Validate LLM judges against human labels using TPR/TNR metrics and train/dev/test splits. Use after writing a judge prompt to verify it agrees with human judgment.
安装
npx skills add growthxai/output --skill output-eval-validate-judgenpx skills add growthxai/output --skill output-eval-validate-judge -a claude-code -gnpx skills add growthxai/output --skill output-eval-validate-judge -a codex -ggemini skills install https://github.com/growthxai/output.git --path coding_assistants/claude/plugins/outputai/skills/output-eval-validate-judgegh skill install growthxai/output output-eval-validate-judgegit clone --depth 1 https://github.com/growthxai/output /tmp/growthxai__output
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/growthxai__output/coding_assistants/claude/plugins/outputai/skills/output-eval-validate-judge ~/.claude/skills/output-eval-validate-judge # Claude Code
cp -R /tmp/growthxai__output/coding_assistants/claude/plugins/outputai/skills/output-eval-validate-judge ~/.agents/skills/output-eval-validate-judge # Codex, Cursor, Gemini CLI, Copilot, OpenCode