This skill should be used when the user asks to "diagnose context problems", "fix lost-in-middle issues", "debug agent failures", "understand context poisoning", or mentions context degradation, attention patterns, context clash, context confusion, or agent performance...

当用户询问“诊断上下文问题”“解决中间信息丢失”“调试代理失败”“理解上下文污染”或提及上下文退化、注意力模式、上下文冲突、上下文混淆或代理性能时，应使用此技能