Command approvals
When the AI Operator is running, you can require it to pause and ask for your approval before executing each command. The agent will show you the exact command it plans to run and wait for you to approve or reject it before continuing.
To enable this, turn on command approvals in the Agent tab before you start the task. This is recommended any time you are working in a sensitive environment, testing against a production system, or working under rules of engagement that require you to document every action taken.
Stealth mode
The stealth level dropdown in the top right of the app controls how aggressively the agent scans. Lower stealth means less network noise, fewer probes, and a smaller chance of triggering IDS alerts or rate limiting. If your rules of engagement require low-impact testing, set the stealth level before running the agent.
If something goes wrong
If the agent runs a command that hits something unexpected or goes outside what you intended, stop the task immediately. Open the Activity tab to see a full log of every command that was run and what it touched. If you accidentally accessed an out-of-scope system, stop testing and notify your client contact right away.