openai/gpt-oss ? reverse-engineered prompt

I want an open source language model I can use for powerful reasoning and agent-like tasks kind of like ChatGPT but that I can run myself without restrictions. It should be possible to run it locally if I have good enough hardware (with versions for either high-end GPUs or more modest ones) or connect to it easily from other apps. I want to see examples of how to use it in Python and get information on how to integrate it with things like Transformers, vLLM, or common tools for text generation. It would be great if it can show its reasoning process and let me adjust how deeply it thinks for different tasks. If there are simple ways to use it with apps like Ollama or LM Studio, include those instructions too. Look up the latest guides or docs online if you need more details.