@hanzlaa/rcode 4.3.2 → 4.4.0

package/rcode/skills/seo/seo-site-builder/rules/03-keyword-strategy.md

@@ -0,0 +1,82 @@
+# Rule: Keyword Strategy
+
+## When to Run
+After deep research completes and user approves (or solo yolo auto-continues)
+
+## Step 1 — Build Keywords.txt
+Extract ALL keywords from research/02-keyword-research.md
+Format: comma-separated, one line, saved to `/keywords.txt`
+
+```bash
+# Extract and deduplicate all keywords from research file
+grep -E "^\|" research/02-keyword-research.md | \
+  grep -v "Keyword\|---\|keyword" | \
+  awk -F'|' '{print $2}' | \
+  sed 's/^[[:space:]]*//' | sed 's/[[:space:]]*$//' | \
+  grep -v "^$" | grep -v "^[0-9]*$" | \
+  sort -u | tr '\n' ',' | sed 's/,$/\n/' > keywords.txt
+```
+
+**Tell user:**
+"Keywords saved to keywords.txt — please paste into Google Keyword Planner:
+1. Go to ads.google.com → Tools → Keyword Planner
+2. Click 'Discover new keywords' → 'Enter keywords manually'
+3. Paste the contents of keywords.txt
+4. Download CSV with Volume, CPC, Competition columns
+5. Share the CSV here and I'll re-prioritize the article schedule"
+
+## Step 2 — If CSV Provided (Preferred Path)
+Parse CSV with these columns:
+- Keyword, Volume, CPC, AVG DA, Weakspot, Golden Score, Intent, Trending
+
+**Priority sort algorithm:**
+```python
+# Step 1: Filter to AVG DA < 55 (beatable without backlinks)
+# Step 2: Sort by Weakspot DESC (more weak pages = easier)
+# Step 3: Within same Weakspot tier, sort by Volume DESC
+# Step 4: Flag Trending=true keywords with ⭐
+# Step 5: Put Golden Score > 1.5 keywords in priority tier
+```
+
+**Article Schedule Tiers:**
+- Tier 1: AVG DA < 45, any Weakspot → Publish Month 1 (no backlinks needed)
+- Tier 2: AVG DA 45-55, Weakspot ≥ 2 → Publish Month 1-2
+- Tier 3: AVG DA 55-65, Weakspot ≥ 1 → Publish Month 3-4
+- Tier 4: AVG DA 65-75, any → Publish Month 4-5 (need some authority)
+- Tier 5: AVG DA 75+, high volume → Money pages, Publish Month 6+
+
+## Step 3 — If No CSV (Fallback Path)
+Use browser tool to manually check top 10 SERPs for top 30 keywords:
+- Note if top results are: big brands (bad) vs small blogs (good)
+- Check Moz or Ahrefs toolbar data if visible
+- Estimate DA based on: site type, content depth, known brand recognition
+
+## Step 4 — Generate article-schedule.md
+Save to `docs/article-schedule.md` with exact format:
+```markdown
+| # | Title | Keyword | DA | Vol | Weak | Score | Type |
+```
+
+Each article must have:
+- Exact target keyword
+- Suggested URL slug
+- Article type (Review/Comparison/How-To/Use-Case/FAQ/Roundup)
+- Target publish month
+- Why it ranks without backlinks (1 sentence)
+
+## Step 5 — Identify Top 5 "First Week" Articles
+These go live in Week 1-2 for fastest sandbox exit signals:
+1. One troubleshoot/panic search (Reddit/forums dominate SERP)
+2. One specific model review (retailer pages only, zero affiliates)
+3. One comparison nobody wrote (brand A vs brand B gap)
+4. One question-based article (PAA box with weak answers)
+5. One use-case article (health condition + product recommendation)
+
+## Copy Keywords to Clipboard
+Always run at end:
+```bash
+cat keywords.txt | clip.exe  # WSL
+# or
+cat keywords.txt | pbcopy    # Mac
+```
+Tell user: "Keywords copied to clipboard — paste into Google Ads Planner now"

package/rcode/skills/seo/seo-site-builder/rules/04-content-plan.md

@@ -0,0 +1,173 @@
+# Rule: Content Plan & Article Stubs
+
+## When to Run
+After article-schedule.md is finalized with real keyword data
+
+## Step 1 — Create Content Directory Structure
+```
+/content/
+├── reviews/        ← single product/model reviews
+├── comparisons/    ← brand-vs-brand, type-vs-type
+├── guides/         ← buying guides, how-to, educational
+├── health/         ← use-case articles (condition-specific)
+└── best/           ← roundup / best-of lists (money pages)
+```
+
+## Step 2 — Generate MDX Stubs for First 20 Articles
+For each article in Tier 1 + Tier 2 of article-schedule.md, create a stub file.
