Simple Memory MCP

Simple Memory MCP is an MCP server that provides persistent memory storage and retrieval for AI assistants using a local database. It connects MCP-compatible clients such as Claude or other local AI workflows to a local data store, allowing assistants to maintain contextual information across conversations. Developers and everyday AI users employ this server to record notes, instructions, project context, and reusable snippets that can be queried on demand. The architecture relies on standard Node.js scripts to manage title-based memory records, optimized MCP tools to fetch lightweight metadata, and a built-in local web dashboard running on an automatically assigned port. Through this system, users can manage stored records directly via natural language commands or visually through a browser interface that supports full-text searching, editing, reordering, and deleting memories. Simple Memory MCP also features a single-instance design, allowing multiple AI assistants to communicate with the same persistent memory backend concurrently without file locking conflicts or duplicated background server processes.

Category: AI Memory & Context

Tags: context, memory, persistence, sqlite

Visit Simple Memory MCP

How to install and configure Simple Memory MCP

1. Clone the repository and navigate into the folder: bash git clone https://github.com/eragonht1/simple-memory-mcp.git cd simple-memory-mcp 2. Install Node.js dependencies and initialize the database: bash npm install npm run init-db 3. Add the server entry to your MCP client configuration file: json { "mcpServers": { "simple-memory": { "command": "node", "args": ["./start-mcp.js"], "cwd": "/path/to/your/simple-memory-mcp" } } } Replace /path/to/your/simple-memory-mcp with your actual project directory path. 4. Restart your AI assistant. You can also run the web interface directly using npm run web or node start-web.js.

What you can do with Simple Memory MCP

• Storing long-term project guidelines and user preferences by requesting the AI assistant to record titled memory entries into the local database. - Retrieving specific technical documentation or study notes during AI chat sessions using exact title queries to minimize context window usage. - Managing stored memory records visually through the local web interface to edit content, organize entries, or remove outdated context. - Searching previous conversational notes across titles and descriptions using full-text keyword queries inside the web interface or via assistant tools. - Sharing a unified persistent memory store across multiple concurrent AI assistant tools running on the same host machine.

Key facts

• Open Source
• https://github.com/eragonht1/simple-memory-mcp
• AI Memory & Context, Databases & Data Stores
• context, memory, persistence, sqlite

Part of MCP Servers

What can Simple Memory MCP do?

Simple Memory MCP gives AI assistants persistent memory storage and retrieval capabilities. It allows assistants to store new memories under user-defined titles, retrieve titles with timestamps to save context space, fetch full memory contents on demand, and delete obsolete entries. It also provides a browser-based web dashboard for managing and searching memories.

How do I install Simple Memory MCP?

To install Simple Memory MCP, clone the repository from GitHub and install dependencies using npm install. Next, initialize the local database using the command npm run init-db. After initialization, configure your MCP client with the node command pointing to start-mcp.js, set your working directory path, and restart the client.

Which MCP clients work with Simple Memory MCP?

Simple Memory MCP works with any AI assistant client supporting the Model Context Protocol, including Claude Desktop, Augment, and other developer environments. Its single-instance architecture allows multiple different AI clients running on the same computer to connect simultaneously and share the same memory database without process conflicts.

Is Simple Memory MCP open source?

Yes, Simple Memory MCP is open source software hosted on GitHub. Users can inspect the Node.js source code, modify tools, contribute improvements, and run the server locally on Windows, macOS, or Linux machines without relying on remote third-party cloud storage.