Remote Ollama (self-hosted) support

### Problem Statement

Currently Nemoclaw onboard is not presenting the option of remote Ollama (self-hosted) provider as it does not detect a local instance of Ollama.

Using an Open-ai compatible endpoint seems non-functional

There are scenarios where a self hosted centralized LLM would be beneficial. At the moment there is a way to acomplish this by using `openshell provider create ` but it doesnt seem to be reliable.

### Proposed Design

Add an option in `nemoclaw onboard` such as Ollama(LAN) which would enable to easily configure lan based self-hosted Ollama.

### Alternatives Considered

_No response_

### Category

enhancement: provider

### Checklist

- [x] I searched existing issues and this is not a duplicate
- [x] This is a design proposal, not a "please build this" request