AI Summarization

What is AI Summarization?

Generate intelligent summaries of your notes using GitHub Copilot’s AI capabilities. Analyze single notes, batch process multiple notes, and extract key insights, action items, and keywords from your knowledge base.

Quick Start

Single Note:

Right-click any note → Summarize Note

Batch Operations:

Command Palette → Noted: Summarize This Week
Command Palette → Noted: Summarize This Month
Command Palette → Noted: Summarize Custom Date Range

Requirements

• GitHub Copilot Extension: Must be installed and active
• VS Code: Version 1.80.0 or later
• Active Copilot Subscription: GitHub Copilot or Copilot Pro

Overview

AI Summarization provides three main capabilities:

1. Single Note Summarization: Quick summaries of individual notes
2. Batch Summarization: Analyze multiple notes across time periods
3. Smart Caching: Performance optimization with automatic cache management

All summaries are generated locally using VS Code’s Language Model API, ensuring your notes remain private on your machine.

Single Note Summarization

From Tree View

1. Right-click any note in the notes tree
2. Select “Summarize Note”
3. Wait for generation (cancellable progress shown)
4. Summary opens in new editor tab

From Active Editor

1. Open any note in editor
2. Open Command Palette (Cmd+Shift+P)
3. Type “Noted: Summarize Current Note”
4. Summary generated for active note

Features

Cancellable Progress:

• Click X to cancel during generation
• Immediate feedback on operation status
• Safe to cancel without side effects

Summary Display:

# Summary: 2025-11-07.md

Generated: 11/7/2025, 10:30:00 AM
Source: 1 note

## Summary

[AI-generated summary content here]

---
Generated by GitHub Copilot • Noted Extension

Batch Summarization

Summarize Recent (Last 7 Days)

Command: Noted: Summarize Recent Notes (Last 7 Days)

Analyzes all notes modified in the past week.

Use Cases:

• Weekly review sessions
• Quick catch-up after time away
• Track recent progress

Summarize This Week

Command: Noted: Summarize This Week

Analyzes all notes from the current week (Sunday-Saturday).

Use Cases:

• End-of-week reviews
• Weekly standup preparation
• Project status updates

Summarize This Month

Command: Noted: Summarize This Month

Analyzes all notes from the current calendar month.

Use Cases:

• Monthly retrospectives
• Progress reports
• Long-term pattern analysis

Summarize Custom Range

Command: Noted: Summarize Custom Date Range

Pick specific start and end dates for analysis.

Steps:

1. Run command
2. Enter start date (YYYY-MM-DD format)
3. Enter end date (YYYY-MM-DD format)
4. Wait for analysis

Use Cases:

• Project-specific reviews (e.g., 3-week sprint)
• Event analysis (e.g., conference period)
• Arbitrary time periods

Batch Processing Modes

Small Batches (≤10 notes):

• Single optimized prompt
• Fast processing
• Cohesive narrative summary

Large Batches (>10 notes):

• Individual note processing
• Detailed “Processing X of Y” progress
• Combined summary output
• Cancellable at any point

Example progress:

Analyzing 25 notes from this month...
Processing 12 of 25: 2025-11-15.md

Smart Caching System

How It Works

Automatic Caching:

• Summaries cached after generation
• File modification timestamp tracked
• Cache key includes summary options

Cache Invalidation:

• Automatic when note file changes
• Configuration changes trigger miss
• Manual clearing available

LRU Eviction:

• Maximum 100 cached summaries
• Oldest summaries removed first
• Most-used summaries retained

Benefits

Performance:

• Instant retrieval for cached summaries
• No redundant API calls
• Reduced wait times

Cost Efficiency:

• Fewer Copilot API requests
• Smart reuse of existing summaries
• Automatic optimization

Cache Management

Clear All Cache:

Command: Noted: Clear Summary Cache

Removes all cached summaries with confirmation dialog.

When to Clear:

• After changing summary settings globally
• Testing different summary options
• Troubleshooting issues

Configuration Settings

All settings under noted.ai.* prefix:

Enable/Disable AI

Setting: noted.ai.enabled Type: Boolean Default: true

Master switch for all AI features.

{
  "noted.ai.enabled": true
}

Summary Length

Setting: noted.ai.summaryLength Type: Enum Options: short, medium, long Default: medium

Controls target word count:

• Short: ~50 words - Quick overview
• Medium: ~150 words - Balanced detail
• Long: ~300 words - Comprehensive

{
  "noted.ai.summaryLength": "medium"
}

Summary Format

Setting: noted.ai.summaryFormat Type: Enum Options: paragraph, bullets, structured Default: structured

Output formatting style:

• Paragraph: Flowing prose, narrative style
• Bullets: Bulleted list, scannable
• Structured: Sections with headers, organized

{
  "noted.ai.summaryFormat": "structured"
}

Include Action Items

Setting: noted.ai.includeActionItems Type: Boolean Default: true

Extract and highlight action items, TODOs, and next steps.

{
  "noted.ai.includeActionItems": true
}

Include Keywords

Setting: noted.ai.includeKeywords Type: Boolean Default: false

Generate 3-5 relevant keywords/tags from content.

