Every AI application needs robust memory to avoid forgetting users, past decisions, or crucial context. Remembra offers a self-hostable, open-source memory layer designed to empower AI chatbots, agents, and assistants with continuous recall. It features advanced capabilities like automated entity resolution, temporal decay, contradiction detection, and graph-aware recall, ensuring AI applications maintain a consistently accurate and evolving understanding of their interactions and stored facts. Integrate it into your projects in minutes with SDKs for Python and TypeScript, or leverage its MCP server for seamless integration with popular AI coding tools.
主な機能
01Bi-Temporal Relationships with point-in-time querying
02Automated Entity Resolution across varied mentions
03Hybrid Search combining semantic and keyword recall
04Automatic Contradiction Detection with full history preservation
05Built-in PII Detection and AES-256-GCM Field Encryption
066 GitHub stars
