About
The Search Console tool we built for ourselves.
QueryScope is an SEO agent that lives in your AI IDE, $149/month with everything included. It connects a remote Model Context Protocol server to your AI IDE (Claude Code, Cursor, Cline, or Windsurf) so the agent you already code with can read your own Google Search Console. It starts with your agent reading your repo to work out what you sell and proposing the searches your buyers type. QueryScope sizes each one and shows where you stand on it and who ranks above you. Its job from there is to get you ranking for the ones you pick: it holds the plan that declares them, hands you one change at a time, and grades every change against how much of that plan you hold. It is built for post-revenue technical teams whose website is a codebase: the search traffic already matters, and the people who change the pages work in an IDE.
It is the productized version of a workflow the team already runs daily. We dogfood it on AppScreenshotStudio, a live SaaS, and ship the same toolset to other operators.
What this is
A Model Context Protocol server for Search Console you install into Claude Code, Cursor, Cline, or Windsurf, paired with a hosted dashboard for connecting your sites and keeping history.
You authorize Google Search Console once. After that, your agent asks what the site should be going after, and proposes searches read out of your repo. QueryScope sizes each one: how many people search it, where you stand on it, and who ranks above you. You decide which of them are your business, and that becomes the plan.
From then on it hands back one target at a time: the search, the page that owns it, the move, and what is ranking above you. Your agent makes the change. QueryScope starts a clock on it and tells you later whether it worked. Your IDE's own LLM does all the reasoning, so there is no separate AI fee on top of your plan.
Why it exists
Search Console is the best free data Google gives you, and it is stuck in a browser tab you forget to open. It also drops everything past 16 months, so you can never see a real trend.
The paid tools point the other way: Ahrefs and Semrush charge a lot for third-party estimates of data you already own for free. Nothing simply puts your own Search Console, and the analysis, where you already work. That is the gap QueryScope fills.
Who it's for
Post-revenue product teams
The site already earns from search, the pages live in a repo, and whoever changes them works in Claude Code or Cursor. Search is worth a real budget here, and the person deciding what to change is the same person who can ship it.
In-house SEO and growth, on a technical team
Responsible for search on a product they do not have to file a ticket to change. The plan lives in the same repo as the pages, so deciding and shipping stop being two different jobs.
It needs three things to be worth the price: a site that already earns something from search, pages that live in a repo, and somebody on the team who works in an IDE. Without all three there are cheaper tools that fit better, and we would rather say so here than take the first month.
How it works
Four steps, kept deliberately boring:
Install the MCP with one line, authorize Google Search Console, and pick a site. Revoke any time from your settings.
Ask what should this site go after? Your agent proposes searches from the repo, and QueryScope sizes each one and shows who holds it. You pick the ones that are your business.
Ask what should I work on? and you get one target, not a backlog. Anything that changes what a visitor reads is written into queryscope.md for you to commit first.
Your agent records what it changed. The clock starts when Google recrawls the page, and that target is not handed back while it settles, so the result stays measurable.
The architecture mirrors Sentry and PostHog: a free local agent in your environment, a paid hosted layer for history and analysis. You can leave any time; it is your Search Console data, read on your own grant.
Where it stops
-
It reads your real Search Console, on a grant you approve and can revoke, rather than a third-party estimate of it.
-
Your agent writes and ships. QueryScope decides what is worth doing and measures whether it worked.
-
The LLM runs inside your own IDE, so there is no separate AI fee on top of the plan.
-
It stops at the click. Search data is all it sees, so it will never tell you a ranking change earned you a signup. It is independent of Google.
How to reach the team
Email [email protected] or reply to any onboarding email. Every message lands in a real inbox the team reads daily. There is no support ticket queue and no AI chatbot between you and the people building this. The contact page lists what to email about.
For account, billing, or privacy requests, see /privacy and /subprocessors.
Ready to try it?
$149/month, everything included. Up to 10 sites.
7-day free trial · nothing due today · cancel in one click