World map of locations
The map shows where your servers stand and colours each location by its state: healthy, warning or critical.
Features
Seven areas, each with views of the app. The data in them is sample data.
Overview
The overview shows load, locations on the world map, open vulnerabilities and the latest events, live. Search with ⌘K finds every server, every package and every finding.

The map shows where your servers stand and colours each location by its state: healthy, warning or critical.
CPU, memory and disk of every server share one timeline over the last 24 hours.
New vulnerabilities appear with their severity, the server and the version that fixes them. “Show fix” leads straight to the fix.
Inventory
Phalanx lists your servers, their containers and the apps from Docker, Coolify, Dokploy, Portainer and Kubernetes. Proxmox clusters appear with their nodes, machines and backups. The inventory lists every installed package on every server and in every container.

Every server with its state and the load of CPU, memory and disk, filtered by critical, warning, offline or healthy.
Containers with their image and vulnerabilities, apps from Coolify, Dokploy, Portainer and Kubernetes with redeploy from the app.
Proxmox clusters with their machines and PBS backups. The inventory records every package, Python projects, Snap, Flatpak and .NET too.
Security
A continuously updated vulnerability feed is matched against the installed packages and containers. For each finding Phalanx names the step, such as a package update or a rebuilt image. Phalanx checks hardening and integrity too. You install updates with approval; on Proxmox the command post takes a snapshot first and can roll the server back.

A finding names what fixes it: update the system packages, raise a library to a new version or rebuild an image.
Hardening checks against recognised benchmarks, plus file integrity, rootkit hints and exposed secrets.
Pending updates per server, installed after approval. On Proxmox the command post takes a snapshot first.
Operations
Alerts follow your rules, on-call and quiet times. Phalanx also monitors domains, certificates, heartbeats and the reachability of services and shows their state on status pages. The logs of every server and container are searchable in one place. Flows run scripts and updates on several servers, if you like only after a second person approves. Costs are shown per server.

Alerts reach you over more than 60 kinds of channel, SMS, calls and ticket systems among them, with on-call, quiet times and noise control.
Checks of websites, domains, certificates and heartbeats, plus public status pages that show the state of your services.
Logs of every server and container with histogram and patterns, flows with schedules and signed scripts, costs per server and app.
Administration
Every person in the team has a role: owner, admin, operator, developer or viewer. Service providers keep their customers apart in tenants. People sign in with a password and a second factor, with a passkey or through single sign-on.

Admins change everything, operators run actions and install updates, developers look after findings, viewers only look.
Password with a second factor and a lock after failed attempts, passkeys and single sign-on through OpenID Connect.
Tenants keep a service provider’s customers apart. Policies with inheritance hold for a server, a group or the whole organization.
AI if you want it
Phalanx explains an alert with the AI provider you set up yourself: OpenAI, Anthropic, Google Gemini, Mistral, Groq or a local model through Ollama or LM Studio. Without that setup none of it is active, and nothing goes to a provider.

OpenAI, Anthropic, Google Gemini, Mistral or Groq, each with a key of your own.
Ollama, LM Studio or any server that speaks the OpenAI format. The request then stays in your network.
The cause of the alert, the services it touches and a suggestion how to fix it.
Self-hosted
The command post updates itself and backs up its data on a schedule, encrypted with a key of its own. It works without internet too; you then bring in new vulnerability data as a file.

The command post only takes versions signed with the Phalanx key, backs up its database first and goes back if the new one does not start.
On a schedule to a second volume or S3 storage, encrypted with a key of their own and restorable from the app.
Without a connection the command post keeps working. You then bring in new vulnerability data as a file.
Get the app, set up the command post and connect your first server.