Zlearning 是面向零基础、跨领域学习者和企业技术培训场景的 GitHub Copilot Skill。它把“专业上正确但新手看不懂”的知识，重构为准确、渐进、可复述、可练习、可验收的课程内容。
- 通俗解释概念、原理、区别和关系；
- 梳理知识地图、端到端机制链路与前置依赖；
- 改写或整合培训材料、课程讲义和学习资料；
- 生成中英文独立 Word 课程，并按需派生 PDF/PPT；
- 设计案例、练习、答案和分层能力测评；
- 审计技术准确性、新手友好度、双语一致性和 Office 交付质量。
尤其适合云计算、虚拟化、存储、网络、灾备、AI，以及 TAC/FAE 技术支持与企业培训。
Zlearning 使用 MAPS 作为内部推演框架，但不要求正文机械套用模板：
- Map：建立知识地图，明确定位和概念关系；
- Aim：说明没有该技术时的问题及其价值；
- Picture：按认知风险选择生活类比或直接演示；
- Steps：拆解输入、判断、动作、输出和交接；
- Scenario：放入真实业务、运维或排障场景；
- Seal：补充辨析、成立边界与记忆锚点。
- 术语依赖与首次定义：先建立依赖图和首次出现台账，避免循环解释。
- 学习主线与价值入口：先让学习者看见课程走向、现实问题和学习价值。
- 解释模式选择：支持
concept_first_analogy、analogy_first和direct_demonstration。 - 语义层隔离：生活类比、正式机制和真实业务案例不混写。
- 机制与覆盖台账：区分名称被提及和真正完成教学。
- 能力进阶：覆盖 Recognize、Explain、Reconstruct、Troubleshoot。
- 范围回归：改善呈现不等于扩大知识范围。
- 双语等值与交付验收：检查技术密度、结构、Office 可用性、视觉质量、版本和哈希。
最低只需提供 主题或任务，例如“解释 RAID 5”或“把这些资料做成新手课程”。为了让结果更贴合实际，建议同时提供：
- 主题与目标：希望解释、改写、整合、审计还是生成完整课程；
- 学习者：受众角色、已有基础及工作场景；
- 来源材料：Word、PDF、PPT、网页、现有课程或指定参考资料；没有材料时也可直接说明主题；
- 范围与边界：必须保留、需要删除、不要展开的内容，以及适用产品/版本；
- 语言与交付格式：中文、英文或双语，以及 Word、PDF、PPT 或仅聊天回答；
- 深度与用途：快速认识、系统自学、课堂培训、认证考试或技术支持；
- 特别要求：案例、练习、答案、页数、品牌风格、截止时间等。
信息不完整时，Zlearning 会先使用已知内容推进；只有缺失项会显著改变课程范围或交付方式时才提问。涉及完整课程且用户未指定格式时，默认规则如下。
请使用 Zlearning 处理以下任务：
主题/任务：[要解释或制作的内容]
学习者：[受众及基础]
来源材料：[文件或链接；没有可写“无”]
必须保留：[核心范围]
不要展开：[排除范围]
语言与格式：[例如：英文 Word]
用途与深度：[例如：新人自学课程]
其他要求：[案例、练习、产品版本等]
完整课程默认生成两个独立的 Word 主版本：中文版和英文版。正式生成前确认是否同时需要 PDF 与 PPT/PPTX。不同语言和格式应共享同一内容基线、术语表、技术结论、案例参数和练习范围。
Zlearning is a GitHub Copilot Skill for beginners, cross-domain learners, and enterprise technical training. It turns technically correct but difficult material into courses that are accurate, progressive, explainable, practice-ready, and verifiable.
- Explain concepts, mechanisms, differences, and relationships in accessible language.
- Build knowledge maps, prerequisite chains, and end-to-end technical flows.
- Rewrite or consolidate training materials, course notes, and learning resources.
- Produce separate Chinese and English Word courses, with optional PDF/PPT derivatives.
- Design scenarios, exercises, answer keys, and progressive competency assessments.
- Audit technical accuracy, beginner accessibility, bilingual equivalence, and Office delivery quality.
It is especially useful for cloud computing, virtualization, storage, networking, disaster recovery, AI, TAC/FAE support, and enterprise enablement.
Zlearning uses MAPS as an internal reasoning framework without forcing a repetitive writing template:
- Map — locate the concept and establish its relationships.
- Aim — explain the problem it solves and the value it provides.
- Picture — choose a safe analogy or direct demonstration.
- Steps — trace input, decision, action, output, and handoff.
- Scenario — apply the mechanism to a realistic work situation.
- Seal — clarify boundaries, distinctions, and a memorable anchor.
- Term dependencies and first definitions prevent circular explanations.
- Learning journey and value entry show learners where the course is going and why it matters.
- Explanation-mode selection supports
concept_first_analogy,analogy_first, anddirect_demonstration. - Semantic-layer separation keeps analogies, formal mechanisms, and business scenarios distinct.
- Mechanism and coverage ledgers distinguish a mentioned term from a concept that has truly been taught.
