I built this because I was frustrated with traditional AI assistants that forget my context every time. I wanted something that understands my code (AGN), my learning (Vidnitive), and my emotions (Encura) as one integrated story.
What makes it different: 1. Runs locally on 8GB RAM (Rust + Python/MLX) 2. Monitors code and diary files simultaneously 3. Provides empathetic suggestions like "You're tired debugging calculate_score. Take a break?" 4. Transparent references (Humble UI principle)
Tech details: - Hub: Rust with notify crate (50MB) - Engine: Python with rule-based context fusion (1.2GB) - Total: ~1.4GB, leaving room for actual work
This is Phase 1. Next steps: - Phase 1.5: Terminal UI - Phase 2: Flutter app + E2EE sync - Phase 3: Full ecosystem integration
Open sourcing soon. Happy to answer questions!