+
+File naming: `content/[type]/[slug].mdx`
+
+**MDX Frontmatter Template:**
+```mdx
+---
+title: "[Full SEO Title Here]"
+slug: "[url-slug]"
+type: "review|comparison|guide|health|best"
+keyword: "[exact target keyword]"
+metaTitle: "[Title under 60 chars for Google]"
+metaDescription: "[Description 140-160 chars with keyword]"
+publishDate: "[YYYY-MM-DD based on schedule]"
+status: "stub"
+wordCountTarget: [1500-5000 based on type]
+affiliateLinks:
+  - brand: "[Brand Name]"
+    url: "[affiliate URL placeholder]"
+    commission: "[X%]"
+internalLinks:
+  - "[related article slug 1]"
+  - "[related article slug 2]"
+imageRequired: true
+imagePomptFile: "assets/image-prompts/[slug]-images.md"
+---
+
+## [H2: Main Section Title]
+
+[STUB: Write 2000-word expert review covering: specs, EMF levels, wood quality, heating performance, assembly, verdict]
+
+## Pros & Cons
+
+**Pros:**
+- [STUB]
+
+**Cons:**
+- [STUB]
+
+## Verdict
+
+[STUB: Final recommendation with affiliate CTA]
+```
+
+## Step 3 — Image Handling (Two Strategies)
+
+For EVERY article, handle images using ONE of these two strategies:
+
+### Strategy A — Download Real Product Images (Preferred for Reviews)
+Use browser tool to find and download actual product images:
+1. Open Google Images: search `[product name] official product photo`
+2. Search brand's official website for press/product images
+3. Search Amazon listing for the product — use main product photos
+4. Download to: `public/images/[slug]/`
+   - `hero.jpg` — main product exterior shot
+   - `detail.jpg` — close-up or interior shot
+   - `specs.jpg` — specs sheet or comparison image if available
+
+**When to use Strategy A:**
+- Product review articles (real product photos = more trust + E-E-A-T)
+- Brand overview articles
+- Any article where real product imagery exists online
+
+**Browser tool commands:**
+```
+1. Navigate to Google Images
+2. Search: "[Brand] [Model] product photo official"
+3. Filter: Large size, labeled for reuse if possible
+4. Save image to correct path
+```
+
+### Strategy B — AI Generation Prompt File (For Infographics, Cards, Pinterest)
+Create `/assets/image-prompts/[slug]-images.md` with detailed prompts:
+
+```markdown
+# Image Prompts: [Article Title]
+# Place generated images at the exact paths specified below
+
+## Image 1 — Hero (if no real product photo found)
+**Place at:** public/images/[slug]/hero.jpg
+**Prompt:** Professional product photography of [product description],
+warm cedar wood interior background, soft diffused natural lighting,
+premium lifestyle feel, no text, 1200x630px, high-end editorial quality,
+warm amber and cream tones
+
+## Image 2 — Verdict/Rating Card
+**Place at:** public/images/[slug]/verdict-card.png
+**Prompt:** Rating card graphic showing [X.X]/10 score,
+product name "[Product Name]" in Playfair Display serif font,
+warm Nordic minimal design, burnt amber (#C9622F) accent,
+cream background (#FAF6F1), star rating icons, clean professional layout
+
+## Image 3 — Pinterest Pin
+**Place at:** public/images/[slug]/pinterest.jpg
+**Prompt:** Tall Pinterest 2:3 ratio graphic (1000x1500px),
+"[Article Headline]" text in Playfair Display,
+warm sauna interior background, burnt amber banner bottom with site name,
+premium editorial feel, high contrast for mobile readability
+
+## Image 4 — Comparison Infographic (for comparison articles)
+**Place at:** public/images/[slug]/comparison.png
