克劳德代码会话的正式评估框架,实施评估驱动开发(EDD)原则
56%
Does it follow best practices?
Impact
99%
1.70xAverage score across 3 eval scenarios
Passed
No findings from the security scan
Fix and improve this skill with Tessl
tessl review fix ./docs/zh-CN/skills/eval-harness/SKILL.mdScanned
xu-xiang/everything-claude-code-zh
dfbf946
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.