ai-tutorials: companion code for CloudYeti YouTube tutorials — uneven completeness

This repository hosts the code, configs, and guides that accompany CloudYeti’s CloudYeti YouTube tutorials. It is structured around a table of contents that maps each tutorial video to its companion folder, covering local inference (llama.cpp, Ollama, LangChain), coding agents (OpenCode, Codex, Claude Code), MCP servers, multi-modal agents, and an AIOps debugging lab. Who it is for. Builders who want to reproduce a specific tutorial locally — picking RAM, choosing a model file, wiring an agent — rather than reading a generic getting-started guide. The repo is reference-oriented: you run the steps, copy the config, and adapt. Readiness evidence. Several entries already ship working configs and scripts: the llama.cpp Qwen setup, the Hermes+Kimi K3 OpenRouter walkthrough, the Codex /goal Ralph loop, and the Ollama MLX benchmark. Others are explicitly marked Coming soon and contain no code folder at all, including Google’s Gmail MCP server, OpenCode on OpenRouter, and the Vibe Coding Guardrails checklist. What may need further validation or production