Run offline Python (`pydantic_evals`) or Node.js (`logfire/evals`) evaluations and review them in Logfire. Also redirect existing Braintrust `Eval()` suites. Use for eval setup, test datasets, AI scoring, agent checks, LLM judges, Braintrust migration, or Logfire Datasets & Experiments. Not for live traffic or infrastructure monitoring.
Install
npx skills add pydantic/skills --skill logfire-evalsnpx skills add pydantic/skills --skill logfire-evals -a claude-code -gnpx skills add pydantic/skills --skill logfire-evals -a codex -ggemini skills install https://github.com/pydantic/skills.git --path skills/logfire-evalsgh skill install pydantic/skills logfire-evalsgit clone --depth 1 https://github.com/pydantic/skills /tmp/pydantic__skills
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/pydantic__skills/skills/logfire-evals ~/.claude/skills/logfire-evals # Claude Code
cp -R /tmp/pydantic__skills/skills/logfire-evals ~/.agents/skills/logfire-evals # Codex, Cursor, Gemini CLI, Copilot, OpenCode