Zettelkasten Note Connector
Transforms raw notes into atomic knowledge units, identifies connections between ideas, and builds a navigable personal knowledge network with tags and index entries.
Category: learning
Difficulty: intermediate
Platforms: chatgpt claude
Tags: zettelkasten note-taking knowledge-management pkm connections
Prompt Template
You are a knowledge management expert specializing in the Zettelkasten method. Help me process and connect my notes.
## Input
Knowledge domain: {{knowledge_domain}}
Note content to process: {{note_content}}
Existing notes or context (if any): {{existing_notes_context: none}}
## 1. Atomic Note Extraction
Break the note content into atomic ideas — each capturing exactly one concept:
- Give each atomic note a unique descriptive title
- Write each note so it is self-contained and understandable without the source
- Mark the source reference for each note
- Indicate the note type: concept, fact, argument, question, or method
## 2. Connection Suggestions
For each atomic note, identify potential links:
- Direct connections: notes that share the same concept or extend each other
- Lateral connections: notes from different sub-topics that inform each other
- Contradictions: notes that present opposing views worth reconciling
- For each connection, explain WHY these notes relate — don't just list titles
## 3. Tag Recommendations
Suggest a tagging taxonomy:
| Tag | Category | Notes Covered |
|-----|----------|---------------|
| ... | Theme / Method / Source / Status | ... |
- Use hierarchical tags where useful (e.g., #statistics/bayesian)
- Limit to 2-4 tags per atomic note to avoid over-tagging
## 4. Index Entries
Create structured index entries:
- Hub notes: broad topics that serve as entry points
- Outline notes: sequences that tell a story or build an argument
- Map of Content (MOC) entries grouping related notes
- Suggested next questions or gaps in the knowledge network
Tips
- Paste the full text of a recent note or highlight — longer input produces richer atomic notes
- Include summaries of 3-5 existing notes so the AI can find genuine cross-connections
- Revisit the suggested connections weekly and prune ones that feel forced
- Use the index entries as starting points for your own Maps of Content
- Combine with a tool like Obsidian or Logseq that supports bidirectional links