StackPlanner: A Centralized Hierarchical Multi-Agent System with Task-Experience Memory Management

arXiv:2601.05890v2 Announce Type: replace Abstract: Multi-agent systems based on large language models, particularly centralized architectures, have recently shown strong potential for complex and knowledge-intensive tasks. However, central agents often suffer from unstable long-horizon collaboration due to the lack of memory management, leading to context bloat, error accumulation, and poor cross-task generalization. To address both task-level memory inefficiency and the inability to reuse coor