Ollama: long running requests without informing user

[MichaelAnders]

By design, it seems, Claude CLI waits for the response of it's send request from the LLM. Running on Ollama, requests may take minutes until they complete. The user will not see anything during that time and might wonder what is happening. Technically, there is probably no way to display this in the CLI.

However, log files do show progress of the model, so Lynk "knows" about them.

Idea: Maybe we could have another small tool which "listens" on some port and receives "model response" updates from Lynkr?

[Plaidmustache]

I was also curious about this. it seems lynkr hides the "thinking" and "tool calling" happenings silently. its a bit difficult to guide long running processes. I may look into how to make this visibile. but anyone else got ideas? ps, the token savings are nuts on lynker. great work.

[vishalveerareddy123]

@MichaelAnders that seems to be a good idea to have another tool to monitor on the port to show the progress of the server
I wonder how claude code is handling it
I did fix one of the issues in my Open pr which would show thinking etc as suggested by @Plaidmustache

[vishalveerareddy123]

#47
I was able to get the ollama local models working
Probably we can try this out