serps.io

MCP

Ask for a ranking without opening a dashboard

An MCP server that exposes your positions, AI citations and audit findings as tools your assistant can call directly. Point Claude or Cursor at it once and the data is in the conversation.

Your rankings, as tools the model can call

Model Context Protocol is the standard way an assistant calls a tool directly. Point Claude or Cursor at the SERPs server and ask for a keyword's position, its citation count across the engines, or the open findings from the last crawl, and the real number comes back in the conversation. No copying figures out of a chart, and no screenshot of a dashboard pasted into a brief.

AI engines checked

5

Samples per prompt

3

Technical audit rules

87

Read only, scoped to your projects

The server is read only by design: it answers questions and changes nothing. Access is scoped to the projects on your account, so an assistant can report a position or a citation but cannot add keywords, start a crawl or touch billing. Revoke access from settings and every client loses it at once.

Projects, Pro plan

10

Projects, Agency plan

50

The same data the dashboard shows

Positions with full history, AI citations per engine with the sources named, audit findings ranked by severity, and Search Console query history kept past Google's own window. The server reads the project you already track, so there is no second dataset to reconcile and nothing to keep in sync.

AI engines checked

5

Samples per prompt

3

Technical audit rules

87

MCP FAQ

Start ranking without the enterprise tax

Every plan starts with a 7 day trial. Connect Search Console and see where you stand in minutes.

Start your trial

7 day trial, card required, cancel any time.