skill
职业分类
描述
更新
run-and-summarize
软件质量保证分析师与测试员
Run a Promptfoo eval for a config and produce the day-01-style reflection table (per-provider verdicts, key differences, latency). Use when the user says "run and summarize", "reflection table", "compare models".
2026年8月16日
new-eval-suite
软件开发工程师
Scaffold a new Promptfoo triplet (prompt reference + tests file + config) for a workshop lesson or bot, wired to the Groq default providers. Use when the user says "new suite", "scaffold a lesson", "add an eval".
2026年8月16日