Security
Supply-chain attacks, exploited zero-days and the trust model behind the software we use. SoilTune's security coverage, start here.
Last updated November 21, 2026
Security coverage at SoilTune follows the parts of the stack that fail quietly: a vendor’s signup flow, an extension inside a remote-management tool, a model that finds a zero-day before the vendor can patch it. We are interested in what actually happened, which trust relationship broke, and what defenders can do about it.
Our approach is deliberately narrow. We read the primary advisories, the vendor statements and the technical write-ups. We separate what is confirmed from what is alleged, and we say when a key question is still unanswered. We do not publish exploit code or instructions that would help someone attack a system. When we get something wrong, we correct it in public on our corrections page.
Start with the supply-chain incidents below. They share a pattern: the victims’ own code was fine, and the trust they placed in someone else’s was the vulnerability. Then follow the zero-day and policy coverage for the other side of the same problem, how attackers find weaknesses, and how governments and vendors respond.
Start here
- Optimizely’s signup flaw put malware on 178 companies’ sites, How a flaw in a vendor’s signup flow put a card skimmer and crypto clipper on major brand sites, and why a three-hour containment claim does not describe a 16-hour delivery window.
- SCKit malware: ScreenConnect attacks aim at MSP technicians, A worm-like campaign that abused ScreenConnect’s own tooling to reach managed service providers and the technicians who administer them.
- ConnectWise ScreenConnect attacks hit nearly 50 organisations, The core product was patched. A third-party extension became the way in.
- Amazon says China-linked AI agents found a medical-imaging zero-day in 11 minutes, What Amazon published about the Cairn campaign, what it did not, and why the speed changes the defender’s calculus.
More security coverage
- Where was Nvidia’s security team? The CSO question, explained, What a chief security officer does, and why a missing one becomes a story after a breach.
- TP-Link’s Wi-Fi 8 router ban waiver, explained, The regulatory fight over routers, firmware and the security of consumer network gear.
- Dutch privacy watchdog warns about AI camera glasses, The surveillance risk that arrives in a normal-looking pair of glasses.
- GTA 6 hacker Arion Kurtaj pleads guilty days before launch, A criminal case that shows how stolen data moves through the games industry.
- Claude picked a bombed Iranian school as a target in tests, What a targeting evaluation found, and what it does not prove about deployed systems.
How we cover security
- Primary sources first. We link to vendor advisories, court documents, regulatory filings and the researchers’ own write-ups. If a claim rests on a secondary report, we say so.
- No exploit code. We describe impact, timeline and defense. We do not publish step-by-step instructions for attacking a system or evading a control.
- Updates are labeled. Refreshed articles carry a “What changed” block and a “Last reviewed” date. Substantive corrections are logged on the corrections page.
- Tips are welcome. Use the contact form and choose Editorial. We do not promise anonymity.
Related hubs
Security work rarely stays in one lane. For the policy side, see Policy. For the companies and products behind the incidents, see Big Tech. For the models and tools being attacked or used as attack tools, see AI. For patches, releases and developer tooling, see Software.
You can also browse the full Security archive, newest first.
Have a tip? Email [email protected] or use the contact form and choose Editorial. We do not promise anonymity.