optimized-memory-mcp-server
5
This project is a fork of a Python Memory MCP Server that uses a knowledge graph with SQLite backend to create persistent user memories. It supports various API operations for entity and relation management, tailored for integration in AI chat workflows.
optimized-memory-mcp-server
A fork of a Python Memory MCP Server demonstrating AI workflows and prompt design. It employs a knowledge graph to persistently store user information, allowing for personalized chat experiences.
Core Concepts
- Entities: Nodes in the knowledge graph with unique identifiers and types, storing observations.
- Relations: Directed connections between entities.
- Observations: Information pieces attached to entities.
API
- create_entities: Add new entities.
- create_relations: Define relations.
- add_observations: Attach observations to entities.
- delete_entities: Remove entities and relations.
- delete_observations: Remove specific observations.
- delete_relations: Delete relations.
- read_graph: Retrieve full graph.
- search_nodes: Search for nodes.
- open_nodes: Retrieve nodes by name.
Usage with Claude Desktop
Add configurations to claude_desktop_config.json for Docker or NPX setup.
Building
Use Docker to build the server image.