agent-conductor-topology-evolution

【效率与协作】Optimize multi-agent collaboration by learning task-specific interaction topologies. Use an LLM orchestrator to generate layered DAG topologies that adapt to inferred problem difficulty, treating agent interactions as a learned graph structure rather than fixed patterns.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准