Pull clicks, impressions, CTR and position for any date range, explain a traffic change, find striking-distance and low-CTR queries, inspect a batch of URLs at once, and submit or remove sitemaps.
PopMCPGoogle Search Console
Every Search Console API call, the analyses PopMCP runs on top of them, and the traffic changes PopMCP can watch for on the connected property.
31 tools in the full catalog: 27 read and 4 write. 31 load by default, and the rest are found with tool search.
PopMCP
gsc_batch_Run Google's URL Inspection on up to 20 URLs in one call (or on the site's top pages by clicks) and get a compact indexing audit: indexed or not and why (coverage state, robots.txt, noindex, fetch errors), last crawl, Google-selected vs declared canonical (mismatches flagged), sitemaps listing the URL, rich result types and issues, plus a summary count.
gsc_brand_Split search performance into BRANDED queries (people who already know you) and NON-BRANDED queries (discovery), with each side's clicks, impressions, CTR, position and share of clicks, the change vs the previous period, and the top queries on each side.
gsc_compare_Compare search performance between two periods (the previous period of equal length, or the same dates last year) and explain the change: totals with absolute and % change, then, for a chosen dimension (query, page, country, device or searchAppearance), the biggest gainers and losers by clicks plus rows that are NEW or LOST.
gsc_content_Find decaying content: pages (or queries) that lost clicks period after period, not just once.
gsc_device_Find pages (or queries) that rank far worse on one device than the other: compares mobile and desktop average position and CTR row by row and ranks the gaps by the impressions at stake, next to the site-wide device totals.
gsc_keyword_Detect keyword cannibalization: queries for which two or more of the site's pages each take a meaningful share of the impressions, so the pages compete instead of one ranking strongly.
gsc_low_Find queries whose click-through rate is far below what this site normally gets at the same ranking position: they rank well but searchers skip them, so the title, meta description or rich result is the problem, not the ranking.
gsc_page_Everything about ONE page in one call: its clicks, impressions, CTR and position against the previous period, the queries it ranks for (with each query's movement), the device and country split, the daily trend, and Google's index status for it (indexed or not and why, last crawl, canonical, rich results).
gsc_performance_Search performance over time, as a series: clicks, impressions, CTR and position per day, week or month (up to 16 months).
gsc_position_The ranking distribution: how many queries (or pages), and how many of the clicks and impressions, sit at positions 1-3, 4-10, 11-20, 21-50 and 51+, against the previous period or the same dates last year.
gsc_query_Everything about ONE search query in one call: its clicks, impressions, CTR and position against the previous period, the pages that rank for it (with each page's share of impressions and its movement), the device and country split, the daily trend, and close variants (other queries that contain it).
gsc_question_Queries phrased as questions (who, what, when, where, why, how, can, does, is, should ...) that the site already appears for, with the page that ranks for each.
gsc_search_Where the site's Google traffic comes from: clicks, impressions and CTR for web search, image search, video search, the News tab, Discover and Google News side by side, each with its share of clicks and its change from the previous period.
gsc_site_One-call SEO overview of the connected property: clicks, impressions, CTR and position vs the previous period, plus the top queries, top pages, device split, top countries, search-appearance (rich result) breakdown and sitemap health (errors, warnings, pending).
gsc_sitemap_Compare the site's submitted sitemaps with what Google actually shows.
gsc_striking_Find 'striking distance' keywords: queries (and the page ranking for each) sitting just off the top positions (default positions 4-20) that already earn impressions, ranked by the extra clicks they would get at this site's own top-3 click-through rate.
gsc_discover_Google Discover performance: which pages earned clicks and impressions in the Discover feed (no search query involved).
gsc_google_Google News performance (news.google.com and the Google News apps): which pages earned clicks and impressions there.
gsc_hourly_Hour-by-hour search performance (clicks, impressions, CTR, position) for the last few days, including today's still-arriving data.
gsc_pages_Which of the site's pages rank for ONE search query, with each page's clicks, impressions, CTR and average position.
gsc_queries_Which search queries bring people to ONE page: every query the page ranked for, with its clicks, impressions, CTR and average position, top queries first.
gsc_query_Query search analytics (clicks, impressions, CTR, position) for the connected site, grouped by any dimensions and filtered freely.
gsc_add_Add (claim) the connected site to Search Console.
gsc_delete_Remove the connected site from Search Console.
gsc_delete_Delete (unsubmit) a sitemap from the connected site.
gsc_get_Retrieve the connected site's verification + permission level.
gsc_get_Retrieve information about a specific submitted sitemap.
gsc_list_List sitemaps submitted for the connected site.
gsc_list_Return the connected Search Console property and its permission level.
gsc_submit_Submit (or resubmit) a sitemap for the connected site.
gsc_inspect_Inspect a URL's index status, coverage, and mobile usability for the connected site.
Authorize once, then an endpoint your client signs into. Nothing to deploy, and nothing to install on the Google Search Console side.
Bring your own Google OAuth client: enable the Search Console API in that Cloud project, add the OAuth Client ID and OAuth Client Secret, then authorize as the account owning the properties and pick which to connect.
1234567890-abc.apps.googleusercontent.com••••••••••••3f7aThe four write tools run as soon as the AI calls them, so removing a sitemap or the site is immediate. Tick Google Search Console on the PopMCP consent screen only for clients that should change it; keep reporting teammates read-only.
