What problem does it solve?
It gives an AI agent persistent local memory, so user preferences, facts, tasks, and important notes survive across conversations instead of being forgotten after each session.
Core Features & Use Cases
- Memory Storage with Auto-Classification: Saves user-provided content into a local SQLite database and automatically tags it as preference, fact, task, important, or general.
- Semantic Search and Prioritized Recall: Retrieves relevant entries using keyword similarity ranked by recency, access frequency, and category weight.
- Statistics and Context Injection: Reports knowledge base status (entry counts, storage size, top-accessed items) and builds a relevant-memory context block for the current conversation.
- Use Case: A user says "save to my knowledge base: I prefer concise answers in Chinese", and later asks "what do you know about my preferences?" — the agent stores the note, then retrieves and applies it in future replies.
Quick Start
Ask the agent to save a piece of information to your local knowledge base, then later ask it to search or summarize what has been stored.
