How I built my Proxmox homelab by talking to Claude
Description and use
The Proxmox homelab was fully re‑engineered through AI‑driven SSH commands, resulting in a secure, documented, and automated environment with a unified dashboard, reliable storage, and remote access.
Problem
A Proxmox VE mini‑PC was abandoned with misconfigured DNS, orphaned disks, and unmaintained services, making it difficult to manage and secure.
Outcome
The Proxmox homelab was fully re‑engineered through AI‑driven SSH commands, resulting in a secure, documented, and automated environment with a unified dashboard, reliable storage, and remote access.
Workflow
- 1. Claude logged in via SSH and inventoried hardware, software, disks, network, VMs and containers.
- 2. It created a knowledge bundle in OKF markdown and stored the inventory.
- 3. It identified six problems (DNS, hosts, Pi‑hole gateway, repos, orphaned disks, and a 3.6 TB volume).
- 4. It fixed DNS, hosts, switched repos, upgraded the system, and logged each change.
- 5. With approval, it wiped the 4 TB NVMe, created an LVMthin pool, and migrated all VMs and containers.
- 6. It handled the Windows VM disk reclamation by enabling discard and performing an offline copy, reducing usage from 100 % to 28.65 %.
- 7. It configured a static IP and RDP for an Ubuntu 26.04 VM, resolving dual‑IP and client bugs with an SSH tunnel and a helper script.
- 8. It set up the 1.4 TB USB drive as a torrent server, passing it through as a USB device, running qBittorrent only when mounted, and adding an eject‑torrents command.
- 9. It installed copyparty for web file access, Samba for Finder access, and Glance for a dashboard.
- 10. It tested all changes with real downloads, simulated unplug, reboots, and network card hangs, and added watchdogs and documentation updates.
- 11. It documented every step in the knowledge bundle and validated with the OKF gem.
- 12. It reviewed and corrected errors such as wrong dates and misdiagnosed issues.
- 13. It maintained secrets in the Mac Keychain and never exposed them in chat.
- 14. It provided actionable scripts and an app icon for RDP access.
- 15. It monitored the system for future hangs and corrected fstab issues.
- 16. It finalized the dashboard to show correct USB drive status.
- 17. It reflected on lessons learned and improved the process.
- 18. It concluded with a summary of the method and outcomes.
Evidence
The first thing Claude did was not change anything. It looked. It logged into the server over SSH and took a full inventory: hardware, Proxmox version, disks, network, VMs and containers.
Right away it found six problems I didn't even know I had: The server couldn't resolve names. DNS still pointed to the old network's router (10.0.0.1), which is why apt update hung.
With my approval, the 4 TB Kingston NVMe was wiped and turned into a 3.57 TB lvmthin storage called vmdata. All VMs and containers were migrated to it.
The Windows VM put up a fight. After the migration its disk ended up 100% allocated, because discard wasn't enabled. Claude turned it on and tried the normal route (a Windows retrim), which didn't work with that version of the VirtIO driver. Instead of giving up, it did an offline round-trip copy through the other storage, and the disk dropped from 100% to 28.65% usage.
I asked for the 1.4 TB USB drive to be used only for torrent downloads, with a web interface. The first thing Claude noticed: the drive was plugged into a USB 2.0 port. I moved it, Claude measured the read speed (117 MB/s at the start of the disk), and we moved on.
For the web, the plan was File Browser. Claude installed it and, when the service started, read the warning the program itself printed: the project had been archived a month earlier, with unpatched security flaws. One of them deleted folders when an upload failed. Claude stopped, explained the risk and offered me three actively maintained alternatives. I chose copyparty.
- Record date: 2026-10-07T14:56:00.252Z