Omni Learning Anything v1.0.1

Omni Learning Anything v1.0.1

技能标识、GitHub仓库、安装包、调用示例和文档统一为 omni-learning-anything。保留已有学习流程与课程内容，不提供别名兼容入口。

Use the complete Skill ZIP below. The identifier and folder are omni-learning-anything; invoke $omni-learning-anything. Existing learning behavior and language support are unchanged.

29 regression tests and Skill metadata validation passed. The existing 24 example PDFs and six bilingual acceptance PDFs passed archive checks. Two synthetic installation PDFs were rebranded and visually reviewed. Local installed resources were renamed; this is not a new full course test on every platform.

Usage · Validation

SHA256: ef64c80cc899bed9dd7024cfb1bc94a2b886a921947df32edf57ddf0bca2cda0

Omni Learning Assistant v1.0.0

Omni Learning Assistant v1.0.0

轻量通用学习 Skill 正式版：从简单主题开始，确认学习目标与时间，先交付 PDF 计划；批准后生成有来源、图示、练习及扩展链接的每日材料。

• 修复重复批准改变暂停/完成状态、时间预算验证与已交付归档核验。
• 29项回归测试通过；真实XeLaTeX失败重试及英文输出检查通过。
• 保留六组主题样例（24份PDF）；新增双语验收样例（6份最终PDF、20页）。
• 按平台记录 Codex、DSH、WorkBuddy、豆包工作界面与OpenClaw的原生安装/发现和资源执行检查。

Bilingual usage · Release validation · Acceptance PDFs

Download the lightweight Skill ZIP below; keep the entire extracted folder. The package contains instructions, scripts and references, without course data, dependencies or fonts. Host research, image and execution tools are required; native scheduling is provided by the host.

Acceptance used simulated learners and receipts, with accelerated consecutive lessons and no live schedules. Codex CLI quota interrupted two stages; the installed Skill was continued in the current Codex desktop session, as disclosed in the validation report. OpenClaw validation covered native installation/discovery and installed helper execution, without a configured model conversation. Study duration is estimated, not measured learner mastery. Host scheduler reliability is outside the Skill test scope.

SHA256: e9064bba37a4bcfa068b33fb639eb3478039baa787c6ccc74cfbd8e3fa632d46

Omni Learning Assistant v0.2.1

Omni Learning Assistant v0.2.1

• 统一模块、锁文件、样例页眉与文档命名，移除旧名兼容说明。
• 安装分为通用对话指令与五个平台的手动操作，依据官方文档及官方客户端界面。
• 版本验收集中在 Skill 流程、PDF 输出和课程状态，移除多天宿主调度测试规划。
• 21 项测试通过；24 份样例、55 页的字体、链接、文字与页面检查通过。来源核验日期保留。

Download omni-learning-assistant-skill-v0.2.1.zip and keep the complete extracted bundle. Use the chat installation prompt or your host's native manual procedure.

Install guide: https://github.com/ShawnRen57/omni-learning-assistant/blob/main/skills/omni-learning-assistant/references/install.md

No new runtime dependencies or scheduler services are added.

Omni Learning Assistant v0.2.0

Omni Learning Assistant v0.2.0

The skill previously called LearnPath is now omni-learning-assistant.

• Install folder and invocation: omni-learning-assistant / $omni-learning-assistant.
• CLI entrypoint: scripts/omni_learning.py.
• Download the ZIP below and copy the complete root folder into your host's skill directory. Preserve scripts and references.
• The repository URL, course format, historical sample PDFs and earlier release assets remain available.
• The legacy .learnpath.lock remains for course compatibility.

本次完成技能改名与安装说明更新，已实测 Codex CLI 与 DSH Desktop 的原生加载及首次 PDF 生成，21 项测试通过。发现并修复了旧版 Python 被误报正常的问题。真实跨日调度不属于本次验证范围。

安装与首次使用报告：https://github.com/ShawnRen57/omni-learning-assistant/blob/main/docs/installation-validation.md

LearnPath v0.1.1 — DeepSeek Harness guide

LearnPath v0.1.1 adds a bilingual DeepSeek Harness installation guide and updates the bundled host instructions to match current upstream's persistent Host Schedule.

• Copy the complete learnpath/ folder into a configured DSH filesystem skill root.
• Daily/weekly/cron support, host availability, pause limitations and inbox-versus-PDF delivery are documented explicitly.
• Skill metadata validation, extracted-package doctor and local guide links checked.
• Runtime code and the 24 sample PDFs are unchanged. Native DSH end-to-end scheduling has not been tested.

下载 learnpath-skill-v0.1.1.zip 安装技能，SHA256SUMS 用于校验。DSH 中英文指南：https://github.com/ShawnRen57/omni-learning-assistant/blob/main/docs/deepseek-harness.md

Unofficial community project; no DeepSeek endorsement is claimed.

LearnPath v0.1.0

LearnPath 0.1.0 turns a learning topic into an approved PDF curriculum and source-grounded daily PDF lessons.

LearnPath 0.1.0 将学习主题转化为需先确认的 PDF 学习计划，以及带来源、教学插图和练习的每日 PDF 材料。

• Install: download learnpath-skill-v0.1.0.zip and import/copy the complete learnpath/ folder into your agent's supported skills directory. Dependencies and platform limits are documented in the bilingual README.
• 安装：下载下方 Skill ZIP，解压后将完整 learnpath/ 文件夹导入 Agent 的技能目录；依赖和平台限制见中英文 README。
• Examples: six complete 30-day plans plus Day01–Day03, 24 PDFs / 55 pages, available in the repository's examples/ directory and source archive.
• 样例：3 个宏观主题和 3 个微观主题，完整 30 天计划及前三天材料，共 24 份 PDF、55 页。
• Validation: 21 tests passed, real XeLaTeX integration, all sample artifact checks and independent review passed.

Examples simulate approval and delivery; no live schedules were created. WorkBuddy/OpenClaw support is documentation-backed, DeepSeek Harness has session-local scheduling limits, and native Doubao support is unverified. No claim of fully tested operation across all clients or multi-day timed delivery.

样例使用模拟批准与交付，不代表真实定时通知实测。请先查阅 README 中的平台兼容性表。