@hanzlaa/rcode 4.3.3 → 4.4.0

package/rcode/skills/seo/seo-content-writer/references/title-formulas.md

@@ -0,0 +1,339 @@
+# Title and Headline Formulas
+
+Proven title structures that drive clicks and rankings. All formulas target 50-60 characters for optimal SERP display.
+
+## Numbered List Formulas
+
+**Pattern**: [Number] [Adjective] [Topic] [Qualifier/Benefit]
+
+### Examples
+- "7 Ways to Improve Your SEO Rankings in 2024"
+- "15 Content Marketing Strategies That Actually Work"
+- "10 Free Tools for Keyword Research (Beginner-Friendly)"
+- "21 Email Marketing Tips from Industry Experts"
+- "5 Quick Wins for Better Site Performance"
+
+**Why it works**: Numbers promise specific value, create curiosity, set clear expectations.
+
+**Best practices**:
+- Odd numbers (7, 9, 15) often outperform even numbers
+- Use specific numbers (13 vs "over 10")
+- Keep total under 30 for credibility
+- Match number to actual content sections
+
+---
+
+## How-To Formulas
+
+**Pattern 1**: How to [Achieve Goal] in [Timeframe]
+- "How to Rank on Google in 30 Days"
+- "How to Write SEO Content in Under 2 Hours"
+- "How to Build Backlinks in 2024"
+
+**Pattern 2**: How to [Achieve Goal] Without [Common Objection]
+- "How to Learn SEO Without Spending Money"
+- "How to Increase Traffic Without Paid Ads"
+- "How to Write Blog Posts Without Writer's Block"
+
+**Pattern 3**: How to [Achieve Goal] (Even If [Limitation])
+- "How to Rank #1 (Even If You're a New Website)"
+- "How to Get Backlinks (Even With Zero Audience)"
+- "How to Write SEO Content (Even If You Hate Writing)"
+
+**Pattern 4**: How to [Achieve Goal] Like [Authority/Expert]
+- "How to Write Email Copy Like a Professional Copywriter"
+- "How to Do Keyword Research Like an SEO Pro"
+- "How to Create Content Like Top Marketing Agencies"
+
+---
+
+## Question Formulas
+
+**Pattern 1**: What Is [Topic]? [Benefit/Hook]
+- "What Is Technical SEO? A Complete Guide"
+- "What Is GEO? Everything You Need to Know"
+- "What Is Content Marketing? (And Why It Matters)"
+
+**Pattern 2**: Why [Action] [Result]
+- "Why Most Content Marketing Strategies Fail"
+- "Why Your Website Isn't Ranking"
+- "Why Email Still Beats Social Media for ROI"
+
+**Pattern 3**: Should You [Action]? [Quick Answer]
+- "Should You Buy Backlinks? (Spoiler: No)"
+- "Should You Hire an SEO Agency? A Cost-Benefit Analysis"
+- "Should You Rewrite Old Content? When It Makes Sense"
+
+**Pattern 4**: When Should You [Action]
+- "When Should You Publish Blog Posts for Maximum Traffic"
+- "When Should You Update Your SEO Strategy"
+- "When Should You Remove Content from Your Website"
+
+---
+
+## Comparison Formulas
+
+**Pattern 1**: [Option A] vs [Option B]: Which Is Better?
+- "WordPress vs Webflow: Which Is Better for SEO?"
+- "Ahrefs vs SEMrush: Complete Comparison 2024"
+- "In-House SEO vs Agency: Which Should You Choose?"
+
+**Pattern 2**: [Option A] vs [Option B] (and the Winner Is...)
+- "Free Tools vs Paid SEO Software (and the Winner Is...)"
+- "Long-Form vs Short-Form Content (and the Winner Is...)"
+
+**Pattern 3**: [Option A] or [Option B] for [Audience/Goal]
+- "Shopify or WooCommerce for Small Business SEO"
+- "YouTube or TikTok for B2B Marketing"
+
+---
+
+## Ultimate Guide Formulas
+
+**Pattern 1**: The [Complete/Definitive] Guide to [Topic]
+- "The Complete Guide to Technical SEO"
+- "The Definitive Guide to Link Building in 2024"
+
+**Pattern 2**: [Topic]: The Ultimate Guide for [Audience]
+- "Content Marketing: The Ultimate Guide for Beginners"
+- "Local SEO: The Ultimate Guide for Small Businesses"
+
+**Pattern 3**: Everything You Need to Know About [Topic]
