RC RANDOM CHAOS

detection engineering

24 posts

Article

OpenCV 5.0 made adversarial perturbations transferable

OpenCV 5's bit-exact numerics and expanded encoder control shrink the attacker's modelling error against deepfake detectors. The exposure is structural.

Article

Sixty-three days to patch a forked parser

Technical breakdown of the FrontierOS RCE: a forked XML parser, an unpatched two-year-old CVE, and the fork-tracking failure that shipped it.

Article

CVE-2024-3400 shipped exploited before the advisory

Why the gap between CVE disclosure and production detection is structural - and where attackers operate inside it.

Article

The integration is the attack surface

Pentagon raised Israeli collection risk to top tier. The technical exposure is supply chain privilege inherited from vendor software, not espionage.

Article

The .docx in your webmail preview pane

Browser-side OOXML rendering converts trusted document parsers into renderer-context exploit primitives. The detection stack does not see the boundary cross.

Article

Thirty years of weaponizing fork-exec

fork+exec inherits file descriptors, environment, and capabilities by default. That inheritance is the bug class behind Shellshock, runc CVE-2019-5736, and Symbiote.

Article

Spanish police flagged GrapheneOS as suspicion

Authorities treating GrapheneOS as a targeting signal inverts threat intel logic and exposes the wrong population to scrutiny. The mechanism breakdown.

Article

340 million records, unverified seller

Technical analysis of plausible attack vectors behind the claimed OnlyFans 340M record leak, with detection signatures for each path.

Article

Your AI security tool blocks nothing

A red team operator's breakdown of why AI cybersecurity tools are sold as controls but function as telemetry with a verdict attached.

Article

Harvard.edu among 141 hosts serving ClickFix lures

Technical analysis of the campaign that weaponised harvard.edu and 140 other legitimate sites - entry vectors, TDS chain, MITRE mapping, EDR telemetry.

Article

Megalodon hijacked 55,000 GitHub repos via token replay

Megalodon compromised 55,000+ GitHub repositories through PAT harvesting, pull_request_target abuse, and OAuth scope inheritance. Technical breakdown.

Article

Your valid credentials are the breach.

Technical analysis of a coordinated GitHub Actions workflow compromise across 5,561 repositories, with detection guidance for audit log and EDR telemetry.