismissed()) return;
+    if (this.respectsReducedMotion()) return;
+    
+    const type = this.config.trigger.type;
+    
+    switch (type) {
+      case 'scroll': this.initScroll(); break;
+      case 'time': this.initTime(); break;
+      case 'exit': this.initExitIntent(); break;
+      case 'click': this.initClick(); break;
+      case 'return': this.initReturnVisitor(); break;
+      case 'interaction': this.initInteraction(); break;
+      case 'dual': this.initDual(); break;
+    }
+  }
+  
+  fire() {
+    if (this.triggered) return;
+    this.triggered = true;
+    this.cleanup();
+    if (typeof this.config.onTrigger === 'function') {
+      this.config.onTrigger();
+    }
+  }
+  
+  cleanup() {
+    this.listeners.forEach(({ el, event, fn }) => el.removeEventListener(event, fn));
+    this.listeners = [];
+  }
+  
+  destroy() {
+    this.cleanup();
+    this.triggered = true;
+  }
+}
+```
+
+---
+
+## 📜 Scroll Trigger
+
+```javascript
+initScroll() {
+  const threshold = this.config.trigger.value || 0.3; // 30% default
+  const delay = this.config.trigger.delay || 0;
+  
+  const onScroll = () => {
+    const scrollPct = window.scrollY / 
+      (document.documentElement.scrollHeight - window.innerHeight);
+    
+    if (scrollPct >= threshold) {
+      if (delay > 0) {
+        setTimeout(() => this.fire(), delay);
+      } else {
+        this.fire();
+      }
+    }
+  };
+  
+  window.addEventListener('scroll', onScroll, { passive: true });
+  this.listeners.push({ el: window, event: 'scroll', fn: onScroll });
+  
+  // Check immediately (user may already be scrolled)
+  onScroll();
+}
+```
+
+**Recommended thresholds:**
+
+| Page Type | Scroll % | Rationale |
+|-----------|----------|-----------|
+| Homepage | 30% | After hero + intro |
+| Service page | 25% | After service description |
+| Blog/article | 40% | After first section |
+| Product page | 20% | After product image |
+
+---
+
+## ⏱️ Time Trigger
+
+```javascript
+initTime() {
+  const delay = this.config.trigger.value || 15000; // 15s default
+  
+  const timer = setTimeout(() => this.fire(), delay);
+  
+  // Store for cleanup
+  this._timer = timer;
+}
+
+// Override cleanup to clear timer
+cleanup() {
+  super.cleanup();
+  if (this._timer) clearTimeout(this._timer);
+}
+```
+
+**Recommended delays:**
+
+| Page Type | Delay | Rationale |
+|-----------|-------|-----------|
+| Homepage | 20-25s | Let user explore first |
+| Landing page | 10-15s | Faster engagement needed |
+| Service page | 15-20s | After reading key info |
+| Blog | 25-30s | Don't interrupt reading |
+
+---
+
+## 🚪 Exit Intent Trigger (Desktop Only)
+
+```javascript
+initExitIntent() {
+  // Only works on desktop (mouse-based)
+  if ('ontouchstart' in window) return;
+  
+  const onMouseLeave = (e) => {
+    // Mouse moved above viewport (toward browser chrome)
+    if (e.clientY <= 0) {
+      this.fire();
+    }
+  };
+  
+  document.addEventListener('mouseleave', onMouseLeave);
+  this.listeners.push({ el: document, event: 'mouseleave', fn: onMouseLeave });
+}
+```
+
+**Mobile alternative:** Use `time` or `scroll` trigger instead. Exit intent doesn't exist on touch devices.
+
+**Enhancement — debounce:**
+
+```javascript
+// Prevent rapid-fire on mouse-near-top
+let exitTimeout;
+const onMouseLeave = (e) => {
+  if (e.clientY <= 0) {
+    clearTimeout(exitTimeout);
+    exitTimeout = setTimeout(() => this.fire(), 300);
+  }
+};
+```
+
+---
+
+## 🖱️ Click Trigger
+
+```javascript
+initClick() {
+  const selector = this.config.trigger.value; // CSS selector
+  
+  const elements = document.querySelectorAll(selector);
+  elements.forEach(el => {
+    const onClick = (e) => {
+      e.preventDefault();
+      this.fire();
+    };
+    el.addEventListener('click', onClick);
+    this.listeners.push({ el, event: 'click', fn: onClick });
+  });
+}
+```
+
+**Use case:** Manual CTA buttons that open bottom sheets.
+
+```html
+<button data-trigger="booking-sheet">📅 Đặt Lịch</button>
