Docsbook
Overview

Get mentions

MCPget_mentionsREAD$0.008 / call

Read the mention checks for the docs: for each query the owner watches, whether Google's AI Overview names them (surface 'ai_overview'), and where they sit on Google's or Bing's results page ('google' / 'bing') — plus which OTHER sites on that page name the product. Answers the half Search Console cannot: a query the docs do not rank for returns nothing there, and Bing is not in it at all. 🔴 EVERY QUERY COMES BACK WITH ITS OWN EARLIER READINGS in history, newest first — this is the before-and-after on a search position, and the only one this product has. It is what makes a claim about ranking judgeable: take the reading before the change, name it as the hypothesis's baseline, and on the check date read the same query here again. A verdict written without looking at history is a verdict on one number with nothing behind it. Cached; pass refresh=true to run the check now (a metered fetch, rate-limited to once every few minutes per surface). Use configure_mentions to choose the queries first — with none saved this returns an empty, switched-off surface rather than an error. This answers WHAT, not what to do about it. If you have not already got the method from docsbook_expert, get it first: it names which readings answer this question, what to compare them against, and what would make the conclusion wrong. One call, changes nothing.

Arguments
workspace_idstringoptional

Workspace ID (optional when MCP endpoint is auto-scoped to a workspace). Numeric workspace id — OR the project as the user names it: 'owner/repo', the repo name alone, the site's display name, its docs URL or custom domain. Text is resolved server-side; an ambiguous name returns the candidates instead of guessing, so pass what the user said rather than calling list_workspaces first.

surfacestringoptional

Which engine to read. Omit for all of them. One of: ai_overview, google, bing.

refreshbooleanoptional

Run the check now before answering. Requires surface, costs one metered fetch, and is refused inside the cooldown (the cached answer comes back with a note).

Call it#

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_mentions",
    "arguments": {
      "surface": "ai_overview"
    }
  }
}
curl -X POST 'https://docsbook.io/api/mcp/server' \
  -H 'Authorization: Bearer dbk_YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -H 'Accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"get_mentions","arguments":{"surface":"ai_overview"}}}'

Try it over REST#

The same tool is callable as a plain HTTP request, no MCP client required. It runs on the same server, at the same price.

Your workspace is resolved from the API key, so workspace_id is decided server-side here and anything you send for it is ignored.

POST/api/v1/tools/get_mentions
Authorization
Authorizationstringrequired

Your API key, sent as Authorization: Bearer dbk_YOUR_API_KEY.Sent as the Authorization header. Your key is used only by your browser for this request — it is never sent to Docsbook or stored.

Body
argsobject

The arguments above, as one JSON object.

Request
curl -X POST 'https://docsbook.io/api/v1/tools/get_mentions' \
  -H 'Authorization: Bearer dbk_YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{"args":{"surface":"ai_overview"}}'

Updated

Was this page helpful?