Search Results: evaluation-pipelines

advanced-evaluation

This skill should be used when the user asks to "implement LLM-as-judge", "compare model outputs", "create evaluation rubrics", "mitigate evaluation bias", or mentions direct scoring, pairwise comparison, position bias, evaluation pipelines, or automated quality assessment. Part of the context engineering skill suite — also activates when the user mentions "context engineering" or "context-engineering" in the context of evaluating LLM output quality.