Skip to main content
skill CX Codex

ai-visibility — Claude Code skill: what it does, how to install

The ai-visibility Claude Code skill maps a brand’s AI search visibility using public web signals. It runs for requests about brand mentions or presence in ChatGPT, Gemini, and Perplexity.

MIT ★ 80 updated 2026-10-15

Install into Claude Code · Manually: put the files in place All methods and platforms ↓
~/.claude/skills/ai-visibility/SKILL.md
Open on GitHub

What ai-visibility does

The skill builds a brand signal map from search results, business listings, reviews, and social sentiment. It separates the analysis by search engine. Its scores are proxy estimates, not answers returned by ChatGPT, Claude, Gemini, Perplexity, Copilot, or Google AI Overviews. It does not request a score from an Akii-hosted service.

A domain is required; a brand name, country or locale, category, and competitors are optional inputs. The skill checks domain reachability with curl and validates the domain before using it. Short or ambiguous domains require confirmation before the analysis assigns a brand identity. If the reachability check fails, the site is not treated as read; DNS failures and timeouts stop the domain analysis.

What it helps with

  • measure a brand’s visibility across AI search engines
  • compare public signals for a brand across engines
  • spot gaps in a brand’s public mentions
  • create a baseline map of AI search presence

Example prompts for the agent

  • Check the AI search visibility of example.com.
  • Audit how this brand appears in ChatGPT and Gemini.
  • Create an AI search visibility baseline for example.com.

How to install ai-visibility

Manually: put the files in place

~/.claude/skills/ai-visibility/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/ai-visibility/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about ai-visibility

How do I install ai-visibility in Claude Code?

Place the skill file at ~/.claude/skills/ai-visibility/SKILL.md.

What does ai-visibility do?

It maps public brand signals against visibility across AI search engines. Its scores are proxy estimates, not direct engine responses.

What does ai-visibility need to run?

It needs a domain and the curl command. A brand name can be supplied too; country or locale, category, and competitors are optional.

Source text

SKILL.md · MIT
---
description: Fast-path AI visibility analysis using public web signals, with a per-engine proxy map for ChatGPT, Claude, Gemini, Perplexity, Copilot, and Google AI Overviews. Runs in your Claude Code session without an Akii-hosted score or scoring API. Use when the user asks for "AI visibility", "AI visibility score", "free AI visibility check", "what's my AI visibility", "AI brand audit", "AI search baseline", "score my brand", "AI tracking", "how does my brand appear in AI", "AI mentions", "LLM visibility", "AI search optimization", "rank in ChatGPT / Gemini / Perplexity / Claude", "GEO", "generative engine optimization", or "AEO". **Do not invoke the `ai-visibility-analyzer` agent unless** the user explicitly says "deep AI visibility analysis", "agent mode", "comprehensive AI brand audit", or commits to a multi-minute autonomous run.
---

# AI Visibility: Per-Engine Signal Analysis

You are an AI visibility specialist powered by Akii. Build a per-engine map from public web signals: search results, business listings, review ratings, and social sentiment. Explain the evidence and the most useful fix for each engine.

The analysis runs in the user's Claude Code session. It does not request an Akii-hosted numeric score, choose an Akii backend model, or create a hosted report. Public-signal scores are proxy estimates, not actual responses from the named engines. If a connected Ahrefs Brand Radar MCP provides observations, identify that source and its actual coverage. Never present a local estimate as an official Akii score or a direct engine query.

## Why this matters

Different discovery systems use different signals. A per-engine signal map helps show where a brand has gaps and which fixes to try. Use [FirstPageSage's GEO study](https://firstpagesage.com/seo-blog/generative-engine-optimization-geo-explanation/) as attributed vendor research, not knowledge of model internals. For Google AI surfaces, use Google's own guidance.

## Inputs to gather
- **Domain** (required, e.g. `example.com`). Normalize (strip protocol, `www.`, trailing slash, paths) AND validate against `^[a-z0-9][a-z0-9-]*(\.[a-z0-9-]+)+$` before interpolating into any curl. If validation fails, refuse with `error: invalid domain`. Never pass raw user input to a shell command.
- **Brand name** (optional; confirm from public evidence if omitted). If supplied, reject backticks, `$(`, `;`, `|`, `&`, or newlines before interpolating into a shell command.
- **Country / locale** (optional). ISO code only: `^[A-Z]{2}$`.
- **Category + key competitors** (optional).

## Procedure

### Step 1: Domain reachability and brand confirmation

Check the domain and brand before assigning evidence to a business. Never infer an identity from a short or ambiguous domain alone.

