Knowledge Graph Memory Server

Knowledge Graph Memory Server

MCP server enabling persistent memory for Claude through a local knowledge graph. This lets Claude remember information about the user across chats.

Overview

The Knowledge Graph Memory Server is an enhanced implementation of persistent memory utilising a local knowledge graph, designed to enable Claude to retain user information across conversations. This fork, developed by Shane Holloman, focuses on local development by allowing a customisable `--memory-path` for storing memory data, moving away from the ephemeral memory installation method.

Created by shaneholloman

By integrating this server, Product Designers, Project Managers, and Design Leaders can ensure that Claude maintains context and personalisation throughout user interactions, leading to more coherent and tailored experiences.

Key Features

  • The memory MCP server is hosted locally on the user's device, allowing for unlimited memories
  • Persistent user information retention across chats.
  • Local knowledge graph implementation for enhanced data control.

Capabilities

  • Creation of multiple new entities in the knowledge graph.
  • Establishment of relations between entities.
  • Addition of new observations to existing entities.
  • Deletion of entities and their associated relations.
  • Removal of specific observations from entities.
  • Deletion of specific relations from the graph.
  • Retrieval of the entire knowledge graph structure.
  • Search functionality across entity names, types, and observation content.
  • Retrieval of specific nodes by name, including their relations.

How It Works

  • Entities are defined with a unique name, entity type, and a list of observations.
  • Relations are established to define directed connections between entities, stored in active voice.
  • Observations are discrete pieces of information about an entity, stored as strings and attached to specific entities.
  • The server provides tools for creating, retrieving, updating, and deleting entities and relations, as well as searching the memory based on natural language queries.

New to MCP or wondering why it matters?
Read how I’m using it in real workflows—with Claude, Figma, and Ghost:
👉 What Is Model Context Protocol (MCP)? It Changed How I Design—and Blog
About the author
Subin Park

Subin Park

Principal Designer | Ai-Driven UX Strategy Helping product teams deliver real impact through evidence-led design, design systems, and scalable AI workflows.

Ai for Pro

Curated AI workflows, prompts, and playbooks—for product designers who build smarter, faster, and with impact.

Ai for Pro - Curated AI workflows and Product Design guides—built for Product Designers, PMs, and design leaders.

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to Ai for Pro - Curated AI workflows and Product Design guides—built for Product Designers, PMs, and design leaders..

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.