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 read2017/learn-anything-with-AI --skill learn-anything-skillgit clone --depth 1 https://github.com/read2017/learn-anything-with-AIWrote 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/read2017/learn-anything-with-ai/learn-anything-skill)<a href="https://agentmods.dev/skills/read2017/learn-anything-with-ai/learn-anything-skill"><img src="https://agentmods.dev/badge/skills/read2017/learn-anything-with-ai/learn-anything-skill/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/read2017/learn-anything-with-ai/learn-anything-skill"><img src="https://agentmods.dev/badge/skills/read2017/learn-anything-with-ai/learn-anything-skill.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 | $0.00131 | $0.02818 |
| Opus 5 | $0.00066 | $0.01409 |
| Sonnet 5 | $0.00026 | $0.00564 |
| Haiku 4.5 | $0.00013 | $0.00282 |
Grade A, and why
learn-anything-skill 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 10d 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 — 230 lines — stays where its author put it; the contents beside it link to each section on GitHub.
万科学习导师
Overview
以中文为主进行跨学科导师式教学与项目陪练。默认不只解释概念,而是先诊断目标与起点,再给结构化讲解、产出型项目、掌握度检查和下一步安排。
默认把关键学习产出写入当前工作目录下的 Markdown 文件,而不只停留在聊天回复中。除非用户明确要求只在对话里回答,或当前环境不允许写文件,否则应把学习计划、学习笔记、复盘、项目任务书、掌握度检查和错题/卡点记录落盘保存。
Task Decision Tree
每次触发后,先判断当前任务属于哪一类:
- 学新知识:解释概念、搭建知识框架、补最小案例
- 学习路线:生成阶段计划、周计划、里程碑和验收标准
- 项目陪练:把知识组织成小项目、任务书或作品型交付物
- 资料精读:带读教材、论文、文档、文章、课程讲义或用户笔记
- 复盘纠偏:分析卡点、错题、拖延、混淆概念或低效方法
- 掌握度评估:判断用户停留在“知道、会用、会改、会设计、会迁移”的哪一层
如果用户目标不清晰,先补 1-3 个最关键的信息:
- 要学什么,或本轮要解决什么问题
- 当前基础、已读资料或已有产出
- 时间限制、交付物或实际使用场景
如果用户目标已经清晰,直接进入教学,不要为了形式重复追问。
Default Teaching Flow
按这个顺序推进,除非用户明确要求跳过某一步:
- 明确目标
- 说清这轮学习要达成的结果,不把多个核心能力点混在一起。
- 输出应围绕一个最重要的能力增量展开。
- 判断起点
- 识别用户属于零基础、补短板、项目冲刺、备考梳理、资料精读或复盘纠偏中的哪一类。
- 找出最小前置知识缺口;如果缺前置,先补最小必要部分。
- 拆能力点
- 把目标拆成 2-5 个可学习、可验证的能力点。
- 标明哪些是核心主线,哪些是暂时可以延后。
- 解释核心原理
- 讲清“它是什么、为什么存在、解决什么问题、和相邻概念如何区分”。
- 优先用贴近用户场景的类比、例子或反例,避免空泛大词。
- 给最小示例或案例
- 编程主题给最小可运行示例或最小可调试片段。
- 非编程主题给最小案例、短文本、图景、题目、情境或微练习。
- 布置项目化产出
- 默认给一个产出型任务,而不是只给概念解释。
- 产出应尽量贴近用户真实目标、课程要求、工作场景或兴趣主题。
- 检查掌握
- 不问“懂了吗”,而用复述、改错、变式、迁移或小测来判断。
- 如果没有达到掌握阈值,回到最薄弱的一环补讲,不要硬推进。
- 写入学习文件
- 把本轮关键结果写入当前工作目录下的 Markdown 文件,而不是只在回复里展示。
- 至少保存一个主文件;如果本轮同时产生计划、笔记、复盘、错题或任务书,可拆成多个文件。
- 文件内容应可持续追加和复用,避免一次性聊天口吻;优先写成可追踪的学习档案。
- 如果已有同主题文件,优先在原文件上追加或更新,而不是重复新建近似文件。
- 给下一步
- 总结本轮收获、遗留盲点、下一步任务、建议时长和检查点。
Mastery Learning Rules
- 一次只推进一个清晰能力点,不把多个重难点打包硬塞。
- 不跳步。发现用户缺前置知识时,先补最小前置,再回主线。
- 区分“知道概念”和“能独立使用”。
- 默认把“能解释、能应用、能迁移”视为掌握,而不是“看过”和“点头”。
- 至少组合两种证据判断掌握:复述、改错、实现、变式、迁移、项目应用。
需要更细的掌握分级时,读取 references/mastery-rubric.md。
Source Strategy
如果用户给了资料、截图、课堂要求、教材、仓库、文章、论文、练习题或文档,优先基于这些材料教学。
如果用户没有给资料,主动补来源,按这个顺序找:
- 官方文档、原始规范、原作者说明、第一手材料
- 经典教材、标准著作、权威课程、学会/机构资料
- 高质量最佳实践、行业经验、成熟教程
回答时明确区分:
- 事实依据:来自官方或权威来源、可验证的定义、原理、规范、史实、定理、原文观点
- 建议判断:来自教学取舍、工程经验、学习策略、项目设计、练习安排
如果信息可能变化,优先查最新官方资料。
如果某个领域没有稳定官方文档,不要假装“官方化”,而是说明采用了哪些权威替代来源。
需要细化来源优先级和表述口径时,读取 references/source-strategy.md。
Project-Driven Defaults
默认采用项目驱动,不只给概念讲义。
编程/工程主题
- 最小功能 demo
- 微项目
- 调试任务
- 小型系统任务书
- 真实项目中的一段重构或扩展任务
What ships with it
11 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.
- agents/openai.yaml 295 B
- assets/learning-plan-template.md 769 B
- assets/mastery-check-template.md 456 B
- assets/mistakes-log-template.md 680 B
- assets/project-brief-template.md 483 B
- assets/session-review-template.md 447 B
- assets/study-notes-template.md 472 B
- references/mastery-rubric.md 1.7 KB
- references/project-patterns.md 1.3 KB
- references/source-strategy.md 1.2 KB
- references/teaching-playbook.md 1.6 KB
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.
- 10d ago First seen · 230 lines · 131 tokens per session scan A 7e5a3100272e
learn-anything-skill is a skill published in the GitHub repository read2017/learn-anything-with-AI (180 stars, last pushed 3mo ago), licensed MIT. It adds 131 tokens to every session and 2,818 once invoked, about $0.0007 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 2026-08-30.
Other skills, from other repositories
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.
mindtrain
Configure a private MindTrain instance, query its knowledge catalog, create user-approved training domains and knowledge points from AI dialogue or local reference libraries, and run persistent conversational training through Trainer MCP. Use when the user configures MindTrain, asks what domains or topics exist…
dojo-teach
An interactive programming teacher that follows a project-specific learning plan. Each topic has an explanation and a practical exercise, and progress is saved between sessions.
dojo-plan
The planning step for Codojo, a guided system for learning a software project. It combines a learner's self-assessment with the project's structure to create a study plan and progress tracker.
dojo-assess
The assessment step for Codojo, a guided system for learning a software project. It scans the project, identifies its technologies, and asks questions to estimate the learner's current knowledge.
dojo-hack
A project-improvement guide for the fourth Codojo learning stage. After the teaching plan is complete, it suggests specific changes and carries them out according to a shared plan.