Cuba Memorys is an open-source Model Context Protocol (MCP) server designed to provide long-term, multi-modal memory for AI coding agents. By employing knowledge graphs and supporting four types of memory (semantic, episodic, procedural, and working), it enables agents to search, reason over, and correct their knowledge of a codebase across sessions. The system is written in Rust, backed by PostgreSQL with pgvector, offers hybrid search and local claim verification with ONNX-based models, and is suitable for developers seeking persistent, reliable, and verifiable contextual memory for code agents.
Visit Cuba Memorys's official website for product details and getting started.
Comprehensive API reference and guides for using Cuba Memorys.
Join discussions and get support from other users and developers of Cuba Memorys.
Insights and updates related to the development and usage of Cuba Memorys.