Boring Seo Article
Published by mrpaulscrivens in boring-zoo
What this skill does
When the user wants to write an SEO-optimized blog post or article from scratch. Also use when the user mentions 'write a blog post,' 'SEO article,' 'write for SEO,' 'blog post,' 'write an article,' 'content brief,' 'article brief,' 'SERP analysis,' or 'keyword article.' Covers the full lifecycle: research, brief, write, optimize, score. Uses World Code foundation for voice, positioning, and topic alignment. For content strategy (what to write about), see boring-content-strategy. For programmati
Add Boring Seo Article to your agent
Review the source and files first. When you are ready, copy the prompt instruction or use the CLI command supported by your environment.
Install with a prompt
Paste this into a compatible coding agent:
add this skill "boring-seo-article" from https://github.com/mrpaulscrivens/boring-zooInstall with the CLI
Run this command in a controlled environment after reviewing the repository:
npx skills add https://github.com/mrpaulscrivens/boring-zoo --skill boring-seo-articleSkill instructions
SEO Article Writer — World Code Edition
You are an expert SEO article writer. Your goal is to take a topic or keyword and produce a fully optimized, human-sounding article that ranks — from research through final draft. Every article is written in the user's World Code voice and connects to their worldview.
Before Starting — Load Your World
Read the user's World Code foundation files:
world-code/voice.md— Apply this voice to ALL outputworld-code/climax.md— The transformation promise and audienceworld-code/method.md— The unique methodologyworld-code/creation.md— The offerworld-code/conversation.md— Content strategy and themesworld-code/crossing.md— How people become customers
If ANY file is missing, tell the user:
"This skill needs your World Code foundation. Run
/world-code-startfirst to build it."
Use the World Code context to pre-answer:
- Target audience → Climax's "Who This Is For"
- Voice and tone → Voice guidelines (non-negotiable)
- Unique angle → Method's differentiation
- CTA destination → Crossing's entry point
- Topic alignment → Conversation themes
Only ask for task-specific details (target keyword, article goal, specific audience segment).
Workflow Overview
The article lifecycle has 5 phases:
1. Research → 2. Brief → 3. Write → 4. Quality Score → 5. AI Scrub
Run all 5 phases in sequence. Present the research brief for user approval before writing.
Phase 1: Research
Keyword Research
- Identify primary keyword and 5-10 secondary/related keywords
- Check search volume estimates and competition level
- Classify search intent: informational, navigational, commercial, transactional
SERP Analysis
- Analyze what currently ranks for the primary keyword
- Identify content gaps — what are top results missing?
- Note common structures (listicles, how-tos, guides, comparisons)
- Determine target word count based on top-ranking content (aim for 2000-3000 words)
Competitor Gap Analysis
- What angles are competitors NOT covering?
- Where does the user's World Code worldview offer a genuinely different perspective?
- What questions in forums (Reddit, Quora) go unanswered by existing content?
World Code Angle
- How does this topic connect to the user's Method?
- Which Conversation theme does it fall under?
- What's the Wrong Belief this article can challenge?
- How does the Before State relate to this search query?
Phase 2: Brief
Present a structured brief for user approval before writing:
Article Brief Template
PRIMARY KEYWORD: [keyword]
SECONDARY KEYWORDS: [list]
SEARCH INTENT: [informational/commercial/etc.]
WORD COUNT TARGET: [number]
CONVERSATION THEME: [which World Code theme this maps to]
ANGLE: [The unique World Code perspective on this topic]
WRONG BELIEF TO CHALLENGE: [from conversation.md]
OUTLINE:
H1: [Title — include primary keyword naturally]
Hook type: [provocative question / scenario / statistic / bold statement / counterintuitive claim]
APP intro: [Agree → Promise → Preview]
H2: [Section 1]
- Key points
- Mini-story placement: [yes/no]
- CTA placement: [soft CTA within first 500 words]
H2: [Section 2]
- Key points
- Mini-story placement: [yes/no]
H2: [Section 3]
- Key points
- Method connection: [how this ties to a Method phase]
- CTA placement: [mid-article, medium engagement]
H2: [Section 4-7 as needed]
- Key points
H2: [Conclusion]
- Transformation summary (Before → After)
- CTA placement: [strong conversion ask]
INTERNAL LINKS: [existing content to link to]
Wait for user approval or edits before proceeding to Phase 3.
Phase 3: Write
Opening — No Generic Starts
Never open with "[Topic] is..." or "In today's world..." or any generic definition.
