[Bug]: kubectl-ai will ollama says starting k8s agent: initializing chat session: Initialize not yet implemented for ollama

[jbrinnand]

Environment (please complete the following):

• OS: macos Version Sequoia 15.6 (24G84)
• kubectl-ai version: 0.0.22
  commit: f151614
  date: 2025-08-14T18:00:57Z
• LLM provider: ollama
• LLM model: mistral:latest

Describe the bug
Check if ollama is running:

curl http://localhost:11434/; echo
Ollama is running

export OLLAMA_HOST=http://localhost:11434/
kubectl-ai --llm-provider ollama --model mistral:latest --kubeconfig '</path/to/my/kubeconfig>

Here is the error:
.....

.....
Use "kubectl-ai [command] --help" for more information about a command.

starting k8s agent: initializing chat session: **Initialize not yet implemented for ollama**

To Reproduce
Steps to reproduce the behavior:

1. Go to 'CLI'
2. Run command kubectl-ai --llm-provider ollama --model mistral:latest --kubeconfig '</path/to/my/kubeconfig>'
3. See error: starting k8s agent: initializing chat session: Initialize not yet implemented for ollama
   Expected behavior
   What you expected to happen. I want to go to the prompt and query for nodes, pods, namespaces and in general interact with k8s agent.

Additional context
Add any other context or logs here.

[droot]

@jbrinnand we have addressed #490 and will be available in the next release (tomorrow).

If you are blocked and can't build and run from the head, pl. use v0.0.20 version.

[jbrinnand]

Many thanks for your timely response. I will wait for the next release. 👍

[nisranjan]

I did implement the Intialize() method for ollama in #567, hope this is sorted anyhow

[jbrinnand]

I just tested it. It's working now. Thanks!