Feature Request: Inline RBI

[allfro]

Hi there,

First of all, this is a great project! Thank you! I wanted to ask if there are any plans to make this support inline RBI. For example, if I type youtube.com in the browser bar, it would be nice if that redirected me to a new brave browser session within Sealskin. That way the experience is seamless. It would be nice to also be able to pass parameters and paths to the isolated session as well.

[thelamer]

How would you propose achieving this goal?

[NotChristianGarcia]

I don't personally need this feature, but I could imagine it being useful for commercial and automation stuff.

Libredirect does something similar in their extension (all open-source). I'm eyeballing it, but looks like they use a listener here which waits for change in url and redirects against list of urls user specifies. If you already have storage for this extension's sshkey config, you might be able to store a list of sites which should redirect to selkie session (which would route to a default browser image and give redirect_url via docker command arg).

That could be mega. kasm only seems to support forward proxies, firewall, right-click menu, not an auto redirect.

[thelamer]

No I get that but this is kind of it's own thing and I apologize for being rhetorical here but I wanted you to dig into it a bit heck maybe you know something I do not know.

In order to to this you would need to fork chromium is the TLDR.

In a browser extension there is no perm for take over the URL bar and navigation buttons which is what you would need in order to actually have RBI:

You need to fork, maintain, and build a chromium fork (chromium is like an all day build when building clean it is up there with AOSP as one of the heaviest things to compile)
This fork would need to be able to swap control modes where the nav bar sends api commands instead of actually controlling the session.
On the server side again another CEF or chromium/electron fork or custom app to display web contents and feed back stuff like open in new window, open in new tab, basically any window actions.
When you opened a new window the rendering stack would need to be fundamentally altered to support many independent displays as sessions would need to be linked to the same process group to be logged in the same etc.

A company of people even with AI agents would need to build this and also maintain it while paying for the build infrastucture.

I know you look at this as look the browser is in the window lets use the native bar and not double render it, but that is just simply not supported through any standard browser API and would need to be built along with the server browser.

I think you can just full-screen the remote session, this seems much easier, ta da RBI!

[allfro]

I think this could be achieved using the following API: https://developer.chrome.com/docs/extensions/reference/api/webRequest