memory-manager

Overview: The Memory Manager skill provides a professional-grade memory architecture for AI agents, ensuring never to lose context again. It implements a three-tier memory system for compression detection, auto-snapshots, and semantic search. Key Features: - Compression detection and auto-snapshots - Semantic search across episodic, semantic, and procedural memories - Three-tier memory system: episodic, semantic, and procedural - Manual categorization and override available How It Works: The skill monitors all memory types, estimates total context usage, and detects compression risk. It organizes memories automatically and provides manual override options. The retrieval strategy includes time-based search, topic-based search, and workflow lookup. Use Cases: - Detect compression risk before memory loss - Save context snapshots - Search historical memories - Track memory usage patterns