How to See Which AI Engines Cite Your Brand Using Content Hub MCP
Retrieve per-engine coverage for ChatGPT, Claude, Gemini, and Perplexity over Content Hub MCP, then see which of your pages each engine cites, which third-party pages it cites alongside them, and how each engine treats your prompts.
AI engines do not cite the same sources. A prompt that ChatGPT answers with your blog post can come back from Claude with a directory page. Your overall citation rate averages these differences away. The engines breakdown keeps them apart, so you can see how strong you are on each engine and which pages each one trusts. When Claude Desktop is connected to Infrasity's Content Hub over MCP, you can retrieve it in a few messages.
Before you start, confirm the following:
- Claude Desktop is connected to Content Hub over MCP. See the Content Hub MCP setup guide.
- Your organization has GEO clusters with prompts that ran on several engines.
Fetch engine coverage
Request the engines breakdown for the Infrasity organization over the last 30 days.
Claude returns one row per engine, showing how many of your tracked prompts the engine cites you on at least once. For Infrasity, 98 prompts are tracked across 4 active clusters:
- ChatGPT cites you on 84 of 98 prompts (86%).
- Perplexity cites you on 81 (83%).
- Gemini cites you on 74 (76%).
- Claude cites you on 69 (70%).
Claude adds the page-level summary. ChatGPT cites you on the most prompts. Perplexity sends the most citations to your site, 211, mostly to your blog.
A second table shows each cluster against each engine as cited prompts out of tracked prompts. Developer Marketing reads 37 of 40 on ChatGPT and 30 of 40 on Claude, so every engine cites it on at least 30 prompts. Content Marketing Agencies is the newest cluster and is already cited on 9 to 11 of 16 prompts per engine.
If a tracked cluster ever shows zero cited prompts on every engine, the table tags it Fix target. That usually means the cluster's target domain does not match the domain the engines cite, so it is a setup check before it is a content question.
See which pages each engine cites
Request the pages Perplexity cites for Infrasity, and the third-party pages it cites alongside them.
Claude starts with your own pages. Perplexity cites Infrasity 211 times, 22% of all of its citations. Your five most-cited pages are three blog posts, one service page, and the homepage, with an average position between #1 and #2. By page type, the 211 citations split into 158 blog, 31 service page, and 22 homepage.
The second table lists the third-party pages Perplexity cites most. Each row carries a type and a suggested action:
- Listicle: pitch the author for a listing.
- Directory and review site: get listed.
- Community: cross-post or comment.
- Competitor: no action, because there is nothing to pitch.
The top third-party source overall is reddit.com, with 77 citations. Each of these pages is a way to extend a lead you already hold.
Check the prompt-by-engine grid
Request the prompt by engine grid for the Developer Marketing cluster.
Claude returns one row per prompt and one column per engine. Each cell has one of three values:
- Cited. The engine's latest answer cited you.
- No. The engine answered and did not cite you.
- Uncounted. The engine has no answer for that prompt in the window. This is not the same as a 0% citation rate. Re-run the prompt before drawing a conclusion.
In the sample, the agency-recommendation and "best of" prompts are cited on all four engines. The ROI, pricing, and cost prompts are cited on most engines, with Claude and Gemini the ones still to win on pricing and cost.
Request takeaways
Request takeaways on the engine data.
Claude returns four findings:
- Strong on every engine. 84 and 81 of 98 prompts on ChatGPT and Perplexity, and 69 on Claude, your lowest engine.
- Your core prompts are fully covered. Agency-recommendation and "best of" prompts are cited on all four engines.
- Pricing and cost are the room to grow. Claude and Gemini are the engines left to win there.
- Your blog does the heavy lifting. 158 of Perplexity's 211 citations land on blog posts.
Claude then offers to check which listicles already mention Infrasity.
Check the engines breakdown after any change that should affect one engine, such as a new comparison page or a directory listing. A change that moves one engine and leaves the others flat is easier to trace than the same change hidden in a blended citation rate.
Open app.infrasity.com and retrieve the engines breakdown for your own organization to see which engines already cite you and which pages they use.
