agentleFS
6t9xstar6t9xstar/one-step-seo/skills/geo-optimization/SKILL.md

geo-optimization

Skill5 starsChanged 35 days agoMITView on GitHub

Audit and optimize pages for AI search and Generative Engine Optimization (GEO/AEO), including direct answer blocks, factual density, entity consistency, question headings, author attribution, and snippet eligibility.

4 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 0 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 4 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run the dedicated one-step-seo ai <url> command or inspect the AI Visibility (0–100) score and citability checklist.

    line 13
  2. Very specific:

    Verify robots.txt does not block major AI crawlers (G02-ai-blocked P1).

    line 35
  3. Very specific:

    Verify <meta name="robots"> and X-Robots-Tag do not set nosnippet or max-snippet: 0 (G09-snippet-controls P2), which prevent Google AI …

    line 36
  4. Very specific:

    Place a direct, self-contained 40–60 word answer immediately after an informational question (What/Why/How/When/Where/Which/Who ...?) or an …

    line 38
  5. Very specific:

    Add structured <table> (with <th> headers) or <ul>/<ol> blocks (C03-extractable, +20 pts).

    line 40
  6. Very specific:

    Ensure pages with 300+ words include concrete dates, units, or quantified measurements (G03-facts, G08-stat-density) and phrase key <h2> …

    line 41

“Common” is measured: kept by at least 6 in 10 of the 4 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

geo-optimization — Generative Engine Optimization (GEO / AEO)

1. Description & Activation Criteria

Activate geo-optimization when the user wants to:

  • Improve how a page is quoted, summarized, or cited by AI answer engines (ChatGPT Search, Claude, Perplexity, Google AI Overviews).
  • Run the dedicated one-step-seo ai <url> command or inspect the AI Visibility (0–100) score and citability checklist.
  • Remediate G01-llms through G11-semantic-content and C02-answer / C03-extractable / C04-faq.

2. Inputs & Prerequisites

  • Target: URL, local HTML file, or build directory.
  • AI Visibility Scoring Model (lib/score.mjs -> GEO_WEIGHTS, max 100):
    • Answer-first opening (hasAnswerBlock): 25 pts
    • Extractable tables or lists (tables > 0 || lists > 0): 20 pts
    • Substantive depth (wordCount >= 300): 15 pts
    • FAQ coverage (faqDetected): 10 pts
    • Single clear <h1> (h1s.length === 1): 10 pts
    • Section <h2> headings (h2s.length >= 1): 5 pts
    • Title + meta description present: 5 pts
    • /llms.txt published (llms.found): 5 pts
    • AI crawlers allowed in robots.txt (!robots.aiBlocked): 5 pts

3. Step-by-Step Workflow

  1. Run the dedicated AI Visibility audit:
    • Execute npx one-step-seo ai <url> --out ./seo-report (or npx one-step-seo page <file> --skills geo-ai-search --offline).
  2. Check snippet and crawler eligibility gates (G02-ai-blocked, G09-snippet-controls):
    • Verify robots.txt does not block major AI crawlers (G02-ai-blocked P1).
    • Verify <meta name="robots"> and X-Robots-Tag do not set nosnippet or max-snippet: 0 (G09-snippet-controls P2), which prevent Google AI Overviews and search engines from lifting passages.
  3. Optimize the 25-point Answer-First Opening (C02-answer):
    • Place a direct, self-contained 40–60 word answer immediately after an informational question (What/Why/How/When/Where/Which/Who ...?) or an explicit summary lead-in (TL;DR:, In short:) within the first 120 words of body copy.
  4. Strengthen factual extractability (C03-extractable, G03-facts, G07-question-headings, G08-stat-density):
    • Add structured <table> (with <th> headers) or <ul>/<ol> blocks (C03-extractable, +20 pts).
    • Ensure pages with 300+ words include concrete dates, units, or quantified measurements (G03-facts, G08-stat-density) and phrase key <h2> headings as natural user questions (G07-question-headings).
  5. Verify entity, freshness, author, and landmark signals (G04-entity, G06-freshness, G10-author, G11-semantic-content):
    • Align og:title with <title> (G04-entity), include <time datetime="..."> or schema dateModified (G06-freshness), declare author attribution (G10-author), and wrap long-form copy in <main>/<article> (G11-semantic-content).

4. Tools & CLI Commands

# Dedicated AI Search Visibility & Citability audit
npx one-step-seo ai https://example.com --out ./seo-report

# Machine-readable JSON output for agent pipelines
npx one-step-seo ai https://example.com --json

# Generate starter /llms.txt and AI crawler robots.txt snippet
npx one-step-seo llms https://example.com

5. Evidence Requirements

  • Report the AI Visibility score (report.scores.ai.score 0–100, report.scores.ai.band A–F) separately from Search SEO (report.scores.search.score).
  • Cite the citability checklist from report.md (## Make this page more citable, built by citableChecklist(report)), the signals object from one-step-seo ai <url> --json (signals.robotsAiBlocked, signals.blockedBots, signals.llmsFound, signals.llmsBytes, signals.schemaTypes, signals.hasAuthor, signals.hasLandmarks, signals.hasDirectAnswer), and each signal's weight from GEO_WEIGHTS in lib/score.mjs.
  • Cite specific finding IDs (G01–G11, C02–C05) and their evidence strings from report.findings.

6. Expected Outputs

  1. AI Visibility Scorecard & Signal Breakdown: Point-by-point table showing earned vs available points out of 100.
  2. Citable Rewrite Proposals: Drafted answer-first opening paragraph, question-phrased H2s, and structured comparison table grounded strictly in the page's existing content.
  3. Crawler & Snippet Policy Fixes: Exact robots.txt and meta robots adjustments if AI bots or snippets are blocked.

7. Limitations & External Data

  • Readiness signals, never citation guarantees: AI answer engines dynamically select sources based on query intent, retrieval rankings, and provider policies. one-step-seo measures structural citability and crawler access, never guaranteed LLM citations.
  • GEO-only score isolation: G01, G02, G04, G05, G06, G07, G08, G10, and G11 are in GEO_ONLY_IDS and never deduct from the Search SEO score (G03-facts and G09-snippet-controls deduct 3 pts from Search SEO because they also degrade classic search snippets).

8. Safety & Approval Rules

  • Never fabricate statistics or fake citations: To satisfy G03-facts or G08-stat-density, only surface real metrics from the user's product/content or insert explicit [TODO: add verified metric] placeholders.
  • Confirm with the site owner before allowing AI crawlers in robots.txt, as some publishers intentionally restrict AI training or retrieval bots.

9. Examples

# Run an AI visibility audit and inspect the citability checklist
npx one-step-seo ai https://example.com/docs/getting-started --out ./seo-report

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.