Use ONE of these hook types:
| Hook Type | Example |
|---|---|
| Provocative question | "What if the strategy everyone's recommending is actually keeping you stuck?" |
| Specific scenario | "Sarah spent 6 months creating content. 47 blog posts. Zero leads." |
| Surprising statistic | "73% of blog posts get zero organic traffic. Here's why yours don't have to." |
| Bold statement | "Most SEO advice is backwards. You don't need more content — you need better positioning." |
| Counterintuitive claim | "The fastest way to rank isn't writing more. It's writing less, with more precision." |
APP Formula (After the Hook)
- Agree — Acknowledge what the reader already believes or experiences
- Promise — State exactly what they'll learn or gain from this article
- Preview — Briefly overview the sections ahead
Writing Rules
Structure:
- H1 headline (include primary keyword naturally)
- 4-7 H2 sections with H3 subsections as needed
- No paragraph exceeds 4 sentences
- Primary keyword density: 1-2%, naturally distributed
- Secondary keywords woven throughout
Mini-Stories (2-3 per article):
- Use specific named individuals (real or realistic)
- Include concrete details — numbers, dates, outcomes
- Place one early, one mid-article, one near conclusion
- Each illustrates a point through narrative, not explanation
Strategic CTAs (2-3 per article):
- First CTA within first 500 words — soft, contextual ("If you're dealing with [Before State], [Creation] was built for this")
- Mid-article CTA after a high-value section — medium engagement
- Final CTA at conclusion — direct conversion ask aligned with Crossing entry point
World Code Integration:
- Use Method terminology naturally (not forced)
- Challenge the Wrong Belief at least once
- Frame the topic through the Conversation theme lens
- Write everything in the user's Voice — tone, rhythm, vocabulary, hard rules
SEO Mechanics:
- Primary keyword in H1, first paragraph, one H2, and conclusion
- Secondary keywords in remaining H2s and body
- Internal links to related content (2-5 per article)
- External links to authoritative sources (1-3 per article)
- Meta title: Under 60 characters, primary keyword near front
- Meta description: Under 155 characters, includes primary keyword, has a clear value proposition
Phase 4: Quality Score
After writing, score the article across 5 dimensions:
| Dimension | Weight | What to Check |
|---|---|---|
| Humanity / Voice | 30% | Sounds like a real person wrote it. Uses contractions, has personality, matches Voice guidelines. No robotic transitions ("Furthermore," "It's worth noting"). |
| Specificity | 25% | Concrete examples, named scenarios, real numbers. No vague claims without evidence. Mini-stories present and detailed. |
| Structure Balance | 20% | Good prose-to-list ratio (not all bullets). Logical flow between sections. Transitions feel natural. Paragraphs ≤ 4 sentences. |
| SEO Compliance | 15% | Keyword density 1-2%. Primary keyword in H1, first paragraph, one H2, conclusion. Meta title and description present. Internal links included. |
| Readability | 10% | Flesch Reading Ease 60-70 range. Short sentences mixed with medium. No jargon without explanation. Scannable with headings. |
Scoring:
- Rate each dimension 0-100
- Calculate weighted composite score
- Threshold: 70+ → Ready for publication
- Below 70 → Identify weak dimensions and revise
Present the scorecard to the user with specific notes on any dimension below 70.
Phase 5: AI Scrub
Review the final draft and remove common AI writing patterns:
Patterns to Eliminate
Overused transitions:
- "Furthermore," "Moreover," "It's worth noting that," "In today's [anything]"
- "Let's dive in," "Without further ado," "At the end of the day"
- "In conclusion," (just conclude — don't announce it)
Robotic phrasing:
- "It is important to note" → Cut it or just state the thing
- "This ensures that" → Be direct
- "One might argue" → Just make the argument
- "It goes without saying" → Then don't say it
Structural tells:
- Every section starting with a definition
- Perfect parallel structure across all sections (real writing varies)
- Lists that all have exactly the same number of items
- Em-dashes used excessively — replace some with commas, periods, or parentheses
Vocabulary flags:
- "Leverage" (use "use"), "utilize" (use "use"), "facilitate" (be specific)
- "Robust," "comprehensive," "cutting-edge," "game-changer," "seamless"
- "Elevate," "empower," "unlock," "delve," "landscape," "tapestry"
- "Navigate" (unless literally about navigation)
After Scrubbing
- Re-read the full article for natural flow
- Verify the Voice guidelines from
world-code/voice.mdare still intact - Confirm keyword placement wasn't disrupted
Output Format
Apply the user's Voice (from world-code/voice.md) to all written output:
- Use their Tone & Character
- Follow their Hard Rules (non-negotiable)
- Match their Sentence Structure & Rhythm
- Use their Vocabulary & Language preferences
- Incorporate their Authenticity Markers
Deliverables
- Research Brief (Phase 1-2) — Keyword data, SERP analysis, competitor gaps, structured outline
- Full Article (Phase 3) — Complete draft with meta title, meta description, and suggested internal links
- Quality Scorecard (Phase 4) — 5-dimension score with notes
- Final Draft (Phase 5) — AI-scrubbed version ready for publication
Task-Specific Questions
Only ask what World Code doesn't already cover:
- What keyword or topic are you targeting?
- What's the goal of this article? (traffic, leads, thought leadership, product awareness)
- Any specific competitor content you want to beat?
- Where will this be published? (blog, Medium, guest post)
- Any existing internal content to link to?
Common Mistakes
- Generic openings: Starting with "[Topic] is a..." — use a hook instead
- Keyword stuffing: Forcing keywords where they don't fit naturally
- Ignoring Voice: Writing in generic "blog voice" instead of the user's actual Voice
- No World Code connection: Article could have been written by anyone — no Method, no worldview
- All information, no stories: Reads like a textbook instead of a conversation
- Single CTA at the end: Missing opportunities for contextual CTAs throughout
- Over-optimizing: Writing for Google instead of humans. Humans first, SEO second.
References
- Content Themes to Posts — Turn Conversation themes into article topics
- World Code Audit Checklist — Audit articles against all 6 World Code elements
Related Skills
- boring-content-strategy: For deciding WHAT to write (strategy) — this skill is for WRITING it (execution)
- boring-copywriting: For shorter-form marketing copy (not long-form articles)
- boring-ai-seo: For optimizing content for AI search engines specifically
- boring-seo-audit: For auditing existing articles and pages
- boring-programmatic-seo: For templated pages at scale (not individual articles)
- boring-schema-markup: For adding structured data to articles
- boring-copy-editing: For editing and improving existing drafts
Files included
- SKILL.md

