ai-visibility-results
Skill0 starsChanged 4 days agoMITView on GitHub
Read supereo's results for a brand and explain them plainly. How often each AI engine (ChatGPT, Gemini, Google AI Mode, AI Overviews, Perplexity and Claude) recommends it, with sample sizes and ranges; who the engines pick instead; every answer and the pages it cites; the fixes worth making first; and the verdicts of supereo's tests on the changes that shipped. Use when someone asks how their brand is doing in AI answers, who ChatGPT recommends instead of them, which pages AI cites, what to fix first, or whether a change worked.
In this repository4 other files it gives its agents.
- aeo-auditSkill
- ai-visibility-auditSkill
- ship-recommendationSkill
- track-questionsSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 53
Don't round up, don't estimate what supereo didn't measure, and don't promise what a change will do: supereo measures that with a test, …
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Read supereo's results
Tools: list_workspaces, get_visibility, list_questions, list_answers, get_cited_pages, get_fix_candidates, get_brand_brief, list_changes, get_experiment, get_track_record, get_plan, list_recommendations, get_scenario_map, get_pages_to_beat, backlink_gap, list_you_do_items and get_page_reads, from supereo's MCP server. Each one only reads, and all of them are free.
Answer from what supereo holds for the brand, in supereo's own numbers.
1. Get the tools
They come from supereo's MCP server, https://supereo.app/api/mcp, signed in as the person. Some clients load MCP tools only when asked: search your tools for get_visibility before you decide it isn't there. If it isn't, connect the server:
- Claude Code:
claude mcp add --scope user --transport http supereo https://supereo.app/api/mcp, then/mcpto sign in. The tools load in the next session. - Claude (claude.ai, the desktop app, and Claude Code signed in with a Claude account): add it as a connector with this link, then connect and sign in.
- Any other agent: https://supereo.app/install has the steps for each one.
Sign-in happens in the person's browser. There is no API key, so never ask for one.
2. Pick the site
list_workspaces lists the person's workspaces, one per site. When they have more than one, pass the site as site to every tool below. A site with no audit has no results yet: run the ai-visibility-audit skill first.
3. Know which kind of number you have
Each tool says which, and so must you:
- A site supereo tracks (on a supereo plan): its panel of buyer questions is asked on all six engines three times a week. Each engine's number is a rate out of 100 over the last 4 weeks, with its 95% range and a badge. Give the range with the rate, every time.
- Any other site: its free audit asked each question once on each engine. Those are counts of single answers ("recommended in 14 of 17 answers"), not rates: never turn them into percentages, and never compare them with a tracked rate as if they were one.
- Demand on a question is a directional read of the Google searches closest to it, never a count of questions people ask AI.
- A test's early signal is never a verdict. Only a written verdict says whether a change worked.
4. Read what the person asked about
- The headline:
get_visibility: each engine's number, with its sample size or range, and who it recommends instead. Visitors confirmed from AI come with it once the site is connected at Cloudflare or Vercel in supereo; before that it says not connected, which is no zero. - The questions:
list_questions: the buyer questions, numbered, with the engines that recommended the brand on each; on a tracked site also each question's id and its demand. - The answers:
list_answers: each answer in full, with the pages it cites and supereo's judgment (mentions the brand, recommends it, cites its site). Narrow it:list_answerswithquestion_idorquestion,engine, orsinceanduntilfor dates. Ten come at a time, up to 40, andtotalsays how many match. - The pages to beat:
get_cited_pages: the pages each engine cites, the brand's own against its competitors' and everyone else's. Narrow it:get_cited_pageswithquestionorengine. Aquestionis an audit's question number, whichlist_questionswithaudit_idgives. - What to fix:
get_fix_candidates: every recommendation of the saved audit, best first, with the exact change, where it goes and the evidence behind it. A strong one rests on hard evidence in the audit itself. - The brand's facts:
get_brand_brief: what supereo knows about the brand, every fact with the page that states it and the date it was checked. - Changes and tests:
list_changes(every change on record, and the recommendations no change answers yet),get_experiment(one test: its early signal while it runs, its verdict with its range once written) andget_track_record(verdicts by fix type, and whether past wins are holding). - What supereo is doing:
get_plan(supereo's plan for the brand and the steps it asked the team to take) andlist_recommendations(the changes it plans, each with its page, its wording and its status). - The first-run analysis, on a plan:
get_scenario_map(the buyers' scenarios, their demand and priority, and who leads each),get_pages_to_beat,backlink_gap(the "best X" and "X alternatives" pages worth a pitch) andlist_you_do_items(the steps only the team can take). - AI reads of the site, once Cloudflare or Vercel is connected in supereo:
get_page_reads(which AI companies read each page, confirmed apart from unconfirmed and spoofed).
5. Report
Lead with the headline numbers and what kind they are. Then who the engines pick instead, and the pages they cite for them. Then the first fixes, each with its evidence. Quote supereo's numbers as they come, with their sample sizes and ranges, and link the report or the change's page when an answer gives one.
Don't round up, don't estimate what supereo didn't measure, and don't promise what a change will do: supereo measures that with a test, against control questions.
Where to go from here
- Change which buyer questions are tracked: the track-questions skill.
- Make one of supereo's recommended changes yourself: the ship-recommendation skill.
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
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.

