
Local encrypted memory engine for personal AI — MCP and agent ready.
Visit Cortex Memory EngineCortex Memory Engine is a local-first Model Context Protocol (MCP) server designed to provide persistent, structured, and privacy-preserving memory for personal AI assistants, including Claude, Cursor, and Windsurf. It implements a 4-tier memory architecture (working, episodic, semantic, procedural), a cross-channel people graph for identity resolution, and a Bayesian belief system for dynamically updating knowledge with confidence scores. Cortex integrates with various AI assistants that are MCP-compatible, stores data locally or syncs via your own cloud (e.g., iCloud, Google Drive), and is aimed at users who need advanced, encrypted long-term memory for their AI agents—ideal for privacy-conscious individuals, researchers, and developers building intelligent assistants.
Visit Cortex Memory Engine's official website for product details and getting started.