The first question on the defense side will be where the spec is going. Thanks to AI_API_BASE_URL, it seems to be working with a local model in a closed network, is that correct, have you tested it? If so, it should be at the top of the README.
Yes that is way I have added this. It could work any OpenAI compatible model or router. I have tested it with Kimi and Qwen. You can bring your own tuned model which is better if you are doing specific work.