- Capability progression covers Recognize, Explain, Reconstruct, and Troubleshoot.
- Scope regression ensures presentation improvements do not expand the approved knowledge scope.
- Bilingual and delivery validation checks technical depth, structure, Office usability, visual quality, versions, and hashes.
At minimum, provide a topic or task, such as “Explain RAID 5” or “Turn these references into a beginner course.” For a more targeted result, include:
- Topic and objective — explain, rewrite, consolidate, audit, or create a complete course;
- Learners — audience roles, prior knowledge, and work context;
- Source material — Word, PDF, PPT, web pages, existing courses, or named references; a topic alone is also acceptable;
- Scope and boundaries — required content, exclusions, depth limits, and applicable product/version;
- Language and deliverable — Chinese, English, or bilingual; Word, PDF, PPT, or chat only;
- Depth and use — quick introduction, self-study, instructor-led training, certification, or support enablement;
- Special requirements — cases, exercises, answer keys, length, brand style, deadline, and similar constraints.
If information is incomplete, Zlearning proceeds with what is known and asks only when a missing choice would materially change the scope or deliverable.
Use Zlearning for this task:
Topic/task: [what to explain or create]
Learners: [audience and prior knowledge]
Source material: [files or links; write “none” if unavailable]
Must include: [core scope]
Do not expand: [excluded scope]
Language and format: [for example, English Word]
Purpose and depth: [for example, beginner self-study course]
Other requirements: [cases, exercises, product version, and so on]
A complete course normally produces two independent Word baselines: one Chinese and one English. PDF and PPT/PPTX outputs are confirmed before production. Every language and format must share the same content baseline, terminology, technical conclusions, case parameters, and exercise scope.
无需手动执行命令。将下面整段 Prompt 复制到具备文件和终端操作能力的 AI 编程工具中即可：
请把 https://github.com/550YU/Zlearning 的最新 main 分支安装为当前项目的项目级 AI Skill。先检查当前项目已有的 Skill 目录规范；如果没有明确规范，就安装到 .github/skills/Zlearning。下载并保留完整的 SKILL.md 和 references 目录。若目标位置已有不同内容，先创建带时间戳的备份，禁止直接覆盖。安装后核对文件列表，并验证所有文件可读取、SKILL.md 存在且 SHA-256 校验无复制差异。不要删除用户级原件，也不要把临时下载目录或嵌套的 .git 目录留在项目中。最后报告安装路径、文件数、版本或提交号和验证结果。
Copy the following prompt into an AI coding tool that can access files and run terminal commands:
Install the latest main branch of https://github.com/550YU/Zlearning as a project-level AI Skill in the current project. First detect and follow any existing project Skill-directory convention; if none exists, install it at .github/skills/Zlearning. Preserve the complete SKILL.md file and references directory. If the destination already contains different content, create a timestamped backup before replacing anything. After installation, compare the file list, confirm that every file is readable, verify that SKILL.md exists, and use SHA-256 hashes to ensure the copied files match. Do not remove any user-level installation, and do not leave a temporary download directory or nested .git directory in the project. Finally report the installation path, file count, installed version or commit, and validation result.
The AI tool must have permission to access the project files and GitHub. / AI 工具需要具备项目文件与 GitHub 访问权限。
- “用 Zlearning 给零基础学员解释 RAID 5 和 RAID 10 的区别。”
- “把这些存储资料整合成一套中英文独立 Word 课程。”
- “检查课程是否存在术语前置依赖、机制断链或双语密度不一致。”
- “Use Zlearning to explain RAID 5 and RAID 10 to first-time learners.”
- “Turn these storage references into separate Chinese and English Word courses.”
- “Audit this course for missing prerequisites, broken mechanism chains, and bilingual depth gaps.”
Zlearning/
├── SKILL.md
├── LICENSE
├── README.md
└── references/
├── answer-template.md
├── course-audit-checklist.md
├── course-delivery-checklist.md
├── examples.md
└── quality-checklist.md
answer-template.md— 标准解释结构 / standard explanation structurecourse-audit-checklist.md— 技术、认知与一致性审计 / technical, cognitive, and consistency auditcourse-delivery-checklist.md— Word/PDF/PPT 交付验收 / Word, PDF, and PPT delivery validationexamples.md— 解释模式与机制示例 / explanation-mode and mechanism examplesquality-checklist.md— 回答级质量检查 / response-level quality checks
- 通俗不等于删除必要术语。Accessibility does not mean removing essential terminology.
- 类比建立直觉，但不替代正式定义。Analogies build intuition but never replace formal mechanisms.
- 先讲主干，再讲影响判断的边界。Teach the main path before the boundaries that affect decisions.
- 产品事实应绑定版本、配置、兼容范围和可靠来源。Product claims must be tied to versions, configurations, compatibility scope, and reliable sources.
- 测试结果是有限证据，不是超范围证明。A test result is bounded evidence, not proof beyond its scope.
基于 MIT License 发布。Released under the MIT License.