Local Memory

概要

Designed as a centralized long-term memory pool, Local Memory allows multiple AI agents to share knowledge and maintain consistent information across various workflows. It transforms ephemeral agent memories into a robust, structured knowledge base, enabling persistent intelligence and evolving understanding within your local ecosystem. This server operates locally, ensuring privacy and resource efficiency while supporting advanced features like hybrid search, knowledge graph traversal, automated entity extraction, and integrated task management.

主な機能

• Hybrid Search (Semantic Vector + Keyword FTS5)
• Knowledge Graph with structured entities, relations, and deep traversal
• Local Embeddings using transformers.js (ONNX) for privacy
• Integrated Todo System with automatic context injection and memory archival
• The Archivist for configurable, automatic knowledge graph building strategies
• Time Tunnel for natural language date querying and temporal recall
• 1 GitHub stars

ユースケース

• Enabling shared long-term memory across diverse AI agents (e.g., coding, CLI, IDE extensions).
• Managing persistent user preferences and project context for AI workflows.
• Automating task tracking and context injection for AI-assisted task management.