
obsidian-notes
by kaishin
These are my dotfiles.
SKILL.md
name: obsidian-notes description: This skill should be used when the user asks to "create a note", "add a note to my vault", "write a note about", "save this to Obsidian", "make a note", "document this in Obsidian", or mentions creating notes in their Obsidian vault. Handles note creation with proper naming, folder organization, and wiki-linking.
Obsidian Notes Skill
Create and organize notes in the user's Obsidian vault with proper naming conventions, folder placement, and wiki-links to related notes.
Vault Location
The vault path is available via the $VAULT environment variable.
Note Creation Workflow
1. Determine Note Title
Convert the topic to Title Case with spaces:
- "knowledge graphs" → "Knowledge Graphs.md"
- "the DSR memory model" → "The DSR Memory Model.md"
- "swift 6 concurrency" → "Swift 6 Concurrency.md"
2. Select Target Folder
Match the topic to one of these folders based on content:
| Folder | Subfolders | Topics |
|---|---|---|
Development/ | Swift/, Android/, Web/, Tools/ | Programming, frameworks, SDKs |
AI-ML/ | Concepts/, Techniques/, Platforms/, Philosophy/ | Machine learning, AI, neural networks |
Finance/ | Trading/, Investing/, Markets/ | Trading, investing, markets |
Knowledge/ | — | Knowledge management, learning, algorithms |
Languages/ | — | Natural languages (Chinese, Swedish, etc.) |
Life/ | — | Health, hobbies, personal development |
Reference/ | — | Technical concepts, databases, business |
Selection logic:
- Swift/SwiftUI/visionOS →
Development/Swift/ - Android/Kotlin/Compose →
Development/Android/ - Web frameworks →
Development/Web/ - ML concepts (transformers, neural nets) →
AI-ML/Concepts/ - ML techniques (RAG, embeddings) →
AI-ML/Techniques/ - AI tools/platforms →
AI-ML/Platforms/ - AI ethics/philosophy →
AI-ML/Philosophy/ - Trading strategies →
Finance/Trading/ - Investment topics →
Finance/Investing/ - Market-specific →
Finance/Markets/ - If unclear, use
Reference/
3. Write Note Content
Use simple markdown format:
# Note Title (optional, filename is usually sufficient)
Brief introduction or definition.
## Section Heading
Content with proper formatting:
- Bullet points for lists
- `code` for inline code
- Code blocks with language tags
- **Bold** for emphasis
- [[Wiki Links]] for related concepts
## Glossary (if applicable)
**Term** - Definition of the term.
4. Find and Add Wiki-Links
Search for related existing notes and add wiki-links:
# Search for related notes by keyword
grep -ril "keyword" "$VAULT" --include="*.md" | head -10
Add links using Obsidian wiki-link syntax: [[Note Name]]
Example: A note about "Vector Databases" might link to:
[[Embeddings]][[Semantic Search]][[RAG]]
5. Create the Note
# Write the note to the appropriate folder
cat > "$VAULT/Folder/Subfolder/Note Title.md" << 'EOF'
Note content here...
EOF
Example: Creating a Note
User request: "Create a note about attention mechanisms in transformers"
Analysis:
- Topic: Attention mechanisms (AI/ML concept)
- Title: "Attention Mechanisms.md"
- Folder:
AI-ML/Concepts/ - Related notes: Search for "transformers", "neural", "deep learning"
Execution:
# Check for related notes
grep -ril "transformer\|attention\|neural" "$VAULT" --include="*.md"
# Create the note
cat > "$VAULT/AI-ML/Concepts/Attention Mechanisms.md" << 'EOF'
The attention mechanism allows models to focus on relevant parts of the input sequence when producing output, rather than relying on a fixed-length context vector.
## Key Concepts
**Self-Attention** - Mechanism where each position in a sequence attends to all positions in the same sequence.
**Multi-Head Attention** - Running multiple attention operations in parallel, allowing the model to attend to information from different representation subspaces.
## How It Works
1. Create Query (Q), Key (K), and Value (V) vectors from input
2. Compute attention scores: $\text{Attention}(Q,K,V) = \text{softmax}\left(\frac{QK^T}{\sqrt{d_k}}\right)V$
3. Apply softmax to get attention weights
4. Multiply weights by values to get output
## Related
- [[Transformers]]
- [[Neural Networks]]
- [[Deep Learning Notions]]
EOF
Quick Reference
Naming Rules
- Use Title Case: "Word Word Word.md"
- Include articles: "The DSR Memory Model.md"
- Use spaces, not hyphens or underscores
Folder Decision Tree
Is it about programming? → Development/
Swift/iOS/visionOS? → Development/Swift/
Android/Kotlin? → Development/Android/
Web tech? → Development/Web/
Other tools? → Development/Tools/
Is it about AI/ML? → AI-ML/
Core concepts? → AI-ML/Concepts/
Techniques/methods? → AI-ML/Techniques/
Tools/platforms? → AI-ML/Platforms/
Ethics/philosophy? → AI-ML/Philosophy/
Is it about money? → Finance/
Trading? → Finance/Trading/
Investing? → Finance/Investing/
Markets? → Finance/Markets/
Is it about learning/knowledge? → Knowledge/
Is it about languages? → Languages/
Is it about life/health/hobbies? → Life/
Otherwise → Reference/
Wiki-Link Tips
- Link to conceptually related notes
- Use exact note names (without .md extension)
- Check that linked notes exist with grep/ls
- Add 2-5 relevant links per note
Additional Resources
For detailed vault organization, see:
references/vault-structure.md- Complete folder structure and categorization rules
Scripts
-
scripts/find-related-notes.sh- Find related notes by keywords# Usage: find-related-notes.sh "keyword1" "keyword2" ... ./scripts/find-related-notes.sh "transformer" "attention"
Score
Total Score
Based on repository quality metrics
SKILL.mdファイルが含まれている
ライセンスが設定されている
100文字以上の説明がある
GitHub Stars 100以上
3ヶ月以内に更新がある
10回以上フォークされている
オープンIssueが50未満
プログラミング言語が設定されている
1つ以上のタグが設定されている
Reviews
Reviews coming soon