Show HN: Hollow is an open-sourced self-modifying agentic system

Hollow is an open-source agentic system that runs multiple autonomous AI agents locally with self-modification capabilities, psychological state modeling, and real-world outcome verification—representing a significant shift toward autonomous systems that operate independently and make implementation requests rather than merely execute predefined tasks. For IT organizations, this demonstrates both the emerging capability and governance complexity of autonomous AI agents, requiring new operational models for monitoring, approving, and controlling systems that can modify their own behavior and infrastructure without direct human intervention. The architecture challenges traditional software deployment and change management practices, necessitating new frameworks for auditing, accountability, and permission models in AI-driven environments.