AI Skill works using UI but doesn't work when calling programmatically

I am trying to debug the Python code that a colleague wrote following the example given here:

 

https://learn.microsoft.com/en-us/fabric/data-science/ai-skill-scenario#use-the-ai-skill-programmatically

 

the code is really simple.  We just plugged in our URL and question and ran it.  Here is the code:

 

 

 

import requests
import json
import pprint
from synapse.ml.mlflow import get_mlflow_env_config


# the URL could change if the workspace is assigned to a different capacity
url = "https://msitapi.fabric.microsoft.com/v1/workspaces/65c73be1-bfdf-4683-812e-cab91cc98cbeNaN"

configs = get_mlflow_env_config()

headers = {
    "Authorization": f"Bearer {configs.driver_aad_token}",
    "Content-Type": "application/json; charset=utf-8"
}

#print (configs)
#question = "{userQuestion: \"Please list all providers\"}"
question = "{userQuestion: \"Please give me total number of claims in the claims.csv table\"}"
# question = "{userQuestion: \"Could you give me total cost of procedures done in year 2023 grouped by procedure code?\"}"

response = requests.post(url, headers=headers, data = question)
print ("I am here now")

print("RESPONSE: ", response)

#print("")

responsebody = json.loads(response.content)

print (responsebody)

#print(response["result"])

 

 

 

 

Here is the response:

I am here now
RESPONSE:  <Response [405]>
{'requestId': '9c739b23-d366-4c71-8fe2-4107e09cbf29', 'errorCode': 'UnknownError', 'message': 'An error occurred while processing the operation'}

 

the log shows no errors.  How do we go about debugging this script.  If anyone can help, that would be great.