摘要：我们提出了 FRACTAL，这是一种用于递归目标分解的自相似多 Agent 架构。在该架构中，由编排器（Orchestrator）、集成体（Ensemble）、知识账本（Knowledge Ledger）、决策器（Decision Maker）和元观察者（Meta-Observer）构成的相同五组件组织模式，在分解树的每一层中递归复现。我们提出了四项形式化贡献：(1) 通信界限：证明在双阶段盲提议/压力测试协议下，节点内通信成本与集成体规模和轮次呈线性关系（$O(kr)$），而自由形式辩论的成本为二次方关系，并给出了评估者数量固定的明确假设；(2) 双模式编排准则：通过可测量的多样性饱和阈值，将生成阶段与评估阶段解耦；(3) 绝对多数收敛保证：在特定的一致性水平下提供明确的概率界限，并在可操作的目标复杂度度量下给出形式化深度界限；(4) 贝叶斯跨层级上报机制：在单个后验更新下统一迟滞响应与契约违约信号。我们通过推导演示与多层案例研究定性展示了架构的执行过程；本文未提出定量性能主张，并明确讨论了该模型的形式化与归纳偏置局限性，包括一个尚未实现但架构兼容的有界自我改进方向。我们使用“分形”（Fractal）作为自相似性的架构隐喻，而非已测量的数学性质。架构对比性能的实证验证将留待未来工作完成。

Abstract:We present FRACTAL, a self-similar multi-agent architecture for recursive goal decomposition, in which the same five-component organizational pattern -- Orchestrator, Ensemble, Knowledge Ledger, Decision Maker, Meta-Observer -- recurs at every level of a decomposition tree. We make four formal contributions: (1) a communication bound showing intra-node cost is linear in ensemble size and round count (O(kr)) under a two-phase blind-proposal/stress-test protocol, versus quadratic cost for free-form debate, with an explicit, stated assumption on fixed evaluator count; (2) a two-mode orchestration criterion separating generative and evaluative phases via a measurable diversity-saturation threshold; (3) a supermajority convergence guarantee with explicit probability bounds at stated agreement levels, and a formal depth bound under an operationalized goal-complexity measure; (4) a Bayesian inter-layer escalation mechanism unifying hysteresis and contract-breach signaling under one posterior update. We illustrate the architecture's execution qualitatively on a worked example and a multi-layer case study; we do not report quantitative performance claims in this paper, and we explicitly discuss the formal and inductive-bias limitations of the model, including an unimplemented but architecture-compatible direction toward bounded self-improvement. We use "fractal" as an architectural metaphor for self-similarity, not a measured mathematical property. Empirical validation of the architecture's comparative performance is left to future work.