CTO Vincent Berg mirrored more than 180,000 Ruby gems and built a scanner to analyse them at scale. What he uncovered offers a fascinating look at the hidden risks and exposed data scattered throughout the Ruby ecosystem.

Director of Research Tao Sauvage breaks down how Anvil uses LLMs during secure code review, keeping human judgment at the center. He compares two approaches across three open-source targets to show why a structured workflow catches more vulnerabilities.

From threat modeling and attack surface mapping to jailbreaks and prompt injection, Security Engineer George Damiris walks through how Anvil Secure tests LLM and GenAI systems for security vulnerabilities.

In Part 2 of a three-part series, Security Engineer Matteo Giordano explores passkey architecture, covering origin binding, synced vs. device-bound credentials, authenticator types, and deployment quirks like key wrapping and signature counters.