+**Prompt:** Clean two-column comparison infographic,
+maple amber (#C9622F) vs dark forest (#1C2B2D) column headers,
+cream background (#FAF6F1), Inter font body text,
+professional data visualization, no stock photo people
+```
+
+**When to use Strategy B:**
+- Comparison infographics (no real product photo needed)
+- Rating/verdict cards
+- Pinterest pins
+- Any custom graphic that doesn't require a real product photo
+
+### Decision Rule Per Article Type
+| Article Type | Hero | Detail | Card | Pinterest |
+|-------------|------|--------|------|-----------|
+| Product Review | Strategy A | Strategy A | Strategy B | Strategy B |
+| Comparison | Strategy B | Strategy A (both products) | Strategy B | Strategy B |
+| How-To / FAQ | Strategy B | Strategy A (if product shown) | — | Strategy B |
+| Roundup | Strategy B | Strategy A (top pick) | Strategy B | Strategy B |
+| Use-Case/Health | Strategy B | Strategy A (product rec) | — | Strategy B |
+
+## Step 4 — Internal Linking Map
+Create `docs/internal-linking-map.md`:
+
+```markdown
+# Internal Linking Map
+
+## Rule: All roads lead to money pages
+
+### Money Pages (receive most links)
+- /best/infrared-sauna-for-home ← linked from: all reviews, all comparisons
+- /best/sauna-blanket ← linked from: all blanket reviews, portable comparisons
+
+### Hub Pages (give and receive links)
+- /brands/[brand-name] ← linked from all model reviews of that brand
+- /guides/buyers-guide ← linked from all Tier 1-2 articles
+
+### Spoke Pages (give links upward)
+- /reviews/[model] → links to: brand hub, most relevant roundup, 1-2 comparisons
+- /health/[condition] → links to: 2-3 model reviews, buyers guide
+
+### Anchor Text Rules
+✅ Use: natural phrases ("best for arthritis", "see our full review", "compared here")
+❌ Avoid: exact match anchors ("best infrared sauna") — triggers over-optimization
+❌ Avoid: generic anchors ("click here", "read more")
+```
+
+## Step 5 — Interactive Gate
+After stubs + image prompts created:
+"Content stubs ready for [N] articles.
+Image prompts saved to assets/image-prompts/ — get them generated and place at paths specified.
+
+Want to:
+1. Continue to site build (Next.js setup)
+2. Research the next niche idea
+3. Write full content for Article #1 now"

package/rcode/skills/seo/seo-site-builder/rules/05-site-build.md

@@ -0,0 +1,187 @@
+# Rule: Site Build — Next.js + Velite + BMAD
+
+## When to Run
+After content plan approved and user wants to build the site
+
+## Tech Stack (Always Use Unless User Says Otherwise)
+- **Framework:** Next.js 16.2.1 (App Router) — `pnpm create next-app@16.2.1`
+- **Content Layer:** Velite (MDX + JSON → typed content)
+- **Styling:** Tailwind CSS + shadcn/ui
+- **Hosting:** Vercel (free tier)
+- **Package Manager:** pnpm (ALWAYS — never npm or yarn)
+
+## Step 1 — Init Project
+```bash
+cd /home/hanzla/development/
+pnpm create next-app@16.2.1 [project-slug] \
+  --typescript \
+  --tailwind \
+  --eslint \
+  --app \
+  --src-dir \
+  --import-alias "@/*"
+cd [project-slug]
+```
+
+## Step 2 — Install Dependencies
+```bash
+# Content layer
+pnpm add velite
+
+# UI components
+pnpm dlx shadcn@latest init
+pnpm dlx shadcn@latest add button card badge separator table
+
+# SEO
+pnpm add next-sitemap
+
+# Utilities
+pnpm add date-fns reading-time
+```
+
+## Step 3 — Directory Structure
+Create exactly:
+```
+src/
+├── app/
+│   ├── layout.tsx              ← root layout with SEO defaults