+```
+
+```javascript
+new TriggerManager({
+  trigger: { type: 'click', value: '[data-trigger="booking-sheet"]' },
+  onTrigger: () => bookingSheet.show()
+});
+```
+
+---
+
+## 🔄 Return Visitor Trigger
+
+```javascript
+initReturnVisitor() {
+  const VISIT_KEY = 'eng_visit_count';
+  const minVisits = this.config.trigger.value || 2; // minimum visits to trigger
+  
+  try {
+    let visits = parseInt(localStorage.getItem(VISIT_KEY) || '0');
+    visits++;
+    localStorage.setItem(VISIT_KEY, visits.toString());
+    
+    if (visits >= minVisits) {
+      // Add time delay so page loads first
+      setTimeout(() => this.fire(), this.config.trigger.delay || 3000);
+    }
+  } catch { /* localStorage blocked */ }
+}
+```
+
+---
+
+## 🎮 Interaction Trigger
+
+```javascript
+initInteraction() {
+  const { selector, minCount } = this.config.trigger.value;
+  // e.g., { selector: '.checklist input[type=checkbox]', minCount: 3 }
+  
+  const elements = document.querySelectorAll(selector);
+  
+  const checkCount = () => {
+    const interacted = document.querySelectorAll(selector + ':checked').length;
+    if (interacted >= minCount) {
+      this.fire();
+    }
+  };
+  
+  elements.forEach(el => {
+    el.addEventListener('change', checkCount);
+    this.listeners.push({ el, event: 'change', fn: checkCount });
+  });
+}
+```
+
+**Use case:** After user checks 3+ symptoms → show "Đặt lịch khám" sheet.
+
+---
+
+## 🔀 Dual Trigger (Time + Scroll)
+
+Both conditions must be met before firing:
+
+```javascript
+initDual() {
+  let timeReady = false;
+  let scrollReady = false;
+  
+  const tryFire = () => {
+    if (timeReady && scrollReady) this.fire();
+  };
+  
+  // Time leg
+  const delay = this.config.trigger.value?.time || 15000;
+  setTimeout(() => { timeReady = true; tryFire(); }, delay);
+  
+  // Scroll leg
+  const threshold = this.config.trigger.value?.scroll || 0.25;
+  const onScroll = () => {
+    const pct = window.scrollY / 
+      (document.documentElement.scrollHeight - window.innerHeight);
+    if (pct >= threshold) {
+      scrollReady = true;
+      window.removeEventListener('scroll', onScroll);
+      tryFire();
+    }
+  };
+  window.addEventListener('scroll', onScroll, { passive: true });
+  this.listeners.push({ el: window, event: 'scroll', fn: onScroll });
+  onScroll();
+}
+```
+
+---
+
+## 💾 Session Management
+
+```javascript
+// Check if user dismissed this engagement
+isDismissed() {
+  const key = this.config.session?.dismissKey || 'eng_dismissed';
+  const storage = this.config.session?.storage === 'localStorage' 
+    ? localStorage : sessionStorage;
+  try { return storage.getItem(key) === '1'; } 
+  catch { return false; }
+}
+
+// Mark as dismissed
+setDismissed() {
+  const key = this.config.session?.dismissKey || 'eng_dismissed';
+  const storage = this.config.session?.storage === 'localStorage' 
+    ? localStorage : sessionStorage;
+  try { storage.setItem(key, '1'); } 
+  catch { /* noop */ }
+}
+
+// Respect prefers-reduced-motion
+respectsReducedMotion() {
+  return window.matchMedia('(prefers-reduced-motion: reduce)').matches;
+}
+```
+
+### Storage Strategy
+
+| Behavior | Storage | Result |
+|----------|---------|--------|
+| Show once per tab session | `sessionStorage` | Resets on browser close |
+| Show once per device forever | `localStorage` | Persistent |
+| Show once per day | `localStorage` + timestamp check | Custom expiry |
+
+---
+
+## ❌ Anti-Patterns
+
+| ❌ Don't | ✅ Do |
+|----------|------|
+| Trigger popup on page load without delay | Minimum 10s or 20% scroll |
+| Use exit-intent on mobile | Fallback to scroll/time trigger |
+| Trigger on checkout/payment pages | Exclude conversion-critical pages |
+| Trigger multiple sheets at once | Queue or prioritize |
+| Ignore `prefers-reduced-motion` | Skip trigger for such users |
+| Trigger every page visit | Respect session dismiss state |
+| Use `setInterval` for scroll check | Use passive scroll event listener |

