Meet our updated MCP: AI visibility in your AI assistant

We know the old routine: you learn a new tool, then click through it by hand, page by page. But AI has changed how people work, and your first stop is now often an assistant.

Our updated MCP brings Rankin AI into the AI assistant you already use. Ask a question in plain language, get your real numbers back, then tell your assistant what to do next. It now covers 161 actions across the platform. So you can set up a brand, dig into your results, and prepare a report in one chat.

What's new in the Rankin AI MCP

We shipped our first MCP in July, so you could connect an AI client to your account with an API key. This update widens what that connection can do. Your assistant can now work across brands, prompts, results, sources, content, site audits, reports and workspace settings.

MCP stands for Model Context Protocol. Think of it as a plug between your assistant and your Rankin AI account. Once it's in, your assistant reads from your live data instead of waiting for you to export a file.

What you can do from your AI assistant

You can ask for all of this in chat today.

Area	What your assistant can do	Try asking
Visibility	Pull visibility, share of voice and citation numbers by date, AI engine, topic or competitor	"How did our visibility change on Gemini this month?"
AI answers	Read full responses with their sources, the brands mentioned and sentiment	"Show me what ChatGPT answered for this prompt."
Sources	Find the pages AI cites, spot competitor gaps and request publisher quotes	"Which pages cite our competitors but not us?"
Prompts and topics	Suggest, add, edit and pause prompts, up to 50 at once	"Suggest prompts for this topic and add the ones I approve."
Brand perception	Compare the attributes AI links to you and to your competitors	"Which traits does AI give competitors but not us?"
Hallucinations	List false claims, add brand facts and resolve alerts	"Which false claims about us are still open?"
Content	Turn recommendations into briefs, draft articles, rewrite pages for AI search and track citations	"Turn our top recommendation into a brief."
Audits and reports	Run E-E-A-T and technical audits, then create PDF reports	"Audit our pricing page and list the issues."

Work on your strategy

Your assistant can also help you decide what to do next. Because it reads your real data, you can ask it for patterns, insights, gaps, and ideas to improve your visibility and content. Here are some examples of questions you can ask your AI assistant.

Patterns & Insights

→ "What patterns do you see in the prompts where we get mentioned?"

→ "Which types of pages appear most often in answers that mention us?"

→ "What changed in our visibility [this month], and which prompts or sources explain it?"

→ "Which topics are we strongest and weakest in right now?"

Strategy

→ "Based on our data, what should our AI visibility priorities be for the next [30] days?"

→ "Where are we weakest against competitors, and what should we fix first?"

→ "If we could only fix [three] things this month, which should they be, and why?"

Sources & Link Building for AI Visibility

→ "Which sources should we target to get mentioned more on [ChatGPT] for our BOFU prompts?"

→ "Which sources should we target to show up for our [X] topic?"

→ "Which [listicles and comparison] pages name our competitors, but not us, and appear across the most prompts? We want to use the "get a quote" feature for them?"

→ "Does [ChatGPT] rely on different domains than [Gemini] in our category? Build a placement list for each."

→ "Which domains are gaining citations this month, and are we on them yet?"

→ "Which prompts and AI engines cite this page, and which competitors does it name?"

→ "How many quote requests do we have left this month, and which eligible pages should get them first?"

→ "Request a quote for [this] page and link it to our [pricing] page with a note for the Rankin AI team [X]."

Sentiment

→ "How does AI describe us compared with [competitor]?"

→ "Which answers talk about us negatively, and which sources are they based on?"

→ "Which sources should we target to improve our negative sentiment?"

Content

→ "What should we publish next to close our biggest content gaps?"

→ "Which content recommendation has the highest impact? Turn it into a brief."

→ "Which of our published articles get cited by AI, and which don't?"

Site audits

→ "Run a site audit for [page URL] in Rankin AI. When it's done, tell me the three things to fix first on that page."

→"Re-run the site audit for [page]. Compare this audit report with the one before it, іo I can see what still wasn't fixedю"

How onboarding gets easier with an AI assistant connection

You no longer have to learn what each feature and filter means, or where to find them. Ask your assistant in plain language, Claude or ChatGPT, and it pulls the right data for you. Anything already in your account comes back right away, in the same chat.

New to AI visibility terms? Ask about those too. For example, "What does citation rate mean, and how are we doing on it?" gets you an explanation next to your own number.

Setup gets simpler as well. Your assistant can develop a prompt strategy from your goals and any context it already has about your business. It can also add your brand facts, run site audits, and finish the setup in the same chat.

How daily work gets easier with MCP

Your dashboard is still there when you want it. But plenty of daily questions are faster to ask than to click through. Perks include:

Custom metrics in seconds. Ask for the exact number you need, filtered by date, AI engine, topic, region, or competitor. You skip the filter menus.
Reports built your way. Don't like how our report is laid out? Tell your assistant which metrics, structure and format you want, and it can build your own version.
Gap, fix and results in one place. Spot a gap, then fix it with a content brief or a quote request for an eligible page. Check the results later in the same place.
Your context and skills included. Your assistant brings what it already knows about your business, plus any skills and tools you've set up. So the output fits how you work.

How to connect your assistant

MCP is included on all paid plans. You can connect Claude Code, Cursor or Claude Desktop in three steps.

Step 1: Create an API key

In Rankin AI, open Settings, then API Keys. Name your key and click Create. You only see the key once, so copy it right away and store it somewhere safe.

Step 2: Add Rankin AI to your client

The same page shows you how to connect under Connect an MCP client. Choose Claude Code, Cursor or Claude Desktop, then click Copy setup. Replace rnk_<your-key> with the key you just created and follow the short instructions on that tab.

Using Claude on the web? Its Custom Connector form doesn't accept Rankin AI's API key header, so use Claude Code or Claude Desktop instead.

Step 3: Check the connection

Ask your client to run list_brands. If your brands appear, you're connected.

Your requests count toward your plan's API rate limit.

Try it

Connect your assistant and start with one question: "Which pages cite my competitors but not my brand?"

New here? Start your 7-day trial and test it on your own brand. Or book a call, and we'll help you set it up.