皮亚杰建构写作法
v1.0.1皮亚杰建构主义(constructivism)写作与教学表达工作流。用于把知识点写成“读者可建构理解”的文章/讲义/课程脚本:从旧知出发,制造可控认知冲突,同化/顺应逐步搭建概念结构,输出概念地图、学习路径、分层讲解、练习题、误解纠偏与目标达成评估。
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name, description, and the included templates/references all align with a writing/teaching-workflow skill. There are no unrelated environment variables, binaries, or install steps that would be disproportionate to the stated purpose.
Instruction Scope
SKILL.md directs the agent to ask clarifying questions, use internal presets/templates, produce concept maps, exercises, and evaluations. It does not instruct reading system files, accessing environment variables, or sending data to external endpoints. The only open behavior is making reasonable assumptions when input is incomplete—those assumptions must be recorded as '待确认项', which bounds the agent's discretion.
Install Mechanism
No install spec and no code files — instruction-only. This minimizes file-system and network risk.
Credentials
No required environment variables, credentials, or config paths. The skill does not request secrets or unrelated tokens.
Persistence & Privilege
always is false; the skill is user-invocable and allows model invocation (the platform default). It does not request permanent presence or modify other skills/configurations.
Assessment
This skill appears coherent and low-risk: it only contains pedagogical instructions and local templates and asks for no credentials or installs. Before installing or using it, consider: (1) test with the included sample requests to verify style and accuracy; (2) if you provide real student examples or sensitive content to the skill, treat that data as potentially visible to the agent — avoid sharing personally identifying information; (3) review outputs for factual correctness and cultural/bias issues (the agent may make assumptions when inputs are incomplete); (4) because the skill can be invoked autonomously (platform default), only enable it in agents you trust and monitor initial outputs. If you want extra assurance, ask the publisher for provenance/homepage information (source is listed as unknown).Like a lobster shell, security has layers — review code before you run it.
latest
皮亚杰建构写作:让读者自己“搭起来”
触发方式
- 用户说“用建构主义/皮亚杰方式写/面向新手写讲义/想写成循序渐进的科普/老师备课/要从旧知引出新概念”时使用本 Skill。
- 适用:教学讲义、科普文章、知识体系笔记、课程脚本、训练营作业讲解。
输入模板(复制填空)
缺信息时先做合理假设,并把假设写入“待确认项”。如果用户只给了 Topic,没有给读者画像,优先问 1-2 个问题补齐。
主题(Topic):
读者(Audience):年级/背景/是否完全新手
写作目标(Goal):理解/能复述/能应用/能迁移
读者旧知锚点(Prior Knowledge):他们已经会什么/相信什么
常见误解(Misconceptions):3-5 条(没有就写“未知”)
必须包含/避免(Must include/avoid):
限制(Constraints):字数/时长/讲义 or 文章/是否允许公式
Quick Start(开箱即用)
当用户输入不完整时,按顺序补齐并开始产出:
- 选择场景预设:科普文章 / 课程讲义 / 学习笔记(见
references/presets.md) - 用
references/rubric.md生成 3-7 条 Success Criteria(成功标准) - 若用户想直接测试,提供
references/sample-requests.md的示例输入 - 读取
assets/output-template.md选择模板:短讲解用assets/compact-template.md;讲义/课程用assets/lesson-template.md - 按模板输出,并在最后做“目标达成评估”
工作流(建构主义写作)
按顺序执行;除非用户指定输出形式,否则使用“默认输出格式”。
1) 建立概念地图(Concept Map)
- 列出 5-10 个核心概念与关系(包含:上位/下位、因果、对比、前置依赖)。
- 选出 1 条“主线因果链/理解链”,后续段落围绕它展开,避免百科式罗列。
2) 旧知锚点与认知冲突(Prior Knowledge -> Cognitive Conflict)
- 用读者已有经验作为锚点(他们已懂的类比/直觉/场景)。
- 设计一个“可控冲突”:让旧知解释不了某个现象/题目/反例,从而引出新概念的必要性。
- 冲突必须可验证:用一个小例子或问题触发,而不是纯口号。
3) 同化与顺应支架(Scaffolding)
- 同化(Assimilation):先在旧框架下解释新信息能解释的部分,建立连续性。
- 顺应(Accommodation):明确指出旧框架的边界,再引入新框架,补齐解释力缺口。
- 每次引入新术语时,给一句话定义 + 一个最小例子(不要一次堆很多术语)。
4) 迁移练习(Transfer)
- 提供 3 组练习:
概念辨析、应用题、迁移题(从熟悉场景迁移到新场景)。 - 每题都给“为什么这么想”的解题思路,而不是只给答案。
5) 误解纠偏(Misconceptions)
- 列出 3-7 条常见误解:误解是什么 -> 为什么容易误解 -> 正确模型是什么 -> 如何自检。
- 需要证据边界时要写清:哪些是普遍结论,哪些取决于条件。
6) 目标达成评估(Evaluate Against Goal)
- 对照 Success Criteria 打分并给理由。
- 给出结论:
已达成/部分达成/未达成(需澄清)。 - 若输出仍显得“像灌输/跳步/术语先行”,读取
references/anti-patterns.md做二次修订后再给最终版。
默认输出格式(建议)
概念地图:概念与关系学习路径:旧知锚点 -> 冲突 -> 新概念 -> 迁移分层讲解:短讲解 or 讲义结构(按模板)练习题(含思路与答案)误解纠偏待确认项目标达成评估
资源文件(references/ 与 assets/)
- 预设:
references/presets.md(场景选择与默认结构) - 评分量表:
references/rubric.md(Success Criteria 与评估口径) - 示例输入:
references/sample-requests.md(复制即用) - 完成品样张:
references/example-outputs.md(对齐颗粒度与风格) - 反模式库:
references/anti-patterns.md(输出不清楚时用于纠偏) - 模板索引:
assets/output-template.md(选择短讲解/讲义模板)
Comments
Loading comments...
