https://www.metachris.dev/tags/security/Recent content in Security on Chris HagerHugo -- gohugo.ioenTue, 10 Feb 2026 00:00:00 +0000https://www.metachris.dev/2026/02/safe-yolo-mode-running-llm-agents-in-vms-with-libvirt-and-virsh/Tue, 10 Feb 2026 00:00:00 +0000https://www.metachris.dev/2026/02/safe-yolo-mode-running-llm-agents-in-vms-with-libvirt-and-virsh/This is a guide for isolating LLM agents in virtual machines, using libvirt and virsh on Linux servers. Running LLMs in VMs isolates them from the host system, mitigating numerous security risks such as destructive operations or unauthorized file access (i.e. private keys, secrets, credentials for communication tools). This is particularly important when granting LLM agents broad permissions, like auto-approving tool use (“yolo mode”). It’s also useful to keep sessions running for extended periods of time, and to interact with agents from the phone / on the go.https://www.metachris.dev/2025/11/sandbox-your-ai-dev-tools-a-practical-guide-for-vms-and-lima/Tue, 25 Nov 2025 00:00:00 +0000https://www.metachris.dev/2025/11/sandbox-your-ai-dev-tools-a-practical-guide-for-vms-and-lima/AI coding assistants, npm, pip, and other development tools can run arbitrary code and scripts on your machine, potentially stealing SSH keys, API tokens, wallet keys, sensitive credentials and other private data without you noticing. This guide shows you how to sandbox these tools in isolated VMs using Lima, so you can experiment and develop freely without putting your sensitive data at risk. Jump straight to the guide, or read on for a bit of personal context.