+- "Everything You Need to Know About Core Web Vitals"
+- "Everything You Need to Know About Schema Markup"
+
+---
+
+## Power Word Formulas
+
+**Pattern**: [Power Word] [Topic] [Qualifier]
+
+### Power Word Categories
+
+**Authority words**:
+- Proven, Tested, Verified, Certified, Official
+- Example: "Proven Strategies to Double Your Organic Traffic"
+
+**Urgency words**:
+- Now, Today, Fast, Quick, Instantly, Immediately
+- Example: "Quick Wins: Improve SEO Performance Today"
+
+**Value words**:
+- Free, Complete, Essential, Ultimate, Comprehensive
+- Example: "The Essential SEO Checklist for 2024"
+
+**Exclusivity words**:
+- Secret, Hidden, Little-Known, Insider, Exclusive
+- Example: "7 Little-Known SEO Tactics That Actually Work"
+
+**Emotional words**:
+- Amazing, Powerful, Effortless, Simple, Easy
+- Example: "Simple SEO Strategies for Busy Entrepreneurs"
+
+---
+
+## Before/After Formulas
+
+**Pattern 1**: From [Bad State] to [Good State]
+- "From Zero to 10K Monthly Visitors: A Case Study"
+- "From Page 5 to Position 1: How We Ranked a Competitive Keyword"
+
+**Pattern 2**: How I/We [Achieved Result] in [Timeframe]
+- "How I Increased Organic Traffic by 300% in 6 Months"
+- "How We Built 100 Backlinks in 30 Days"
+
+**Pattern 3**: [Bad State] → [Good State]: [Method]
+- "500 Visitors → 50K: Our Content Strategy Revealed"
+- "Struggling → Thriving: SEO Transformation Story"
+
+---
+
+## Mistake/Problem Formulas
+
+**Pattern 1**: [Number] [Topic] Mistakes [Qualifier]
+- "5 SEO Mistakes Killing Your Rankings"
+- "10 Content Marketing Mistakes to Avoid in 2024"
+- "3 Technical SEO Mistakes Even Pros Make"
+
+**Pattern 2**: Stop [Wrong Action]. Do [Right Action] Instead
+- "Stop Keyword Stuffing. Do This Instead"
+- "Stop Buying Backlinks. Try These 7 Tactics Instead"
+
+**Pattern 3**: Why Your [Topic] Isn't Working (And How to Fix It)
+- "Why Your Content Isn't Ranking (And How to Fix It)"
+- "Why Your SEO Strategy Isn't Working (And How to Fix It)"
+
+---
+
+## Checklist Formulas
+
+**Pattern 1**: [Topic] Checklist: [Number] Must-Have [Items]
+- "Technical SEO Checklist: 23 Must-Check Items"
+- "Blog Post Checklist: 15 Things to Check Before Publishing"
+
+**Pattern 2**: The [Timeframe] [Topic] Checklist
+- "The 2024 SEO Audit Checklist"
+- "The Pre-Launch Website Checklist"
+
+---
+
+## Curiosity Gap Formulas
+
+**Pattern 1**: The [Adjective] Reason [Unexpected Fact]
+- "The Surprising Reason Long Content Ranks Better"
+- "The Real Reason Your Blog Posts Aren't Getting Traffic"
+
+**Pattern 2**: [Number] Things About [Topic] That [Surprising Fact]
+- "7 Things About SEO That Google Doesn't Tell You"
+- "5 Things About Content Marketing That Changed in 2024"
+
+**Pattern 3**: What [Expert/Group] Know About [Topic] (That You Don't)
+- "What Top SEO Agencies Know About Link Building (That You Don't)"
+- "What Professional Copywriters Know About Headlines"
+
+---
+
+## Benefit-Driven Formulas
+
+**Pattern 1**: [Action] to [Specific Benefit]
+- "Optimize Your Title Tags to Double Your CTR"
+- "Write Better Meta Descriptions to Increase Organic Traffic"
+
+**Pattern 2**: [Action] for [Outcome] Without [Objection]
+- "Scale Content Production for Better Rankings Without Hiring Writers"
+- "Improve Site Speed for Higher Rankings Without Technical Skills"
+
+---
+
+## Authority/Credibility Formulas
+
+**Pattern 1**: [Expert/Authority] Says [Action/Tip]
+- "Google's John Mueller Says Internal Linking Is Underrated"
+- "SEO Experts Reveal Their Top Ranking Factor for 2024"
+
+**Pattern 2**: [Number] [Experts/Companies] Share [Topic]
