memory-visualizer
Interactive visualizer for Anthropic's Memory MCP knowledge graphs. Instantly explore, debug, and analyze entities, relations, and observations from memory.json files in the Model Context Protocol.
claude mcp add --transport stdio mjherich-memory-visualizer npx -y @modelcontextprotocol/server-memory \ --env MEMORY_FILE_PATH="/path/to/custom/memory.json"
How to use
Memory Visualizer is an interactive frontend for exploring Anthropic's Memory MCP knowledge graphs stored in memory.json files. It renders the graph with D3.js, allowing you to filter by entity and relation types, search across entities and relationships, and open an info panel to view detailed observations attached to each entity. You can drop a memory.json file into the app or paste JSON directly to load a graph, making it convenient for debugging memory structures produced by the Memory MCP server. The visualization helps you navigate complex networks visually and understand how entities relate through various observations, making it a valuable tool for analyzing memory graphs and debugging AI memory structures.
How to install
Prerequisites:
- Node.js (v14+ recommended) and npm/yarn
- A working memory.json file generated by the Memory MCP server
Installation steps:
-
Clone the repository (or download the project files): git clone https://github.com/modelcontextprotocol/servers.git cd servers/src/memory/memory-visualizer
-
Install dependencies: npm install
-
Run the development server: npm run dev
-
Build for production (optional): npm run build
If you prefer using a prebuilt package, you can also install via npm if the project is published as a package, e.g.: npm install @modelcontextprotocol/server-memory-visualizer --save npm run dev
Note: The exact repository path and npm script names may vary slightly depending on project structure. Refer to the project's package.json for the authoritative scripts.
Additional notes
Tips and common considerations:
- MEMORY_FILE_PATH environment variable: point this to the memory.json you want to visualize. If omitted, the app may prompt for a file upload or expect a default location.
- Ensure memory.json conforms to the Memory MCP structure (entities, relations, and observations) expected by the visualizer.
- The dev server typically runs on localhost with a port like 5173 (or as configured by your Vite setup); check the console during npm run dev for the exact URL.
- If you integrate with Claude Desktop or other MCP clients, you may provide this visualizer as a helper tool to inspect and debug memory graphs generated by Memory MCP servers.
- You can drop memory.json files or paste JSON directly into the app to quickly load and explore graphs; this is useful for ad-hoc debugging and testing.
Related MCP Servers
mcp-knowledge-graph
MCP server enabling persistent memory for Claude through a local knowledge graph - fork focused on local development
mcp-manager
simple web ui to manage mcp (model context protocol) servers in the claude app
mcp-memory-libsql
🧠 High-performance persistent memory system for Model Context Protocol (MCP) powered by libSQL. Features vector search, semantic knowledge storage, and efficient relationship management - perfect for AI agents and knowledge graph applications.
Devmind
DevMind MCP provides **persistent memory capabilities** for AI assistants through the Model Context Protocol (MCP). It enables AI to remember context across conversations, automatically track development activities, and retrieve relevant information intelligently.
mcp-turso
MCP server for interacting with Turso-hosted LibSQL databases
agentgate
Approval workflows for AI agents