Issue When Analyzing any resource

[ShoaibQureshi01]

Hi there,
i have installed the kube-copilot using helm with Azure OpenAI gpt-35-turbo-16k model. Now every other function is working fine except the analyze it is showing some python syntax error like this

here's the screenshot of the error

Got this in Logs of kube-copilot pod :-
"""
root@master:~/service# kubectl logs kube-copilot-6ff7b996c9-f2wbt
You can now view your Streamlit app in your browser.
Network URL: http://172.16.194.186:8501
External URL: http://110.235.219.135:8501

Entering new AgentExecutor chain...
Question: What is the YAML for Pod web in namespace default?
Thought: I can use the kubectl command to retrieve the YAML for the Pod.
Action:

{
  "action": "python",
  "action_input": "kubectl get -n default Pod web -o yaml"
}

Observation: SyntaxError('invalid syntax', ('', 1, 9, 'kubectl get -n default Pod web -o yaml\n', 1, 12))
Thought:Could not parse LLM output: There seems to be a syntax error in the Python code. I need to fix the syntax error and try again.
Observation: There seems to be a syntax error in the Python code. I need to fix the syntax error and try again.
Thought:I need to fix the syntax error in the Python code and try again. The error is likely due to the fact that the command is not enclosed in quotes. I will enclose the command in quotes and try again.
"""
!!!
For reference here is the logs of diagnose function which is working fine

""""

Entering new AgentExecutor chain...
Question: What is the issue with Pod mypod in namespace default?
Thought: To diagnose the issue with the Pod, we need to retrieve information about its status, logs, and events using the Kubernetes Python SDK.
Action:

{
  "action": "python",
  "action_input": "from kubernetes import client, config\n\nconfig.load_kube_config()\nv1 = client.CoreV1Api()\n\n# Retrieve Pod status\npod_status = v1.read_namespaced_pod_status('mypod', 'default')\n\n# Retrieve Pod logs\npod_logs = v1.read_namespaced_pod_log('mypod', 'default')\n\n# Retrieve Pod events\npod_events = v1.list_namespaced_event('default', field_selector=f'involvedObject.name=mypod')\n\nprint(pod_status)\nprint(pod_logs)\nprint(pod_events)"
}

Observation: {'api_version': 'v1',
'kind': 'Pod',
'metadata': {'annotations': {'kubectl.kubernetes.io/last-applied-configuration': '{"apiVersion":"v1","kind":"Pod","metadata":{"annotations":{},"labels":{"app":"httpd"},"name":"mypod","namespace":"default"},"spec":{"containers":[{"image":"httpd","name":"myhttpd","ports":[{"containerPort":80}]}],"nodeSelector":{"node":"mynode"}}}\n'},
'creation_timestamp': datetime.datetime(2023, 12, 14, 6, 24, 3, tzinfo=tzlocal()),
'deletion_grace_period_seconds': None,
'deletion_timestamp': None,
'finalizers': None,
'generate_name': None,
'generation': None,
'labels': {'app': 'httpd'},
'managed_fields': [{'api_version': 'v1',
......Output Ommited

From what i can think of is that the diagnose function is converting the prompt into python code but the analyze function is not, it is using the simple bash kubectl command. But i am not a developer and that is just a guess.

Any reply would be of help.
Thanks.