Remote Memory

You deploy Remote Memory MCP Server to synchronize a knowledge graph with GitHub repositories, enabling remote storage, collaboration, and automated conflict-aware syncing. It exposes CRUD operations for entities, relations, and observations, with real-time or manual synchronization and advanced query tools to explore your memory graph.

How to use

Set up your MCP client to connect to the Remote Memory MCP Server. You will perform CRUD operations on memory graph elements (entities, relations, observations), and the server will synchronize changes with your GitHub repository. Use the enhanced query tools to list entities, quickly search names, and view entity type statistics. Customize commit messages, enable backups, and review commit history as you evolve your memory graph.

How to install

Prerequisites: ensure you have Node.js installed on your system. You can verify with node -v and npm -v.

Then run the following steps to install and build the MCP server.

Installation steps

cd C:\YOUR_PATH\remote-memory-mcp
npm install
npm run build

Configuration and runtime parameters

You must provide GitHub credentials and repository information to enable remote storage and collaboration.

Claude Desktop configuration (MCP server setup)

{
  "mcpServers": {
    "remote_memory": {
      "command": "node",
      "args": ["C://YOUR_PATH//remote-memory-mcp//dist//index.js"],
      "env": {
        "GITHUB_TOKEN": "YOUR_GITHUB_TOKEN_HERE",
        "GITHUB_OWNER": "YOUR_GITHUB_USERNAME", 
        "GITHUB_REPO": "YOUR_GITHUB_REPO",
        "GITHUB_BRANCH": "main",
        "SYNC_INTERVAL": "0",
        "AUTO_PUSH": "false"
      }
    }
  }
}

Usage and capabilities

• Perform CRUD operations for entities, relations, and observations. - Use real-time synchronization with GitHub repositories or manual sync via your chosen interval. - Enable conflict detection and resolution based on the latest modification timestamps. - View and filter data with advanced entity queries, including lists, quick name lookups, type statistics, and date range filtering. - Customize commit messages, create backups, and review commit history. - Optional auto-push after every change via the AUTO_PUSH setting.

Data structure overview

Memory data is stored in the memory/graph.json file in your GitHub repository, containing entities, relations, and metadata about versions and last sync times.

Notes on synchronization and configuration

• You must provide a GitHub Personal Access Token with repo permissions. - The server enforces API rate limits for authenticated users and requires network access. - You can choose auto-push (AUTO_PUSH) or manual synchronization (SYNC_INTERVAL = 0). - Maintain a clean, well-formed memory graph to maximize the effectiveness of conflict resolution and incremental syncing.

Troubleshooting

If you encounter permission or network errors, validate your GitHub token, repository ownership, and network connectivity. Check logs for synchronization status and any conflict resolution prompts.

Available tools

list_entities

Retrieve a list of entities with optional filtering, sorting, and pagination.

get_entity_names

Quick lookup for entity names.

get_entity_types

Statistics by entity type.