GEO Visibility
Checks how visible a brand is in AI-generated answers (ChatGPT/Perplexity/Google AI Overviews-style citation behavior) — the newest, least-measured axis of search visibility. Most SEO tools, including Cogny's skill pack, have no equivalent to this.
Usage
/geo-visibility example.com
Steps
1. If the Seobase MCP is connected (real data, preferred)
Call mcp__seobase__semrush_ai_visibility with the domain. It returns:
aiVisibility— how often the domain is cited by LLMsaiVisibilityBenchmark— the category average, for contextcitedPages— how many distinct pages get citedmentionStats— a per-LLM breakdown
Call mcp__seobase__semrush_competitors for the same domain to see which competitors
to compare against, then consider running semrush_ai_visibility for 1–2 of them too
if the user wants a competitive read, not just an absolute number.
2. If not connected (free-tier fallback — manual spot-check, not a measurement)
Be explicit that this is a rough proxy, not real citation-frequency data. Use WebSearch for a handful of natural questions a buyer would ask an AI assistant in this category (e.g. "best for ") and note whether the brand appears in the search results Google surfaces as AI-Overview-style summaries. Do not present this as equivalent to the MCP-connected number — label it clearly as a manual spot-check.
3. Report
GEO Visibility: example.com
Source: live Semrush AI-visibility data | manual WebSearch spot-check
AI visibility score: X (category benchmark: Y)
Cited pages: N
Per-LLM breakdown:
- ChatGPT: N mentions, N cited pages
- Perplexity: N mentions, N cited pages
- ...
Verdict: [above/below benchmark] — [one line on what that means]
Suggested next step: [e.g. "run /write-article on the 2 topics where competitors get cited but you don't"]