A Mem MCP runs on your own machine — the client starts it, so there is no endpoint to ping. 157 installs a week from pypi. Last commit 17 Jan 2026.
Self-evolving memory system for AI agents
What the code does
We read the source, 4 d ago · rules 995398eba69f
What this server is able to do. For an MCP server this is often the job itself — a terminal server runs commands because that is what it is for. Listed so you know what you are plugging in, not as an accusation.
document.getElementById('legend-items').innerHTML = categories.map(cat => `
Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.
This code can reach further than it looks
We found places where it runs commands, builds paths or queries from values it is given. None of that is a flaw by itself — it becomes one when the code changes, and code changes quietly between releases. We re-read it on every one.
Connect this server
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add a-mem-mcp -e 'OPENAI_API_KEY=<OPENAI_API_KEY>' -- uvx a-mem{
"mcpServers": {
"a-mem-mcp": {
"command": "uvx",
"args": [
"a-mem"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>"
}
}
}
}[mcp_servers.a-mem-mcp]
command = "uvx"
args = ["a-mem"]
env = { OPENAI_API_KEY = "<OPENAI_API_KEY>" }{
"mcpServers": {
"a-mem-mcp": {
"command": "uvx",
"args": [
"a-mem"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>"
}
}
}
}{
"servers": {
"a-mem-mcp": {
"type": "stdio",
"command": "uvx",
"args": [
"a-mem"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>"
}
}
}
}Put your own value in place of
<OPENAI_API_KEY> (OpenAI API key (required if LLM_BACKEND=openai)).
The author’s registry entry marks it as required.
Optional settings from the registry entry, left out of the block above:
LLM_BACKEND (LLM backend to use (openai, ollama, sglang, or openrouter)), LLM_MODEL (LLM model name (e.g., gpt-4o-mini, llama2, etc.)), EMBEDDING_MODEL (Sentence transformer model for embeddings), CHROMA_DB_PATH (ChromaDB storage directory path).
Alternatives to A Mem MCP
same job, measured the same wayA self-improving, verifiable memory layer for AI coding agents.
Shared debugging memory for AI coding agents
Memory system for AI agents with semantic search. Store and recall memories with ease.
Undisk MCP: safe file memory for AI agents.
CogMemory MCP Server — Unified context subsystems for AI coding agents
Local-first memory system for AI agents with hybrid search and graph reasoning
MCP server for persistent, searchable memory for AI agents.
Local-first persistent memory MCP server for AI coding agents.
A Mem MCP — questions
Answers built from our own checks of this server.