# Memlo [Health: Active]

**Category:** 🧠 Knowledge & Memory  
**Repository:** https://memlo.ai/?utm_source=allmcps&utm_medium=directory  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/memlo

## Description
Hosted memory for AI agents over MCP. One memory shared by all your agents, with search, version history and sharing. Invitation-only beta, free during the beta.

## Claude Desktop Quick Installation
No generated config for this server: it installs through its own setup steps.

The project's own setup steps (from https://memlo.ai/agent), copied verbatim. Run them in order; they are not an mcpServers config.

1. Add to Claude Code:

```sh
claude mcp add --transport http memlo https://memlo.ai/mcp
```

2. Add to Claude Code:

```sh
claude mcp add --transport http --scope user memlo https://memlo.ai/mcp
```

## Documentation

## What Memlo does

Memlo provides hosted memory for AI agents through MCP. Instead of keeping separate memory in each agent or client, Memlo supports one shared memory that multiple agents can use. Its documented capabilities include searching memory, reviewing version history, and sharing memory.

The shared-memory model is the main reason to consider Memlo. It fits workflows where several agents need access to the same accumulated information, rather than isolated context maintained by a single agent. Because the memory is hosted, agents connect to a common service instead of relying only on local state.

## How it works

Agents interact with Memlo over the Model Context Protocol. MCP provides the connection between compatible AI agents and Memlo’s hosted memory service. Through that connection, agents can search the shared memory and work with its version history and sharing features.

The service is designed around a single memory shared by all of your agents. This makes Memlo relevant for multi-agent workflows in which information created or retained by one agent should remain available to others.

## Tools and capabilities

- Hosted memory for AI agents
- Shared memory accessible to all connected agents
- Memory search
- Version history
- Memory sharing

## Limitations and notes

Memlo is currently available by invitation only, so access depends on receiving an invitation to the beta. The service is free during the beta period. Its focus is hosted, shared agent memory rather than a broad collection of unrelated MCP tools.

Before adopting Memlo for a production workflow, confirm that the beta access model and service availability fit the intended use. The shared-memory design also makes it important to choose deliberately which agents should connect to the same memory.

_Full upstream README: https://allmcps.com/mcp/memlo/readme_