Drop the PopMCP URL into Claude, ChatGPT, Claude Code, Codex or any other supported client. It signs in through PopMCP with OAuth, so your Google Search Console credential never leaves the server.
app.popmcp.com/mcp
Claude wants to connectApproveAsk for something harmless before the agent changes anything. Writes run as soon as it calls them, so confirm it reads the right account first.
Which pages lost the most clicks versus the previous 28 days, and what does each one rank for now?
gsc_compare_periodsok
Then just ask for the work
“Which pages lost the most clicks versus the previous 28 days, and what does each one rank for now?”
“Resubmit every sitemap Google hasn't read in over a month”
“Which keywords just outside the top three would earn the most extra clicks, and do any have two of my pages competing?”
“Are my top 10 pages by clicks all indexed, and is the site still verified?”
One hosted endpoint, registered once per client. The same tools and the same permissions, whichever window the request comes from.
Add the PopMCP URL under Customize → Connectors and the tools appear in every chat.
The same endpoint, registered in the browser. Nothing to install on the machine you are working from.
Register PopMCP as a custom connector, then let ChatGPT read and update the account.
Wire the endpoint into Codex or the Codex CLI when a task needs live business context, not a fixture.
Register the remote MCP server in OpenCode's config for terminal-first and scripted agent runs.
Add PopMCP as an MCP server in Z.ai's desktop app and work the account from the same window.
Every one of them runs on the same hosted endpoint, with the same controls.
Google Search Console is one of the connections in the Analytics and SEO MCP stack and the Content and social MCP stack.
Used by: MCP for ecommerce, MCP for agencies, AI for SaaS companies.
FAQ
What teams ask before putting Google Search Console behind an AI client.
Yes. The connection uses your own Google OAuth client, so the connect form asks for an OAuth Client ID and an OAuth Client Secret from that project. Enable the Search Console API under APIs & Services then Library in the same project first, then authorize: that is the order the connect flow expects.
Authorize with the Google account that owns the properties. After the OAuth redirect you pick which properties to connect and which workspace each one belongs in; properties where that account is only an unverified user are left out, because every call on them would fail. Each connection is then locked to its property: the site URL is filled into every request, analysis tools included, and the AI cannot swap in another one, even one the same Google account can see.
Clicks, impressions, CTR and average position, broken down by query, page, country, device, search appearance, date or hour. query_search_analytics is the general report, queries_for_page and pages_for_query answer the one-page or one-query question, and hourly_performance covers the last 10 days hour by hour, including today's still-arriving data. discover_performance and google_news_performance cover Discover and Google News, which report no queries or positions. Dates default to the 28 days ending 3 days ago, and a code like US is converted to the country code Search Console expects.
Yes. inspect_url checks one URL's index status and coverage. batch_inspect_urls runs Google's URL Inspection on up to 20 URLs in one call, or on the site's top pages by clicks, and reports whether each is indexed and why not (coverage state, robots.txt, noindex, fetch errors), the last crawl, Google-selected versus declared canonical with mismatches flagged, and rich result issues, plus a summary count. One failing URL does not fail the batch. Google allows 2,000 inspections per day and 600 per minute per property.
Yes. Each analysis tool runs several Search Console requests and does the arithmetic in one call. striking_distance_queries ranks queries in positions 4 to 20 by the extra clicks they would earn at the site's own top-3 CTR. low_ctr_queries finds queries whose CTR trails what the site normally gets at that position. keyword_cannibalization lists queries where several pages split the impressions. compare_periods explains a change with gainers, losers, new and lost rows, brand_vs_nonbrand separates branded from discovery traffic, and site_overview is the one-call health check. All are reads.
Yes. There are threshold triggers for clicks dropping or spiking week over week, impressions dropping week over week, CTR falling below a level you set, average position slipping past a level you set, a top page losing most of its clicks, a tracked keyword reaching the top three, and a full day with no search clicks at all. You set the threshold, and PopMCP watches the connected property against it.
No. PopMCP hosts it and gives you a Streamable HTTP MCP URL. There is nothing to deploy, no callback URL to register, and no infrastructure to keep running or patch.
Any client that supports a remote Streamable HTTP MCP server: Claude Desktop, Claude Web, Claude Code, ChatGPT custom connectors, Codex and the Codex CLI, OpenCode and Z.ai's ZCode, as well as editors such as Cursor, Windsurf and Zed. Your client signs in to PopMCP with OAuth; the provider credential never leaves the server.
Connect Google Search Console in PopMCP first, then run claude mcp add --transport http popmcp https://app.popmcp.com/mcp in a terminal. Inside Claude Code, run /mcp and sign in to PopMCP in the browser window it opens, ticking Google Search Console on the consent screen. The Google Search Console tools are then available in that session, and no Google Search Console credential is stored on your machine.
Yes. The Google Search Console catalog is small enough that all 31 tool definitions load into every conversation, so your AI never has to search for one before it can call it.
A tool is something your AI calls, and there are 31 here. A trigger is something PopMCP watches for you: 8 Google Search Console events that can start work on the PopMCP side. Triggers never enter your AI client's context, and none of them ask you to paste a callback URL into Google Search Console.
Launch a hosted MCP endpoint for Google Search Console, keep tool access under control, and start testing real workflow prompts right away.