{
  "noted.ai.includeKeywords": false
}

Cache Enabled

Setting: noted.ai.cacheEnabled Type: Boolean Default: true

Enable/disable summary caching system.

{
  "noted.ai.cacheEnabled": true
}

Understanding Summaries

What AI Looks For

Key Topics:

• Main subjects discussed
• Core themes and concepts
• Technical details

Decisions:

• Conclusions reached
• Choices made
• Direction set

Action Items:

• Tasks to complete
• Follow-up needed
• Next steps identified

Outcomes:

• Results achieved
• Problems solved
• Progress made

Multi-Note Summaries

Additional Analysis:

• Common themes across notes
• Patterns and trends
• Chronological narrative
• Open questions
• Unresolved issues

Example Output:

# Summary: This Month (November 2025)

Generated: 11/7/2025, 10:30:00 AM
Source: 15 notes

## Source Notes
  - 2025-11-01.md
  - 2025-11-02.md
  [... more notes ...]

## Summary

**Common Themes:**
The month focused heavily on API development, with particular
emphasis on authentication and rate limiting. Multiple meetings
discussed performance optimization strategies.

**Key Accomplishments:**
- Completed OAuth2 integration
- Implemented Redis caching layer
- Deployed to staging environment

**Action Items:**
- [ ] Review security audit findings
- [ ] Update API documentation
- [ ] Schedule load testing session

**Keywords:** API, authentication, performance, caching, deployment

---
Generated by GitHub Copilot • Noted Extension

Use Cases

Daily Review

Scenario: Quick catch-up on recent work

Approach:

1. Summarize Recent (last 7 days)
2. Read summary in 1-2 minutes
3. Identify follow-ups

Benefit: Stay current without re-reading all notes

Weekly Retrospective

Scenario: End-of-week review

Approach:

1. Summarize This Week
2. Extract action items
3. Identify patterns
4. Plan next week

Benefit: Structured reflection and planning

Project Documentation

Scenario: Documenting completed project

Approach:

1. Summarize Custom Range (project dates)
2. Extract key decisions
3. Document outcomes
4. Share with team

Benefit: Automatic project history

Knowledge Transfer

Scenario: Onboarding new team member

Approach:

1. Summarize relevant time period
2. Extract technical decisions
3. Highlight key concepts
4. Provide context

Benefit: Quick ramp-up documentation

Meeting Prep

Scenario: Preparing for status meeting

Approach:

1. Summarize This Month
2. Extract accomplishments
3. Identify blockers
4. Generate talking points

Benefit: Data-driven status updates

Privacy & Security

Data Handling

Local Processing:

• Notes read from local file system
• Sent to GitHub Copilot API only
• No third-party services
• No data storage outside VS Code

API Usage:

• VS Code Language Model API
• GitHub Copilot backend
• Same security as Copilot chat
• Covered by GitHub privacy policy

Cache Storage:

• Stored in VS Code global state
• Local to your machine
• Not synced or shared
• Cleared on extension uninstall

Best Practices

Sensitive Information:

• Review summaries before sharing
• Don’t include credentials in notes
• Be mindful of confidential data
• Use judgment when summarizing

Opt-Out:

{
  "noted.ai.enabled": false
}

Troubleshooting

“GitHub Copilot is not available”

Cause: Copilot extension missing or inactive

Solutions:

1. Install GitHub Copilot extension
2. Sign in to GitHub account
3. Verify active Copilot subscription
4. Restart VS Code

Slow Summary Generation

Cause: Large notes or slow API response

Solutions:

• Wait patiently (normal for large notes)
• Check internet connection
• Try again during off-peak hours
• Cancel and retry if stalled

“AI summarization is disabled”

Cause: Feature disabled in settings

Solution:

{
  "noted.ai.enabled": true
}

Unexpected Summary Content

Cause: AI interpretation varies

Solutions:

• Try different format (paragraph vs. structured)
• Adjust length setting
• Regenerate (results may vary)
• Clear cache and retry

Cache Not Working

Cause: Cache disabled or cleared

Solutions:

1. Verify cache enabled:

   {
     "noted.ai.cacheEnabled": true
   }
   

2. Check cache stats in output
3. Summaries cache after first generation

Tips & Best Practices

Writing Notes for Better Summaries

Clear Structure:

• Use headers for sections
• Organize thoughts logically
• Separate topics clearly

Explicit Action Items:

• Use TODO, ACTION, or checkboxes
• Be specific about next steps
• Include context

Descriptive Content:

• Avoid abbreviations
• Write complete thoughts
• Include reasoning

Optimizing Performance

Cache Utilization:

• Don’t change settings frequently
• Let cache build over time
• Clear only when needed

Batch Operations:

• Prefer specific time ranges
• Use week/month for regular reviews
• Custom ranges for projects

Progress Management:

• Cancel if wrong operation selected
• Wait for progress indicators
• Don’t interrupt final processing

Summary Management

Save Important Summaries:

• Copy to permanent notes
• Use as meeting agendas
• Archive for later reference

Iterate on Settings:

• Try different lengths
• Test format options
• Adjust to preferences

Combine with Other Features:

