CueScout.

See which conversations AI answers actually read about your category, and get into them.

AI clients

CueScout in ChatGPT

The honest version first: you cannot add the CueScout MCP server to ChatGPT. Its connectors talk to remote MCP servers over HTTP, and ours is a stdio process a local client launches. Those are different transports and no amount of configuration bridges them.

What does work is the layer underneath. The five tools are thin wrappers over public HTTP endpoints with no authentication, so a custom GPT action calls them directly — same checks, same answers, one less hop. That is what this page documents.

Where the config lives: Not applicable — ChatGPT connectors need a remote server, and ours is stdio

Setting it up

  1. 1.Call the endpoint directly to see the shape

    Every endpoint takes JSON and returns JSON. No key, no account. This one runs the full eight-check readiness catalogue against any public site.

    In a terminal

    curl -s https://cuescout.com/api/public/ai-readiness \
  -H 'content-type: application/json' \
  -d '{"url":"example.com"}'

  2. 2.Wire it into a custom GPT as an action

    In the GPT builder: Configure → Create new action, then describe these endpoints in the OpenAPI schema with no authentication. The five paths are /public/geo-check, /public/ai-readiness, /public/ai-crawler-check, /public/structured-data and /public/sitemap-check, all POST, all under https://cuescout.com/api.

  3. 3.Rate-limit yourself before we do

    The endpoints are limited per IP, and the visibility check costs us a real model call. A GPT that runs it on every turn will start getting 429s, which reads to your users as the tool being broken.

What this does not do

  • The MCP server does not work with ChatGPT. It is stdio; ChatGPT connectors are remote HTTP. This is a transport mismatch, not a setting.
  • Custom GPT actions are a ChatGPT Plus feature, and the GPT you build is yours — this is not a CueScout connector in the ChatGPT directory.
  • The endpoints are rate-limited per IP and store nothing.

Frequently asked questions

Will there be a remote MCP server?

It is the obvious next step and it is not built. When it exists this page will document it instead of the workaround.

Is there an official CueScout GPT?

No. What is public is the API, which anybody can point a GPT at.

Does this need a CueScout account?

No. The five tools call public endpoints with no authentication — the same ones behind the free tool pages. There is no key to paste and nothing to sign up for.

What does it cost?

Nothing to you. Four of the five tools are HTTP fetches against the site you name. The visibility check puts a real question to a search-grounded engine, which costs us a model call, so it is rate-limited per IP.

Is any of this stored?

No. The public endpoints write nothing to a database. If you want the same checks tracked over time against a site you own, that is the paid product.

Related

AI clients

Claude Desktop

Paste one block into claude_desktop_config.json and ask about your site in ordinary chat.

AI clients

Claude Code

One `claude mcp add` line, then ask about your site's AI visibility mid-session.

AI clients

Cursor

One block in ~/.cursor/mcp.json, and the agent editing your site can check what engines see of it.

Every integration we publish

Run the checks on your own site

The readiness check is free, needs no signup, and takes about half a minute. It works out what your site is built on and gives you the fixes written for it.

Run the free checkTrack it over time