+- "50 SEO Experts Share Their #1 Ranking Tip"
+- "10 Successful Brands Share Their Content Strategy"
+
+---
+
+## CTR Optimization Patterns
+
+### Elements That Increase CTR
+
+**Year/Timeframe** (signals freshness):
+- "SEO Trends for 2024"
+- "Best Practices [Updated January 2024]"
+
+**Brackets/Parentheticals** (adds extra context):
+- "Email Marketing Tips (That Actually Work)"
+- "SEO Tools [Free & Paid Options]"
+
+**Numbers** (promise specific value):
+- "7 Ways...", "15 Tools...", "3 Reasons..."
+
+**Power words** (emotional trigger):
+- Ultimate, Essential, Complete, Proven, Secret
+
+**Target audience** (relevance signal):
+- "...for Beginners", "...for Small Businesses", "...for B2B"
+
+### Elements That Decrease CTR
+
+- All caps (looks spammy)
+- Excessive punctuation (!!!, ???)
+- Clickbait without substance
+- Misleading promises
+- Generic titles ("SEO Tips", "Marketing Guide")
+
+---
+
+## Title Length Guidelines
+
+### Google SERP
+- **Optimal**: 50-60 characters
+- **Maximum**: ~70 characters before truncation
+- **Mobile**: ~55 characters (truncates earlier)
+
+### Social Media
+- **Twitter/X**: 70-100 characters for best engagement
+- **Facebook**: 40-50 characters (shorter performs better)
+- **LinkedIn**: 70-100 characters
+- **Pinterest**: 40-60 characters
+
+### Email Subject Lines
+- **Optimal**: 40-50 characters
+- **Mobile**: 30-40 characters (truncates earlier)
+- **Desktop**: Up to 60 characters
+
+---
+
+## Before/After Title Examples
+
+### Generic → Optimized
+
+**Before**: "SEO Tips"
+**After**: "7 SEO Tips That Increased Our Traffic by 300%"
+- Added: Number, specificity, proof
+
+**Before**: "How to Do Keyword Research"
+**After**: "How to Do Keyword Research (Free Tools & 5-Step Process)"
+- Added: Benefit, specificity, process indicator
+
+**Before**: "Content Marketing Guide"
+**After**: "The Complete Content Marketing Guide for Small Businesses [2024]"
+- Added: Completeness indicator, target audience, year
+
+**Before**: "What Is Technical SEO"
+**After**: "What Is Technical SEO? A Beginner's Guide to Better Rankings"
+- Added: Audience level, benefit
+
+**Before**: "WordPress vs Shopify"
+**After**: "WordPress vs Shopify: Which Is Better for SEO? (2024 Comparison)"
+- Added: Specific comparison angle, year
+
+**Before**: "Link Building Strategies"
+**After**: "15 White-Hat Link Building Strategies That Still Work in 2024"
+- Added: Number, quality signal, time relevance
+
+---
+
+## Title Writing Checklist
+
+Before finalizing any title, check:
+
+- [ ] Includes primary keyword naturally
+- [ ] Length is 50-60 characters
+- [ ] Matches search intent (informational/transactional/navigational)
+- [ ] Contains power word or emotional trigger
+- [ ] Specific and concrete (not vague)
+- [ ] Promises clear value or benefit
+- [ ] Accurate representation of content
+- [ ] No clickbait or misleading claims
+- [ ] Unique compared to competitors
+- [ ] Works when truncated on mobile
+
+---
+
+## Testing and Optimization
+
+### A/B Test These Variables
+- Number variations (7 vs 10 vs 15)
+- Power word placement (beginning vs end)
+- Question vs statement format
+- With/without year
+- With/without brackets
+- Different benefit framing
+
+### Track These Metrics
+- Click-through rate (CTR) in Search Console
+- Average position in SERPs
+- Time on page (engagement indicator)
+- Bounce rate (relevance indicator)
+
+### When to Rewrite Titles
+- CTR below 2% for top 3 positions
+- CTR below 5% for positions 4-10
+- Content is updated (add [Updated 2024])
+- Competitors have better titles
+- Ranking but not getting clicks

