> Each agent runs inside a dedicated microVM with your dev environment What's a "microVM" and what's the security model here compared to using real virtual machines with actual constraints on breakouts? Is it marketing fluff? Incus/LXD has had VM's for a long time now. incus launch images:ubuntu/26.04 my-ubuntu-vm --vm incus exec my-ubuntu-vm -- bash
Your example is not complete, you have to show how it will run claude/codex, you have to do extra things to install run and mount folders there, this one does that with less config, also with this agents can run docker, lxd doesn't allow you to do that
You create your own image first rather than blank ubuntu.
Packer and Incus setup:
packer {
required_plugins {
incus = { version = ">= 1.0.0", source = "github.com/lxc/incus" }
}
}
source "incus" "ubuntu" {
image = "images:ubuntu/24.04"
output_image = "ubuntu-claude"
type = "virtual-machine"
}
build {
sources = ["source.incus.ubuntu"]
provisioner "shell" {
inline = [
"curl -fsSL https://claude.ai/install.sh | bash",
"echo 'export PATH=\"$HOME/.local/bin:$PATH\"' >> ~/.bashrc"
]
}
}
packer init ubuntu-claude.hcl
packer build ubuntu-claude.hcl
incus launch ubuntu-claude my-claude-vm
incus exec my-claude-vm -- claude -p "solve the EC discrete logarithm problem, if it doesn't work keep going" --dangerously-skip-permissions