3Claude: Elevated Error Rates for Opus 4.8, Opus 4.7, Opus 4.6, and Sonnet 4.6 (opens in new tab)(status.claude.com)34forks3d ago38Save
4Claude: Elevated errors across many models [resolved] (opens in new tab)(status.claude.com)189forks8d ago160Save
5An Unbiased OSS Benchmark for Code Review Agents (opens in new tab)(codereview.withmartian.com)3forks19d ago0Save
6The human cost of 10x: How AI is physically breaking senior engineers (opens in new tab)(techtrenches.dev)82forks2mo ago71Save
7Zero Day Clock: The gap between disclosure and exploitation is collapsing to 0 (opens in new tab)(zerodayclock.com)3forks3mo ago0Save
9AWS Duvet: a bidirectional link between implementation and specification (opens in new tab)(awslabs.github.io)12forks5mo ago1Save
10Hax: Verifying Security-Critical Rust Software Using Multiple Provers (opens in new tab)(eprint.iacr.org)2forks5mo ago0Save
13Architecting Security for Agentic Capabilities in Chrome (opens in new tab)(security.googleblog.com)1forks5mo ago0Save
14Continuously hardening ChatGPT Atlas against prompt injection attacks (opens in new tab)(openai.com)3forks5mo ago0Save
