Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    KiteGEO: AI-Readiness Audit & Fixes

    1

    Score how well ChatGPT, Claude, Gemini and Perplexity can read your site, then get ready-to-paste robots.txt, llms.txt, schema and FAQ fixes.

    Free

    1 useSecurity scanned
    KiteGEO: AI-Readiness Audit & Fixes

    KiteGEO: AI-Readiness Audit & Fixes

    Example session with this skill installed

    is lihaila.si ready for AI search? fix what's wrong

    • Read your context and instructions
    • Compiled the kitegeo

    lihaila.si scores 88/100 for AI readiness. The important things are in place: no AI search crawler is blocked, 774 words are readable without JavaScript, and there's schema (WebApplication, Organization, FAQPage), a clear title, description and H1, a FAQ, a sitemap and HTTPS. Two checks fail:

    CheckWeightResult
    llms.txt file8Missing: AI agents have no short guide to the site
    About / company page4No About page linked from the home page, which weakens entity trust

    Fix 1: add /llms.txt

    Save this at the site root as https://lihaila.si/llms.txt:

    
    # HitPost
    
    > Write X posts that hit. Grok reads what X is talking about right now, Brave checks the facts, DeepSeek and Qwen polish the draft, and you post in one click.
    
    HitPost is a web app by Lihai La LLC for writing posts, threads and X Articles for X (Twitter).
    
    ## Key pages
    - [HitPost (English)](https://lihaila.si/): the app, pricing and FAQ
    - [HitPost (中文)](https://lihaila.si/cn/): the Chinese version
    - [Privacy](https://lihaila.com/en/privacy)
    - [Terms](https://lihaila.com/en/terms#hitpost)
    - [Lihai La LLC](https://lihaila.com/): the company behind HitPost
    

    Fix 2: link an About page

    The site has no About link. The simplest fix is a footer link to the company page, plus a matching sameAs in the existing Organization schema:

    <a href="https://lihaila.com/">About Lihai La</a>
    
    "sameAs": ["https://lihaila.com/", "TODO: your X and LinkedIn profile URLs"]
    

    I left the social profile URLs as a TODO because they aren't on the page. Wrong sameAs links get repeated by AI assistants.

    What this audit can't tell you

    Readiness is only half of GEO. Whether AI assistants like ChatGPT actually

    name HitPost when people ask "what's the best AI tool for writing X posts?", and who they name instead, takes a live check of the assistants themselves.

    Next steps

    • Write these files into the site repo for you
    • Re-run the audit after deploying to confirm 100/100
    • Draft FAQ entries phrased the way buyers ask assistants

    Connects securely to your tools. The creator never sees your data.

    About this skill

    More buyers now ask an AI assistant ("what's the best X for Y?") instead of searching Google, and being named in the answer is the new page one. The first step is being readable: AI search crawlers that aren't blocked, text that's in the HTML, and clear facts about your business. This skill checks that and writes the fixes. It uses the same 11-point audit and weights as KiteGEO (lihaila.com/geo), an AI-visibility checker in production.

    What it does

    • Audits your site with 11 weighted checks and scores it out of 100:
      • AI search crawlers allowed in robots.txt (OAI-SearchBot, Claude-SearchBot, PerplexityBot). Blocking training crawlers (GPTBot, ClaudeBot, Google-Extended, Applebot-Extended, CCBot) is reported as an opt-out and costs no points. Blocking ChatGPT-User is reported as a note and costs no points.
      • Text readable without JavaScript
      • schema.org JSON-LD
      • llms.txt
      • Title and description
      • One clear H1
      • Q&A content
      • Sitemap
      • About page
      • HTTPS
      • Response speed
    • Explains the failures in order of impact. A blocked AI search crawler matters most, because it can keep your site out of that assistant's search answers.
    • Writes the fixes:
      • if you approve the change, a corrected robots.txt that keeps your existing rules and training opt-outs
      • an llms.txt with your real pages
      • Organization/Product/FAQPage JSON-LD
      • a better title and description
      • FAQ entries phrased the way buyers ask assistants
    • Validates them. The bundled script confirms the robots.txt no longer blocks AI search crawlers and that the JSON-LD parses with @context and @type.
    • Never invents facts about your business: anything it doesn't know becomes a clear TODO, because wrong schema facts get repeated by AI assistants.

    How it works

    1. Ask "is mysite.com ready for AI search?"
    2. The audit script fetches your home page, robots.txt, llms.txt and sitemap. It runs on your machine and talks only to your site.
    3. You get the score, what's wrong and why it matters, and each fix as a ready-to-paste file with where it goes. In a site repo, it can write the files for you.

    Limitations

    • This measures readiness: can AIs read and trust your site? It can't measure what AI assistants actually say about your brand, or which competitors they recommend instead. KiteGEO's free check at lihaila.com/geo does that, and the KiteGEO Pro skill runs it from your agent.
    • It audits the home page plus the site-wide files (robots.txt, llms.txt, sitemap), not every page.

    Free and open source (MIT): github.com/Lihaila/kitegeo-skill

    Measures readiness only: it can't see what AI assistants actually say about your brand or which competitors they recommend (KiteGEO's free check does that). It audits the home page plus site-wide files, not every page. Pages rendered only by JavaScript are judged by their raw HTML, the way most AI crawlers see them.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 14 days ago

    • Free with an Agensi account

    Requires

    Node.js 18+

    Permissions

    Terminal / Shell
    Write Files
    Read Files
    Network Access

    Network: fetches the audited site's home page, robots.txt, llms.txt and sitemap, and nothing else. Files: reads draft files you ask it to validate and writes fix files only when you ask it to.

    Listed24 days ago
    Updated14 days ago

    What's inside

    Frequently Asked Questions

    SEO audits target Google's ranking. This one targets what AI assistants need: their search crawlers (OAI-SearchBot, Claude-SearchBot, PerplexityBot) not being blocked, text readable without JavaScript (most AI crawlers don't run it), schema.org facts, an llms.txt guide, and Q&A content assistants can quote. It uses the same 11 checks and weights as KiteGEO.

    Yes. The audit reads your raw HTML the way AI crawlers do, so it tells you if a JavaScript-rendered page looks empty to them. The fixes (robots.txt, llms.txt, JSON-LD, meta, FAQ) work on any stack.

    Yes, free and open source (MIT). You get SKILL.md and one dependency-free script with three commands: the audit, a robots.txt checker and a JSON-LD validator. The fixes themselves are written by your agent from your page's real content.

    No. It checks whether AI assistants can read and trust your site. What they actually say, whether they name you and who they name instead, needs a live check across the assistants. KiteGEO's free check at lihaila.com/geo does that, and the KiteGEO Pro skill runs it from your agent.

    No. Fixes use only facts from your page or from you. Anything unknown (founding year, address, social profiles) is left as a clearly marked TODO, because wrong schema or llms.txt facts get repeated by AI assistants.

    New in Marketing & SEO