←
Sandfly Security779 posts
Sandfly Security avatar

Sandfly Security

@SandflySecurity

Agentless Linux security. We post threat hunting, DFIR and hardening for people running Linux in production. No agents, no kernel modules.

◷ Joined March 2017▧ 185 Photos and videos
Sandfly Security avatar
Sandfly Security@SandflySecurity·2h

Sandfly 6.0 is our biggest release in two years 💥 🛡️ Alerts on the host, not individual results. Automatic alert triage, machine learning network implant detection  and faster scanning. Works with or without AI. Watch the demo: ap1.hubs.ly/H01w83d0

Sandfly Security avatar
Sandfly Security@SandflySecurity·21h

AI gets evidence, not raw logs. Triage Linux alerts at scale automatically and do deep analysis on demand.  Bring your own model through an OpenAI-compatible API, on-prem included. Or leave AI off, and everything still works. Read our new blog ap1.hubs.ly/H01w83j0

Post attachment
Sandfly Security avatar
Sandfly Security@SandflySecurity·Nov 17

Noisy cryptominers, C2 frameworks, BPFDoor, PAM backdoors that steal passwords, stealth rootkits. @CraigHRowland walks through how each one works on Linux, and why some are far harder to catch than others. Read the blog ap1.hubs.ly/H01xcSQ0

Sandfly Security avatar
Sandfly Security@SandflySecurity·Nov 16

Passive backdoors. Magic-packet implants. Fileless loaders. New in 6.0: a machine learning classifier that hunts Linux malware on any host. Agentless, works air-gapped.  Read our blog on the new release: ap1.hubs.ly/H01w86K0

Post attachment
Sandfly Security avatar
Sandfly Security@SandflySecurity·Nov 16

Your Linux alert queue has had its last good day. 🪦 Sandfly 6.0 does the triage before you log in. Every finding is grouped by the process, file or user behind it, and ranked to identify what’s real. Log in to answers, not alerts. Watch the demo: ap1.hubs.ly/H01w8710

Post attachment
Sandfly Security avatar
Sandfly Security@SandflySecurity·Nov 12

One view per artifact: work the evidence, not the alert queue. Every detection on a process, file or user is grouped into one artifact, with full process context, live state, system context and a single severity score. Read our blog on the new release: ap1.hubs.ly/H01w6vV0

Post attachment
Sandfly Security avatar
Sandfly Security@SandflySecurity·Nov 10

Only 13 of 51,620 Linux malware samples write to syslog, journald or the kernel log by default. If your Linux detection relies on logs, you’re waiting for malware to report itself. Craig Rowland on what to check instead 👇 ap1.hubs.ly/H01w1360

Post attachment
Sandfly Security avatar
Sandfly Security@SandflySecurity·Nov 10

Log in to answers, not alerts. Sandfly 6.0 is live! Your team opens the dashboard to a ranked list of real threats, with the triage already done. Agentless, and it works with or without AI. Read our blog on the new release: ap1.hubs.ly/H01wk6B0

Post attachment
Sandfly Security avatar
Sandfly Security@SandflySecurity·Nov 9

If you've been following our Linux threat hunting series - that's a wrap. If you'd like to see what those same checks turn up across your own system, we offer a free trial you can run yourself. Request a Sandfly Free Trial ap1.hubs.ly/H01w8w_0

Post attachment
Sandfly Security avatar
Sandfly Security@SandflySecurity·Nov 5

Share your best war story: what's the best fake process name you've run into on a compromised Linux box? We've seen [kworker/1:2], systemd-journalb (one letter off), and one that just went with [khelper] and sat there unnoticed for months. Also BPFDoor on a system with an uptime measured in years. Curious what else is out there. #ThreatHunting #Linux