Remembra – Persistent memory for AI agents with local, secure storage

Summary: Remembra provides persistent memory for AI applications with entity resolution, temporal features, and graph-aware recall. It achieves 100% accuracy on the LoCoMo benchmark and supports self-hosting with local data storage and encryption.

What it does

Remembra enables AI agents to retain information over time using entity resolution and temporal features, running fully locally with AES-256-GCM encryption and PII detection to prevent sensitive data storage.

Who it's for

It is designed for developers integrating persistent memory into AI tools like Claude Code, Cursor, and JetBrains, requiring secure, local data handling.

Why it matters

It solves the problem of AI agents forgetting prior interactions by providing accurate, secure, and persistent memory storage.