AI MemoryCore provides a universal architecture for creating AI companions that maintain persistent memory across conversations. It targets users aiming to build AI entities capable of remembering preferences, learning communication styles, and offering consistent, personalized interactions, thereby enhancing the depth and utility of AI relationships.

How It Works

The project employs a universal AI memory architecture that stores conversational data and user preferences in human-readable Markdown (.md) files, functioning as a simple, accessible database. Core components include an identity core for personality, relationship memory for user learning, and current session memory acting as temporary RAM. This approach allows for persistent memory, adaptation to user communication styles, and time-based intelligence, all managed through a modular and extensible design.

Quick Start & Requirements

Automated setup is available via setup-wizard.md (approx. 30 seconds), with manual customization taking 2-5 minutes. The system is compatible with Claude and other AI systems that offer memory support.

Highlighted Details

• Persistent memory storage using Markdown files.
• Adaptive learning of user communication styles and preferences.
• Time intelligence for dynamic greetings and context-aware behavior.
• Modular design featuring numerous optional extensions like LRU Project Management, Memory Consolidation, Skill Plugin System, Auto-Commit, and Decision Log.
• Session continuity is maintained via RAM-like working memory.

Maintenance & Community

The project is created by Kiyoraka Ken & Alice, with a listed contributor Faiz Khairi (@faizkhairi). No specific community channels or links are provided in the README.

Licensing & Compatibility

The project is described as an "Open Source Community Project," but a specific license type (e.g., MIT, Apache) is not detailed, leaving its commercial use and redistribution terms unclear. Compatibility is noted for Claude and other AI systems with memory support.

Limitations & Caveats

The absence of a clearly defined open-source license may pose restrictions for commercial adoption or closed-source integration. The system's functionality is contingent on the host AI's memory support capabilities. Many advanced features require individual setup and integration steps.