Urgent.News

What's breaking now, across thousands of outlets.

Tech

How I stopped getting paged at 3am for raw SSH logs (and built a host agent for Linux)

If you run your own Linux boxes, you already know what /var/log/auth.log looks like five minutes after spinning up a fresh VPS. Nobody knows the IP. You haven't pointed a domain at it yet. Still, automated scanners find it almost immediately and start hammering port 22 with root , admin , ubuntu , postgres , and deploy . Most security tools handle this in one of two ways: They dump thousands of…

If you run your own Linux servers, you're probably familiar with the /var/log/auth.log file becoming flooded with connection attempts shortly after deploying a new virtual private server (VPS). Automated scanners target commonly used credentials like "root," "admin," and "ubuntu" to gain access to the system. Some security tools simply dump unrelated logs into dashboards, leaving users to piece together the attack sequence, often at a steep cost and requiring dedicated security expertise.

To solve this issue, Al and his team developed Vigilhold, a lightweight host agent for Linux and Docker servers that monitors processes, network activity, file changes, and SSH signals, providing plain English explanations and enforcing actions within predefined limits.

Vigilhold's demo page, notorious for its canned, sanitized data, was removed in favor of a live public lab server accessible at vigilhold.com. Here, visitors can observe real bot attacks against the Vigilhold agent, averaging around 1,600 failed SSH login attempts daily from over 100 different IP addresses. The agent records these signals, correlates them, and blocks malicious activities while maintaining a tamper-evident audit log of all actions.

The agent operates without opening inbound ports, relying solely on outbound TLS connections. It runs with minimal privileges, using a small root helper for necessary tasks. Installation is straightforward, requiring just a single curl command, and supports Docker for monitoring containers as well as the underlying host. The agent can operate in three modes: Observe (alert-only), Assist (automatically block obvious threats after your approval), or Autopilot (automatic actions for confirmed high-confidence threats).

Every action is recorded and reversible, providing peace of mind for administrators concerned about making unauthorized changes.

Pricing is designed to be affordable for small teams and individuals. The Developer plan ($39/month) covers up to 5 servers and 3 users, offering real-time alerts, SSH configuration monitoring, encrypted credential storage, and audit logging. The Business plan ($149/month) supports up to 25 servers and 20 users, adding features like Autopilot mode, automated threat response, CVE scanning, webhook integrations, and compliance reporting.

A 14-day free trial is available without a credit card, allowing users to experience the system's capabilities firsthand.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in Tech

Scoping a first version: What to build first, and what to leave out

Originally published on ayurchenko.xyz . A first version is scoped well when it proves one thing to one kind of user, and leaves out everything that does not help prove it.

  • Single claim validated by one specific user type
  • Essential features: end-to-end job and trust-building elements
  • First version excludes non-contributing features

More from Monday 16 November →