package/rcode/skills/seo/seo-growth-orchestrator/SKILL.md

@@ -0,0 +1,79 @@
+---
+name: seo-growth-orchestrator
+description: Orchestrator for growing organic traffic, rankings, and revenue with Claude — condensed from a 1000+ hour SEO practitioner. Use when the user wants to "grow traffic", "rank a local business", "get backlinks", "find guest post opportunities", "audit my Google Business Profile / GBP categories", "mine Google Search Console for page-2 keywords", "build service + city landing pages", "fix NAP / citations", "build a content engine", "automate blog + social", "add MRR with SEO", or asks for "SEO prompts", "SEO workflow", or mentions Arvo/Blotato automation. Also use when the user wants the "Goals Protocol" (slash /goal) for higher-quality structured outputs. This is the orchestration layer — it calls the granular skills (keyword-research, backlink-analyzer, serp-analysis, schema-markup-generator, internal-linking-optimizer, geo-content-optimizer, on-page-seo-auditor) rather than replacing them.
+metadata:
+  version: 1.0.0
+  source: "SEO Claude Masterclass (5 trainings / 6h condensed) — 1000+ hr practitioner"
+---
+
+# SEO Growth Orchestrator
+
+You are an SEO growth operator. Your job is not to "describe SEO" — it is to **run repeatable, evidence-backed workflows** that move a real business's traffic, rankings, and revenue, using Claude (plus the user's browser, GSC, and optional paid tools) as the engine.
+
+This skill is the **orchestration layer**. It encodes the *workflows, prompts, and judgment* of a high-volume practitioner and routes to the granular skills for deep mechanics (see Integration). Do not duplicate their internals here.
+
+## When to Use
+
+- "Grow my traffic / rankings / organic revenue"
+- "Get backlinks", "find guest post opportunities", "steal competitor backlinks"
+- "Rank my local business", "fix my Google Business Profile", "GBP categories"
+- "Mine Google Search Console", "page-2 keywords", "low-hanging keyword wins"
+- "Build service + city landing pages", "location page stack"
+- "Citation audit", "NAP consistency", "directory cleanup"
+- "Build a content engine", "automate blog and social", "30-day SEO plan", "$10k MRR with SEO"
+- "Use the Goals Protocol" / "/goal" for a high-quality structured deliverable
+
+## The Five Plays (and where each lives)
+
+| # | Play | Goal | Rules file |
+|---|------|------|-----------|
+| 1 | **Backlink acquisition** | Guest-post finder + competitor-backlink mining (+ optional exchange) | `rules/backlinks.md` |
+| 2 | **Local SEO 6-prompt stack** | Rank any local business (GBP, GSC, reviews, city pages, citations, content) | `rules/local-seo-stack.md` |
+| 3 | **Goals Protocol** | Make ANY output 5–20× better via worker→evaluator loop | `rules/goals-protocol.md` |
+| 4 | **Content + automation engine** | SEO blog → RSS → social, optionally fully automated | `rules/content-engine.md` |
+| 5 | **30-day MRR sprint** | Pick clusters, role-model study, day-by-day plan, ruthless cut | `rules/mrr-sprint.md` |
+
+The non-negotiable judgment layer — what to ALWAYS do and what to NEVER do — is in `rules/dos-and-donts.md`. **Read it before running any play.**
+
+## Operating Principles (the through-line of all 5 plays)
+
+1. **Clarify until 95% confident, then act.** Every strong prompt in this orchestrator ends with "ask me clarifying questions until you are ≥95% confident." Generic prompts → generic results. This is the single highest-leverage habit.
+2. **Evidence over vibes.** Pull real data (GSC, the live SERP, the actual competitor pages, real reviews) before recommending. AI-estimated volume/KD is a *starting hypothesis*, not ground truth — see don'ts.
+3. **GBP is the homepage for local.** For local-service businesses, the Google Business Profile is the money asset; the website is the backup. Sequence work accordingly.
+4. **Ship in 30 days, compound over 180.** New blog content rarely ranks in 30 days. GBP + service/city pages + reviews pay in 30; cornerstone blog content is the 60–180 day layer. Set this expectation explicitly with the user.
+5. **Pages rank, not sites.** No dedicated service×city page → no ranking for that query, however good the homepage is.
+6. **Automate the manual, never the judgment.** Use tools to remove copy-paste/formatting toil; keep the strategy, category choices, and quality bar human-reviewed.
+
+## Quick Start
+
+1. Ask which play fits, or infer from the request.
+2. Open `rules/dos-and-donts.md` + the relevant play file.
+3. Grab the matching prompt from `templates/` and fill the `{{placeholders}}`.
+4. Run it. If the deliverable must be airtight, wrap it in the **Goals Protocol** (`/goal`).
+5. Hand off concrete artifacts (CSV, page copy, schema, 30-day plan) — not advice.
+
+## Integration
+
+This skill ORCHESTRATES; it does not replace the granular skills. Call them for the deep mechanics:
+
+- **Related skills**: `keyword-research`, `backlink-analyzer`, `serp-analysis`, `competitor-analysis`, `content-gap-analysis`, `on-page-seo-auditor`, `technical-seo-checker`, `schema-markup-generator`, `internal-linking-optimizer`, `meta-tags-optimizer`, `geo-content-optimizer`, `rank-tracker`, `seo-content-writer`, `seo-site-builder`, `seo-audit`
+- **Plugin**: `claude-seo:seo-local`, `claude-seo:seo-maps`, `claude-seo:seo-backlinks`, `claude-seo:seo-cluster`, `claude-seo:seo-geo`, `claude-seo:seo-schema`
+- **Browser**: GSC/GBP scraping needs the Claude Chrome extension or `browser-harness` with the user logged in.
+
+## File References
+- **Dos & Don'ts (read first)**: `rules/dos-and-donts.md`
+- **Backlinks**: `rules/backlinks.md`
+- **Local SEO 6-prompt stack**: `rules/local-seo-stack.md`
+- **Goals Protocol**: `rules/goals-protocol.md`
+- **Content + automation engine**: `rules/content-engine.md`
+- **30-day MRR sprint**: `rules/mrr-sprint.md`
+- **Tools (Arvo / Blotato / etc.)**: `rules/tools.md`
+- **Prompt templates**: `templates/`
+
+## Validation Checklist
+- [ ] Identified which of the 5 plays the request maps to
+- [ ] Read `rules/dos-and-donts.md` before recommending
+- [ ] Used a clarifying-questions-to-95% prompt where applicable
+- [ ] Pulled real data (GSC / SERP / reviews / live pages) — not just AI estimates
+- [ ] Set the 30-day vs 180-day expectation honestly
+- [ ] Delivered a concrete artifact, not generic advice

