This repository provides step-by-step documentation for self-hosting an LLM server on Debian, covering Docker containerization, inference engine selection, and integration of complementary services. It includes configuration for multiple inference backends (Ollama, llama.cpp, vLLM), a chat interface (Open WebUI), a RAG pipeline for document retrieval, MCP proxy and MCPJungle for extending model capabilities, ComfyUI for image generation, and Kokoro for text-to-speech. The guide also covers system hardening, networking, and secure remote access via Tailscale.