Ai Seo

Prompts

Use when the user wants to audit a URL for AI citation eligibility, check why AI systems don't cite their pages, score AI Overview or citation worthiness, generate or validate an llms.txt file, rewrite content for AEO/GEO, extract named entities, or diff two page versions for SEO regressions. No API keys required.

Install

openclaw skills install automatelab-ai-seo

ai-seo

Pairs with the @automatelab/ai-seo-mcp server (14 tools). Audits why AI systems do or do not cite a page — structured answer extraction, schema completeness, crawler access, llms.txt, canonical hygiene — and generates fixes.

When to use which tool

Start here: audit_page composes all 14 checks in one call and returns a ranked list of citation-blockers with specific fixes. Use it as the default entry point.

Drill into a specific dimension:

ToolUse when
audit_schemaUser asks about structured data, JSON-LD, or schema.org coverage
audit_canonicalChecking canonical URL, og:url, hreflang, or noindex traps
check_robotsVerifying GPTBot / ClaudeBot / PerplexityBot / OAI-SearchBot are allowed
check_sitemapChecking lastmod freshness signals and sitemap coverage
check_technicalCore Web Vitals signals, crawl accessibility, redirect chains
score_ai_overview_eligibility"Will this page appear in Google AI Overviews?"
score_citation_worthinessCross-engine citation probability score
generate_llms_txtCreate an llms.txt for the site
validate_llms_txtValidate an existing llms.txt against the spec
rewrite_for_aeoRewrite a page section for Answer Engine Optimization
rewrite_for_geoRewrite a page section for Generative Engine Optimization
extract_entitiesCheck named entity density and sameAs coverage
diff_pagesCompare two versions of a page for SEO regressions

Default workflow

User: "Audit https://example.com/my-post"
→ audit_page(url)          # full citation-blocker report
→ Read findings, prioritize by severity
→ For fixable issues, call the specific tool (e.g. rewrite_for_aeo) or draft inline edits

Server setup

Add to your MCP config:

Claude Code (.claude/mcp.json):

{
  "mcpServers": {
    "ai-seo": {
      "command": "npx",
      "args": ["-y", "@automatelab/ai-seo-mcp"]
    }
  }
}

Claude Desktop (claude_desktop_config.json):

{
  "mcpServers": {
    "ai-seo": {
      "command": "npx",
      "args": ["-y", "@automatelab/ai-seo-mcp"]
    }
  }
}

Requires Node 20+. No API keys needed — all checks are read-only HTTP fetches against the target URL.


Developed by AutomateLab. Source: github.com/AutomateLab-tech/ai-seo-mcp.