package/rcode/skills/seo/seo-growth-orchestrator/rules/backlinks.md

@@ -0,0 +1,29 @@
+# Play 1 — Backlink Acquisition
+
+Goal: earn high volume of relevant, higher-DR backlinks → authority → traffic. Two methods; run both, they compound.
+
+## Method A — Claude backlink research workflow (free, evidence-based)
+
+Two sub-workflows in one run. Prompt: `templates/backlink-research.md`.
+
+**Workflow A — Guest-post finder.** For the user's niche, find sites that accept guest posts / contributions. For each: site, guest-post/contact URL, and a 0–100 quality score (site structure, blog presence, social presence, topical relevance). Output a ranked table + CSV. Then (optional, with Gmail connector) draft — *never auto-send* — a personalized outreach email per opportunity.
+
+**Workflow B — Competitor backlink mining.** For each competitor, list where they earn backlinks; surface the highest-authority, replicable ones for the user's niche, scored and ranked. Output a ranked table + CSV. Logic: if a competitor earned it, the user likely can too.
+
+**Data source:** connect Ahrefs API for accurate DR/traffic if available. If not, tell the model to *drop the Ahrefs connection and score natively from its own analysis* — works, just label scores as estimates (see don'ts). Re-run every 2–4 weeks; backlink profiles drift.
+
+Delegate the heavy mechanics to `backlink-analyzer` / `competitor-analysis` / `claude-seo:seo-backlinks`. This play is the orchestration + outreach layer on top.
+
+## Method B — Automated backlink exchange (optional, paid)
+
+Some platforms (e.g. Arvo, see `tools.md`) run a backlink exchange: you register your URL once; as members publish content, the platform inserts relevant, higher-DR external links pointing to matching member sites — including yours. Effectively hands-off daily backlinks.
+
+Caveats to surface to the user:
+- It's a paid third-party tool — present as optional, not required.
+- Only relevant + higher-DR placements are worthwhile; confirm the exchange enforces that.
+- Link exchanges sit in a gray area of Google's guidelines; keep volume natural and relevance tight. **This is not "buy backlinks"** (a hard don't) but treat it with the same caution — quality and relevance over volume.
+
+## Hand-off
+- `outreach-targets.csv` (guest posts: site, URL, score, draft email)
+- `competitor-backlinks.csv` (source, target competitor, DR/est, replicability score)
+- A prioritized "go-get-these-10-first" shortlist

package/rcode/skills/seo/seo-growth-orchestrator/rules/content-engine.md

@@ -0,0 +1,27 @@
+# Play 4 — Content + Automation Engine
+
+Goal: SEO-rich blog content → RSS → repurposed social, optionally fully automated. The chain: **Claude (brain) → blog/RSS (SEO-rich) → social posts → publish**.
+
+## The pipeline
+1. **Business context** — Claude asks for site URL, industry, niche, audience (clarify to ≥95%).
+2. **Competitive + opportunity research** — competitor gaps, weak spots, keyword clusters, ranking opportunities. (Delegate to `competitor-analysis`, `content-gap-analysis`, `keyword-research`, `serp-analysis`.)
+3. **Blog plan** — titles + primary/supporting keywords, mapped to search intent, interlinked with service pages and content clusters.
+4. **Generate content WITH the on-page SEO layer** — title/meta, H2/H3, internal links to money pages (from sitemap), external links to trusted sources, alt text, FAQ schema. **Raw Claude prose lacks this layer — you must add it.** Either:
+   - Use the granular skills: `seo-content-writer` + `on-page-seo-auditor` + `internal-linking-optimizer` + `schema-markup-generator`, OR
+   - Use a tool that bakes SEO structure in on publish (see `tools.md`).
+5. **RSS** — published posts populate the RSS feed.
+6. **Social repurposing** — pull the RSS feed, generate platform social posts that link back to the blog/money pages, attach images (pulled from the article or generated).
+7. **Publish/schedule** — manually, or via a publishing API on a weekly/monthly schedule.
+
+## Automation tiers (let the user pick by budget)
+- **Manual**: Claude produces plan + content; user pastes/publishes. $0 tooling.
+- **Semi**: Claude plans; a content tool builds SEO-structured articles; user reviews + auto-posts to site.
+- **Full**: content tool API auto-publishes articles → RSS → Claude builds social posts → social publishing API schedules them. Hands-off weekly run.
+
+## Save it as a skill / reusable run
+Once the workflow prompt is built, save it in Claude as a reusable skill (e.g. "SEO content autopilot") so it's one command each week. The practitioner runs it every Monday to schedule the week's blog + social.
+
+## Honesty guardrails (see dos-and-donts.md)
+- Tool APIs (Arvo content, Blotato social) are **paid third parties** — optional, disclose cost.
+- Human-review the plan and any factual claims before auto-publish.
+- Don't mass-publish thin posts to hit a number; quality + intent match.

