Google Search Console MCP server for your AI

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, so the agent's reach is limited to the properties you selected. Every tool acts on the connected site, so a request resolves against that property rather than your whole Search Console account.

gsc_query_search_analytics returns clicks, impressions, CTR and position for the connected site, which is what a query, page or date-range report is built from. It is the search analytics module in full; the rest of the catalog covers sites, sitemaps and URL inspection. Ask in your own words for a date range and the agent shapes the query for you.

gsc_inspect_url returns a URL's index status, coverage and mobile usability for the connected site. It handles one URL per call, so a question like "are my top pages indexed" runs search analytics first to get the list, then inspects each URL in turn. Keep the list tight if you want a fast answer.

There are 10 operations across 4 API modules: 6 read and 4 write. The writes are gsc_add_site, gsc_delete_site, gsc_submit_sitemap and gsc_delete_sitemap. Set the connection to read-only if you only want reporting, or keep full access with approval required so unsubmitting a sitemap or removing the site pauses for your review.

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, 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.

Loading 10 tool definitions into every conversation burns context before the work starts. PopMCP loads a curated 10-tool set covering the everyday workflows and keeps the rest searchable, so your AI pulls a specialist tool only when a task needs it. You can switch a connection to the full catalog at any time.

A tool is something your AI calls — there are 10 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.