ScaffoldAgent: Utility-Guided Dynamic Outline Optimization for Open-Ended Deep Research
This work addresses the problem of scaffold drift in open-ended deep research for systems that generate long-form reports, offering a structured decision process for outline evolution.
ScaffoldAgent introduces a utility-guided dynamic outline optimization framework for open-ended deep research, improving long-form report generation and factual grounding over existing methods on DeepResearch Bench and DeepResearch Gym.
Open-ended deep research (OEDR) requires systems to acquire knowledge through multi-round retrieval and generate coherent long-form reports. The outline plays a central role as a structural scaffold that coordinates retrieval, evidence organization, and generation. However, existing methods either fix the outline before writing or refine it with local heuristics, leading to scaffold drift under continuous information accumulation and delayed feedback for evaluating outline modifications. We propose ScaffoldAgent, a utility-guided dynamic outline optimization framework for OEDR. ScaffoldAgent models outline evolution as a structured decision process with three operations: Expansion, Contraction, and Revision, enabling controlled updates to the report scaffold. It further introduces a utility-guided feedback mechanism that estimates the downstream value of each outline operation from retrieval gain, structural coherence, and trial-generation quality. The resulting utility signal guides node selection, operation scheduling, and termination during inference. Experiments on DeepResearch Bench and DeepResearch Gym show that ScaffoldAgent consistently improves long-form report generation and factual grounding over existing deep research agents.