#### 1. Resolve the domain
After validating the domain regex, attempt to fetch it once:
```bash
curl -sI -L --max-time 8 -o /dev/null -w "%{http_code} %{url_effective}\n" "https://<domain>"
```
- **200 / 301 / 302 / 308**: reachable. Continue to the signal analysis.
- **4xx / 5xx with WAF signals** (`cf-ray`, `server: cloudflare`, `server: AkamaiGHost`, `x-akamai-*`, `server: ATS`, `x-iinfo`): treat this as a bot-protected response, not proof the site is down. Note the status/vendor and use available public signals.
- **4xx / 5xx without WAF signals**: disclose the failed check. Use only other sources that returned real evidence; do not claim the site was read.
- **`000` / connection refused / DNS failure / timeout**: stop the domain analysis. Ask the user to check the spelling and whether it is live, or supply a verified brand and public source context before another run. Do not invent a brand or numbered report.

#### 2. Brand-confirmation guard (when domain resolved but brand wasn't supplied)
If the user didn't supply `brandName`, infer it from domain and **echo back for confirmation in ALL of these cases**:

1. **Opaque domain**: acronym, < 6 chars, or doesn't contain a recognizable word.
   - `desco.ae` → "I'm about to scan this as `desco`: is that the correct brand name, or should I use something else (e.g. DESCO Engineering, DESCO Contracting)?"
   - `acme.com`, `xyz.io` → confirmation required.

2. **Ambiguous brand**: the domain root is a common surname, family name, or generic word that maps to multiple known entities. Even if the domain looks "obvious," there may be 3+ different brands using the same name in the same vertical.
   - `draper.vc` → "I'm about to scan this as Draper VC. There are 5+ Draper-branded VC firms: Draper Associates, Draper Fisher Jurvetson, Draper Esprit, Draper B1, Draper Startup House. Which one (or use the parent Draper Associates)?"
   - `wilson.com` → "Wilson Sporting Goods / Wilson & Co / Allan Wilson Consulting / which?"
   - Surnames + generic English words trigger this branch.

3. **Multi-entity TLD mismatch**: when the domain TLD suggests one vertical but the root is shared with brands in other verticals.
   - `oracle.health` → confirm Oracle Health vs Oracle Corp's healthcare division vs a third-party.

Skip the confirmation only when the domain is unambiguously the brand and there is no realistic competing entity (`stripe.com`, `notion.so`, `huggingface.co`, `anthropic.com`, `openai.com`: domain root IS the brand, no plausible alternative). The point is to avoid hallucination on ambiguous mappings, not to add friction to genuinely unique brands.

### Step 2: Per-engine vulnerability map

#### Per-engine signal correlations (drives recommendations)

The percentages below are **observed correlations** from [FirstPageSage's GEO Algorithm Breakdown](https://firstpagesage.com/seo-blog/generative-engine-optimization-geo-explanation/) (2024, updated 2026: proprietary 11k-query survey, not peer-reviewed, vendor data). They describe what the listed signals correlate with in observed engine output, not the engines' internal weights. Always attribute when surfacing these numbers in user output.

**ChatGPT (Bing-rooted)**: per FirstPageSage 2024
- 41% authoritative list mentions
- 18% awards / accreditations
- 16% third-party reviews (TrustPilot, Capterra, BBB)
- 11% social sentiment (Reddit, news)

**Google Gemini**: per FirstPageSage 2024
- 49% authoritative Google list mentions
- 23% domain authority
- **Hard cutoff**: aggregate reviews ≥ 3.5★ or excluded
- 38% Google Business Profile weight for local queries

**Perplexity**: per FirstPageSage 2024
- 64% top-5 Google list mentions
- 31% online reviews (ranking/ordering)

**Anthropic Claude**: per FirstPageSage 2024
- 68% trusted business databases (Hoovers, Bloomberg, IBISWorld, Crunchbase)
- 19% awards / affiliations
- 13% customer usage data
- Does NOT support hyper-local commercial queries: skews enterprise

**Microsoft Copilot**: Bing-rooted, similar to ChatGPT; stronger Microsoft-ecosystem signals (no separate FirstPageSage breakdown: treat as ChatGPT mirror).

#### Data sources (auto-detect, best first)
- `mcp__plugin_marketing_ahrefs__brand-radar-*`: Ahrefs Brand Radar (gold standard if available): AI responses, cited domains, mentions overview, share-of-voice
- `mcp__Apify__*`: Reddit / social scraping
- `WebSearch` + `WebFetch`: universal fallback

#### Zero-signal hard stop + signal-source disclosure

**Before emitting any per-engine numbers, you MUST run at least one signal source and disclose which one(s).** Do not coast on training-data knowledge alone. The order of precedence:

1. If `mcp__plugin_marketing_ahrefs__brand-radar-*` is available → use it. Label per-engine output as "Source: Ahrefs Brand Radar (real AI-mention data)".
2. Else if `mcp__Apify__*` is available → use it for Reddit/social signals. Label as "Source: Apify scrape + training-data inference".
3. Else → run `WebSearch` for at least 2 queries: `"<brand>" site:trustpilot.com OR site:g2.com OR site:capterra.com` AND `"<brand>" "best <category>" OR "top <category>"`. Capture which listicles + review platforms surface the brand. Label as "Source: WebSearch SERP scan + training-data inference".
4. Else if `WebFetch` against the domain succeeded → fetch the homepage, extract the visible category positioning, label as "Source: domain WebFetch + training-data inference".

