add onstart check for LLM communication

import threading

from django.apps import AppConfig

from django.conf import settings

def test_llm_connection():

    from common_lib.logger import logger

    from . import llm_classes as client_module

    try:

        client_module.LLM_CLIENT.query_llm(

            [{"role": "user", "content": "pinging, respond with pong"}]

        )

        logger.info("LLM connection test successful.")

    except Exception as e:

        logger.error(f"LLM connection test failed: {e}")

class LlmConfig(AppConfig):

    name = "llm"

    def ready(self):

        from common_lib.logger import logger

        from . import llm_classes as client_module

        from django.conf import settings

        if client_module.LLM_CLIENT is not None:

            return

        if settings.LLM_MODEL == "":

            logger.info(

                "LLM model not configured, LLM client will not be initialized."

            )

            return

        # TODO: implement switch for OpenAI child class and possibly other

        client_module.LLM_CLIENT = client_module.OllamaLLM(

            model=settings.LLM_MODEL,

            llm_url=settings.LLM_URL,

            timeout=settings.LLM_TIMEOUT,

        )

        threading.Thread(target=test_llm_connection).start()