AOrchestra：为代理编排自动化子代理创建
AOrchestra: Automating Sub-Agent Creation for Agentic Orchestration
AI总结：
AOrchestra通过统一的代理抽象实现自动化子代理创建，提升多轮任务解决的适应性和效率。
摘要：
语言代理在任务自动化方面展现了强大的潜力。为了实现这一承诺，针对日益复杂和长周期任务，出现了以子代理为工具的多轮任务解决范式。然而，现有设计仍然缺乏对子代理的动态抽象视图，从而影响了适应性。我们通过一种统一的、框架无关的代理抽象来解决这一挑战，该抽象将任何代理建模为一个元组：指令、上下文、工具、模型。这个元组充当能力的组合配方，使系统能够按需生成专门的执行器。基于这一抽象，我们引入了一个代理系统AOrchestra，其中中央协调器在每一步都将元组具体化：它精选与任务相关的上下文，选择工具和模型，并通过即时自动代理创建进行委托。这种设计减少了人类工程努力，同时保持了框架无关性，并支持多样代理作为任务执行器的即插即用。它还实现了可控的性能-成本权衡，使系统能够接近帕累托最优。在三个具有挑战性的基准测试（GAIA、SWE-Bench、Terminal-Bench）中，AOrchestra在与Gemini-3-Flash配对时，相对于最强基线实现了16.28%的相对改进。代码可在：https://github.com/FoundationAgents/AOrchestra上获得。
英文摘要：
Language agents have shown strong promise for task automation. Realizing this promise for increasingly complex, long-horizon tasks has driven the rise of a sub-agent-as-tools paradigm for multi-turn task solving. However, existing designs still lack a dynamic abstraction view of sub-agents, thereby hurting adaptability. We address this challenge with a unified, framework-agnostic agent abstraction that models any agent as a tuple Instruction, Context, Tools, Model. This tuple acts as a compositional recipe for capabilities, enabling the system to spawn specialized executors for each task on demand. Building on this abstraction, we introduce an agentic system AOrchestra, where the central orchestrator concretizes the tuple at each step: it curates task-relevant context, selects tools and models, and delegates execution via on-the-fly automatic agent creation. Such designs enable reducing human engineering efforts, and remain framework-agnostic with plug-and-play support for diverse agents as task executors. It also enables a controllable performance-cost trade-off, allowing the system to approach Pareto-efficient. Across three challenging benchmarks (GAIA, SWE-Bench, Terminal-Bench), AOrchestra achieves 16.28% relative improvement against the strongest baseline when paired with Gemini-3-Flash. The code is available at: https://github.com/FoundationAgents/AOrchestra
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