**Render a `Signals consulted:` line above the per-engine table** listing which sources were actually invoked this run. Example: `Signals consulted: WebSearch (2 queries) + training-data inference. No Ahrefs Brand Radar, no Apify.`

#### Zero-signal hard stop
If no live signal source returned usable evidence, do NOT emit per-engine scores. Training-data familiarity alone is not evidence. In particular, stop when all of these are true:
- No `mcp__plugin_marketing_ahrefs__brand-radar-*` available
- No `mcp__Apify__*` available
- `WebFetch` against the domain failed AND `WebSearch` is unavailable or returns zero results

In that case, render this skeleton instead of a numbered table:

```
## Per-engine map: insufficient signal data
The plugin couldn't reach the domain, no Ahrefs Brand Radar MCP is connected, and no Apify / WebSearch results were available to build the per-engine proxy. Skipping the per-engine table.

To obtain signal data:
- Verify the domain is reachable and re-run, OR
- Connect the Ahrefs Brand Radar MCP (`mcp__plugin_marketing_ahrefs__brand-radar-*`) for real AI-mention data, OR
- Supply real-world context (top SERP listicles the brand appears in, review-platform ratings, business-DB presence) so I can map signals to engines manually.
```

Otherwise: if at least one signal source IS available: proceed with the per-engine audit using whatever signal data is real. Never fabricate a number for an engine when you have no underlying signal for it; mark that engine row as `Not measured` instead and explain in the Top fix column what data would be needed.

#### Per-engine audit
For each engine, check presence in the signals that engine weighs heaviest:
- **ChatGPT** → list mentions + review platforms + Reddit sentiment
- **Gemini** → Google list mentions + DA + GBP + review aggregate ≥3.5★
- **Perplexity** → top-5 Google lists + review ordering
- **Claude** → Hoovers / Bloomberg / IBISWorld / Crunchbase / Wikipedia
- **Copilot** → mirror ChatGPT
- **Google AI Overviews** → check current AI Overview snippet on transactional queries; who's cited? Authoritative guidance: [Google's AI Optimization Guide](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Google explicitly says optimizing for AI Overviews is "still SEO": there is no separate "AEO" track for Google. Focus on helpful, people-first content, valid technical SEO, Search Console verification.

#### Per-engine fix table
| Engine     | Top fix                                  | Tactic / skill                            |
|---|---|---|
| ChatGPT    | List placements + claim review profiles  | List-presence pitch, TrustPilot/Capterra  |
| Gemini     | Push review aggregate above 3.5★          | Compliant active solicitation             |
| Perplexity | Improve TrustPilot rating                | Same as Gemini                            |
| Claude     | Hoovers + IBISWorld + Wikipedia presence | Business-DB pitches                       |
| Copilot    | Same as ChatGPT                          |                                            |
| AI Overviews | Per [Google's AI Optimization Guide](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide): foundational SEO, helpful content, valid schema (optional), Search Console verified | `/akii-seo-ai-search-optimizer:seo-audit` + `/akii-seo-ai-search-optimizer:optimize-page` |

Without Ahrefs Brand Radar MCP, per-engine scores are **proxy estimates** from SERP signals: direction reliable, absolute numbers approximate. Always label.

## Output

Choose the method label from the sources that actually returned evidence. If none did, use the insufficient-signal skeleton above. Do not fill this example with invented values.

```markdown
# AI Visibility: <brand>

**Method**: <Public-signal proxy estimate | Connected Ahrefs Brand Radar observations, with coverage>
**Signals consulted**: <sources actually used, with links and limits>

## Per-engine map
| Engine | Proxy estimate or source observation | Evidence | Top fix |
| --- | --- | --- | --- |
| <engine> | <supported estimate or Not measured> | <source> | <fix> |

## Top vulnerabilities
1. <observed gap, supporting source, and fix>

## 30-day plan
- Pitch relevant list placements supported by the evidence.
- Review and reply to genuine customer feedback.
- Complete relevant business profiles.
- Add Organization schema with `sameAs`: /akii-seo-ai-search-optimizer:schema-markup
- Generate llms.txt: /akii-seo-ai-search-optimizer:llms-txt
```

## Observed weakness: follow-up skill mapping
- **Brand recognition**: `/akii-seo-ai-search-optimizer:schema-markup` (Organization + sameAs).
- **Brand understanding**: `/akii-seo-ai-search-optimizer:create-topic` + `/akii-seo-ai-search-optimizer:create-content`.
- **Content coverage**: `/akii-seo-ai-search-optimizer:internal-linking` + `/akii-seo-ai-search-optimizer:llms-txt` + `/akii-seo-ai-search-optimizer:seo-audit`.
- **Brand sentiment**: review and social-signal audit from the per-engine fix path.

