# open-memex

> open-memex is an open-source MCP server for knowledge and memory. Shared local Markdown memory for coding agents across Copilot, Cursor and Claude Code. It runs locally on your machine and is started by your AI client.

- **Kind:** MCP server
- **Pricing:** Open source
- **Platforms:** CLI, vscode, self-hosted
- **Website:** https://stoneskin.github.io/open-memex/
- **Repository:** https://github.com/stoneskin/open-memex
- **MCP Registry name:** io.github.stoneskin/open-memex
- **Categories:** knowledge and memory, developer tools
- **Listing updated:** 2026-11-14

## About

open-memex targets one frustration: every coding session starts from zero, and decisions made yesterday vanish when the chat ends. It stores memories as plain Markdown files, treated as the source of truth, with a rebuildable SQLite FTS5 index for fast retrieval, so you can read, edit, back up or delete them with ordinary tools.

Because every connected editor reads and writes the same store, something captured in Cursor is respected later in Claude Code, VS Code Copilot, Visual Studio or opencode. `open-memex init` detects installed editors and wires them up. Search is keyword-based with a curated multilingual synonym map, so `ship` finds `deploy`, with no embedding model, and a derived knowledge graph of project memories feeds an outline into session start.

Privacy is local: no account, cloud service or telemetry, common API keys are masked before saving, and text inside `<private>` tags is stripped. Personal memories never sync, while project memories can be proposed, reviewed and published through Git branches and pull requests. It needs Node.js 22.14 or newer and is Apache-2.0 licensed.

## Features

- **Plain Markdown storage:** Memories are files you can open, edit and back up.
- **Shared across editors:** Copilot, Cursor, Claude Code, Visual Studio and opencode use one store.
- **Synonym-aware search:** A curated multilingual map lets related wording find the right memory.
- **Secret masking:** Common keys and tokens are masked and private-tagged text is stripped before saving.
- **Git team sharing:** Project memories go from personal draft to reviewed memory via pull requests.
- **Health check:** `open-memex doctor` verifies storage, editor wiring and the SQLite driver.

## Good for

- Remember deployment rules and environment quirks across sessions
- Record architecture decisions with their reasons
- Share one memory between several coding tools
- Pass reviewed project knowledge to teammates through Git
- Keep personal working preferences available to every agent

## Install

### Claude Code (Terminal)

```sh
claude mcp add open-memex -- npx -y open-memex
```

### Claude Desktop (claude_desktop_config.json)

```json
{
  "mcpServers": {
    "open-memex": {
      "command": "npx",
      "args": [
        "-y",
        "open-memex"
      ]
    }
  }
}
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "open-memex": {
      "command": "npx",
      "args": [
        "-y",
        "open-memex"
      ]
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "open-memex": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "open-memex"
      ]
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "open-memex": {
      "command": "npx",
      "args": [
        "-y",
        "open-memex"
      ]
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "open-memex": {
      "command": "npx",
      "args": [
        "-y",
        "open-memex"
      ]
    }
  }
}
```

## FAQ

### Does it send data to a cloud?

No. There is no account, cloud service or telemetry, and personal memories stay on the machine.

### What does it need?

Node.js 22.14 or newer, installed with `npm install -g open-memex`.

### Is an embedding model required?

No. Retrieval uses SQLite FTS5 keyword search with a synonym map.

### What is the licence?

Apache-2.0.

### Is open-memex free?

Yes. open-memex is open source, so you can use and self-host it for free.

### How do I add open-memex to Claude Code?

Run this command in your terminal: claude mcp add open-memex -- npx -y open-memex

### What are the best alternatives to open-memex?

The closest alternatives to open-memex are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [GitHub](https://ai-powered.com/mcp/github-2): GitHub is an open-source MCP server for developer tools.
- [Serena MCP: the IDE for your agent](https://ai-powered.com/mcp/serena-mcp-the-ide-for-your-agent): Serena MCP: the IDE for your agent is an open-source MCP server for developer tools.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.
- [Ouroboros](https://ai-powered.com/mcp/ouroboros): Ouroboros is an open-source MCP server for developer tools.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [Nautilus Compass Memory](https://ai-powered.com/mcp/nautilus-compass-memory): Nautilus Compass Memory is an open-source MCP server for knowledge and memory.

---
Source: https://ai-powered.com/mcp/open-memex
