From isolated snippets → full workflows. Proud to partner with @Stanford and @LaudeInstitute on #TerminalBench 2.0 — helping redefine agent evaluation. Thanks @Mike_A_Merrill and @alexgshaw for leading the charge and allowing our researchers the opportunity to contribute.

→ View original post on X — @snorkelai