Security TrustWinning the Opir Challenge by testing the edgesI was one of the first-place winners of Wordcab's Opir Challenge after probing where semantic guardrails get thin.June 29 2026//7 min read#opir#ai security#red teamingOpen post→
Model WatchOpir made guardrails look like classifiers againOpir put safety filtering back in the small-model conversation with encoder guardrails for toxicity, jailbreaks, and policy categories.May 29 2026//6 min read#opir#classification#guardrailsOpen post→