Your AI assistant typically forgets everything when you close a chat. Local Memory solves this by storing learnings, decisions, people, and projects in a single SQLite file directly on your machine. It features a built-in Knowledge Graph, duplicate detection, and full-text search, all without needing any cloud services or API keys. It allows your AI to pick up where it left off, learn from past interactions, and build a rich understanding of your work environment over time.
Key Features
01Structured Decision Tracking with reasoning and alternatives
02Knowledge Graph with entities and relations
03Duplicate Guard using FTS5 similarity to prevent redundant storage
04Automatic Session Context loading past interactions
05Full-Text Search (FTS5 + bm25 ranking) across all stored data
061 GitHub stars
Use Cases
01Building a local knowledge graph of people, projects, and tools encountered by the AI
02Providing AI assistants with continuous context across conversations and projects
03Storing and recalling technical patterns, insights, and mistakes for future reference