+│   ├── page.tsx                ← homepage
+│   ├── reviews/
+│   │   └── [slug]/page.tsx     ← single product review
+│   ├── compare/
+│   │   └── [slug]/page.tsx     ← comparison articles
+│   ├── guides/
+│   │   └── [slug]/page.tsx     ← buying guides
+│   ├── health/
+│   │   └── [slug]/page.tsx     ← use-case articles
+│   ├── best/
+│   │   └── [slug]/page.tsx     ← roundup pages
+│   └── sitemap.ts              ← auto-generated sitemap
+├── components/
+│   ├── ui/                     ← shadcn components
+│   ├── ArticleLayout.tsx       ← shared article wrapper
+│   ├── ReviewCard.tsx          ← product review card
+│   ├── ComparisonTable.tsx     ← side-by-side comparison
+│   ├── AffiliateButton.tsx     ← CTA button with disclosure
+│   ├── RatingStars.tsx         ← star rating display
+│   ├── ProsCons.tsx            ← pros/cons list
+│   ├── KeywordBox.tsx          ← target keyword highlight
+│   └── SchemaMarkup.tsx        ← JSON-LD schema injection
+└── lib/
+    ├── velite.ts               ← content queries
+    └── utils.ts                ← helpers
+```
+
+## Step 4 — Velite Config
+Create `velite.config.ts` with schemas for each content type:
+
+```typescript
+// velite.config.ts
+import { defineConfig, defineCollection, s } from 'velite'
+
+const reviews = defineCollection({
+  name: 'Review',
+  pattern: 'reviews/**/*.mdx',
+  schema: s.object({
+    title: s.string(),
+    slug: s.string(),
+    keyword: s.string(),
+    metaTitle: s.string().max(60),
+    metaDescription: s.string().max(160),
+    publishDate: s.string(),
+    wordCountTarget: s.number().optional(),
+    affiliateLinks: s.array(s.object({
+      brand: s.string(),
+      url: s.string(),
+      commission: s.string(),
+    })).optional(),
+    rating: s.number().min(1).max(10).optional(),
+    price: s.number().optional(),
+    status: s.enum(['stub', 'draft', 'published']).default('stub'),
+  }).transform(data => ({ ...data, slugAsParams: data.slug }))
+})
+
+// Same schema for: comparisons, guides, health, best
+export default defineConfig({
+  root: 'content',
+  output: { data: '.velite', assets: 'public/static' },
+  collections: { reviews, comparisons, guides, health, best }
+})
+```
+
+## Step 5 — Core SEO Components
+
+### Metadata Generation (app/reviews/[slug]/page.tsx)
+```typescript
+export async function generateMetadata({ params }): Promise<Metadata> {
+  const article = reviews.find(r => r.slug === params.slug)
+  return {
+    title: article.metaTitle,
+    description: article.metaDescription,
+    openGraph: { title: article.metaTitle, description: article.metaDescription },
+    alternates: { canonical: `https://[domain].com/reviews/${params.slug}` }
+  }
+}
+```
+
+### JSON-LD Schema per Page Type
+- Review pages: `Review` + `Product` schema
+- Comparison pages: `Article` + `FAQPage` schema
+- How-To pages: `HowTo` schema
+- FAQ pages: `FAQPage` schema
+- Roundup pages: `Article` + `ItemList` schema
+
+## Step 6 — Install BMAD in Project
+```bash
+# Copy BMAD core from global install
+BMAD_SRC="/home/hanzla/.nvm/versions/node/v24.7.0/lib/node_modules/bmad-method/bmad"
+cp -r "$BMAD_SRC/core" ./bmad/
+cp -r "$BMAD_SRC/_cfg" ./bmad/
+cp -r "$BMAD_SRC/docs" ./bmad/
+```
+
+Create `bmad/config.yaml`:
+```yaml
+user_name: Hanzla
+communication_language: English
+document_output_language: English
+output_folder: "{project-root}/docs"
+project_name: [SiteName]
