stacklok/brood-box ? reverse-engineered prompt
Reverse engineered prompt
Build me a CLI called Brood Box that lets me run coding agents inside a locked down microVM instead of directly on my machine.
I want to be able to start something like bbox claude-code, bbox codex, or bbox opencode, and have it boot an isolated session, mount a safe copy of my current workspace, and then show me a diff review when the agent is done so I can accept or reject changes before they touch my real files.
Please include support for basic config files, per workspace overrides, resource settings like CPU and memory, file exclude patterns, and a simple way to choose how open the network access should be. It should also support forwarding the env vars and git credentials an agent needs, plus a way to disable that if I want.
If anything is unclear, check the current docs online and keep the whole thing focused on security, review, and a smooth terminal workflow.
Are you gonna build this?
make sure you review the code using coderabbit