Anthropic Memory MCP Visualizer

Interactive visualizer for Anthropic's Memory MCP knowledge graphs. Instantly explore, debug, and analyze entities, relations, and observations from memory.json files generated by the Knowledge Graph Memory Server.

Try it now: This tool is freely available at https://memviz.herich.tech

Features

• Interactive knowledge graph visualization using D3.js
• Filter by entity and relation types
• Search across entities and relationships
• View detailed observations in the info panel
• Drop memory.json files or paste directly from clipboard
• Navigate complex entity networks visually
• Debug AI memory structures and connections

Development

# Install dependencies
npm install

# Start development server
npm run dev

# Build for production
npm run build

Tech Stack

• React + TypeScript
• D3.js for visualization
• Vite
• TailwindCSS

Using with Claude Desktop App

This tool is intended for use with the Claude Desktop app, which supports the Memory MCP protocol for knowledge graph storage and retrieval.

Configuring the Memory MCP Server

1. First, ensure you have the modelcontextprotocol-servers repository cloned locally.

2. Configure the Claude Desktop app to use the Memory MCP server by adding the following to your mcp_config.json file:

"memory": {
  "command": "npx",
  "args": [
    "-y",
    "@modelcontextprotocol/server-memory"
  ],
  "env": {
    "MEMORY_FILE_PATH": "/path/to/custom/memory.json"
  }
}

The above example shows a generic configuration. You should replace the paths with the actual locations on your system where you want to store your memory file.

💡 Personal tip: I store my memory.json in a Google Drive folder which automatically backs it up and makes it accessible across devices. Consider using a cloud storage solution for your memory file too!

3. Start Claude Desktop with this configuration to enable the memory features.

4. Use this Memory Visualizer to explore and debug the knowledge graph created by Claude.

Knowledge Graph Structure

Visualizes Anthropic's Memory MCP format:

• Entities: Nodes with name, type, and observations (e.g., people, organizations)
• Relations: Directed connections between entities in active voice (e.g., "works_at")
• Observations: Atomic facts attached to entities

Use Cases

• Debug entity relationships in AI memory
• Analyze memory.json structures
• Explore entity connections
• Visualize graph growth over time