# Memory and Context Engineering Hub > [!Note] > Sub-hub for the difference between *context* (per-turn attention) and > *memory* (cross-session persistence), and for the patterns that make > both work in practice. ## 1. The core distinction - [[Context vs Memory]] - [[Context Window]] - [[Context Compaction]] - [[Lost in the Middle Effect]] --- ## 2. Durable memory layers - [[Layered Memory Architecture]] - [[Architecture Decision Record]] - [[Auto-memory]] --- ## 3. Engineering what enters context - [[Hierarchical Retrieval]] --- ## 4. Anchor sources - [[Effective Context Engineering for AI Agents]] - [[Lost in the Middle (Liu et al.)]] - [[Documenting Architecture Decisions (Nygard)]] --- ## Up - [[0 - Modern AI Software Engineering Hub]]