We built evaluator templates around the most common eval requirements when putting agents into production. 30+ customizable templates are now available, including: LLM-as-judge evaluators with tuned prompts Rule-based code evaluators Also available in openevals v0.2.0,

→ View original post on X — @langchain,