Heartbleed's ancient cousin, hiding in Squid since 1997.

Reading Time: 8 minutesXem bản tiếng Việt tại đây / Read the Vietnamese version Field Detail CyStack Advisory ID CSA-2026-NGINX-01 CVE IDs CVE-2026-42530 […]

Probably no, right?

That would be very dangerous, especially if the phishing website serving it is an exact copy of the thing you was looking for.

Well that's what happened to me!


TL;DR

I found a fake tronscan blockchain explorer duplicate on the #1…

A two-phase evaluation of fourteen open-source deepfake detectors on an SDXL + InstantID benchmark, with a robustness study against four social-media ingest pipelines.

Every line of client-side bot detection runs on hardware the attacker fully owns. Here's the engineering philosophy we use to defend code we can never actually hide.

I pay for 10 VPN connections. I use all 10 VPN connections.

Threat intelligence report: QuietBoost: Silent Activation of a Commerce-Tracking SDK . Research by MalExt Sentry.

Achieving remote code execution on every player in a game lobby through a malicious Steam Workshop map.

Offensive C development has been stuck in the dark ages. SindriKit is what happens when you apply thirty years of software engineering to the problem.

CVE-2026-25860 write-up on OpenClinic GA, covering a reflected XSS flaw in DICOM image uploads and its real-world security impact.

Indirect prompt injection poses a critical threat to LLM agents that interact with untrusted external data, yet automated attack methods--proven effective for jailbreaking--remain underexplored in...

Summary A vulnerability in Cisco Unified Communications Manager (CUCM) allows unauthenticated attackers to arbitrarily write files in the server which in turn can be used to run arbitrary commands/code on the server. Vendor Response The vendor has issued…

When a critical Linux kernel privilege escalation was publicly disclosed, Cloudflare's security and engineering teams detected, investigated, and mitigated the threat across our global fleet, confirming zero customer impact and no malicious exploitation.

Can AI actually hack? 2,639 real challenge attempts, 10 models, 5 categories

LucidBit Labs research shows how Samsung's FIVE integrity subsystem exposed an app-reachable kernel use-after-free with multiple exploit-relevant primitives.

A leaked dataset from 4vps[.]su shows a hosting environment where proxy networks, criminal infrastructure, and targeted attack activity coexist without requi...

Campus Recruiting vulnerability exposed student information, and Audit Tracking Management System vulnerability exposed confidential internal audit data.

In my last post, I argued that the Axios compromise was catchable because the malicious versions had no provenance — they were pushed by hand with a stolen t...

Discover how we hatch ventures with founders and apply.