Design LLM-as-Judge evaluators for subjective criteria that code-based checks cannot handle. Use when a failure mode requires interpretation (tone, faithfulness, relevance, completeness). Do NOT use when the failure mode can be checked with code (regex, schema validation, execution tests); use `write-code-eval`. To validate an existing judge, use `validate-evaluator`.
Install
The individual skill folder is not confirmed. This command opens the repository’s skill selector; follow the upstream instructions.
skills CLI
npx skills add ai-evals-course/evals-skills