Attach judges to config variations for automatic LLM-as-a-judge evaluation. Create custom judges, configure sampling rates, and monitor quality scores.
安装
npx skills add launchdarkly/ai-tooling --skill online-evalsnpx skills add launchdarkly/ai-tooling --skill online-evals -a claude-code -gnpx skills add launchdarkly/ai-tooling --skill online-evals -a codex -ggemini skills install https://github.com/launchdarkly/ai-tooling.git --path skills/agentcontrol/online-evalsgh skill install launchdarkly/ai-tooling online-evalsgit clone --depth 1 https://github.com/launchdarkly/ai-tooling /tmp/launchdarkly__ai-tooling
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/launchdarkly__ai-tooling/skills/agentcontrol/online-evals ~/.claude/skills/online-evals # Claude Code
cp -R /tmp/launchdarkly__ai-tooling/skills/agentcontrol/online-evals ~/.agents/skills/online-evals # Codex, Cursor, Gemini CLI, Copilot, OpenCode