MCP server · Notes
agent-memory MCP server
by xChuCx
Give your AI a project memory that lives in your repo as plain Markdown files you can review.

agent-memory is a small helper that gives your AI coding assistant a lasting memory for one project. Instead of forgetting everything between chats, the AI can save notes about your project's rules, decisions, and tricky spots into plain text files inside your repository. It is handy if you use a coding assistant like Claude Code, Cursor, or Codex and want it to remember how your project works.
What is an MCP server? The 30-second version
On its own, your AI can only chat. An MCP server is a small helper program that gives your AI a new skill or a connection to something else. This one connects your AI to a memory store kept inside your own project folder, so the AI can look up what it already knows and write down new things when you ask. You do not need any cloud account or database for it.
What this MCP server does
You ask your AI something about your project, or ask it to remember something. The AI uses this helper to read or write notes in a special folder called .agent-memory inside your repo. Important notes like rules and decisions get held back for your review first, so you can see exactly what would change. You look at the difference, approve it, and the note becomes part of the project. Later, the AI can find those notes again when it needs them.
Click to zoomWhat you can do with it
- Save project rules, decisions, and pitfalls as plain text files
- Look up what the AI already knows about a topic like auth or build steps
- Stage important changes so you can review them before they land
- See a clear before-and-after difference of any memory change
- Keep separate memory for each branch of your project
- Block secrets and personal information from being written into memory
- Let your AI fill in the memory by reading your existing codebase
Try asking your AI
- “Remember that we always run go test ./... before committing.”
- “What do you already know about how authentication works in this project?”
- “Look through this repo and write down the build commands and main modules.”
- “Show me the staged memory changes so I can review them.”
What it gives back to you
You get back notes, lists, and short summaries pulled from your project memory, shown right in the chat. When something new is saved, you see a clear difference showing exactly which lines were added or changed. Nothing important is written until you approve it, so you always know what changed.
Before you start
What you need
- A project folder stored with git
- A coding assistant that can use MCP servers, like Claude Code, Cursor, or Codex
- Node.js if you want to run it with npx, or Go 1.25 or newer if you build it yourself
Good to know
It can change files in your project, so review staged changes before you approve them, and never let secrets or passwords get written into memory.
Install it with your AI
Add agent-memory MCP server to your AI, no technical skills needed
You don't install anything by hand. You copy one prompt, paste it into an AI that can work on your computer, and it checks, installs and connects the server for you, asking you when it needs something.
Sign in to get the install prompt
Members get a ready-made prompt that lets the Claude desktop app check agent-memory MCP server, install it and connect it for them, step by step. You don't need any technical skills: you copy, paste and answer a few questions. Your connected AI can also find and install any of the 4,066 MCP servers here for you.
Who it's for
Developers and technical teams who use AI coding assistants and want them to remember project details between sessions.