package/rcode/skills/seo/seo-growth-orchestrator/rules/dos-and-donts.md

@@ -0,0 +1,40 @@
+# Dos & Don'ts (read before any play)
+
+The judgment layer. The plays tell you *what to run*; this tells you *how not to wreck the account or mislead the client*. Distilled from the practitioner's repeated warnings + standard SEO safety.
+
+## ✅ DO
+
+- **DO clarify to ≥95% confidence before producing.** End every research/strategy prompt with "ask me clarifying questions one at a time until you are ≥95% confident, then proceed." This is the #1 quality multiplier in the whole playbook.
+- **DO pull real data first.** GSC for keywords already ranking, the live SERP for competition, the actual competitor pages for gaps, real reviews for voice-of-customer. Recommendations ride on evidence, not the model's prior.
+- **DO treat GBP as the primary asset for local.** Categories, posts (weekly), reviews, and NAP consistency pay faster than any blog post.
+- **DO stagger GBP category changes.** Add the highest-correlation / lowest-risk categories first, then more a week later. Bulk-changing categories overnight looks manipulative and risks a flag. Keep the original primary category.
+- **DO build one dedicated page per service × city.** Pages rank, not sites. A location page stack is the highest-ROI build for multi-area local businesses.
+- **DO mine page-2 keywords (avg position 8–20).** They are the cheapest wins — already ranking, just need title/H1/first-100-words/internal-link nudges.
+- **DO use the customers' own language.** Mine 100+ reviews per competitor for fear/outcome/recommendation phrases and put that exact vocabulary in copy, GBP description, FAQs, and review-request scripts.
+- **DO set honest timelines.** GBP + service/city pages + reviews → results in ~30 days. New cornerstone blog content → 60–180 days. Say this out loud; never imply a brand-new post ranks in 30 days.
+- **DO keep a human in the loop on strategy.** Auto-format and auto-publish are fine; auto-deciding categories, claims, and what's "true about the business" is not.
+- **DO verify keyword volume/difficulty against a real source** (GSC, Ahrefs/SEMrush, or an API) before betting budget on it.
+- **DO scope the Goals Protocol to ONE deliverable** with output spec + numeric acceptance criteria + constraints + stop-and-verify (see `goals-protocol.md`).
+- **DO disclose tool costs and that some are paid third parties** (Arvo, Blotato) — present them as optional, not required.
+
+## ❌ DON'T
+
+- **DON'T trust AI-estimated search volume / KD as ground truth.** The transcript's own demos run "without Ahrefs connected — usually pretty close." *Usually close* ≠ correct. Label any non-API number as an estimate and verify before spending money or time on it.
+- **DON'T buy backlinks.** Explicitly called out as a don't. Earn them (guest posts, citations, local links, genuinely linkable content / exchanges within ToS).
+- **DON'T mass-publish thin blog posts.** 50 low-quality posts < a few strong cornerstone pages. Quality and search-intent match win.
+- **DON'T chase national/head keywords for a local business.** Stay local; you can't out-authority national players, and local intent converts.
+- **DON'T obsess over keyword tools** or endless research. Cap the research phase (≤48h in the sprint) and start publishing/optimizing.
+- **DON'T bulk-edit GBP** (categories, name, address) all at once or stuff keywords into the business name — flag/suspension risk.
+- **DON'T let NAP drift.** Name/Address/Phone must be byte-identical across every directory; abbreviations ("Ste" vs "Suite") count as inconsistencies.
+- **DON'T auto-send outreach email.** Draft only; the human reviews and sends. (Also a deliverability/spam safeguard.)
+- **DON'T ship AI blog content without the on-page SEO layer** (title/meta, H1–H3, internal links to money pages, alt text, schema). Raw Claude prose lacks it — add it via `on-page-seo-auditor` + `schema-markup-generator` + `internal-linking-optimizer`, or a tool that bakes it in.
+- **DON'T jam multiple goals into one Goals Protocol run.** #1 cause of failure. One goal, sequenced.
+- **DON'T present a tool as a silver bullet or shill it.** Recommend a workflow; name tools as one option with their tradeoffs and price.
+- **DON'T promise rankings.** SEO is probabilistic. Promise process and leading indicators (indexation, impressions, position movement, leads), not guaranteed positions.
+- **DON'T scrape GSC/GBP without the user's own logged-in session.** Use their authenticated browser tab; never enter credentials from a screenshot.
+
+## The 3 prompt-quality tests (apply to every prompt you write)
+
+1. **Stranger test** — could a stranger read the prompt and know exactly what "done" looks like? If they'd need follow-ups, the criteria are weak.
+2. **Spreadsheet test** — could every acceptance criterion be a spreadsheet column answered yes/no or a number? Fuzzy words fail.
+3. **Runway test** — if the agent misinterprets in the worst way, would your constraints catch it within minutes? If not, add guard rails.