+project_type: affiliate-content-site
+domain: [domain].com
+niche: [Niche Description]
+```
+
+## Step 7 — Create CLAUDE.md
+```markdown
+# [SiteName] — Project Instructions
+[standard project instructions with niche, tech stack, content paths]
+```
+
+## Step 8 — Hand Off to BMAD
+After site scaffolding is complete, activate BMAD master agent:
+"BMAD is now installed. To continue building:
+1. Open `/bmad/core/agents/bmad-master.md` in your Claude session
+2. BMAD will guide you through: component building, content writing, deployment
+3. Tell BMAD: 'Build the [SiteName] affiliate review site based on the research in /research/ and content stubs in /content/'"
+
+## Design Theme — Maple Light
+Every site built with this skill uses the Maple Light aesthetic:
+
+**Color Palette:**
+- Primary: `#C9622F` (burnt amber / maple)
+- Dark: `#1C2B2D` (deep forest)
+- Background: `#FAF6F1` (warm cream)
+- Accent: `#E8A96A` (golden amber)
+- Text: `#2D2D2D` (near black)
+
+**Typography:**
+- Headings: Playfair Display (serif, authoritative)
+- Body: Inter (clean, readable)
+
+**Design Principles:**
+- Warm, natural, premium feel
+- Cedar wood photography in heroes
+- Clean white space
+- Trust signals prominent (author bio, testing methodology, disclosure)
+- No dark mode (warm cream is the vibe)

package/rcode/skills/seo/seo-site-builder/rules/06-yolo.md

@@ -0,0 +1,127 @@
+# Rule: Yolo Mode — Full Autonomous Execution
+
+## Trigger Phrases
+User says any of:
+- "yolo"
+- "do all the work and report after you finish"
+- "do everything and report back"
+- "do it all"
+- "handle everything"
+- "just do everything"
+- "report when done"
+
+## NOT Triggers (do NOT activate yolo on these)
+- "solo yolo" — this was a one-off phrase, not a real trigger
+- "run autonomously" — too vague, ask for clarification
+- "full auto" — ask: "Do you mean yolo mode (full autonomous, report when done)?"
+
+## What Changes in Yolo Mode
+
+### Decision Rules (No User Approval Needed)
+| Decision | Yolo Behavior |
+|----------|-------------------|
+| Niche selection | Auto-pick #1 scored niche |
+| Domain name | Auto-pick top recommendation |
+| Research findings | Continue without showing intermediate results |
+| Article schedule | Auto-finalize based on DA + Weakspot sort |
+| Site tech stack | Always use Next.js 16.2.1 + Velite |
+| Content stubs | Generate all Tier 1+2 stubs automatically |
+| BMAD install | Install automatically |
+
+### What Yolo NEVER Skips
+Even in full auto mode, always:
+1. **Save all research files** — every agent writes its output file
+2. **Generate keywords.txt** — user needs this for Ads Planner
+3. **Tell user about keywords.txt** — they must manually run Ads Planner
+4. **Create image-prompt files** — user generates images separately
+5. **Report completion** — final summary of everything created
+
+### Announcement at Start
+When yolo is triggered, say:
+```
+🔴 YOLO MODE ACTIVATED
+
+I will now run the complete workflow autonomously:
+✓ Pick best niche automatically
+✓ Launch 10 research agents in parallel
+✓ Build complete master business plan
+✓ Generate article schedule from keyword data
+✓ Create all content stubs + image prompts
+✓ Init Next.js site + install BMAD
+
+I'll only stop if I hit a critical blocker.
+Two things you'll need to do manually:
+1. Paste keywords.txt into Google Keyword Planner
+2. Generate images from the prompt files
+
+Starting now...
