Toward Lifecycle Governance for Multi-Agent Systems: Principles and Research Agenda

Position paper proposing four principles for lifecycle governance in multi-agent systems: maturation before reproduction, selective knowledge transfer, reallocation over termination, and trackable identity. Identifies open challenges including decentralized knowledge stores, adversarially robust valuation, regulatory alignment, and empirical validation.