Difficulty Accessing JSON Data from Endpoint in Custom GPT

GPT builders
1

Hello everyone,

I’m currently working on a custom GPT model that is designed to analyze JSON data retrieved from an endpoint on my server. The server, for example, is located at https://x.y.z/ and the specific endpoint I’m targeting is /data. This endpoint returns JSON data in the following format:

{
    "data": "Today my factory X produced 140 pieces of glasses"
}

The objective of my GPT model is to reference this data when asked questions such as “how is production going?”

Here’s the configuration I’m using:

info:
  description: API for getting information from your factory
  license:
    name: MIT
  title: Swagger Factory
  version: 1.0.0
openapi: 3.1.0
paths:
  /data:
    get:
      description: Endpoint for retrieving information from the factory
      operationId: get_data_data_get
      responses:
        '200':
          content:
            application/json:
              schema: {}
          description: Successful Response
      summary: Get all data
      tags:
      - data
servers:
- url: https://x.y.z/

However, when I trigger the GPT model, I receive the following response:

[debug] Calling HTTP endpoint
{
  "domain": "x.y.z",
  "method": "get",
  "path": "/data",
  "operation": "get_data_data_get",
  "operation_hash": "XXX",
  "is_consequential": false,
  "params": {}
}
[debug] Response received
{}

It returns a ClientResponseError without any status or HTML errors. I’ve confirmed that the server is online by testing GET requests using tools like Postman or cURL from external machines, and it returns the data correctly. Additionally, there are no connections opened on the API server.

I’m puzzled by this behavior and would appreciate any insights or suggestions on how to troubleshoot and resolve this issue. Thank you in advance for your help!

2

Error in SSL certificate to https://x.y.z/, solved

3

FYI: deleting and recreating the GPT always solves these issues :wink: