SkillPublic
eval-dataset-design
设计 Agent 评估数据集或单条评估任务时使用——解剖评估任务的四个组成部分、设计验证器与验收标准（含防敷衍性修复）、划分任务难度与陷阱任务、防范训练/评估数据泄漏（参数化实例、canary）、建设评估集的质量控制与长期维护、选择公开基准/自建业务集/生产轨迹回流三个来源。覆盖 τ²-bench、SWE-bench、AndroidWorld、OSWorld、Terminal-Bench、GAIA 的可迁移设计手法。
Quality
87%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Securityby
Passed
No findings from the security scan
No evaluations available
This skill hasn't been evaluated yet
tessl scenario generatetessl eval run- Repository
bojieli/ai-agent-book
- Revision
dbc046e
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.