SkillAI Search

AI Search Visibility

Measures whether AI assistants name and cite your brand, which sources they trust instead, and what to fix.

Price Free to install

By SkillRune

Add to your agent

How it works

Runs 5 tools in order, in one call. You only pay for the tool calls it makes.

1AI Mentions$0.180
See whether a brand is named and cited in AI answers across ChatGPT and Google AI Overviews, with the sources those answers cite.
2Read URL$0.003
Turn any URL into clean markdown, ready for an LLM.
3AI Search Volume$0.020
Estimated monthly prompts mentioning a term inside AI tools, with a 12-month trend.
4Ask an LLM$0.004
Get a live LLM answer for a prompt, to check who gets named and recommended.
5Technical Page Audit$0.003
Technical SEO for one page: on-page score, meta and headings, load timing, and the checks it fails.

Typical run~$0.21

The playbook

The exact instructions your agent follows. Connect SkillRune and ask it to run this skill, or have it read the playbook and work through the steps itself.

AI Search Visibility (GEO)

Measure whether AI assistants recommend a brand, and why not. Target: a domain (e.g. stripe.com). This is the AI-era counterpart to an SEO audit: buyers now ask ChatGPT and read Google AI Overviews instead of scrolling ten blue links.

Steps

Baseline the brand — ai_mentions with the domain. Read: total mentions, the byPlatform split (google = AI Overviews, chat_gpt = ChatGPT), and citedSources, the domains those answers actually cite. If the brand's own domain is not near the top of citedSources, AI is talking about them using someone else's pages.
Understand the brand — web_read on the homepage. Extract what they sell, for whom, and the 3-5 buying questions a customer would ask an AI assistant (e.g. "best payment API for startups").
Is there AI demand? — ai_keyword_volume on the 2-3 most commercial questions from step 2. A term with real AI search volume and no mentions for this brand is the gap worth attacking.
Read the actual answer — ai_ask with each buying question, phrased the way a customer would. Note precisely: who gets named, in what order, which sources are cited, and whether the brand appears at all. Quote it.
Compare with a rival — ai_mentions on the strongest competitor named in step 4. The delta in mentions and cited sources tells you what winning looks like in this category.
Check the crawlability of the truth — page_audit on the brand's homepage and any key product page. AI answers are grounded in pages that can be fetched and parsed: a thin, slow or badly structured page cannot be cited.

Report

Verdict — one paragraph: does AI know this brand, and what does it say?
The answer, verbatim — what an AI tells a buyer today, with who wins.
Cited-source gap — which domains AI trusts for this topic vs. the brand's own. Getting cited on those sources is the lever.
Top 5 fixes — ranked by impact × effort, each tied to evidence above. Favor: being present on the sources AI already cites, answering the exact buying questions in fetchable pages, and structured, quotable content.
What to track — the 3 numbers to re-run monthly (mentions, platform split, AI search volume for the money questions).

Skip a step whose tool fails and say so rather than guessing.

Run this skill

Give it a target and watch it orchestrate its tools, step by step. You only pay for the tool calls it makes.