원클릭으로
"深度反思学习过程,巩固知识,发现盲点,优化学习策略。基于笔记内容生成反思报告、学习建议、知识图谱。"
npx skills add https://github.com/Lcollection/AcademiClaw --skill learning-reflector이 명령을 Claude Code에 복사하여 붙여넣어 스킬을 설치하세요
"深度反思学习过程,巩固知识,发现盲点,优化学习策略。基于笔记内容生成反思报告、学习建议、知识图谱。"
npx skills add https://github.com/Lcollection/AcademiClaw --skill learning-reflector이 명령을 Claude Code에 복사하여 붙여넣어 스킬을 설치하세요
| name | Learning Reflector |
| description | "深度反思学习过程,巩固知识,发现盲点,优化学习策略。基于笔记内容生成反思报告、学习建议、知识图谱。" |
Use this skill when you want to:
Triggers:
反思今天的学习内容
生成本周学习反思报告
分析我的学习盲点并给出建议
python {workspace}/PaperVault/scripts/learning-reflect.py
Analyzes your notes to generate insights:
Helps you remember and understand:
Identifies what you might have missed:
Personalized suggestions:
---
type: reflection
date: YYYY-MM-DD
period: daily|weekly|monthly
---
# 🧠 学习反思报告 - YYYY-MM-DD
## 学习概览
- **论文数量**: X 篇
- **核心概念**: X 个
- **理解深度**: ⭐⭐⭐☆☆ (3/5)
## 核心收获
### 1. 概念理解
- **Transformer 架构**:
- 理解了自注意力机制
- 掌握了位置编码原理
- 需要深入: 多头注意力的细节
### 2. 方法学习
- **RAG 优化**:
- 学习了检索增强策略
- 了解了重排序方法
- 待探索: 混合检索方案
## 知识连接
```mermaid
graph TD
A[Transformer] --> B[Attention]
B --> C[Self-Attention]
B --> D[Multi-Head]
A --> E[Position Encoding]
F[RAG] --> G[Retrieval]
F --> H[Generation]
C --> H
| 维度 | 评分 | 说明 |
|---|---|---|
| 理解深度 | ⭐⭐⭐☆☆ | 基本概念清晰,细节需加强 |
| 知识连接 | ⭐⭐⭐⭐☆ | 能发现概念间联系 |
| 实践能力 | ⭐⭐☆☆☆ | 理论多,实践少 |
| 持续性 | ⭐⭐⭐⭐☆ | 坚持每日学习 |
今天学习了 Transformer 和 RAG 的基础知识。虽然理解了主要概念,但感觉深度不够,特别是数学推导部分。明天需要花更多时间在公式理解上,而不是只看文字描述。
实践方面明显不足,需要找时间动手实现。计划本周完成一个简单的 RAG demo,这样能更好地理解理论。
整体来说,学习方向正确,但需要调整理论和实践的比例。
## Reflection Types
### 1. Quick Reflection (5 minutes)
**When**: After reading each paper
**Focus**:
- Main takeaway (1 sentence)
- Questions (1-2)
- Next steps (1-2)
### 2. Daily Reflection (15 minutes)
**When**: End of each day
**Focus**:
- What I learned today
- What confused me
- What to review tomorrow
### 3. Weekly Reflection (30 minutes)
**When**: End of each week
**Focus**:
- Weekly themes
- Knowledge connections
- Progress towards goals
### 4. Monthly Reflection (1 hour)
**When**: End of each month
**Focus**:
- Monthly achievements
- Knowledge map evolution
- Strategy adjustment
## Prompt Templates
### Concept Understanding Check
基于我的笔记,请帮我反思对 "{concept}" 的理解:
我理解的层次:
### Knowledge Connection Discovery
请分析以下概念之间的联系: {concept1}, {concept2}, {concept3}
### Learning Strategy Optimization
基于我过去 {period} 的学习记录:
时间分配分析:
效率评估:
## Advanced Features
### Spaced Repetition
Automatically suggests concepts to review based on:
- Ebbinghaus forgetting curve
- Your understanding depth
- Concept importance
### Knowledge Graph Generation
Creates visual maps of:
- Concept relationships
- Learning paths
- Knowledge gaps
### Comparative Analysis
Compares your learning with:
- Previous periods
- Recommended paths
- Expert roadmaps
## Integration with Other Skills
- **paper-fetcher**: Source of papers to reflect on
- **paper-summarizer**: Summaries to analyze
- **pdf-reader**: Deep paper understanding
- **self-improving**: Long-term memory
## Troubleshooting
### Not Enough Data
Error: Insufficient notes for reflection
**Solution**: Use paper-fetcher to get more papers first.
### Superficial Reflection
Warning: Reflection too shallow
**Solution**:
1. Add more detailed notes
2. Use pdf-reader for deeper analysis
3. Increase reflection time
## Best Practices
1. **Reflect Daily**: 15 minutes at end of day
2. **Be Honest**: Admit what you don't understand
3. **Take Action**: Act on the suggestions
4. **Track Progress**: Review reflections over time
5. **Stay Curious**: Follow the questions that arise
## Related Skills
- `paper-fetcher` - Paper retrieval
- `paper-summarizer` - Progress tracking
- `pdf-reader` - Deep understanding
- `self-improving` - Long-term memory
"自动化论文检索、翻译、导入和学习闭环系统。每日从 arXiv 检索论文、自动翻译标题和摘要、导入 Zotero、生成 Obsidian 笔记和学习报告。"
"生成各种类型的图表和流程图:流程图、时序图、类图、甘特图、思维导图等。支持 Mermaid、PlantUML、Graphviz 等格式。"
"GitHub CLI (gh) 自动化操作。管理仓库、Issues、Pull Requests、Workflows、Releases 等。支持所有 gh 命令。"
"从 arXiv 自动检索论文、翻译标题摘要、导入 Zotero、生成 Obsidian 笔记。支持关键词检索、自动去重、中文翻译。"
"生成论文学习报告:每日总结、每周总结、月度报告、季度报告、年度总结。统计学习进度、分析研究方向、追踪知识积累。"
"深度阅读和分析 PDF 论文。支持 PDF 转 Markdown、智能摘要、关键信息提取、问答式学习、笔记生成。"