Hi,
Under no circumstances from a security perspective do I want to allow my users to use an agent within their environment. How do I disable this for my team? I also would like to hard disable any code being executed or commands being executed by AI. Everything should be vetted by the developer locally prior to running. Is there a setting for this as well?
Hi @William_MacDonald , yes, there is an option to Disable Auto-Run, and each developer on your team can enact this setting Cursor Settings -> Agents -> Auto-Run -> Ask Every Time.
That requires the developer to explicitly approve tools and terminal commands before anything runs. So Cursor can still suggest actions, but it will not execute them automatically without review.
What is not available on Teams today is a centrally enforced version of that policy for the whole team. If you need org-wide enforcement, that level of control lives in our Enterprise admin controls, where admins can manage Auto-Run centrally and disable Run Everything. However, we typically recommend self-service team plans for as long as feasible, so the Teams plan is probably still the best fit for you.
Thank you Kevin, I do hope that some of these admin functions are extended to Teams instead of being held up in enterprise. Security is a huge concerns around agents and AI usage, it is very difficult to have good governance over these tools currently and it would be very helpful to small and medium sized operations.
Noted and understood!