Tidianez

Building SafeRun — inline reliability layer for AI agents in production. Stops wrong/dangerous actions before they happen and lets you replay failures exactly. https://saferun.dev