Last stop on the panel. There are three things here: the command that protects you, the list of the ones you don't need to touch, and who to turn to if something goes sideways.
/approve — your safety net 🟡
When your agent wants to do something sensitive (touch a file, run an important action…), it stops and asks your permission. You decide:
- allow-once: give it permission just this once.
- allow-always: always let it through in that case.
- deny: no way.
Before you approve, look at exactly what it's going to do. It's your seatbelt: you're the one in charge here.
This is security by default: your hablo agent doesn't run sensitive actions without your go-ahead. Nothing happens "behind the scenes".
The "do not touch" zone 🔴
There are powerful admin commands —/config, /restart, /plugins, /bash, /exec…— that configure the system from the inside. They can break things or affect security.
Good news: you don't have to use them. Your hablo agent comes set up, configured and with its own plugins, and we maintain it. If something here ever needs touching, we do it ourselves or support walks you through it step by step.
Do I tell this to my agent, to the platform, or to BotFather?
So you don't type a command in the wrong place, hold on to this distinction:
- Your agent's commands (
/status,/model,/new…): in your chat or your panel, talking to your agent. - Platform commands (
/supportand/telegram): used in @equipo360_bot, which is where you sign up and log in directly./supportto ask for help;/telegramto connect or upgrade your Telegram. - BotFather commands (
/newbot,/setprivacy,/setjoingroups): in @BotFather, Telegram's official bot for creating and configuring bots.
In short: for human support don't type /support in your agent's chat; go to @equipo360_bot. 🙂
You've got one last block left: the commands that are hablo's own for managing your account, your usage and your support. 👉