Locki – AI sandboxing for real-world projects

Name: Locki – AI sandboxing for real-world projects
Availability: InStock
Author: PufPufPuf

by PufPufPuf·Apr 3, 2026·3 points·4 comments

Visit Project View on HN

AI Analysis

●●●BangerSolve My ProblemWizardry

Single VM with LXC isolation beats spawning full VMs per worktree—starts in 10 seconds.

Strengths

•VM is the security boundary, not LXC—honest about what's actually protecting you.
•VSCode worktree integration shows agent changes natively in sidebar.
•Runs Docker, Kubernetes, systemd services inside sandboxes without host risk.

Weaknesses

•Linux/macOS only, no Windows support mentioned.
•First VM boot takes minutes before the 10s sandbox starts work.

Post Description

Hello HN! I was bothered that no existing sandboxing solution can run my $job project, due to the limitations of running inside an OCI container, or some sort of limited process like landlock / bubblewrap etc. My options were spawning a new VM per worktree, which takes minutes to boot and allocates a chunk of RAM... or build a custom solution which uses just a single VM plus LXC containers -- the kind of containers with a full init system, capable of running systemd services, OCI containers and even full Kubernetes distributions.

So here it is. Fresh container start takes <10s. Works best with VSCode, which will neatly show changes from all worktrees in the sidebar, letting you review, edit and commit them easily. Let me know what you think!