February 12, 2026
featured post
Web3 Audit Benchmark: GPT-5.2 and Claude 4.5 vs. Sherlock AI
A Controlled Benchmark comparing ChatGPT 5.2 and Claude Sonnet 4.5 against Sherlock AI v2.2 on the Flayer + Moongate repo, scored by an independent security researcher for validated findings, false positives, and code-grounded evidence.