agentleFS
RomeoSEORomeoSEO/seo-skills/skills/seo-blog-writer/SKILL.md

seo-blog-writer

Skill7 starsChanged 14 days agoMITView on GitHub

Write an SEO blog post, how-to, listicle, comparison, refresh or localized article from an approved brief, answer-first and without generic AI prose. Use for "write this blog post", "draft the article from this brief", "refresh this page", "write a listicle", "make this sound less like AI".

8 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use content-repurposer.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.

The file

Show its rules and text

We split this skill into the rules it gives an agent

38 rules. 0 are common, 35 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 35

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write for one reader, not a generic audience segment.

    line 61
  2. Distinctive:

    Use simple words and short paragraphs.

    line 62
  3. Distinctive:

    Do not write filler transitions between every section.

    line 64
  4. Distinctive:

    Do not sound like a neutral encyclopedia when the reader needs a decision.

    line 65
  5. Very specific:

    Do not invent pricing, capabilities, testimonials, review scores, privacy claims, competitor claims, or legal/medical/financial claims.

    line 67
  6. Distinctive:

    Before output, run a de-robot pass:

    line 71
  7. Distinctive:

    Remove generic openings like "In today's fast-paced world", "Let's dive in", "Unlock", "Harness", "Game-changer", "Ever-evolving", "It is …

    line 73
  8. Distinctive:

    Remove repetitive sentence starts.

    line 74
  9. Distinctive:

    Use honest drawbacks for tools and comparisons.

    line 77
  10. Very specific:

    Avoid hedging every claim with may/might/could unless uncertainty is real.

    line 78
  11. Distinctive:

    Keep conclusions useful, not a rephrased intro.

    line 79
  12. Distinctive:

    Use punctuation accepted by the site's style guide.

    line 80
  13. Distinctive:

    Every draft must include:

    line 84
  14. Distinctive:

    Use tables/bullets only when they make scanning easier.

    line 105
  15. Distinctive:

    Add FAQPage markup only if the brief requires it.

    line 112
  16. Distinctive:

    Use when the reader wants instructions.

    line 118
  17. Distinctive:

    Start steps with verbs.

    line 134
  18. Distinctive:

    Keep background after the core process.

    line 135
  19. Distinctive:

    Add screenshots where steps involve software UI.

    line 136
  20. Distinctive:

    Use when the reader is comparing multiple tools/options.

    line 140
  21. Distinctive:

    Every featured tool needs a homepage/relevant link.

    line 159
  22. Distinctive:

    Every featured tool needs an image/screenshot unless the asset policy says otherwise.

    line 160
  23. Distinctive:

    Do not rank the site's own product #1 by default.

    line 161
  24. Distinctive:

    Use when the reader is choosing between named products or searching competitor alternatives.

    line 166
  25. Distinctive:

    Do not claim market superiority without evidence.

    line 183
  26. Distinctive:

    Use careful language for reviews, Reddit sentiment, accuracy, speed, CPU/RAM, storage, privacy, and pricing.

    line 184
  27. Distinctive:

    Avoid defamatory or aggressive claims.

    line 185
  28. Distinctive:

    Use examples from that job, not generic productivity copy.

    line 201
  29. Distinctive:

    Avoid medical/legal claims unless sourced and approved.

    line 202
  30. Distinctive:

    Use when improving an existing ranking page.

    line 206
  31. Distinctive:

    Do not restructure ranking-sensitive pages beyond the brief.

    line 213
  32. Distinctive:

    Use researched local keywords and slugs, not direct translations.

    line 220
  33. Distinctive:

    Use the approved locale voice, pronouns, CTA language, currency, and claims.

    line 222
  34. Distinctive:

    No fake testimonials, review counts, benchmarks, awards, or customer logos.

    line 245
  35. Distinctive:

    No unsupported regulated claims.

    line 246
  36. Distinctive:

    Use approved asset libraries before fresh generation/capture.

    line 251
  37. Distinctive:

    If assets are missing, output a clear asset queue rather than pretending they exist.

    line 253
  38. Very specific:

    Run seo-content-qa with the draft, the brief and the fact-check notes.

    line 315

“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

SEO blog writer

Write SEO blog, how-to, listicle, comparison, refresh, and localized drafts from a detailed, approved brief, and hand back a publishable draft or refresh package with its meta, schema, links, asset list and fact-check notes. This skill does not pick topics from scratch. If the brief is weak, contradictory, or missing critical facts, stop and return a brief defect report instead of producing generic content.

Romeo's rule: Answer the question, admit the drawbacks, and cut anything a reader would skim past; nobody ever wished an article had more throat-clearing.

Use this when

  • A detailed brief is approved and you need a publishable draft or refresh package.
  • You need a how-to, listicle, comparison or alternatives page, profession/use-case guide, landing-style article or localized article written to a brief.
  • An existing ranking page needs a careful refresh that leaves its URL and ranking-sensitive parts alone.
  • A draft reads like generic AI prose and needs a de-robot pass.

