Skip to content
everloot
Get the addon

fetch

Fetch one document (ChatGPT connector contract)

One knowledge record by the id a search result gave, whole: title, the full text in the source's words, the URL to cite, and metadata with the source, licence and date. The text block is the same JSON as a string.

Parameters#

idstring, up to 20 charactersrequired
An id from a search result.

The answer#

Text for the assistant to read, and structuredContent with these top-level fields. The JSON Schema is at /schema/mcp/fetch.json.

idstringrequired
The record id as a string.
titlestringrequired
The record's name.
textstringrequired
The whole record, in the source's words.
urlstringrequired
Where the record is read, as in search.
metadataobjectrequired
No description.

Without an MCP client#

The same answer is a GET away: /api/tools/fetch with the arguments in the query string. See the endpoint and the recipe.

Call it
curl https://everloot.app/mcp/public \
  --header "content-type: application/json" \
  --header "accept: application/json, text/event-stream" \
  --data '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"fetch","arguments":{"id":"146235"}}}'

Arguments show the shape of a call; check the parameters for real values.