Hello,
I’m a bit surprised that there isn’t yet any integration with AI.
Is anything planned on this topic? Is there a public roadmap that mentions it?
The potential seems quite significant. For example, ChatGPT already has a connector for iTerm2, and something similar for Royal TS would make a lot of sense in my opinion.
Curious to hear if this is something being considered. 
Hi!
Thanks for the feedback. We are exploring options and are actually working on some things in this area but it would be helpful to have more precise feature requests, rather than vague ones like “do something with AI” 
If you have something specific in mind, let us know.
Thanks!
Hello,
Yes, I understand. I’m an MSP/sysadmin, and the best example I can think of (for now) is when I’m connected via SSH and need a system command either to ask a question or get help figuring out how to achieve a specific goal.
Idea 1 :
I could imagine a button or a small window that lets me ask a question to my AI assistant, which would understand the context from the console. (Note: it should be possible to configure your own API key and choose the provider you want.)
Idea 2 :
For example, ChatGPT for Mac includes something similar with “Work with apps” connectors, which automatically capture the content of the active window and attach it to the prompt. Several terminal applications are integrated with ChatGPT Desktop, but Royal TSX is not.
https://help.openai.com/en/articles/10119604-work-with-apps-on-macos
Kind regards.
I think the only connection type where this would make sense, is the terminal connection - providing we have full access to the data which is sent and has been received. I have to investigate if this is possible with iTerm2 but it will definitely not be possible with the PuTTY based plugin on Windows. I think the use case is clear to me. We also have to consider certain guard-rails to ensure nothing bad/destructive can happen easily, so that the user is always in charge of what is sent to the remote system. I think this may be the biggest challenge.
I am very happy that RoyalTS ha no AI integration at all. I personally see no real use case, but many abuse cases. I do not want to have any automation/remote hands/it-pumuckl on my server connections. If I need to ask AI for help/advice (which I also do) I am the proxy between my servers and some ai-reply. For a reason!
So for me @devs: if you find a nice use-case, please go ahead and implement it. but don´t implement something-ai just to be able to slap another AI label on your product(s). and - most importantly - implement a proper kill switch to disable all things ai (and have it disabled by default). Not having a kill switch would be a reason to switch products for me.
I completely agree with the “kill switch”. It’s a must have with AI-feature.
One thing was clear from the beginning: we certainly do not plan anything which allows AI agents interact with a remote session without user consent/confirmation. We’re not trying to build Royal OpenClaw