Skip to content

The definitive guide to Claude's memory_user_edits - an officially documented but practically undiscovered feature. Based on systematic research and controlled testing with the 5Levels platform. Results: 62% error reduction, up to 100% accuracy improvement. Includes testing methodology and production best practices.

License

Notifications You must be signed in to change notification settings

frmoretto/claude-memory-user-edits-guide

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

The Complete Guide to Claude's memory_user_edits

The undocumented power tool for reliable AI memory

Systematic testing proves: 62% fewer errors • Up to 100% accuracy improvement

License: CC BY 4.0 Research PRs Welcome


📖 Read the Full Guide Online

→ Visit the GitHub Pages site for the best reading experience

This repository contains the source code. For optimal navigation, search, and formatting, please use the GitHub Pages version above.

Alternative Formats:


🎯 The Core Insight

Memory edits work for FACTS, not BEHAVIORS.

✅ Works ❌ Doesn't Work
"Backend stores 7KB per user" "Always check docs first"
"Architecture: two-layer model" "Prioritize architecture over details"

This one insight separates failed experiments from 62% error reduction.


📊 Proven Results

From controlled testing with 5Levels (48 documentation files, 35 database tables, 11 modules):

Metric Improvement
Architectural errors ↓ 62% reduction
First-try accuracy ↑ 60% → 100%
Context rebuilding ↓ Zero needed
Steering required ↓ 3x less

🚀 Quick Start

5-Minute Test:

  1. Open a Claude Project
  2. Ask Claude: "Please add a memory edit: 'User works on [YourProject]'"
  3. Start a new conversation
  4. Ask: "What project am I working on?"

Claude will remember—even though you never mentioned it in this conversation.

TL;DR Version (2-minute read)


📖 The Guide

The /guide folder contains the complete methodology:

Chapter Topic
Chapter 1 The Discovery
Chapter 2 Facts vs. Behaviors (The Key Insight)
Chapter 3 The Four Commands
Chapter 4 What Works (And What Doesn't)
Chapter 5 Real-World Case Study: 5Levels
Chapter 6 Best Practices
Chapter 7 Managing Your 30-Edit Limit
Chapter 8 Troubleshooting
Chapter 9 The Honest Limitations

Prefer a single file?Complete Guide (all chapters)


📋 Templates

The /templates folder contains starter sets by project type:

Template Use Case
Web Development Full-stack web projects
Data Science ML/AI projects
Mobile App iOS, Android, React Native, Flutter
E-Commerce Online stores, marketplaces
Content Writing Blogs, social media, documentation

🔧 Claude Integration

The SKILL.md file is designed for Claude's Skills feature.

How to use:

  • Claude Web/Desktop: Upload via Skills interface (Settings → Features → Skills)
  • Claude Projects: Add to project knowledge

📄 Alternative Formats

Format Best For
Web Version Reading online with navigation
Complete Markdown Single-file, LLM-friendly
PDF Version Printing or tablet reading
Word Version Editing or annotations
TL;DR 2-minute summary

🎓 Who This Is For

Perfect for:

  • 🏢 Enterprise AI teams managing complex projects
  • 👨‍💻 Developers building multi-agent systems
  • 🔧 Solution architects designing AI workflows
  • 📚 Technical leads coordinating AI development
  • 🚀 Power users wanting expert-level Claude performance

Not for:

  • ❌ Casual Claude users (regular memory works fine)
  • ❌ Simple single-conversation tasks
  • ❌ Projects without documentation

🤝 Contributing

Found something that works even better? Discovered a new pattern?

Contributions welcome:

  • 📝 Additional examples
  • 🧪 Testing results from other projects
  • 🐛 Bug reports or clarifications
  • 💡 Best practices from your experience

Contributing Guidelines


📜 License & Attribution

License: Creative Commons Attribution 4.0 International (CC BY 4.0)

You are free to:

  • ✅ Share and redistribute
  • ✅ Adapt and build upon
  • ✅ Use commercially

Under these terms:

  • 📌 Give appropriate credit
  • 📌 Link to license
  • 📌 Indicate if changes were made

👤 About the Author

Francesco Marinoni Moretto
Enterprise AI Leader | Constitutional AI Expert | N1AI Community

Related Work:

  • Stream Coding — The 10-20x methodology for AI-accelerated development
  • 5Levels — LinkedIn relationship intelligence platform

📚 Resources

Official Documentation:

Community:


⭐ Star This Repository

If this guide helped you achieve better Claude performance, star this repository to help others discover it.

Questions? Issues? Ideas?
Open an Issue


Last Updated: December 2025
Version: 3.0
Status: Actively Maintained


Stop fighting Claude's memory. Start controlling it.
Read OnlineTL;DRFull Guide

About

The definitive guide to Claude's memory_user_edits - an officially documented but practically undiscovered feature. Based on systematic research and controlled testing with the 5Levels platform. Results: 62% error reduction, up to 100% accuracy improvement. Includes testing methodology and production best practices.

Resources

License

Contributing

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published