- pentest benchmarkQwen3.6-27BClaude Code skills harness
How is Qwen3.6 for automated pentesting?
Alessio MauroMay 5, 20268 min readQwen3.6-27B solves 75.96% of XBOW pentest challenges with the Claude Code skills harness — benchmarked head-to-head against GPT-5.4, Opus, Sonnet, Haiku, and GPT-5-nano in skills and vanilla modes.
- SecurityOSSecurity OperationsCloud Security
Why SecurityOS?
Venkat PothamsettyMay 5, 20263 min readApps, tools, and dashboards stopped working for me. This post explains why SecurityOS is the model that finally fits real security investigations.
- audit qualityPCI QSASOC 2
Auditor Fatigue Is Real and It's Affecting Your Results
Udit PathakMay 4, 202610 min readWhen I told a CTO his clean SOC 2 probably masked a pretty thin audit, he wasn't surprised, but relieved someone said it out loud. A short, blunt read on why audit quality slips (it's not just tired people), and the tells that separate real testing from checkbox theater.
