Official

Memory MCP Server

Persistent knowledge graph memory enabling AI to recall entities and relations across chats.

Data last scanned today · Reviewed today

713 dangerous tools9 tools88,628TypeScriptNOASSERTION

Overview

The Memory MCP Server implements a local knowledge graph that allows your AI assistant to maintain long-term context. By storing information as discrete entities—such as people, organizations, or events—and defining active-voice relationships between them, it creates a structured database of facts.

Our verdict

This server carries a high-risk profile with a security score of 56/100. The primary vulnerability is a complete lack of authentication, meaning any process with configuration access can manipulate your data, compounded by the fact that the architecture provides full write access without a read-only mode. While it is maintained by the official Model Context Protocol organization and has a strong contributor base, users should exercise caution and avoid using it in environments where unauthorized access to sensitive knowledge graphs poses a significant threat.

  • Maintained by the official Model Context Protocol organization
  • Single direct dependency minimizes supply-chain surface area
  • Highly active repository with frequent commits
  • Lacks any authentication method for protecting stored data
  • No read-only mode restricts access control options
  • License status is marked as NOASSERTION

Setup

Claude Desktop
{
  "mcpServers": {
    "memory": {
      "args": [
        "-y",
        "@modelcontextprotocol/server-memory"
      ],
      "command": "npx"
    }
  }
}

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.

  • No read-only mode — running this grants full read/write access.
  • No built-in authentication — anyone with access to this config can invoke every tool it exposes.

Tools

ToolDescriptionRisk
create_entitiesCreate multiple new entities in the knowledge graphmedium
create_relationsCreate multiple new relations between entitiesmedium
add_observationsAdd new observations to existing entitiesmedium
delete_entitiesRemove entities and their relationshigh
delete_observationsRemove specific observations from entitieshigh
delete_relationsRemove specific relations from the graphhigh
read_graphRead the entire knowledge graphlow
search_nodesSearch for nodes based on querylow
open_nodesRetrieve specific nodes by namelow

Compatibility

ClientLocalDockerRemoteRead-only
Claude Desktop
Cursor
VS Code
Windsurf
ChatGPT

Alternatives

Badge

Memory MCP Server security score, rated on RepoAI

Maintain this server? Add the live badge to your README.