Don't use this for

  • Writing without a brief, or from a thin one → seo-content-brief.
  • Picking the topic or keyword → seo-opportunity-research.
  • Scoring the draft before it ships → seo-content-qa.
  • Putting the finished draft into a CMS → cms-publishing-adapter.

What you need

Required

  • Approved content brief.
  • Content type: how-to, listicle, comparison, alternatives, profession/use-case guide, refresh, landing-style article, localized article.
  • Target keyword cluster and search intent.
  • Brand voice and banned phrases.
  • Product facts and prohibited claims.
  • Required links and conversion page.
  • Required assets or image policy.
  • Schema requirements.
  • Output format: Markdown, HTML, document draft, CMS JSON, or handoff package.

Helpful

  • SERP summaries.
  • Existing draft or URL for refresh.
  • Screenshots/images.
  • Competitor fact sources.
  • Localization glossary.
  • CMS constraints.

No tools?

This skill needs no paid tools: the brief, the product facts and a text editor are enough. If you can't fetch URLs, ask the user to paste the existing page (for a refresh), the competitor pages behind any claim you need to make, and any source you must cite. Get screenshots from the user, or list them in the missing asset queue; never describe an image you don't have. Check any JSON-LD you write by pasting it into the Schema Markup Validator (validator.schema.org) or Google's Rich Results Test.

Ground rules

  • Answer first. Background comes later.
  • Write for one reader, not a generic audience segment.
  • Use simple words and short paragraphs.
  • Be specific early: numbers, exact tools, real situations, concrete tradeoffs.
  • Do not write filler transitions between every section.
  • Do not sound like a neutral encyclopedia when the reader needs a decision.
  • Be honest about product fit. Credibility beats forced promotion.
  • Do not invent pricing, capabilities, testimonials, review scores, privacy claims, competitor claims, or legal/medical/financial claims.

Mandatory anti-AI voice rules

Before output, run a de-robot pass:

  • Remove generic openings like "In today's fast-paced world", "Let's dive in", "Unlock", "Harness", "Game-changer", "Ever-evolving", "It is important to note", "Whether you're a...".
  • Remove repetitive sentence starts.
  • Break perfect parallel structure across list items when it sounds synthetic.
  • Replace vague adjectives like powerful, robust, seamless, cutting-edge with specific benefits or delete them.
  • Use honest drawbacks for tools and comparisons.
  • Avoid hedging every claim with may/might/could unless uncertainty is real.
  • Keep conclusions useful, not a rephrased intro.
  • Use punctuation accepted by the site's style guide. If em dashes are banned, remove all em dashes.

Universal structure

Every draft must include:

  1. TL;DR near the top.
    • Listicle: all items with links and one-line best-fit/value.
    • How-to: 2-3 sentence summary of what to do.
    • Comparison: quick verdict and who each option fits.
    • Refresh: concise answer that improves the existing page without disrupting ranking-sensitive areas.
  2. Answer-first block immediately after TL;DR.
    • 2-4 short sentences.
    • Directly answer the main query.
    • Make it self-contained and quotable.
  3. Short intro.
    • 2-3 short paragraphs max.
    • No generic scene-setting.
    • State the problem and what the reader will get.
  4. Main sections.
    • H2/H3 hierarchy follows the brief.
    • Important H2 sections open with a short direct answer before detail.
    • Use tables/bullets only when they make scanning easier.
  5. Conclusion.
    • Short, decision-oriented, with a clear next action.
  6. FAQs.
    • Concise, direct, matched to search questions.
    • Add FAQPage markup only if the brief requires it; Google retired FAQ rich results in May 2026, so it wins no rich result there.

Content type templates

How-to guide

Use when the reader wants instructions.

Structure:

  • TL;DR.
  • Answer-first summary.
  • Requirements/prerequisites only if needed.
  • Steps near the top. Do not bury steps below long background sections.
  • Each step: action, short explanation, expected result.
  • Troubleshooting.
  • Alternatives or built-in options if SERP expects them.
  • FAQs.
  • Related internal links.

Rules:

  • Start steps with verbs.
  • Keep background after the core process.
  • Add screenshots where steps involve software UI.

Listicle

Use when the reader is comparing multiple tools/options.

Structure:

  • TL;DR with all tools and links.
  • Early verdict: best overall, best budget, best for specific use cases.
  • Methodology: short and credible.
  • Comparison table if 5+ items.
  • H2 per tool:
    • what it is
    • best for
    • key strengths
    • biggest downside
    • pricing when safely known
    • verdict
  • FAQs.

Rules:

  • Every featured tool needs a homepage/relevant link.
  • Every featured tool needs an image/screenshot unless the asset policy says otherwise.
  • Do not rank the site's own product #1 by default. Rank it only where evidence and fit support it.
  • Include at least one honest drawback per tool when relevant.

Comparison / alternatives

Use when the reader is choosing between named products or searching competitor alternatives.

Structure:

  • TL;DR verdict.
  • Who should choose each product.
  • Optional third-option/product framing only if honest and approved.
  • Comparison table with row-level winner/fit where useful.
  • Pricing/value section with clear assumptions.
  • Feature sections.
  • Privacy/security/fact-sensitive section if relevant.
  • Decision scenarios.
  • FAQs.

