{"id":"f8b34f94-6fda-4a97-b8a5-8d727ca4e86a","slug":"clawhub-kokxi-qa-expert-review","name":"qa-expert-review","description":"当 AI 生成的测试用例已经过输出评审和盲区补盲、准备终审上线时使用此技能。由资深测试对 AI 输出的用例做人工抽样校验，从业务有效性、场景完整性、可执行性三个维度做最后把关。⚠️ 如果发现系统性问题（比如遗漏了某个关键模块），需要回退修正并记录到 Prompt 优化反馈库。专家评审不是走形式——发现的问题必须闭环。 触发场景：专家评审、用例审查、校正反馈、评审用例、检查用例、终审、用例上线前需要终审时。 Use when the user asks about: final human-style review of AI-generated test cases before shipping, sampling for business validity, scenario completeness, and executability.","capabilities":[],"protocols":["OPENCLAW"],"safetyScore":84,"overallRank":62,"trustScore":null,"trust":null,"source":"CLAWHUB","updatedAt":"2026-10-11T11:52:17.276Z"}