eval-design
Use when: 要判断 Agent 是否专业、省心或真正改善，需设计判断依据、指标、校准，或修改/审计 eval。Not for: 单次交付(quality-gate)、外部claim(source-audit)、摩擦诊断(code-as-harness)。Output: 测量与校准草案或指标出生证（纵向eval含触发契约），或病名+处置。
No security review available
No security review has been published for this skill
- Repository
- zts212653/clowder-ai
- Commit
3045379
Is this your skill?
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.