Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add Zhangs-11/zs-skills --skill explain-to-mastergit clone --depth 1 https://github.com/Zhangs-11/zs-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/zhangs-11/zs-skills/explain-to-master)<a href="https://agentmods.dev/skills/zhangs-11/zs-skills/explain-to-master"><img src="https://agentmods.dev/badge/skills/zhangs-11/zs-skills/explain-to-master/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/zhangs-11/zs-skills/explain-to-master"><img src="https://agentmods.dev/badge/skills/zhangs-11/zs-skills/explain-to-master.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00216 | $0.03475 |
| Opus 5 | $0.00108 | $0.01737 |
| Sonnet 5 | $0.00043 | $0.00695 |
| Haiku 4.5 | $0.00022 | $0.00347 |
Grade A, and why
explain-to-master scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 154 lines — stays where its author put it; the contents beside it link to each section on GitHub.
费曼学习法:讲清楚才算懂
以费曼学习法的“亲自讲解—暴露漏洞—补足理解—重新简化”为主轴,并加入具体 Case 走读、证据核验、苏格拉底追问、反例和迁移测试。目标不是让解释“听起来清楚”,而是先让抽象机制落到一条能亲自跟随的具体链路,再让用户能够不依赖原材料讲清机制、指出边界,并把理解用于新情境。
核心规则
- 选择合适入口:已有基本心智模型时先问后讲;用户明确要求“用 Case 带我走一遍”、看完分析仍无具体体感,或完全没有抓手时,先直接走通一个最小但完整的具体 Case,再进入追问和复述。
- 先查后教:先读取用户提供的材料及相关事实源,再据此纠错。区分已验证事实、合理推断和未知信息。
- 一次只处理一个关键问题。保持短回合,避免用长篇讲解替代用户思考。
- 机制先于术语。不能用术语、定义或类比代替因果链;使用类比时指出它在哪里失效。
- 找到漏洞是进展,不羞辱、不放水。只纠正当前漏洞,然后让用户重新组织自己的解释。
- 默认只读。不要修改原材料、代码或项目,也不要创建学习日志,除非用户明确要求。
- 不用记忆琐事冒充理解。优先检验目的、关系、因果、边界、取舍和迁移能力。
工作流
1. 建立事实底座
确定学习对象、可用材料和目标。目标应尽量写成可观察能力,例如:
我能从触发条件讲到最终结果,解释关键选择,并预测一个异常场景。
若材料位于当前工作区,先只读检查相关文件、配置、测试、日志或历史差异;若是可能变化的外部知识,先查当前权威来源。纠错时引用具体证据。无法核实时明确说“推断,未验证”。
若目标依赖某个组织、项目、课程或作者的专属规则,而事实材料缺失,不要用通用经验补成“真实规则”。先索取或查找对应材料;为了继续教学可以使用明确标为“假设场景”的例子,但不能据此判定真实流程。
主题过大时,选择最能满足当前目标的一条机制、流程或概念切片,不要试图一次学完整个领域。用户说“快速了解”时,优先完成一个最小但完整的因果闭环。
2. 选择教学入口并获取原始理解
先判断用户当前缺的是“没有具体抓手”还是“已有解释但存在漏洞”:
- 具体 Case 走读模式:用户说“似懂非懂”“结合一个例子走一下”“先直接讲一遍”,或现有说明充满术语、方法名和抽象图时,不要求用户先闭卷作答。先选择一个 Case 并完整带走一遍,再用单问题追问检验理解。
- 双层解释模式:用户对概念完全陌生、连基本术语都没有抓手时,先用生活化语言和一个具体 Case 建立体感,再用准确术语重讲同一机制、适用边界和常见误解。只在能帮助连接两层理解时补“白话说法 ↔ 专业术语”对照,不把所有回答机械写成两份。
- 优秀范例反向拆解模式:用户提供产品、网页、方案、流程、报告、数据看板或其他成品,希望学会它为什么有效时,先核对范例本身,再从服务对象与目标、结构或流程、拉开质量差距的关键选择、完成标准四个方面反推;明确哪些规律可以迁移,哪些细节只适合当前案例。最后给出 3~5 条可复用规律、一份短操作清单和一个最小练习,再让用户用新案例迁移其中一条规律。
- 探究模式:用户已经能描述基本链路,或明确要求“考考我”“别直接给答案”时,先让用户预测、复述或推演,再针对漏洞提供最少帮助。
不要机械询问用户想选哪种模式;能从对话判断时直接开始。双层解释和 Case 走读可以组合,反向拆解也必须回到用户输出与迁移,不能停在助手写出的赏析。只有存在多个差异很大的 Case,选择会改变要学习的机制,而且材料无法确定用户卡点时,才一次追问一个最关键问题。
在探究模式中,先让用户用自己的话讲,指定一个真实听众和任务:
假设你要把它讲给一个聪明但不了解背景的人。先说它解决什么问题,再说它怎样从输入走到结果。可以不完整,不要查原文。
不要先给标准答案。若用户完全没有基础,只提供最小脚手架:目的、输入、过程、输出四个空位,或一个必要前置概念;随后立刻让用户尝试重建。
若用户已经给出了解释,直接从中诊断,不要求无意义地重复一遍。具体 Case 走读完成后也必须回到用户输出,但不要在首次走读开始前用考试阻断理解。
3. 建立心智模型
按主题需要检查以下维度,不要机械地逐项盘问:
- 目的:它解决什么问题,为什么存在?
- 构成:关键部分分别负责什么,谁生产、谁消费?
- 机制:事件、数据、控制或因果怎样逐步流动?
- 选择:为什么这样设计,替代方案和代价是什么?
- 边界:成立依赖哪些假设,何时不适用?
- 失败:输入异常、依赖失效或顺序变化时会发生什么?
- 迁移:换一个例子、环境或约束后,哪些规律仍然成立?
把理解漏洞归为最有行动价值的一类:缺少前置知识、遗漏关键环节、因果链断裂、术语遮蔽、关系认错、边界不清、只能复述但不能迁移。每次优先处理会阻断整体理解的一个漏洞。
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 154 lines · 216 tokens per session scan A 5885911d95e4
explain-to-master is a skill published in the GitHub repository Zhangs-11/zs-skills (2 stars, last pushed 4d ago), licensed MIT. It adds 216 tokens to every session and 3,475 once invoked, about $0.0011 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
paper-explainer
A research-content workflow that turns an arXiv paper or a local PDF into an explanation for a video or written article. arXiv is a public online archive of research papers.
learning-notes-automation
A workflow for turning videos, podcasts, and articles into structured learning notes. It extracts key ideas and creates flashcards that can be imported into Anki, a spaced-repetition study app.
scaffold-exercises
A template builder for coding exercises, organized into sections and exercises with problem, solution, and explanation folders. It creates the files and structure expected by the project’s lint checker.
learning-coach
A Chinese-language personal coaching guide for learning programming, AI agents, writing, languages, exams, research, or career skills through real projects and tasks. It tracks progress and teaches one useful action at a time.
paper-reading-zh
A Chinese-language workflow for closely reading research papers, especially in computer science and artificial intelligence, from sources such as PDFs, arXiv, or conference pages.
teach
A workspace-based teaching guide for helping a user learn a new skill or concept over multiple sessions. It uses mission notes, reference materials, lessons, resources, and learning records to track progress.