+```
+
+## Yolo Execution Flow
+
+```
+1. Run niche discovery
+   → Score all candidates
+   → Pick #1 automatically
+   → Announce: "Selected: [Niche] (Score: XX/35)"
+
+2. Create project directory
+   → /home/hanzla/development/[niche-slug]/
+
+3. Launch ALL 10 research agents simultaneously
+   → Announce each agent as it completes (1-line summary)
+
+4. Fire synthesis agent
+   → Wait for master plan
+
+5. Extract + save keywords.txt
+   → Copy to clipboard
+
+6. Parse keyword data (if CSV exists)
+   → OR use research data directly
+
+7. Generate article-schedule.md
+   → 50+ articles, Tier 1-5
+
+8. Generate MDX stubs (Tier 1+2 articles)
+   → All frontmatter filled
+
+9. Generate image-prompt files
+   → One per article
+
+10. Init Next.js project
+    → Full structure
+
+11. Install BMAD
+
+12. Create CLAUDE.md
+
+13. Final report:
+"✅ SOLO YOLO COMPLETE
+
+Project: [domain].com
+Location: /home/hanzla/development/[slug]/
+Articles planned: [N]
+Stubs created: [N]
+Image prompts: [N] files in assets/image-prompts/
+
+YOUR ACTIONS NEEDED:
+1. 📊 Keywords in keywords.txt → paste into Google Keyword Planner
+2. 🖼️ Generate images from assets/image-prompts/ files
+3. 🚀 Open BMAD: bmad/core/agents/bmad-master.md
+
+Research: research/00-MASTER-BUSINESS-PLAN.md"
+```
+
+## Yolo Blocker Conditions
+STOP and ask user only if:
+- All top 10 niches score below 18/35 (market is oversaturated)
+- No affiliate programs found for selected niche
+- Domain name ideas all appear taken
+- Research agents return empty/error results 3+ times
+
+## Continue with Next Idea
+After completing one full workflow, if more niches exist in the queue:
+
+**Interactive:** "Research complete for [Niche 1]. Want me to research [Niche 2] next?"
+**Yolo:** Automatically start next niche research, announce: "Starting research for Niche #2: [Name]..."

package/rcode/skills/seo/seo-site-builder/templates/article-stub.mdx

@@ -0,0 +1,62 @@
+---
+title: "[Full Article Title with Primary Keyword]"
+slug: "[url-slug-lowercase-hyphens]"
+type: "review"
+keyword: "[exact target keyword]"
+metaTitle: "[Under 60 chars — keyword near start]"
+metaDescription: "[140-160 chars — keyword included, action-oriented]"
+publishDate: "YYYY-MM-DD"
+status: "stub"
+wordCountTarget: 2000
+tier: 1
+avgDA: 0
+weakspot: 0
+volume: 0
+intent: "commercial"
+affiliateLinks:
+  - brand: "[Brand Name]"
+    url: "AFFILIATE_URL_PLACEHOLDER"
+    commission: "X%"
+    cookieDays: 30
+rating: null
+price: null
+imageHero: "/images/[slug]/hero.jpg"
+imagePomptFile: "assets/image-prompts/[slug]-images.md"
+internalLinks:
+  - slug: "[related-article-slug]"
+    anchor: "[natural anchor text]"
+featured: false
+---
+
+{/* STUB — Replace with full article content */}
+{/* Target: [wordCountTarget] words */}
+{/* Primary keyword: [keyword] */}
+{/* Intent: [intent] — write accordingly */}
+
+## Overview
+
+[STUB: 200-word intro covering what this product/topic is, who it's for, and what we'll cover. Include primary keyword naturally in first 100 words.]
+
+## Key Specs
+
+[STUB: Table with key specifications]
+
+## Full Review
+
+[STUB: 800-1000 words covering: build quality, performance, EMF levels, heating, assembly, value]
+
+## Pros & Cons
+
+**Pros:**
+- [STUB]
+
+**Cons:**
+- [STUB]
+
+## Who Should Buy This
+
+[STUB: 150 words on ideal buyer profile]
+
+## Verdict
+
+[STUB: 200-word final verdict with affiliate CTA and rating]