package/skills/cm-how-it-work/SKILL.md

@@ -72,7 +72,7 @@ The **CodyMaster (CM)** kit transforms ideas into production code through 13 spe
     - `cm-tdd`: Red-Green-Refactor cycle. No production code without a failing test first.
     - `cm-execution`: Execute plans intelligently (Manual, Parallel, or Subagent mode).
     - `cm-project-bootstrap`: For new projects — setup repo, i18n, SEO, and deploy pipeline from Day 0.
-    - `cm-git-worktrees`: Isolate different work items to avoid state mixing.
+    - `cm-execution`: Isolate different work items to avoid state mixing.
 
 ---
 
@@ -104,7 +104,7 @@ The **CodyMaster (CM)** kit transforms ideas into production code through 13 spe
 | `cm-tdd` | Strict Red-Green-Refactor TDD — no code without failing test first. |
 | `cm-debugging` | Systematic 5-phase error investigation (root cause first). |
 | `cm-quality-gate` | 6-gate verification: static analysis → blind review → ship. |
-| `cm-test-gate` | Setup 4-layer test infrastructure (unit → integration → e2e → security). |
+| `cm-quality-gate` | Setup and enforce the verification gate before claiming "done". |
 | `cm-code-review` | Manage PR lifecycle: request → receive feedback → complete branch. |
 | `cm-codeintell` | AST knowledge graph + architecture diagrams + smart context (30% fewer tokens). |
 
@@ -113,9 +113,9 @@ The **CodyMaster (CM)** kit transforms ideas into production code through 13 spe
 |-------|-----------------|
 | `cm-safe-deploy` | Multi-gate deploy pipeline with rollback strategy. |
 | `cm-identity-guard` | Verify account before push/deploy to prevent wrong-project incidents. |
-| `cm-git-worktrees` | Isolate feature work in separate worktrees — no state mixing. |
+| `cm-execution` | Isolate feature work in separate workspaces/batches — no state mixing. |
 | `cm-terminal` | Safe terminal execution with progress logging and error capture. |
-| `cm-secret-shield` | Defense-in-depth: pre-commit hooks, repo scanning, token lifecycle. |
+| `cm-safe-deploy` | Defense-in-depth: pre-commit hooks, repo scanning, token lifecycle. |
 | `cm-safe-i18n` | Safe multi-language management with multi-pass batching and 8-gate audit. |
 
 ### 🎨 Product
@@ -125,7 +125,7 @@ The **CodyMaster (CM)** kit transforms ideas into production code through 13 spe
 | `cm-brainstorm-idea` | Strategic analysis gate: 9 Windows + Double Diamond → 2-3 qualified options. |
 | `cm-planning` | Brainstorm intent → write implementation plan → coordinate execution. |
 | `cm-ux-master` | 48 UX Laws + 37 Design Tests + Figma/Stitch + BM25 semantic design search. |
-| `cm-ui-preview` | Live UI concept generation via Google Stitch or Pencil MCP. |
+| `cm-design-system` | Live UI concept generation via design-token and Stitch/Open Design workflows. |
 | `cm-dockit` | Generate complete knowledge bases, SOPs, and API docs from codebase. |
 | `cm-readit` | Web audio engine: TTS reader, MP3 audio player, Voice CRO trigger system. |
 | `cm-jtbd` | JTBD customer discovery: Switch Interview → Canvas → Opportunity Scoring. |
@@ -142,7 +142,7 @@ The **CodyMaster (CM)** kit transforms ideas into production code through 13 spe
 |-------|-----------------|
 | `cm-execution` | Execute plans: Manual / Parallel / Subagent / RARV batch modes. |
 | `cm-continuity` | Working memory protocol: read at session start, update at session end. |
-| `cm-skill-mastery` | Meta-skill: when to invoke which skill, how to create new skills. |
+| `cm-skill-index` | Meta-skill: when to invoke which skill and how to discover the right one. |
 | `cm-skill-chain` | Compose skills into automated multi-step pipelines (5 built-in chains). |
 | `cm-skill-index` | Progressive disclosure index — 90% token savings on skill discovery. |
 | `cm-deep-search` | Semantic search via qmd for large codebases (>200 src / >50 docs files). |
@@ -153,8 +153,7 @@ The **CodyMaster (CM)** kit transforms ideas into production code through 13 spe
 | Skill | Primary Function |
 |-------|-----------------|
 | `cm-start` | Orchestrate full workflow from objective to production code. |
-| `cm-dashboard` | Render Kanban board from cm-tasks.json — visual status overview. |
-| `cm-status` | Ultra-concise 1-2 sentence progress summary (what's done, what's next). |
+| `cm-status` | Ultra-concise 1-2 sentence progress summary; pair with `cm dashboard` for visuals. |
 
