mcp-knowledge-graph
If you are the rightful owner of mcp-knowledge-graph and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
An improved implementation of persistent memory using a local knowledge graph with a customizable memory path.
The Knowledge Graph Memory Server is a sophisticated tool designed to enhance persistent memory capabilities by utilizing a local knowledge graph. This server allows for the storage and retrieval of user information across different chat sessions, enabling a more personalized interaction experience. It is a fork of the original Memory Server, with improvements that eliminate the need for ephemeral memory installations. The server organizes data into entities, relations, and observations, each with version tracking to maintain historical context. This structure supports complex data relationships and ensures that the server can evolve with the user's interactions. The server is particularly useful for applications requiring long-term memory retention and retrieval, such as virtual assistants and chatbots.
Features
- Entities: Primary nodes with unique identifiers, types, and observations.
- Relations: Directed connections between entities with version tracking.
- Observations: Atomic pieces of information attached to entities.
- API Tools: Comprehensive set of tools for creating, deleting, and managing entities, relations, and observations.
- Custom Memory Path: Allows specification of memory file path via command-line or environment variables.
Tools
create_entities
Create a new entity
create_relations
Create relationships between entities
add_observations
Add observations to entities
delete_entities
Delete entities and their associations
delete_observations
Delete specific observations
delete_relations
Delete a specific relationship
read_graph
Read the entire knowledge graph
search_nodes
Search nodes based on query
open_nodes
Retrieve specific nodes by name