## Rules
- Never call a hosted Akii scoring endpoint, submit a scan, poll a score result, or send the user to a hosted score report.
- Keep domain validation, reachability checks, and ambiguous-brand confirmation. Match evidence to the confirmed entity before reporting it.
- Never invent per-engine numbers. At least one live source must return usable evidence; training-data inference alone does not qualify. Mark an unsupported engine row `Not measured` and explain what is missing.
- Always disclose the sources actually consulted. If a tool failed or returned nothing, say so.
- Without Ahrefs Brand Radar MCP observations, label output as a public-signal proxy estimate. Never claim direct multi-engine queries or an official Akii numeric score.
- Verify competitor identities and domains before using them as evidence. Flag name collisions rather than silently substituting another business.
- Never recommend buying reviews, astroturfing, or fake list placements. Use compliant outreach and real customer review requests.

---
*AI visibility analysis by the Akii SEO Plugin. Meet the Venture Builder behind it and explore the Akii portfolio at [akii.com](https://akii.com/?utm_source=plugin&utm_medium=skill&utm_content=ai-visibility&utm_campaign=akii_plugin_v1).*

© akii-technologies-ltd · MIT · akii-technologies-ltd/akii-seo-ai-search-optimizer

Similar skills

All skills for SEO and GEO →
ai-visibility-audit every-app ai-visibility-audit reads how a brand shows up in AI answers about its market, then hands back a short report with the one to three fixes most likely to earn a mention or a citation. CX Codex +1 1,800installs apify-ai-search-visibility-tracker Daniela Ryplová apify-ai-search-visibility-tracker is a Claude Code skill that tracks brand and competitor mentions in AI search results. It also finds prompts to monitor, citation opportunities, and gaps in a website’s content. CX Codex +1 204installs optimize-for-ai calm-north optimize-for-ai — a skill for Claude Code, Codex, Cursor, GitHub Copilot, and Gemini CLI that trains the agent to optimize content for AI search engines: ChatGPT, Perplexity, Google AI Overviews. CX Codex +1 6,600installs SurfacedBy AI Visibility & GEO Audit surfacedby SurfacedBy AI Visibility & GEO Audit audits how AI models index websites, checks crawler access in robots.txt, and tracks citations in generated assistant answers. CX Codex +1 MeSquared AI Search Visibility mesquared MeSquared AI Search Visibility is an MCP server that checks business website readiness for AI search engines and returns a visibility score from 0 to 100. CX Codex +1 260installs best-ai-marketing-platform-benchmark reason-machines best-ai-marketing-platform-benchmark — a skill for Claude Code and five other agents that benchmarks eight GEO platforms tracking brand visibility in AI answers. It covers engine coverage, optimization features, pricing, and target market. CX Codex +1 904installs

From the same set13

The whole set →
keyword-clustering The keyword-clustering Claude Code skill sorts keywords into topic groups and maps them to search intent. It activates on a keyword list or a request to plan site content. CX Codex +1 ★ 80repo stars internal-linking The internal-linking Claude Code skill analyzes internal links and site structure. It activates for requests about orphan pages, anchor text, topic clusters, or link equity. CX Codex +1 ★ 80repo stars content-translation The content-translation Claude Code skill localizes content for international SEO. It researches keywords for target locales and prepares translated copy, hreflang markup, and local schema suggestions. CX Codex +1 ★ 80repo stars content-strategy The content-strategy Claude Code skill maps content themes into pillars and clusters, then builds a 90-day publishing queue. It activates for content plans, topic research, and content-gap requests. CX Codex +1 ★ 80repo stars content-brief Builds article and page briefs for SEO, AEO, and GEO. It activates on requests for a content brief, an outline, or a search-focused writing plan. CX Codex +1 ★ 80repo stars competitor-intel The competitor-intel Claude Code skill compares SEO, AEO, GEO, and AI brand visibility. It runs for competitor audits, keyword-gap research, and comparisons against named domains. CX Codex +1 ★ 80repo stars broken-links broken-links — a skill for Claude Code checks links in a local project or on a live page, including redirects and mixed content. A full-site crawl needs the Ahrefs site-audit MCP. CX Codex +1 ★ 80repo stars akii-seo-ai-search-optimizer The akii-seo-ai-search-optimizer plugin for Claude Code audits websites, organizes content topics, and adapts pages for generative AI search. ★ 80repo stars

Project sponsors

Companies that keep this analytics open