Rules:

  • Source-check competitor claims.
  • Do not claim market superiority without evidence.
  • Use careful language for reviews, Reddit sentiment, accuracy, speed, CPU/RAM, storage, privacy, and pricing.
  • Avoid defamatory or aggressive claims.

Profession / use-case guide

Structure:

  • TL;DR.
  • Why this audience cares.
  • Specific workflows.
  • Setup or recommended tool path.
  • Tips by scenario.
  • Risks/limitations.
  • FAQs.

Rules:

  • Use examples from that job, not generic productivity copy.
  • Avoid medical/legal claims unless sourced and approved.

Refresh

Use when improving an existing ranking page.

Rules:

  • Preserve existing URL.
  • Preserve title/meta by default on pages with meaningful rankings/impressions.
  • Improve top answer block, TL;DR, FAQs, internal links, clarity, examples, and stale facts first.
  • Do not restructure ranking-sensitive pages beyond the brief.
  • Mark any risky title/meta/slug recommendation separately for human approval.

Localized article

Rules:

  • Use researched local keywords and slugs, not direct translations.
  • Match local search intent and SERP page type.
  • Use the approved locale voice, pronouns, CTA language, currency, and claims.
  • Avoid English-first assumptions.

SEO requirements

  • Primary keyword in title/H1 for new content.
  • Primary keyword in first 100 words where natural.
  • Related terms distributed naturally.
  • Question-style headings when they match query language.
  • Internal links to conversion pages first, then supporting content.
  • External links for named tools, platforms, sources, or claims.
  • Meta title/description suggestions for new content or when requested.
  • Schema matching visible content.

Product/factual safety

Before output, check:

  • Pricing matches provided facts.
  • Free/free-trial wording is accurate.
  • Platform support is accurate.
  • Privacy/security/processing claims match approved language.
  • Competitor claims are sourced or softened.
  • No fake testimonials, review counts, benchmarks, awards, or customer logos.
  • No unsupported regulated claims.

Asset rules

  • Listicles/comparisons: every featured tool/product needs link + image/screenshot unless exempted.
  • Use approved asset libraries before fresh generation/capture.
  • Include alt text that describes the actual product or UI, not keyword stuffing.
  • If assets are missing, output a clear asset queue rather than pretending they exist.

Steps

  1. Check the brief. Confirm it gives target intent and job type, carries product facts for product-heavy content, and does not require unsupported claims or duplicate an existing page. If any check fails, stop and return a brief defect report (see Stop and flag if).
  2. Pick the template. Choose the content type template that matches the brief. For a refresh, read the existing page first and note what is ranking-sensitive.
  3. Draft to the universal structure. TL;DR, answer-first block, short intro, main sections in the brief's H2/H3 order, conclusion, FAQs, plus the template's own structure and rules.
  4. Apply the SEO requirements. Title/H1, first 100 words, headings, internal and external links, meta suggestions, schema matching visible content.
  5. Run the de-robot pass. Apply every mandatory anti-AI voice rule to the whole draft, not just the intro.
  6. Run the product/factual safety check. Fix or soften every claim that fails it, and log risky claims in the fact-check notes.
  7. Resolve assets. Apply the asset rules; anything missing goes in the missing asset queue.
  8. Package the output. Produce everything in What to hand back, in the requested format.

What to hand back

Fill in this template. The draft goes in the format the brief asked for; everything else stays in Markdown.

**What I found:** <what was written, for which brief, query and content type>
**What matters most:** <the riskiest claim, deviation or missing input>
**What to do first:** <e.g. supply 3 screenshots, then run QA>

## Draft
<full draft in the requested format: Markdown, HTML, document draft, CMS JSON, or handoff package>

## Meta
- Meta title: <...>
- Meta description: <...>

## Schema JSON-LD
<JSON-LD block, or "not required by the brief">

## Links
- Internal: <URL> | anchor | why
- External: <URL> | what it supports

## Assets
- Present: <file or URL> | alt text
- Missing asset queue: <what is needed> | where it goes | suggested alt text

## Fact-check notes
- <risky claim> | source or "provided by site owner" | softened: yes / no

## QA handoff notes
- <deviations from the brief and why; what QA should check hardest>

## What I could not check
- <e.g. competitor pricing page not reachable; claim softened>

Stop and flag if

Fail instead of writing, and return a brief defect report (what is missing or contradictory, and what is needed to proceed), if:

  • Brief does not specify target intent and job type.
  • Product facts are missing for product-heavy content.
  • The requested angle requires unsupported claims.
  • The draft would duplicate an existing page without a refresh/consolidation brief.
  • Required output format cannot be produced from available information.

Next step

  • Run seo-content-qa with the draft, the brief and the fact-check notes.
  • Once QA passes, run cms-publishing-adapter with the approved draft, meta, schema and asset list.

Part of the free RomeoSEO skill library: https://romeoseo.com/skills

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.