 ## 🚀 Autonomous Workflow System
 
@@ -200,6 +199,6 @@ There are 2 ways to use CodyMaster: **Fully autonomous (Via Workflows)** or **Ma
 4. **Test Before Code:** RED → GREEN → REFACTOR. No exceptions (`cm-tdd`).
 5. **Evidence Over Claims:** Only trust terminal/test results output, not AI "saying" it's done.
 6. **Deploy via Gates:** 8 gates must pass sequentially. Any gate fails = STOP (`cm-safe-deploy`).
-7. **Safe Secrets:** Never commit secrets. Pre-commit hooks protect every push (`cm-secret-shield`).
+7. **Safe Secrets:** Never commit secrets. Pre-commit hooks protect every push (`cm-safe-deploy`).
 8. **Parallel Power:** Use parallel execution for i18n or multi-bug fixes (`cm-execution`).
 9. **Working Memory:** Read CONTINUITY.md at session start, update at session end (`cm-continuity`).

package/skills/cm-identity-guard/SKILL.md

@@ -295,7 +295,7 @@ git config user.email "dev@workdomain.com"
 ## Phase 5: Token Lifecycle Management 🔄
 
 > **NEW — Secrets don't just need to be hidden. They need to be ROTATED.**
-> **Full rotation playbooks in `cm-secret-shield` Layer 5.**
+> **Full rotation playbooks now live in `cm-safe-deploy`.**
 
