Open Superintelligence Stack
Today, we are launching Prime Agent, our self-improving coding harness designed around two abstractions, the Recursive Language Model (RLM) [citation] and Continual Harness [citation]. Modern harness
