What are agent sandboxes missing?

Question

I'm building an agent sandbox platform with opinionated customizable templates for general purpose agents using the OpenCode SDK.The idea is to provide a great CLI experience so it's easy to use an AI client like Claude Code or Codex to launch sandboxes, fetch/retrieve data and prompts, review generated outputs, and manage queues for ask-user questions.I also plan to add an AI gateway layer, similar to Tailscale Aperture(not affiliated but one of their engineers gave a great talk at AI Engineer it's on YT) for monitoring network access, distributing secrets, and controlling what services/systems a sandbox has access to.For people building agent workflows, I'd love to hear feedback if you've tried other sandbox solutions.What was missing and what was the hard part?

aborsy · Accepted Answer

Docker sandboxes do that. They take quite bit of resources though, almost same as a standard VM.
I need to mount directories or remove after creation easily, like I do with containers.
Gateway would be good too.
Firewall needs more functionality.

What are agent sandboxes missing?

Docker sandboxes do that. They take quite bit of resources though, almost same as a standard VM.I need to mount directories or remove after creation easily, like I do with containers.Gateway would be good too.Firewall needs more functionality.

Docker sandboxes do that. They take quite bit of resources though, almost same as a standard VM.
I need to mount directories or remove after creation easily, like I do with containers.
Gateway would be good too.
Firewall needs more functionality.