mcp-shared-memory
Universal Shared MCP Memory Server - Multi-tool compatible memory management for AI assistants across Claude Desktop, Cursor, Windsurf, Cline, and more.
claude mcp add --transport stdio ideas-net-studio-mcp-shared-memory npx -y mcp-shared-memory
How to use
This MCP server provides a shared, structured memory store that can be used across multiple MCP-enabled tools and chat sessions. It is designed to help tools like Claude Desktop, Cursor, Windsurf, Cline, RooCode, Warp, Antigravity, and others maintain project-specific memories, entities, concepts, and sessions in a centralized, organized way. You can run the server directly with npx, or install and run it from a local environment after cloning the repository. Once running, configure your MCP-enabled tools to connect to the shared-memory server so they can read from and write to the memory store, enabling consistent context and memory across tools and sessions. A notable addition in v0.1.2 is version tracking and update notifications, so you can ask the assistant “What’s new in mcp-shared-memory?” to learn about latest changes.
How to install
Prerequisites:
- Node.js and npm installed on your system
- Git for cloning repositories (optional if using npx)
Installation steps:
- Install dependencies and build (if you prefer a local install):
# Clone the repository
git clone https://github.com/Ideas-Net-Studio/mcp-shared-memory.git
cd mcp-shared-memory
# Install dependencies
npm install
# Build for production (if provided by the project)
npm run build
- Run using npx (recommended):
npx -y mcp-shared-memory
- If you clone and want to run locally after build, configure your MCP tool to point to the built entry:
{
"mcpServers": {
"shared-memory": {
"command": "node",
"args": ["/absolute/path/to/mcp-memory/dist/index.js"]
}
}
}
- Optional: add an npm script for convenience in your project, e.g. in package.json:
"scripts": {
"start:mcp-shared-memory": "npx -y mcp-shared-memory"
}
Then run: npm run start:mcp-shared-memory
Additional notes
Tips and recommendations:
- Environment and memory storage: memoryDir, enabled, and description can typically be configured per your project needs (e.g., in a .mcp-memory.json or mcp-memory.json file as shown in the project structure).
- If you build locally, ensure the absolute path to the built entry (dist/index.js) is correct when configuring tools.
- Common issues: ensure Node and npm versions are compatible with the server, and that your MCP tools are configured to point to the correct host/port or file path as required by your setup.
- Version 0.1.2 adds version tracking and update notifications; use the assistant prompt “What’s new in mcp-shared-memory?” for recent changes.
- This server is designed for multi-tool compatibility, so you should be able to connect multiple MCP-enabled tools to the same memory store and share project memories across sessions.
Related MCP Servers
awesome-claude-skills
A curated list of awesome Claude Skills, resources, and tools for customizing Claude AI workflows
octocode
MCP server for semantic code research and context generation on real-time using LLM patterns | Search naturally across public & private repos based on your permissions | Transform any accessible codebase/s into AI-optimized knowledge on simple and complex flows | Find real implementations and live docs from anywhere
BifrostMCP
VSCode Extension with an MCP server that exposes semantic tools like Find Usages and Rename to LLMs
agent-pm
MCP server for the planning and execution of AI-assisted development projects.
envmcp
A lightweight way to use environment variables in your Cursor MCP server definitions.
minime
Universal infinite memory layer for Developer AI assistants. One shared brain across Claude, Cursor, Windsurf & more. 100% local, built on MCP standard. Stop re-explaining context