iflow-mcp_savantskie-persistent-ai-memory

# Persistent AI Memory System [](https://opensource.org/licenses/MIT) [](https://www.python.org/downloads/) > 🌟 **Community Call to Action**: Have you made improvements or additions to this system? We want to include your work! Every contributor will be properly credited in the final product. Whether it's bug fixes, new features, or documentation improvements - your contributions matter and will help shape the future of AI memory systems. Submit a pull request today! **GITHUB LINK** - https://github.com/savantskie/persistent-ai-memory.git --- 🆕 **Recent Changes (2025-09-04)** - **🧠 Enhanced Embedding System**: Implemented intelligent embedding service with primary/fallback providers - **Preservation Strategy**: All existing embeddings (15,000+) are automatically preserved - **LM Studio Primary**: High-quality embeddings for new content via LM Studio - **Ollama Fallback**: Fast local embeddings when LM Studio unavailable - **Real-time Provider Monitoring**: Automatic availability detection and graceful fallback - **⚙️ Standardized Configuration**: Added `embedding_config.json` for easy provider management - **📁 Improved Organization**: Moved all test files to proper `tests/` folder structure - **🔧 Enhanced Core Systems**: Updated `ai_memory_core.py` with intelligent provider selection - **🛡️ Backward Compatibility**: All existing functionality preserved while adding new capabilities - **📊 Performance Optimization**: Better semantic search quality with preserved data integrity Previous Changes (2025-09-01): - Updated `ai-memory-mcp_server.py` to include enhanced tool registration logic for `update_memory` and other tools. - Improved MCP server functionality to dynamically detect and register tools based on client context. - Added robust error handling and logging for tool execution. - Enhanced automatic maintenance t…