 ### Rotation Schedule
 
@@ -321,7 +321,7 @@ npm run deploy:staging
 # 4. Verify: test staging URL
 ```
 
-> For emergency rotation (leaked secret), see `cm-secret-shield` Emergency Rotation Playbook.
+> For emergency rotation (leaked secret), see the emergency rotation guidance in `cm-safe-deploy`.
 
 ---
 
@@ -397,8 +397,8 @@ npm run deploy
 |-------|------|
 | `cm-project-bootstrap` | Identity lock is Phase 0 of every new project |
 | `cm-safe-deploy` | Gate 0 secret hygiene checks wrangler.jsonc |
-| `cm-test-gate` | Phase 4 secret hygiene in test gate setup |
-| `cm-secret-shield` | Layer 5 token lifecycle extends identity management |
+| `cm-quality-gate` | Phase 4 verification and secret hygiene checks |
+| `cm-safe-deploy` | Token lifecycle and release safety extend identity management |
 
 ## The Bottom Line
 

package/skills/cm-jtbd/SKILL.md

@@ -0,0 +1,98 @@
+---
+name: cm-jtbd
+description: "Customer discovery framework using Jobs-To-Be-Done theory — uncover the functional, social, and emotional jobs customers hire products to do. Produces JTBD canvases with job statements, outcome metrics, and competing solutions. Use alongside cm-brainstorm-idea for evidence-based product decisions."
+---
+
+# Jobs-To-Be-Done — Customer Discovery Framework
+
+> **Understand the job, not the customer.**
+> People don't buy products — they hire them to get a job done.
+
+## When to Use
+
+- Before designing a new feature or product
+- When existing features aren't converting or being used
+- Alongside `cm-brainstorm-idea` for deep customer context
+- User mentions: "customer discovery", "JTBD", "what do customers want", "product-market fit", "why are users churning"
+
+## The JTBD Framework
+
+### Job Statement Formula
+
+```
+When [SITUATION], I want to [MOTIVATION], so I can [EXPECTED OUTCOME]
+```
+
+### Three Job Dimensions
+
+| Dimension | Definition | Example |
+|-----------|-----------|---------|
+| **Functional** | The core task to accomplish | "Get from A to B quickly" |
+| **Social** | How the person wants to be perceived | "Be seen as a reliable professional" |
+| **Emotional** | How the person wants to feel | "Feel confident in my decision" |
+
+## Process
+
+### Phase 1: Job Discovery (Interviews)
+
+1. Recruit 5-8 recent customers (ideally within 90 days of purchase)
+2. Use the Switch Interview technique — ask about the moment they decided to switch/buy
+3. Key questions:
+   - "Walk me through the day you decided to [buy/switch/start using X]"
+   - "What were you doing before that solution existed?"
+   - "What was the first thing you tried? Why didn't that work?"
+   - "What almost stopped you from switching?"
+4. Record patterns: triggers → anxiety → progress → outcomes
+
+### Phase 2: JTBD Canvas
+
+For each major job discovered, complete the canvas:
+
+```
+JOB STATEMENT:
+When [situation], I want to [motivation], so I can [outcome]
+
+FUNCTIONAL DIMENSION: [core task]
+SOCIAL DIMENSION:     [perception goal]
+EMOTIONAL DIMENSION:  [feeling goal]
+
+FORCES PUSHING TO HIRE:
+(+) Push: [what makes them switch from current solution]
+(+) Pull: [what attracts them to new solution]
+
+FORCES RESISTING HIRE:
+(-) Anxiety: [fears about new solution]
+(-) Habit: [attachment to old solution]
+
+COMPETING SOLUTIONS CURRENTLY HIRED:
+1. [direct competitor or workaround]
+2. [indirect solution]
+3. [do-nothing option]
+
+OUTCOME METRICS (how customer measures success):
+- Speed: [e.g., "get answer in <5 minutes"]
+- Accuracy: [e.g., "zero errors in the output"]
+- Effort: [e.g., "no manual steps required"]
+```
+
+### Phase 3: Opportunity Scoring
+
+Rate each outcome metric:
+- **Importance** (1-10): How important is this outcome to the customer?
+- **Satisfaction** (1-10): How satisfied are they with current solutions?
+- **Opportunity score** = Importance + max(Importance − Satisfaction, 0)
+
+Scores ≥ 15 = underserved outcomes → highest priority to address.
+
+## Output
+
+Save JTBD canvas to `docs/jtbd/jtbd-canvas-[date].md`.
+
+## Integration
+
+| Skill | Relationship |
+|-------|-------------|
+| `cm-brainstorm-idea` | UPSTREAM: JTBD feeds into strategic analysis |
+| `cm-planning` | DOWNSTREAM: Validated jobs inform feature plans |
+| `cm-cro-methodology` | COMPLEMENT: JTBD objections → CRO objection handling |
+| `cm-dockit` | OUTPUT: JTBD canvases are a document type in DocKit |

package/skills/cm-notebooklm/SKILL.md

@@ -0,0 +1,156 @@
+---
+name: cm-notebooklm
+description: "Sync high-value dev knowledge (skills, decisions, lessons) into Google NotebookLM for cloud recall, podcasts, and flashcards. Pairs with Dockit/deep-search. Use for NotebookLM, nlm, knowledge base, skill sync, or cross-machine AI memory."
+---
+
