Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Security tools don’t usually break. They just slowly stop doing what you think they’re doing. Or perhaps were never set up to do what you needed in the first place. Something got deployed. It worked. Then it drifted. No one noticed. And three years later, you’re questioning the renewal because you’re not even sure what it’s protecting anymore. That’s configuration rot. Thanks to Julian Lee at eChannelNews for the fun, thoughtful and much needed conversation on this topic and more.

This week on the podcast, we cover the WatchGuard Threat Lab's Internet Security Report for the second half of 2025. In this episode, we cover the latest trends in malware at both the network perimeter and endpoints, network attacks, and the top malicious domains from the period before, and end with some tips everyone can use to defend their networks.

We don’t bend the elements, but we do help defend against dodgy AI avatars – meet Avast: The last online protector you’ll ever need.

In this episode, we look at Moltworker (for OpenClaw) and Markdown for Agents — including how converting HTML to markdown can reduce a 16,000-token page to ~3,000.

Is Cursor’s new Composer 1.5 model a major leap forward, or just a marginal update? Today, we’re putting the latest version of Cursor’s agentic AI to the test using our "Production-Ready Note App" prompt. We compare the speed, UI design, and agentic capabilities of 1.5 against version 1.0. Most importantly, we run a full security audit using the Snyk extension to see if the AI-generated code is actually safe for production.