Memory Markdown FAQs

Question 1

Why use a local Markdown file for Claude's memory?

Accepted Answer

Using a local Markdown file ensures your AI's memory is private, easily accessible, and human-readable. It promotes a zero-dependency setup, simplifies backups, and offers full control over your data without relying on external databases or cloud services, making it a robust and transparent solution.

Question 2

What is Memory Markdown and how does it work?

Accepted Answer

Memory Markdown is a lightweight Model Context Protocol (MCP) server written in pure PHP 8.1+. It enables Claude to store and retrieve persistent long-term memory by reading from and writing to a local Markdown file. This eliminates the need for databases or cloud services, keeping your AI's memory local and transparent.

Question 3

What are the key features of Memory Markdown?

Accepted Answer

It offers persistent memory stored in plain Markdown, zero dependencies (pure PHP 8.1+), three focused tools (`knowledge`, `remember`, `append`), a customisable initial memory template, and a real-time memory watcher for live monitoring of changes. It's also fully compatible with Claude Desktop and any MCP-capable client.

Question 4

How do I install and configure Memory Markdown?

Accepted Answer

Installation is simple: clone the Git repository and make `mcp.php` executable. There are no Composer dependencies. For Claude Desktop, you'll need to add an absolute path to `mcp.php` in your `claude_desktop_config.json` file under `mcpServers` and then restart Claude Desktop. All paths and limits are configurable via `Config.php`.

Question 5

What tools does Memory Markdown provide to Claude?

Accepted Answer

Memory Markdown provides three essential tools: `knowledge` to read the entire content of `memory.md` (ideal at the start of conversations), `remember` to replace the entire memory with new text (useful for reorganization), and `append` to add text to the end of `memory.md` without losing existing content.

Memory Markdown

Memory Markdown

Key Features

Use Cases

Key Features

Use Cases