Connect Search Console, Analytics and Bing Webmaster once. Murkuz hands Claude or ChatGPT the real numbers for every site you run, so your agent stops guessing.
Free on one site, no card: Search Console only, and 16 of the 32 tools answer on it. Reading your own data is never metered, on any plan.
Sandbox is a read-only demo account on sample data. Nothing to sign up for, nothing to connect.
Claude, with Murkuz connected
ExampleAcross all my sites, which queries are sitting just off page one?
My_SitesAll_Sites_OverviewLive_Quick_Wins3 sites, no credits spentNine queries across three sites are ranking 8 to 15 with real impression volume. The three worth looking at first:
| Query | Site | Pos. | Impr. |
|---|---|---|---|
| wireless charger stand | your-store.com | 11.4 | 8,210 |
| usb c hub for macbook | your-store.com | 12.1 | 5,940 |
| best desk setup 2026 | blog.your-store.com | 9.8 | 4,175 |
Read live from your Search Console, unmetered. Keyword difficulty on these three is a research call, so it costs 2 credits.
Reads from
Works in
Your own properties, through the official APIs, with your own permissions. Murkuz never asks for a password and cannot see a property you have not connected.
The friction
The data already exists. It is in your Search Console, your Analytics, your Bing account. The work is not analysis. The work is moving it, one export at a time, into the window where you are already thinking.
Open Search Console. Set the window. Filter. Export the CSV. Paste it in. Realise you needed page dimensions too. Start again.
Ask a model about your rankings and it reasons from nothing, because it has no access to your property. It sounds certain either way.
Search Console shows you one property per view. If you run twelve, a portfolio question means twelve exports and a spreadsheet.
How this is different
Every other tool in this category runs the analysis on its own servers and shows you the conclusion. That made sense when the conclusion was the hard part.
It is not the hard part any more. You are already paying for a model that reads a table better than our summariser would, and it can follow up, cross-reference and change its mind. So Murkuz's job is the plumbing: real rows, filtered, sorted and paginated, with the window and the source attached to every answer so nothing is a black box.
Paginated, not truncated
A long answer says how many rows exist and how to ask for the next page, instead of quietly cutting at fifty.
Sourced, every time
Each response carries which property it came from, which window, and how fresh the data is.
Never a fabricated zero
A source that is not connected says so. It does not report zero traffic and let you draw the wrong conclusion.
Setup
About three minutes, most of it waiting on Google's consent screen.
Sign in with Google and pick the Search Console and Analytics properties you want, then connect Bing Webmaster. Read-only, through the official APIs, and revocable from your own Google account at any time.
Free covers Search Console for one site. Paid plans connect all three sources.
Paste one URL into Claude or ChatGPT and approve the connection. Any other MCP client works the same way. Nothing to install, no API key to rotate, no local server to keep running.
One connector covers every site on your account, not one per property.
That is the whole setup. Ask your questions in the window you already work in. The web app is there if you want to change what is connected, and most people do not open it again.
Ask about one site or all of them in the same question.
In practice
Real questions, and the tool your agent reaches for. You never pick one by name.
Never costs a credit, on any plan.
“Which of my pages lost the most clicks in the last 28 days, and to which queries?”
Live_Traffic_Drops“What am I ranking 8th to 15th for where there is real impression volume?”
Live_Quick_Wins“Which queries get impressions but almost no clicks?”
Live_CTR_Opportunities“Are two of my pages competing for the same query?”
Live_Cannibalization“Which landing pages bring search traffic that actually converts?”
Search_To_Revenue“How is this page doing on Bing, not just Google?”
Bing_Search_Performance“Compare every site I run on clicks this month against last month.”
Portfolio_Comparison“Has Google indexed this URL, and is it picking the canonical I expect?”
Inspect_URLCosts credits, because we have to buy the data. Asking what a sweep will cost does not.
“What is the search volume and difficulty on these five keywords?”
Keyword_Metrics2 credits“What does this competitor rank for that I do not?”
Competitor_Gap3 credits“Who is on page one for this query right now?”
SERP_Snapshot1 credit“Give me related keywords and questions around this topic.”
Keyword_Ideas2 credits“What is the Bing volume on these terms?”
Bing_Keyword_Metrics9 credits“How many credits do I have, and what will this sweep cost?”
My_Data_CreditsNo creditBuilt for more than one site
If you run one site, this is a convenience. If you run twelve, it is the reason to bother: the question “where should I spend this week” is unanswerable when the data lives in twelve separate views.
Connect your first site freeAll_Sites_Overview and Portfolio_Comparison take the whole account, so a roster question is one call and one answer rather than a dozen exports joined in a spreadsheet.
You add Murkuz to Claude once. Connect a new site next month and it is simply there in the next question. Nothing to reconfigure on the agent side.
Sites are the only fence, and they start fencing at Operator. Solo is 1 site, because the first paid step buys the other two sources and research credits rather than more properties. 3 sites on Operator, 10 on Portfolio. Reading any of them is unmetered.
The whole surface
Reading your own Search Console, Analytics and Bing data costs us nothing to serve, so we never meter it. Only research spends credits, because only research buys data from outside your accounts.
What is connected, and how fresh it is. Where an agent starts before it asks anything else.
Never metered
Your own Google Search Console data, read live from Google on every call. Never metered.
Never metered
The other two sources, on their own terms. Also your own data, also never metered.
Never metered
The questions neither tool can answer alone, because the answer needs both sides of the visit.
Never metered
The reason the connector is worth having when you run more than one property. One question, every site.
Never metered
Read and change what Google has been told to crawl.
Never metered
The only group that looks outside your own accounts, so the only group that costs credits. Nothing here runs without your say.
Spends credits
No tool is held back for a higher plan. Every account, Free included, is handed the same 32. Two counts are at work here and they measure different things: 24 never spend a credit, on any plan, while what a plan changes is how many have a connected source to answer from. 16 answer on Free, which connects Search Console alone. All 32 answer on any paid plan, which connects all three sources and includes research credits.
Read the full tool reference