# MMW Managed Memory Workspace

> MMW Managed Memory Workspace is an MCP server for knowledge and memory. Shared memory that lets ChatGPT, Codex and Telegram retrieve the same saved notes. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Free trial
- **Starting price:** $18 a month
- **Platforms:** Web, API
- **Website:** https://mmwhub.tech
- **Repository:** https://github.com/msingatullin/mmw-mcp
- **MCP Registry name:** io.github.msingatullin/mmw
- **Categories:** knowledge and memory, AI models and LLM tooling
- **Listing updated:** 2026-11-15

## About

MMW keeps selected decisions and notes in one project so that a fact saved in one assistant can be found from another. Records carry an ID and an optional source_id, so you can check that two clients are looking at the same entry instead of trusting that two answers sound alike. Sources, revisions and conflicting facts are surfaced for review, and the README is explicit that provenance is not proof a fact is true.

Nothing is saved automatically. You ask a client to save something, and MMW does not read all your conversations. An optional local agent, mmw-agent, can upload supported Claude Code and Codex sessions, but only with your permission, and uploading history does not extract facts by itself. Access is controlled with OAuth or keys scoped to a project.

For teams there is an organization layer with roles (admin, unit manager, member, contractor, auditor), memory visibility from author-only drafts up to the whole organization, single-use seven-day invitations, time-boxed contractor access and an audit log that never contains memory text. The repository holds the public connection kit under the MIT licence; the hosted backend is private.

## Features

- **Same record across clients:** Save in ChatGPT, then find the record in Codex or linked Telegram and compare ID and source_id.
- **Source-aware facts:** Records can carry provenance, a hash and a revision, with conflicting facts flagged for review.
- **Save only what you choose:** Notes are stored at your request; MMW does not read all conversations.
- **Scoped access:** OAuth or keys limit each connection to a project, and connections can be revoked.
- **Team roles and visibility:** Organizations get roles, visibility levels, contractor expiry and an audit log without memory text.
- **Offboarding transfer:** When someone leaves, keys are revoked and their knowledge goes to a chosen recipient.

## Good for

- Carry a project decision from ChatGPT into Codex without re-explaining it
- Keep working agreements with a source attached for a small business
- Share technical context between coding assistants on one project
- Give contractors time-limited access to one project's memory
- Review conflicting facts before relying on them

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http mmw https://mcp.mmwhub.tech/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://mcp.mmwhub.tech/mcp
```

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

```json
{
  "mcpServers": {
    "mmw": {
      "url": "https://mcp.mmwhub.tech/mcp"
    }
  }
}
```

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

```json
{
  "servers": {
    "mmw": {
      "type": "http",
      "url": "https://mcp.mmwhub.tech/mcp"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "mmw": {
      "serverUrl": "https://mcp.mmwhub.tech/mcp"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "mmw": {
      "type": "http",
      "url": "https://mcp.mmwhub.tech/mcp"
    }
  }
}
```

## FAQ

### Does MMW read all my conversations?

No. It stores only what you ask a client to save. Uploading Claude Code or Codex session history through mmw-agent is a separate opt-in.

### How do I check that sharing works?

Save a note with a unique marker in one client, then search for the marker in another and compare the record ID and source_id.

### Is the backend open source?

No. The repository contains the public connection kit under MIT, not the private backend.

### What does the team plan cost?

The website lists $18 per seat per month, or $14 per seat billed annually, with a five-seat minimum.

### Is MMW Managed Memory Workspace free?

MMW Managed Memory Workspace is a paid product with a free trial. Paid plans start at $18 a month.

### How do I add MMW Managed Memory Workspace to Claude Code?

Run this command in your terminal: claude mcp add --transport http mmw https://mcp.mmwhub.tech/mcp

### What are the best alternatives to MMW Managed Memory Workspace?

The closest alternatives to MMW Managed Memory Workspace are Codebase Memory, XERJ, Claude Code Ultimate Guide, Nautilus Compass Memory and note1.

## Alternatives

- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.
- [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.
- [note1](https://ai-powered.com/mcp/note1): note1 is an MCP server for productivity.
- [MedXpert 图书馆连接器（通用版）](https://ai-powered.com/mcp/medxpert): MedXpert 图书馆连接器（通用版） is an open-source MCP server for knowledge and memory.
- [Context Nest](https://ai-powered.com/mcp/context-nest): Context Nest is a freemium MCP server for knowledge and memory.
- [Tulsa Ecosystem Map](https://ai-powered.com/mcp/tulsa-ecosystem-map): Tulsa Ecosystem Map is a free MCP server for knowledge and memory.

---
Source: https://ai-powered.com/mcp/mmw-managed-memory-workspace
