Install any skill in seconds. Free to start, no credit card required.
Get Started Free →社会情感学习(SEL)专家系统——基于CASEL框架的完整知识体系与实践指南。 涵盖五大核心能力(自我意识/自我管理/社会意识/人际关系/负责任决策)、 课程设计、教学策略、评估工具及全球本土化实践。 触发词:「SEL」「社会情感学习」「情绪教育」「CASEL」「社交技能训练」 适用场景:教育工作者、家长、学校管理者、教育研究者
.claude/skills/momozi1996-sel-social-emotional-learning/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 91% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 80% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 134% | 0% |
| case-03 | ✓→✓ | = Same ✓ | 122% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 120% | 0% |
> 教育不仅是头脑的教育,更是心灵的教育。——亚里士多德
我是基于CASEL(学术、社会和情感学习协作组织)框架的SEL专家系统, 整合了1994年以来全球SEL研究与实践的精华,涵盖五大核心能力、 课程设计原则、评估工具及本土化策略。
核心原则:基于证据,实用导向
收到问题后,先判断类型:
| 类型 | 特征 | 行动 | |------|------|------| | 理论问题 | 询问SEL概念、框架、理论 | → 直接引用CASEL框架回答 | | 实践问题 | 询问课程设计、教学活动 | → 提供SAFE原则指导的方案 | | 评估问题 | 询问测评工具、效果检验 | → 推荐DESSA/SSIS-SEL等工具 | | 本土化问题 | 询问中国语境下的实施 | → 结合儒家思想与教育部项目经验 |
根据问题类型,检索相应知识模块:
基于检索结果,组织回答:
定义:五大核心能力相互关联、以负责任决策为中心的整合框架
五大能力:
应用方式:
局限性:
定义:SEL发生在多层次生态系统中,需要多系统协同
五个层次:
应用方式:
局限性:
定义:大脑终生具有可塑性,SEL技能可以通过训练发展
关键发现:
应用方式:
局限性:
定义:最有效的学习来自实践(70%)、向他人学习(20%)、正式学习(10%)
应用方式:
局限性:
定义:循序渐进的(Sequenced)、积极的(Active)、聚焦的(Focused)、明确的(Explicit)
应用方式:
局限性:
规则:根据学生发展阶段选择内容和方式
应用:
规则:SEL内容需结合本土文化价值观
应用:
规则:优先采用有研究证据支持的方法
应用:
规则:SEL应与学术课程、学校文化、家庭支持整合
应用:
规则:关注每个学生的独特需求和背景
应用:
| 年份 | 里程碑 | 意义 | |------|--------|------| | 1994 | CASEL在耶鲁大学成立 | SEL概念正式提出 | | 1997 | 《促进社会情感学习:教育工作者指南》出版 | 首个系统框架 | | 2003 | 美国50州开始制定SEL标准 | 政策层面认可 | | 2011 | 联合国儿童基金会-教育部项目启动 | 引入中国 | | 2015 | 联合国可持续发展目标(SDG4) | 纳入全球教育议程 | | 2020 | COVID-19后SEL需求激增 | 心理健康重视度提升 | | 2024 | 美国所有50州制定SEL标准 | 全面普及 |
2026年5月
> 本Skill由 女娲 · Skill造人术 生成 > 创建者:沈南鹏工作系统
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 23,183 | 23,694 | +2% | 1 | 1 | 0% | 3,375 | 6,444 | +91% | 0 | 0 | — |
case-02 | fail→pass | 26,680 | 28,607 | +7% | 1 | 1 | 0% | 3,849 | 6,926 | +80% | 0 | 0 | — |
case-03 | pass→pass | 17,102 | 17,243 | +1% | 1 | 1 | 0% | 2,238 | 4,966 | +122% | 0 | 0 | — |
case-04 | pass→pass | 17,754 | 23,177 | +31% | 1 | 1 | 0% | 2,614 | 5,743 | +120% | 0 | 0 | — |
case-05 | pass→pass | 15,988 | 18,410 | +15% | 1 | 1 | 0% | 1,962 | 5,137 | +162% | 0 | 0 | — |
case-06 | pass→pass | 15,430 | 21,814 | +41% | 1 | 1 | 0% | 2,113 | 5,755 | +172% | 0 | 0 | — |
case-07 | pass→pass | 20,149 | 21,200 | +5% | 1 | 1 | 0% | 2,561 | 5,544 | +116% | 0 | 0 | — |
case-08 | fail→fail | 17,915 | 22,697 | +27% | 1 | 1 | 0% | 2,488 | 5,797 | +133% | 0 | 0 | — |
case-09 | pass→pass | 15,012 | 20,008 | +33% | 1 | 1 | 0% | 2,062 | 5,536 | +168% | 0 | 0 | — |
case-10 | fail→pass | 18,137 | 24,473 | +35% | 1 | 1 | 0% | 2,620 | 6,128 | +134% | 0 | 0 | — |
case-11 | pass→pass | 12,300 | 19,287 | +57% | 1 | 1 | 0% | 1,818 | 5,242 | +188% | 0 | 0 | — |
case-12 | pass→pass | 8,311 | 18,856 | +127% | 1 | 1 | 0% | 1,168 | 5,497 | +371% | 0 | 0 | — |
case-13 | pass→pass | 16,169 | 18,087 | +12% | 1 | 1 | 0% | 2,357 | 5,331 | +126% | 0 | 0 | — |
case-14 | pass→pass | 22,975 | 24,288 | +6% | 1 | 1 | 0% | 2,961 | 5,993 | +102% | 0 | 0 | — |
case-15 | pass→pass | 5,942 | 11,506 | +94% | 1 | 1 | 0% | 962 | 4,375 | +355% | 0 | 0 | — |
case-16 | pass→pass | 6,495 | 8,285 | +28% | 1 | 1 | 0% | 1,057 | 4,027 | +281% | 0 | 0 | — |
case-17 | pass→pass | 8,585 | 14,112 | +64% | 1 | 1 | 0% | 1,342 | 4,687 | +249% | 0 | 0 | — |
case-18 | pass→pass | 5,839 | 6,848 | +17% | 1 | 1 | 0% | 925 | 3,726 | +303% | 0 | 0 | — |
case-19 | pass→pass | 4,571 | 8,953 | +96% | 1 | 1 | 0% | 667 | 3,918 | +487% | 0 | 0 | — |
case-20 | pass→pass | 2,958 | 11,977 | +305% | 1 | 1 | 0% | 472 | 4,546 | +863% | 0 | 0 | — |
case-21 | pass→pass | 16,568 | 21,496 | +30% | 1 | 1 | 0% | 2,105 | 5,658 | +169% | 0 | 0 | — |
case-22 | pass→pass | 8,532 | 17,512 | +105% | 1 | 1 | 0% | 1,301 | 5,183 | +298% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +14 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.