See your Proxmox homelab. Ask it anything.

A live map of every node, VM and container, and a local model that answers questions about them.

The whole lab in one window.

  1. 01

    Every number, per node

    CPU, memory, disk and network at a glance.

  2. 02

    The live topology

    See how your services connect.

  3. 03

    Ask what needs attention

    A local model reads the whole lab.

Edges that mean something.

Each edge takes the color of what runs inside the guest.

  • DNS / Security
  • Storage
  • Network
  • Dashboard
  • Database
  • Container

Click a node, read its vitals.

Usage, history, cluster health and power, live from Proxmox.

Resource usage for nextcloud: CPU 0.1%, memory 284.0 MB of 2.0 GB at 14%, disk 2.8 of 97.9 GB at 3%.

Resource usage

CPU, memory and disk against what each guest was given.

An hour of history for nextcloud: CPU averaging 1% and now at 0%, memory steady at 14%, each with a short spike when it started.

History

An hour, a day or a week, per node.

Cluster overview: 5 nodes, 4 of 5 online, 3 guests running, 7 services, 1 host, 0 VMs, 4 containers, 1 down.

Cluster health

Offline and overloaded nodes surface on their own.

Power controls after one click on Shutdown: the button has turned red and asks Confirm shutdown?

Power controls

Start, stop, reboot or shut down, with a confirmation first.

A model that has read your cluster before you ask.

You ask

Is anything running low on memory?

InfraLens answers

Yes, nextcloud is using 1.8 GB of its 2 GB, about 90%. Consider raising its memory limit before it starts swapping.

Memory it already knew

  • nextcloud1.8 / 2.0 GB
  • homepage142 MB / 1.0 GB
  • pihole98 MB / 512 MB

The live state of every node goes with each question.

Answered by Ollama on your machine. No cloud, no account, no telemetry.

Install InfraLens

One native app. Nothing else to run.

Needs a Proxmox VE host with an API token. Ollama is optional, for the assistant.

Download

macOS Apple Silicon InfraLens_0.2.0_aarch64.dmg26.7 MB Download macOS .dmg
Linux Any distro, AppImage InfraLens_0.2.0_amd64.AppImage120.7 MB Download Linux AppImage
Debian / Ubuntu x86_64, .deb package InfraLens_0.2.0_amd64.deb43.3 MB Download Debian / Ubuntu .deb

First launch on macOS

InfraLens isn't notarized by Apple yet, so macOS may block the first launch.

  1. Move InfraLens to Applications.
  2. Run this command:
    xattr -cr /Applications/InfraLens.app
  3. Open InfraLens and allow Local Network access under System Settings → Privacy & Security, so it can reach your Proxmox host.

Running on Linux

Make the AppImage executable, then run it:

chmod +x InfraLens_0.2.0_amd64.AppImage
./InfraLens_0.2.0_amd64.AppImage

Or install the Debian package:

sudo dpkg -i InfraLens_0.2.0_amd64.deb

Release notes · v0.2.0

First run

  1. Connect your cluster

  2. Detect services optional

  3. Connect local AI optional

Build from source

A Python backend and a React frontend. Run both locally to work on InfraLens itself.

backend

# Clone the repo
git clone https://github.com/AdemYahmadi/InfraLens.git
cd InfraLens/backend

# Set up the environment
python -m venv venv && source venv/bin/activate
pip install -r requirements.txt

# Add your Proxmox API token and SSH credentials
cp .env.example .env

# Run
python main.py

frontend

# Install dependencies
cd InfraLens/frontend
npm install

# Point it at your backend in .env
VITE_API_URL="http://127.0.0.1:8756"

# Start the dev server
npm run dev