Chubby Skills is a collection of 14 agent skills and command-line tools that collects videos, podcasts, articles, and local documents into a searchable personal knowledge base of Markdown files with source information. Content creators and people maintaining personal research libraries use it to save material, find supporting evidence, and export source-based research packs. Its catalogue add-ons are the project's own skills, which let coding agents collect, search, read, and package knowledge-base content, with an optional MCP server for agent access.
Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/chubbyguan/chubbyskillsnpx agentmods add skills/chubbyguan/chubbyskills/learning-notes-automationWrote 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/chubbyguan/chubbyskills/learning-notes-automation)<a href="https://agentmods.dev/skills/chubbyguan/chubbyskills/learning-notes-automation"><img src="https://agentmods.dev/badge/skills/chubbyguan/chubbyskills/learning-notes-automation/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/chubbyguan/chubbyskills/learning-notes-automation"><img src="https://agentmods.dev/badge/skills/chubbyguan/chubbyskills/learning-notes-automation.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00058 | $0.01842 |
| Opus 5.5 | $0.00023 | $0.00737 |
| Sonnet 5.5 | $0.00012 | $0.00368 |
| Haiku 4.5 | $0.00006 | $0.00184 |
Grade A, and why
learning-notes-automation 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 1mo 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 — 255 lines — stays where its author put it; the contents beside it link to each section on GitHub.
学习笔记自动化
核心价值
被动学习 → 主动记忆：看视频 ≠ 学会，生成闪卡 = 记住
输入源
| 输入类型 | 工具 | 输出格式 |
|---|---|---|
| YouTube 视频 | youtube-transcribe |
转录文本 |
| B站视频 | bilibili-transcribe |
转录文本 |
| 播客 | podcast-transcribe |
转录文本 |
| 抖音 | douyin-transcribe |
转录文本 |
| 公众号文章 | wechat-article-ingest |
Markdown |
工作流程
Phase 1: 内容转录
复用本仓库的转录 skill（相对路径，按需替换为你的安装位置）：
# YouTube（自动翻译）
python3 ../youtube-transcribe/scripts/transcribe.py "https://youtube.com/watch?v=xxx" -o ./
# B站
python3 ../bilibili-transcribe/scripts/transcribe.py "https://bilibili.com/video/xxx" ./
# 播客
python3 ../podcast-transcribe/scripts/transcribe.py "https://xiaoyuzhoufm.com/episode/xxx" ./
输出是一份转录 Markdown，作为 Phase 2 的输入。
Phase 2 + 3: 知识点提取 + 闪卡生成（一步到位）
把转录稿交给 make_notes.py，自动提取知识点并生成 Anki 兼容闪卡：
export DEEPSEEK_API_KEY=***
python3 scripts/make_notes.py 转录稿.md --output ./notes --max-cards 20
# 产出：<标题>-学习笔记.md（核心要点+闪卡+关联知识） 和 <标题>-闪卡.csv（直接导入 Anki）
脚本用 DeepSeek 输出结构化 JSON，再渲染成下面约定的笔记与闪卡格式。 下面是各维度的设计说明，供理解和手动微调时参考。
提取维度：
-
核心概念（必须掌握）
- 定义
- 原理
- 应用场景
-
关键事实（需要记忆）
- 数据
- 时间线
- 人物/公司
-
方法论（可以复用）
- 步骤
- 框架
- 最佳实践
-
金句/洞察（值得引用）
- 精辟总结
- 独特观点
Phase 3: 闪卡生成
闪卡格式（Anki 兼容）：
## 闪卡类型
### 1. 概念卡（Cloze Deletion）
Q: {{c1::Transformer}} 架构的核心机制是 {{c2::自注意力（Self-Attention）}}
A: Transformer, 自注意力（Self-Attention）
### 2. 问答卡（Basic）
Q: 什么是 RAG？
A: Retrieval-Augmented Generation，检索增强生成。通过检索外部知识库来增强 LLM 的回答能力，解决幻觉问题。
### 3. 对比卡（Comparison）
Q: Fine-tuning vs RAG 的区别？
A:
| 维度 | Fine-tuning | RAG |
|------|-------------|-----|
| 成本 | 高（需要训练） | 低（只检索） |
| 更新 | 需要重新训练 | 实时更新 |
| 适用 | 特定任务 | 知识问答 |
### 4. 步骤卡（Process）
Q: 如何构建一个 RAG 系统？
A:
1. 文档分块（Chunking）
2. 向量化（Embedding）
3. 存入向量数据库
4. 检索相关片段
5. 拼接 Prompt
6. LLM 生成回答
Phase 4: 知识图谱更新
实体提取：
- 人物（Who）
- 概念（What）
- 工具/产品（Tool）
- 方法论（How）
- 时间（When）
关系映射：
发明了：人物 → 概念/工具属于：概念 → 领域替代了：新工具 → 旧工具依赖于：概念 → 概念
What ships with it
1 file 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.
- 1mo ago Changed · -5 lines 2993cef05af6
- 1mo ago First seen · 260 lines · 58 tokens per session scan A 6a4ac496c461
learning-notes-automation is a skill published in the GitHub repository chubbyguan/chubbyskills (1,276 stars, last pushed 9d ago), licensed MIT. It adds 58 tokens to every session and 1,842 once invoked, about $0.0002 per session on Opus 5.5. 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
transformer-attention
Use when reasoning about Transformer self-attention, multi-head attention, positional encoding, masked decoder attention, or why attention replaced recurrence/convolutions in sequence models; not for generic NLP or unrelated attention topics.
ai-teacher
A guide to teaching about artificial intelligence, including how large language models work, prompt writing, AI agents, tools, and AI ethics.
learning-visualization-skill
Generate single-file HTML visual explanations for learning and review. Use this skill when the user wants concept maps, process diagrams, principle demos, comparison diagrams, timelines, AI/ML model visualizations, or animated teaching pages that make a topic easier to understand,复习, or present.
maieutic-deepseek-adapter
An adaptation of the Maieutic prompt package for DeepSeek and other commonly available Chinese AI platforms. Maieutic uses Socratic questions to help people learn, explore, reflect, and create.
rwkv-architecture
RNN+Transformer hybrid with O(n) inference. Linear time, infinite context, no KV cache. Train like GPT (parallel), infer like RNN (sequential). Linux Foundation AI project. Production at Windows, Office, NeMo. RWKV-7 (March 2025). Models up to 14B parameters.
llm-wiki
Explain and apply the foundational LLM Wiki architecture and knowledge-management strategy. Use for setup, architecture, schema, or organization decisions; operational ingest/query/lint work uses dedicated skills.