Document Summarization Guide
The Document Summarization feature uses AI to generate concise, one-sentence summaries of your notes, automatically storing them in the note's frontmatter for easy access and organization.
Table of Contents
- Overview
- How It Works
- Using Summarization
- Configuration
- Use Cases
- Tips for Better Summaries
- Working with Summaries
- Troubleshooting
Overview
What Summarization Does
- Analyzes your entire note content
- Generates a concise, one-sentence summary
- Automatically adds the summary to your note's frontmatter
- Uses a dedicated AI model optimized for summarization
Benefits
- Quick Overview: See what a note contains at a glance
- Better Organization: Use summaries in dataview queries
- Improved Search: Find notes by their summary content
- Memory Aid: Quickly recall note contents without opening
- Sharing: Easily describe notes to others
How It Works
The Summarization Process
- Content Analysis: The AI reads your entire note
- Key Point Extraction: Identifies main themes and ideas
- Synthesis: Combines key points into one sentence
- Frontmatter Update: Adds summary to note metadata
AI Model
- Uses a fast, efficient model (typically Gemini Flash)
- Optimized for quick, accurate summarization
- Separate from chat model for better performance
Using Summarization
Generate a Summary
- Open the note you want to summarize
- Open command palette (
Ctrl/Cmd + P) - Search for "Gemini Scribe: Summarize Active File"
- Press Enter
What Happens
- A loading notice appears
- The AI processes your note (usually 1-3 seconds)
- Summary is added to frontmatter
- Success notice confirms completion
Example
Before:
# Meeting Notes - Project Alpha
Discussed timeline changes, budget concerns, and new feature requests...
[long content]After:
---
summary: 'Project Alpha meeting covered timeline adjustments due to budget constraints and approved three new feature requests for Q2 development.'
---
# Meeting Notes - Project Alpha
Discussed timeline changes, budget concerns, and new feature requests...
[long content]Configuration
Settings
In Settings → Gemini Scribe:
Summary Model: Choose the AI model for summarization
- Gemini 1.5 Flash (recommended for speed)
- Gemini 1.5 Flash-8B (fastest, good quality)
- Gemini 1.5 Pro (highest quality, slower)
Summary Frontmatter Key: Customize the metadata field
- Default:
summary - Change to:
description,abstract,brief, etc.
- Default:
Model Selection Guide
| Model | Speed | Quality | Best For |
|---|---|---|---|
| Flash-8B | Fastest | Good | Daily notes, quick captures |
| Flash | Fast | Better | Most use cases |
| Pro | Slower | Best | Important documents, publications |
Use Cases
1. Daily Notes
Summarize daily notes for monthly reviews:
---
summary: 'Completed API integration, attended planning meeting, and started documentation for new features.'
---2. Meeting Notes
Quick overview of meeting outcomes:
---
summary: 'Team agreed to extend deadline by two weeks and allocate additional resources to testing phase.'
---3. Research Notes
Capture essence of research findings:
---
summary: 'Study demonstrates 40% improvement in performance using new caching strategy with minimal memory overhead.'
---4. Book Notes
Summarize key takeaways:
---
summary: 'Explores how deliberate practice and focused attention lead to expertise, emphasizing quality over quantity in skill development.'
---5. Project Documentation
Create quick project descriptions:
---
summary: 'Customer feedback system using React frontend and Node.js backend with real-time updates via WebSocket.'
---Tips for Better Summaries
1. Well-Structured Notes
The AI summarizes better when notes are organized:
- Use clear headings
- Include introduction paragraphs
- Group related content
- Use lists for key points
2. Sufficient Content
- Very short notes may get generic summaries
- Include at least 3-4 paragraphs for best results
- Add context about purpose or goals
3. Clear Writing
- Use specific language
- Avoid excessive jargon
- Include concrete examples
- State conclusions explicitly
4. Update Summaries
Regenerate summaries when:
- Note content changes significantly
- You refine the note structure
- Original summary seems inaccurate
Working with Summaries
1. Dataview Queries
List notes with summaries:
TABLE summary
FROM "Projects"
WHERE summary
SORT file.mtime DESC2. Search and Filter
Find notes by summary content:
- Search:
summary:"budget" - Filter in graph view
- Use in Smart Folders
3. Note Templates
Include summary field in templates:
---
created: { { date } }
summary:
tags: []
---4. Index Pages
Create automatic indexes:
LIST summary
FROM "Meetings"
WHERE date(file.name) >= date(today) - dur(7 days)5. Export and Share
Summaries make sharing easier:
- Include in exported PDFs
- Use in email descriptions
- Add to project overviews
Advanced Techniques
1. Batch Summarization
Summarize multiple notes:
- Use Templater or QuickAdd
- Create macro to run command
- Apply to selected notes
2. Custom Summary Styles
Influence summary style by note structure:
- Action-focused: Start with verbs
- Descriptive: Use adjectives
- Technical: Include specific terms
3. Summary Templates
Guide AI with structure:
# Purpose
[What this note achieves]
# Key Points
[Main ideas]
# Conclusion
[Final thoughts]4. Multi-Language Support
The AI can summarize in various languages:
- Writes summary in note's primary language
- Handles mixed-language content
- Maintains technical terms appropriately
Troubleshooting
Summary Not Appearing
Check Frontmatter
- Ensure note allows frontmatter
- Look for YAML syntax errors
- Verify frontmatter position (must be at start)
Check Settings
- Confirm API key is valid
- Verify summary model is selected
- Check frontmatter key setting
Poor Quality Summaries
Too Generic
- Add more specific content
- Include concrete examples
- State main purpose clearly
Too Long
- This is rare but can happen with complex notes
- Consider breaking into multiple notes
- Focus on key message
Missing Key Points
- Reorganize with clear headings
- Put important info early
- Use emphasis for key concepts
Performance Issues
Slow Generation
- Switch to faster model (Flash-8B)
- Check internet connection
- Reduce note length if extreme
Failures
- Very long notes may timeout
- Check for special characters
- Ensure proper markdown syntax
Best Practices
1. Regular Summarization
- Summarize after major edits
- Include in note creation workflow
- Batch summarize weekly
2. Summary Review
- Read generated summaries
- Edit if needed (they're just frontmatter)
- Use as quality check for note clarity
3. Consistent Usage
- Decide on summary style for note types
- Use same frontmatter key throughout
- Document your summarization practices
4. Integration
Combine with other features:
- Use summaries in custom prompts
- Reference in chat conversations
- Include in rewrite operations
Examples of Good Summaries
Technical Note
"Implemented Redis caching layer reducing API response time by 60% through strategic key expiration and lazy loading patterns."
Meeting Note
"Q3 planning meeting established three priority initiatives: mobile app launch, API v2 development, and customer dashboard redesign."
Research Note
"Analysis of 50 user interviews reveals primary pain points in onboarding flow, with 80% citing confusion around initial configuration steps."
Personal Note
"Reflections on productivity experiment show morning writing sessions yield 3x output compared to evening work, suggesting schedule adjustment needed."
Tutorial Note
"Step-by-step guide for configuring GitHub Actions CI/CD pipeline with automated testing, security scanning, and deployment to AWS."
Conclusion
Document summarization is a powerful feature that enhances your note-taking workflow. By automatically generating concise summaries, you can:
- Navigate large vaults more easily
- Find information faster
- Share knowledge more effectively
- Build better organizational systems
Start with your most important notes and expand from there. The more you use summarization, the more valuable your note metadata becomes.