package/rcode/skills/seo/seo-growth-orchestrator/rules/goals-protocol.md

@@ -0,0 +1,34 @@
+# Play 3 — The Goals Protocol (`/goal`)
+
+A meta-technique that makes ANY output dramatically better by adding a **worker → evaluator** loop instead of a one-shot prompt. Invoked in Claude with `/goal`. Use it whenever a deliverable must be airtight (keyword list, page spec, audit, plan).
+
+## How it differs from a normal prompt
+- Normal prompt: prompt in → output out. No check.
+- Goals Protocol: a **worker agent** iterates on the deliverable, then a fresh **evaluator agent** (no memory of the work) audits it against your acceptance criteria one-by-one and either marks complete or kicks it back to the worker (which may ask you for missing info). Result quality jumps because every criterion is independently verified.
+
+## The five components (GOALS)
+- **G — Goal statement**: one sentence, action verb first, names a single deliverable, fits on one line.
+- **O — Output spec**: exact filename, format, columns/schema. The evaluator can only audit what you specify.
+- **A — Acceptance criteria**: numbered checklist; every item answerable yes/no or a number.
+- **L — Limits / constraints**: forbidden actions, scope walls, resource caps (what NOT to do).
+- **S — Stop and verify**: force the agent to list evidence for each criterion before declaring done, then hand to the evaluator.
+
+Template: `templates/goal-protocol.md`. Tip: you can paste the template into Claude and say "here's what I'm trying to do — fill this in correctly," then run it.
+
+## Apply the 3 quality tests (from dos-and-donts.md)
+Stranger test · Spreadsheet test · Runway test. If the prompt fails any, tighten before running.
+
+## Common failure modes → fixes
+| Failure | Fix |
+|--------|-----|
+| Multiple goals jammed into one | Split into sequenced single-goal runs |
+| Too vague / no end state | Add concrete output spec + numeric criteria |
+| Missing acceptance criteria | Add the numbered yes/no checklist |
+| Overpacked (e.g. 47 criteria) | Trim to the few that define "done" |
+| Agent runs out of tokens | Reduce scope per run |
+
+## Good vs bad (illustrative)
+- ❌ "Find me some keywords for my dog grooming business and make sure they're good for SEO." — vague verb, no output spec, no criteria, no constraints, no stop condition → the agent guesses wildly.
+- ✅ Goal: build a local SEO keyword list for `{{business}}` in `{{location}}`. Output: `keywords.csv` with columns keyword, intent, est_volume, est_KD, suggested_title. Acceptance: ≥25 keywords; ≥15 with KD ≤30; every keyword has a local modifier. Limits: no out-of-area keywords; no titles >60 chars; max 50 keywords. Stop & verify: list each criterion with evidence, then hand to evaluator.
+
+This play composes with all others — wrap any of the local prompts or the content engine in `/goal` when the stakes warrant it.

package/rcode/skills/seo/seo-growth-orchestrator/rules/local-seo-stack.md

@@ -0,0 +1,35 @@
+# Play 2 — Local SEO 6-Prompt Stack
+
+Six prompts that rank a local business. Run in this order; each produces a concrete artifact. Full prompts in `templates/local-*.md`.
+
+## 1. GBP category audit  → `templates/local-gbp-categories.md`
+Fastest local win. Analyze the business's Google Business Profile categories, reverse-engineer every category competitors use, and recommend exactly which to add and **in what order** so Google doesn't flag rapid changes.
+- Output: "add today" (highest correlation, lowest risk) vs "add next week" lists; keep the existing primary; categories the model refuses to recommend without live verification; rationale.
+- Why: most listings have a primary category from years ago and zero secondaries — leaving rankings on the table. GBP is what shows first in Maps/Search and what ChatGPT/Perplexity pull from.
+
+## 2. GSC page-2 keyword gold mine  → `templates/local-gsc-goldmine.md`
+Needs the user logged into Google Search Console in Chrome; Claude reads it via the Chrome extension / browser-harness (their session).
+- Filter: avg position 8–20, has impressions, **not** branded, commercial/transactional intent. Sort by impressions, take top 20.
+- Per keyword: forensic audit (keyword in title? H1? first 100 words? word count vs top-5? internal links pointing in? meta description? schema present?).
+- Output: rewritten title + meta per page, plus a **30-day sprint** (week-by-week, current→new state per fix).
+
+## 3. Competitor review sentiment mining  → `templates/local-review-mining.md`
+Read ~100 reviews per competitor (200+ total). Extract: 5-star emotional language, top-10 *before-state fears*, top-15 *outcome phrases*, exact *recommendation language*, and the 5-star vs 3-star *trust-trigger gap*.
+- Output (3 deliverables): rewritten GBP description (A/B/C), homepage headline + subhead variations, FAQ entries + a review-request script — all in the customers' own words.
+
+## 4. Service + city landing page builder  → `templates/local-city-pages.md`
+Pages rank, not sites. Build the full location-page stack in one session.
+- Output: site audit + **gap matrix** (which service×city pages exist vs missing), publishing order (what first + why), full page copy per page, FAQ schema, internal-linking plan, image alt text.
+- Rule: one page per service × each surrounding city/neighborhood the business wants jobs in.
+
+## 5. Local citation audit + directory cleanup  → `templates/local-citations.md`
+Find everywhere the NAP (name/address/phone/website) appears; flag inconsistencies (abbreviations count). Industry-specific directories (Thumbtack, Angi, HomeAdvisor, etc.).
+- Output: high-priority fixes with exact corrected values + where to fix; missing high-value listings to claim (easy authority + backlinks).
+
+## 6. Blog / keyword / content engine  → see `rules/content-engine.md`
+Business context → competitive + opportunity analysis → blog titles + keyword list mapped to search intent, interlinked with service pages and content clusters. Optionally fully automate via tools (Play 4).
+
+## Sequencing logic (Nick-Huber-style 90-day shape)
+Days 1–14: GBP setup + NAP cleanup + first ~10 reviews. Days 15–45: one ultra-strong service/money page + 3–5 neighboring city pages + a trust page + 2 cornerstone posts. Ongoing: review velocity, weekly GBP posts, local link building. The 30-day payoff comes from GBP + pages + reviews — **not** new blog posts (those are 60–180 days).
+
+Delegate mechanics to: `claude-seo:seo-local`, `claude-seo:seo-maps`, `on-page-seo-auditor`, `schema-markup-generator`, `internal-linking-optimizer`.

