There are a lot of sandboxes on the market right now for AI agents, but most of them focus on production deployments, rather than coding agents (e.g. Claude Code).
I always felt uncomfortable running Claude Code or other coding agents directly on my machine, and I didn't want to buy a second computer or rent a VPS just for that.
That's why I made SideKernel: a sandbox that tries to be as transparent as possible to the everyday developer.
I hope you find some value in it, and feedback is more than welcome. Usability is hard to get right without large user studies, so I'd really value hearing about any annoyances or blockers you run into.
Thanks!