• Use with Tags for filtering
• Complement Search results
• Include in Exports

Related Features

• Search: Find notes to summarize
• Tags: Organize summaries by topic
• Smart Collections: Auto-updating note groups
• Calendar View: Visual time-based navigation

Phase 3 Features (v1.38.0)

Search Results Summarization: After performing a search (Cmd+Shift+F), you’ll see a prompt:

Found 15 note(s). Would you like to summarize these results?
[Summarize Results] [Dismiss]

Click “Summarize Results” to generate an AI summary of all matching notes. The summary includes:

• Search query context in the title
• List of all source notes
• Synthesized analysis across all results

Export with AI Summaries: When exporting notes (Cmd+Shift+E), you’ll be prompted:

Include AI summaries in export?
○ No - Export notes without summaries
○ Yes - Include AI-generated summaries for each note

If you select “Yes,” each note in the export will include:

==================================================
2025-11-07.txt
==================================================

## AI Summary

[AI-generated summary here]

## Note Content

[Original note content]

Hover Summary Previews: Hover over any wiki-style link ([[note-name]]) to see a tooltip with:

• ✨ AI Summary header with note name
• First 100 words of cached summary (if available)
• Clickable links: “View full summary” or “Generate summary”
• Only shows cached summaries (no slow generation on hover)

Enhanced Action Item Extraction: Action items are now formatted with rich detail:

## Action Items

- [ ] Review authentication PR (John)
- [ ] Deploy API updates by Friday
- [ ] [URGENT] Fix production bug in payment service
- [ ] Update documentation for new endpoints

Features:

• Checkbox format for task tracking
• Assignees included when mentioned
• Deadlines preserved
• Priority markers ([URGENT], [HIGH], etc.)
• Grouped by category in multi-note summaries

Enhanced Keyword Generation: Keywords are now more intelligent and useful:

## Keywords

#authentication #api-security #oauth2 #backend/auth #production-fix

Features:

• Hashtag format for easy searching
• Hierarchical tags: #project/component
• Focuses on specific, searchable terms
• Technical terms and project names prioritized
• Overarching themes for multi-note summaries

Phase 4 Features (v1.39.0)

Custom Prompt Templates: Create and use your own AI summarization prompts for specific use cases.

Built-in templates include:

• Default - Balanced summary with key points and action items
• Technical Deep Dive - Focus on architecture, implementation details, technical decisions
• Meeting Summary - Emphasize decisions, action items, attendees, outcomes
• Code Review - Highlight code changes, issues found, improvement suggestions
• Brainstorm Session - Capture ideas, possibilities, creative solutions

Commands:

• Noted: Create Prompt Template - Create custom template with variables
• Noted: Edit Prompt Template - Modify existing custom templates
• Noted: Delete Prompt Template - Remove custom templates
• Noted: Duplicate Prompt Template - Copy template as starting point
• Noted: List Prompt Templates - View all available templates
• Noted: View Template Variables Reference - See all available variables

Template picker appears before each summarization, remembering your last selection.

Summary History/Versions: Track every summary version with full metadata and comparison tools.

Each version stores:

• Summary text and generation timestamp
• Options used (length, format, action items, keywords)
• AI model name and prompt template
• Note’s last modified time

Commands:

• Noted: Show Summary History - View all versions for current note
• Noted: Compare Summary Versions - Side-by-side diff of two versions
• Noted: Restore Summary Version - Open old version for review
• Noted: Clear Summary History - Clear history for note or all notes
• Noted: Show Summary History Stats - View storage statistics

Features:

• Automatic version saving (up to 10 per note with LRU eviction)
• Stored in .noted-cache/summaries/ as JSON
• Quick pick with “Version 2 of 5 • 3 days ago • medium/structured” format
• VS Code diff editor for visual comparison

Auto-Tagging from Summaries: Automatically generate and apply tags based on AI summary keyword extraction.

How it works:

1. Generate summary with includeKeywords: true
2. Extract hashtags from keywords section
3. Validate and filter generic words
4. Show preview with current vs. suggested tags
5. Apply with user confirmation

Commands:

• Noted: Auto-Tag Note from Summary - Tag single note (also in context menu)
• Noted: Auto-Tag Current Note - Tag currently open note
• Noted: Batch Auto-Tag Notes - Tag multiple notes with time period picker
• Noted: Suggest Tags for Current Note - Preview only, no modifications

Batch options:

• This Week - Notes from current week
• This Month - Notes from current month
• Last 7 Days - Recent notes
• All Untagged - Notes with fewer than 3 tags

Features:

• Hierarchical tags: #project/frontend
• Smart filtering of generic words
• Tag modes: Append (add to existing), Replace (replace all), Suggest (preview)
• Frontmatter creation/update: tags: [tag1, tag2, tag3]
• Progress tracking with cancellation
• Context menu integration

Future Enhancements

Beyond Phase 4:

• Semantic search using embeddings
• Note version comparison via git integration
• Custom model selection (beyond GitHub Copilot)
• Summary templates marketplace
• AI-powered note linking suggestions

See AI_SUMMARIZATION_PLAN.md for roadmap details.