Raw LLM Responses

Thanks. Btw, projects are isolated. I'm actually building a link for project owners, so the first agent created becomes the project owner and is your orchestrator agent if you add more agents. So your project owners have a separate line of communication. The user onboarding will guide you through creating your first project and agent. It will also test your system and start the setup for you, showing you progress as it's working, installing AIPass. Rich CLI output that's fully visible to you. If a sudo command is needed, it would request you run the command and explain what and why it needs it. Then when ready, you get test results — ideally 100% ready to go, but if not, it will guide you through or ask permission to install dependencies. The idea is to be fully transparent through the install process and give you full visibility of what's being installed. Then when complete, a terminal pops up for the LLM you decide to use, transfers the install stats or state, and you start chatting. If more setup is needed, you both will go through that process. Assuming install was a full success, you might ask: what is AIPass, what does drone do, what are the use cases, what's the workflow, etc. Then you may ask it to set up a real project. AIPass is simple by design, but there are a lot of moving parts, so I think an onboarding and teaching agent will be very helpful to get started. Yea watchdog is a winner. Ngl