Claude Memory Knowledge Graph Visualizer

The Claude Memory Visualizer is a versatile tool designed to transform any Claude user's memory.json data into stunning interactive graphical representations. With this visualizer, users can explore the connections and evolution of their AI assistant's knowledge in an intuitive and immersive way.

Features

• Beautiful Interactive Graph: Utilizes a force-directed layout for dynamic visualizations.
• Smart Filtering: Effortlessly filter data by entity type or search for specific information across the entire dataset.
• Rich Tooltips: Gain deeper insights with detailed entity information available on hover.
• Color-coded Types: Each entity type is distinctively color-coded for easy identification.
• Responsive Design: Fully operational on any device, ensuring accessibility.
• Zero Configuration: Simple setup allows users to clone and run instantly.

What It Looks Like

Experience real Claude memory data visualized as an interactive knowledge graph displaying entities and relationships.

Memory File Location

The visualizer automatically searches for the memory.json file in multiple locations, including:

1. Current directory (./memory.json)
2. User home directory (~/memory.json)
3. Code directory (~/code/memory.json)
4. Config directory (~/.config/claude/memory.json)
5. As a fallback, it provides demo data if no file is found.

Controls

• Filter by Type: Display specific entity types like personalities, tools, and frameworks.
• Search: Quickly locate entities by name or specific content.
• Physics Control: Adjust node repulsion and enhance graph spacing.
• Mouse Controls: Drag to reposition nodes, scroll to zoom, and click to reveal details.
• Keyboard Shortcuts: Press ESC to close modal windows.

Supported Data Format

The tool is compatible with Claude's standard NDJSON memory format:

{"type":"entity","name":"Example Entity","entityType":"example_type","observations":["observation 1","observation 2"]}
{"type":"relation","from":"Entity A","to":"Entity B","relationType":"connects_to"}

Technical Aspects

• Frontend: Built using pure HTML/CSS/JavaScript integrated with Cytoscape.js for graph visualizations.
• Backend: A Python-based HTTP server that requires no external dependencies.
• Performance: Optimized to handle large datasets efficiently.

Perfect For…

• Claude users aiming to visualize their memory data and insights.
• Researchers analyzing complex knowledge graph structures.
• Developers seeking to understand Claude's memory architecture.
• Teams wishing to share and review collective AI interactions.

Contributing

This project is community-driven. Contributions that include additional file format support, export capabilities, advanced filtering, and performance enhancements are highly valued.
Engage with this project to expand its capabilities and network with other Claude users.