Memograph runs on your own machine — the client starts it, so there is no endpoint to ping. 51 installs a week from pypi. Last commit 25 Jun 2026.
Graph-based memory system for LLMs with knowledge graphs and semantic search.
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.
result = subprocess.run(
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 memograph -e 'MEMOGRAPH_VAULT=<MEMOGRAPH_VAULT>' -- uvx memograph{
"mcpServers": {
"memograph": {
"command": "uvx",
"args": [
"memograph"
],
"env": {
"MEMOGRAPH_VAULT": "<MEMOGRAPH_VAULT>"
}
}
}
}[mcp_servers.memograph]
command = "uvx"
args = ["memograph"]
env = { MEMOGRAPH_VAULT = "<MEMOGRAPH_VAULT>" }{
"mcpServers": {
"memograph": {
"command": "uvx",
"args": [
"memograph"
],
"env": {
"MEMOGRAPH_VAULT": "<MEMOGRAPH_VAULT>"
}
}
}
}{
"servers": {
"memograph": {
"type": "stdio",
"command": "uvx",
"args": [
"memograph"
],
"env": {
"MEMOGRAPH_VAULT": "<MEMOGRAPH_VAULT>"
}
}
}
}Put your own value in place of
<MEMOGRAPH_VAULT> (Path to your MemoGraph vault directory (created if missing)).
The author’s registry entry marks it as required.
Optional settings from the registry entry, left out of the block above:
MEMOGRAPH_READONLY (Set to 'true' to block destructive MCP tools (delete_memory, update_memory, bulk_create, batch_delete, batch_update). Recommended for shared or untrusted clients), MEMOGRAPH_PROVIDER (Optional LLM provider for query_with_context tool. Only needed if you want AI-generated answers), MEMOGRAPH_MODEL (Optional LLM model name (e.g. 'claude-sonnet-4-6'). Used together with MEMOGRAPH_PROVIDER), ANTHROPIC_API_KEY (API key for Anthropic Claude. Required when MEMOGRAPH_PROVIDER=claude or when using the [anthropic] extra), OPENAI_API_KEY (API key for OpenAI. Required when using the [openai] extra (embeddings or LLM)), OLLAMA_HOST (Base URL for a local Ollama server (default http://localhost:11434). Only needed when MEMOGRAPH_PROVIDER=ollama or the [ollama] extra is installed).
Alternatives to Memograph
same job, measured the same waySemantic memory for AI agents with hybrid search, knowledge graph, and consolidation
Event-sourced knowledge graph memory for AI coding agents with semantic search
Persistent knowledge-graph memory for AI agents, with local semantic search. No API keys.
Local memory engine for AI agents with knowledge graphs, forgetting, and semantic recall
Persistent semantic memory for AI agents with knowledge graphs and Intelligence Packs.
Persistent semantic memory for AI agents with knowledge graphs and Intelligence Packs.
Persistent memory for AI assistants with semantic search and knowledge graph relationships.
Persistent working memory for AI agents: a knowledge graph with semantic search and decay.
Memograph — questions
Answers built from our own checks of this server.