agentsclimarketplace
Sign in

Seo geo

Skill propiter/landing-craft/skills/seo-geo

Trigger: SEO, posicionamiento web, ranking, meta tags, schema, JSON-LD, structured data, sitemap, robots, Core Web Vitals, GEO, AEO, AI search, get cited by ChatGPT/Perplexity/Google AI, technical SEO, on-page. Make a site rank in search AND get cited by AI answer engines.From its SKILL.md

Install
npx -y skills add propiter/landing-craft --skill seo-geo

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
  • runs commandsInstructs the agent to run 1 command, including `claude plugin install claude-seo@agricidaniel-claude-seo`.

What its file declares

Copied from the file, not written here

The file declares its own license as Apache-2.0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

4.2 KB, ~1.0k tokens by cl100k_base, as published. Nobody here has run it

When to Use

Building or auditing any public page/landing/site that must be FOUND — in Google AND in AI answer engines (ChatGPT, Perplexity, Google AI Overviews). Keywords/intent come from marketing-strategy (the ICP); words from brand-voice; OG/social tags from web-assets. Run the checklist before shipping any page.

THE HONESTY RULE (read this first — it governs all of GEO)

Review / AggregateRating, testimonials, and entity facts (founder, ratings, sameAs, awards, foundingDate) are emitted ONLY from REAL data the research/brand provided. NEVER fabricate them — fake review/rating schema earns a Google manual penalty, is dishonest, and teaches AI engines to distrust the brand. If a fact isn't real, OMIT that schema gracefully. A thinner-but- true entity graph beats a rich lie. When in doubt, leave it out.

The 3 layers (do in order)

  1. Technical — crawlable, fast, structured. Foundation; nothing ranks without it.
  2. On-page — title/meta/headings/content match the searcher's intent.
  3. GEO/AEO — get CITED by AI answer engines (ChatGPT, Perplexity, Google AI Overviews, Claude, Copilot) — the 2026 edge. Five moves: let AI crawlers in (a robots.ts that deliberately welcomes GPTBot/ClaudeBot/PerplexityBot/Google-Extended/CCBot/… with a documented opt-out), feed them clean text (llms.txt + llms-full.txt per llmstxt.org), describe your entities (richer + entity schema: BreadcrumbList, SoftwareApplication, WebSite+SearchAction, Article with dates, entity-deep Organization with real sameAs/founder), write extractable answers (self-contained answer chunks, comparison tables, FAQ, definitions, sourced stats), and measure it (AI-referral analytics by referrer hostname → a GA4 event).

Full checklists, the robots.ts/llms.txt patterns + JSON-LD templates: references/checklist.md.

Hard Rules

  • Validate every JSON-LD block (valid JSON, real schema.org @type, https://schema.org context) BEFORE shipping.
  • Emit schema only for what the page ACTUALLY is, only from REAL data. No Review/AggregateRating unless real; no SearchAction without a search route. Omit gracefully when a fact isn't there.
  • One page = one primary intent / keyword cluster. Answer the query in the first 100 words, in the searcher's words.
  • Core Web Vitals targets: LCP <2.5s, INP <200ms, CLS <0.1 — measure with Lighthouse/PSI, never guess.
  • NEVER keyword-stuff or fabricate. Human first, machine second. <title>/<h1> must match the real content (no bait).
  • GEO: write self-contained answer chunks (a clear claim + its supporting fact together) with real numbers/dates — that's what AI engines quote.
  • Welcome AI crawlers for a marketing site (so it can be cited) but keep /api/ and sensitive paths blocked for every bot; ship the opt-out toggle so a privacy-sensitive client can decline.
  • Keep llms.txt + llms-full.txt IN SYNC with the real pages — a fact in them that contradicts the page is debt.
  • Put real datePublished + dateModified on pages/articles and <lastmod> in the sitemap — freshness strengthens a citation.

Output Contract

Return: the meta/title/canonical/OG block, the JSON-LD schema, the heading outline mapped to intent, the Core Web Vitals check, and any gap (missing sitemap, slow LCP, thin content).

Escalate to the heavy pack when needed

For a deep competitive audit (0–100 score, backlink/keyword data, multi-site rank tracking, PDF/Excel reports) the claude-seo pack (by AgriciDaniel) is registered and one command away: claude plugin install claude-seo@agricidaniel-claude-seo It adds a GLOBAL Edit|Write schema-validation hook (safe, local-only) — enable it for a focused SEO push, then claude plugin disable it. For build-time on-page / technical / GEO work, THIS skill is enough.

What ships with it: 1 file

13.9 KB alongside SKILL.md

references/

Gives 0 of the 12 instructions most marketing audience skills give in ~1.0k tokens

Counted across 690 of the 894 authors here whose files we hold, read 2026-08-07

  • Apply Poppins font to headingsin 41 of 690, across 6 files
  • Apply Lora font to body textin 41 of 690, across 6 files
  • Use Arial fallback for headingsin 39 of 690, across 4 files
  • Use Georgia fallback for body textin 39 of 690, across 4 files
  • Maintain text hierarchy and formattingin 39 of 690, across 4 files
  • Use accent colors for non-text shapesin 38 of 690, across 3 files
  • Use RGB values for precise color matchingin 38 of 690, across 3 files
  • Use brand colors for primary text and backgroundsin 36 of 690, across 1 file
  • Read product marketing context file before asking questions, starting, or auditingin 35 of 690, across 23 files
  • Use active voice instead of passive voicein 26 of 690, across 10 files
  • Implement or generate appropriate JSON-LD structured datain 24 of 690, across 17 files
  • Prioritize clarity over clevernessin 22 of 690, across 8 files

Said here and by no other author read

  • run the checklist before shipping any page
  • build the technical foundation before on-page and GEO layers
  • write self-contained answer chunks with real numbers and dates
  • validate every JSON-LD block before shipping
  • welcome AI crawlers via robots configuration
  • block sensitive paths for every bot

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.