+# Goal
+
+Cloud brain / soul cho AI — lưu trữ kiến thức **giá trị nhất**, đồng bộ qua
+nhiều máy tính. `brain.md` = compiled source of truth, upload 1 lần thay vì 50.
+
+# How It Works
+
+```
+Skills + Lessons + Experiences
+        ↓ compile
+  ~/.codymaster/brain.md  (local file = backup)
+        ↓ upload (1 API call)
+  NotebookLM (cloud = cross-machine sync)
+```
+
+All automation via `scripts/brain-sync.sh` — AI chỉ cần gọi script.
+
+# Instructions
+
+## The Dual-Brain Architecture
+
+CodyMaster uses a **Dual-Brain** approach to scale memory across multiple projects without context collapse:
+
+1. **Master Brain (`codymaster`)**: Global brain storing Meta-Learnings, proven Architecture Decisions, and universal patterns. Evolved via the "Rule of 3" (knowledge graduation).
+2. **Project Brain (Isolated)**: Optional local brain created per project to hold internal PRDs, APIs, and local code context from `./docs`.
+
+| Level | Goal | Query Command |
+|-------|------|---------------|
+| **Master** | Tìm pattern chung | `nlm notebook query codymaster "..."` |
+| **Project** | Tìm context domain | `nlm notebook query $(cat .cm/notebook_id) "..."` |
+
+> 🤖 **AI Rule:** Tự đưa quyết định khi hỏi: Hỏi `codymaster` cho các giải pháp mang tính hệ thống chung. Hỏi qua `.cm/notebook_id` cho các business logic đặc thù dự án đang làm.
+
+## Master Brain Setup (Global)
+
+```bash
+# 1. Install CLI
+uv tool install notebooklm-mcp-cli
+
+# 2. Setup Master Brain
+bash ~/.gemini/antigravity/skills/cm-notebooklm/scripts/brain-sync.sh init
+```
+
+## Project Brain Setup (Local - Optional)
+
+Nếu dự án đủ lớn và nhiều doc:
+```bash
+# Tạo Project Brain riêng cho thư mục hiện tại
+bash ~/.gemini/antigravity/skills/cm-notebooklm/scripts/brain-sync.sh init-project
+```
+
+## Daily Usage
+
+```bash
+SCRIPT=~/.gemini/antigravity/skills/cm-notebooklm/scripts/brain-sync.sh
+
+# Add a lesson learned
+bash $SCRIPT lesson "Tên bài học"
+# → Edit ~/.codymaster/lessons.md → fill in details
+
+# Add coding experience
+bash $SCRIPT experience "Tên pattern"
+# → Edit ~/.codymaster/experiences.md → fill in details
+
+# Sync to Master Brain (Thêm rule of 3)
+bash $SCRIPT sync
+
+# Sync to Project Brain (Up tài liệu local docs/)
+bash $SCRIPT sync-project
+
+# Check status
+bash $SCRIPT status
+
+# Query
+nlm notebook query codymaster "your question"
+```
+
+## Auto-Sync Triggers
+
+AI detect và **hỏi user** (không tự động):
+
+| Trigger | Prompt |
+|---------|--------|
+| Skill mới tạo | "Sync skill mới vào brain?" → `bash $SCRIPT sync` |
+| Bug fixed / post-mortem | "Lưu bài học?" → `bash $SCRIPT lesson "..."` |
+| Architecture changed | "Update brain?" → `bash $SCRIPT sync` |
+| User nói "sync/update" | → `bash $SCRIPT sync` |
+
+## Cross-Machine (Máy Mới)
+
+```bash
+uv tool install notebooklm-mcp-cli
+nlm login
+nlm notebook list          # CodyMaster Brain already there
+nlm alias set codymaster <id>
+# Done! Query ngay: nlm notebook query codymaster "..."
+```
+
+## Content Generation
+
+```bash
+nlm audio create codymaster --format deep_dive --confirm   # Podcast
+nlm report create codymaster --format "Study Guide" --confirm
+nlm flashcards create codymaster --difficulty medium --confirm
+nlm studio status codymaster
+```
+
+# What Gets Indexed
+
+```
+✅ HIGH-VALUE (auto-compiled into brain.md):
+├── Skill Index (names + descriptions — NOT full SKILL.md)
+├── Lessons Learned (~/.codymaster/lessons.md)
+├── Coding Experiences (~/.codymaster/experiences.md)
+└── AGENTS.md (project identity)
+
+❌ NOT INDEXED (use qmd/cm-deep-search instead):
+├── Full SKILL.md files (too many, hard to maintain)
+├── Source code, tests, configs
+└── Duplicated content
+```
+
+# Memory Architecture
+
+```
+Session  → variables        → temporary
+Working  → CONTINUITY.md    → ~500 words/turn
+Local    → qmd              → BM25+vector, offline, stable
+Cloud    → NotebookLM       → AI brain, cross-machine, podcast
+```
+
+# Integration
+
+| Skill | Role |
+|-------|------|
+| `cm-deep-search` | Local search complement (code) |
+| `cm-dockit` | Generate docs → select high-value → feed to brain |
+| `cm-continuity` | Session memory, brain = long-term |
+| `skill-creator-ultra` | TRIGGER: new skill → prompt sync |
+| `cm-debugging` | TRIGGER: bug fixed → prompt lesson |
+
+# Constraints
+
+- 🔌 `nlm` CLI = **third-party** (jacob-bd). May break. Fallback: cm-deep-search.
+- 🧠 **Selective only** — quality > quantity. Don't dump everything.
+- 🔐 NO secrets in NotebookLM — cloud service.
+- ❌ NO delete without user confirmation.
+- 🤖 NO `nlm chat start` — use `nlm notebook query` only.
+- 🔄 Auto-sync = **ask first**, never silent upload.
+
+<!-- Generated by Skill Creator Ultra v1.0 -->