I think we should have sticky post about security and risks and safe practices as agentic become more prominent.

Many started with ollama / llama.cpp and other simple framework / backends that are relatively safe But in recent months agentic ai has became popular and accessible to which in my opinion is very welcoming. But if one is to go watch youtube videos or simple guide they will find simple set of instruction that will simply instruct them to install without mentioning security at all. I think this is where this sub can step in.