package/rcode/skills/seo/seo-growth-orchestrator/rules/mrr-sprint.md

@@ -0,0 +1,27 @@
+# Play 5 — 30-Day MRR Sprint
+
+Goal: add organic-driven monthly recurring revenue fast using SEO + free organic traffic. Three prompts + one ruthless-cut bonus. Prompts in `templates/mrr-*.md`.
+
+## 1. Opportunity finder  → `templates/mrr-opportunity-finder.md`
+Find SEO content angles that could generate the first ~$1,000 MRR within 90 days. Each angle must:
+- target keywords realistically rankable **without high domain authority**,
+- tie to a product/service/affiliate offer the user can monetize,
+- have proven search intent,
+- be publishable within 7 days.
+End with: ask clarifying questions until ≥95% confident (platform, monetization path, trade/niche, geographic scope, client value, existing site). Output: 1–2 angles with target keywords, why it converts, and the math to $1k.
+
+## 2. Role-model study  → `templates/mrr-role-model.md`
+"Who is the single most successful SEO operator/site owner I should study for THIS exact situation? Limit to one. Show exactly what they did in their first 90 days that I should replicate."
+- Output: one named operator + their first-90-day playbook mapped to the user's niche, plus an explicit **"what they do NOT do"** list (don't chase national keywords, don't buy backlinks, don't pump 50 posts, don't obsess over keyword tools) and the single biggest lesson (for local: GBP is the homepage, the website is the backup).
+
+## 3. 30-day money plan  → `templates/mrr-30day-plan.md`
+"Pretend you're that operator. Create a 30-day SEO content plan, <$500 upfront, that gets me to my first ranking or first dollar of organic revenue as fast as possible. Publishing starts day one. Research phase ≤48h."
+- Reality check the model must state: a brand-new blog post will NOT rank in 30 days via pure SEO. In 30 days you rank GBP + service pages locally, get found, close one job (= first dollar). Blog is the 60–180-day layer; do both, but GBP work is what pays in 30.
+- Output: budget allocation + day-by-day plan (morning/afternoon blocks), with weeks: 1–2 foundation, 3–10 reviews + neighborhood pages, 11–20 blog + local link plays, 21–30 close-the-loop + measure (GSC triage). Expected 30-day result stated honestly (e.g. 15–25 reviews, money page indexed, GBP in local 3-pack, N inbound leads, 1–3 closed jobs).
+
+## Bonus — ruthless cut  → `templates/mrr-cut.md`
+"Ask me 3 yes/no questions to identify what will stop me publishing this month, then cut the plan in half and keep only what directly leads to rankings and revenue."
+- Purpose: strip AI's rosy fluff, expose the real blocker (will you commit 1h/day? will you ship an imperfect page without a week of editing? will you text 50 past customers?), and produce a streamlined plan.
+
+## Through-line
+Each prompt feeds the next (90-day → 30-day → cut). Always clarify to 95%, always state the 30 vs 180 day truth, always end with concrete next actions, not encouragement.

package/rcode/skills/seo/seo-growth-orchestrator/rules/tools.md

@@ -0,0 +1,19 @@
+# Tools (optional, mostly paid third parties)
+
+The masterclass leans on three tools. **None are required** — every play has a manual path. Always disclose cost and present as one option. Never shill.
+
+| Tool | Role | Notes / cautions |
+|------|------|------------------|
+| **Claude** (+ Chrome extension / browser-harness) | The brain: research, prompts, Goals Protocol, GSC/GBP reading | GSC/GBP scraping needs the user's own logged-in session. Free path for everything. |
+| **Arvo** (arvow) | SEO content generation with on-page structure baked in (meta, H2/H3, internal links from sitemap, alt text, schema), site integration, auto-publish; also runs an LLM brand monitor, SEO reports, and a **backlink exchange** | Paid; API tier costs more. Value = it adds the on-page SEO layer Claude prose lacks. Verify exchange enforces relevance + higher-DR. Treat exchange with link-scheme caution. |
+| **Blotato** (blotado/blowto) | Social publishing API: schedule posts, auto-generate infographics from blog concepts | Paid. Pulls from RSS to repurpose SEO content into social. |
+
+## Free substitution map (when the user won't pay)
+- Arvo content layer → `seo-content-writer` + `on-page-seo-auditor` + `internal-linking-optimizer` + `schema-markup-generator` + `meta-tags-optimizer`, then paste/publish manually.
+- Arvo backlink exchange → Play 1 Method A (guest-post + competitor mining outreach).
+- Arvo brand/LLM monitor → `claude-seo:seo-geo` + manual prompt checks.
+- Blotato → manual scheduling, or the user's existing social scheduler.
+- Ahrefs API (for backlink/volume accuracy) → native Claude estimation **labeled as estimate** + verify before spending (see dos-and-donts.md).
+
+## Honesty rule
+When recommending any of these, state: it's paid, it's optional, here's the free path, and here's the specific thing it saves you. The user decides.