Residuum | Agentic AI with continuous context

Residuum is a personal AI agent framework that eliminates session boundaries by maintaining continuous memory across all interactions, compressing conversation history into a dense observation log that stays in context. Unlike traditional AI agents that treat conversations as isolated, Residuum allows users to seamlessly switch between channels like CLI, Discord, and webhooks while maintaining the same continuous thread. The system uses structured YAML scheduling for proactive tasks instead of constantly burning tokens, and features two-tier memory compression that keeps recent history accessible without retrieval delays.