Story2Proposal: A Scaffold for Structured Scientific Paper Writing
Abstract
Generating scientific manuscripts requires maintaining alignment between narrative reasoning, experimental evidence, and visual artifacts across the document lifecycle. Existing language-model generation pipelines rely on unconstrained text synthesis with validation applied only after generation, often producing structural drift, missing figures or tables, and cross-section inconsistencies. We introduce Story2Proposal, a contract-governed multi-agent framework that converts a research story into a structured manuscript through coordinated agents operating under a persistent shared visual contract. The system organizes architect, writer, refiner, and renderer agents around a contract state that tracks section structure and registered visual elements, while evaluation agents supply feedback in a generate evaluate adapt loop that updates the contract during generation. Experiments on tasks derived from the Jericho research corpus show that Story2Proposal achieved an expert evaluation score of 6.145 versus 3.963 for DirectChat (+2.182) across GPT, Claude, Gemini, and Qwen backbones. Compared with the structured generation baseline Fars, Story2Proposal obtained an average score of 5.705 versus 5.197, indicating improved structural consistency and visual alignment.
Cite
@article{arxiv.2603.27065,
title = {Story2Proposal: A Scaffold for Structured Scientific Paper Writing},
author = {Zhuoyang Qian and Wei Shi and Xu Lin and Li Ling and Meng Luo and Ziming Wang and Zhiwei Zhang and Tengyue Xu and Gaoge Liu and Zhentao Zhang and Shuo Zhang and Ziqi Wang and Zheng Feng and Yan Luo and Shu Xu and Yongjin Chen and Zhibo Feng and Zhuo Chen and Bruce Yuan and Biao Wu and Harry Wang and Kris Chen},
journal= {arXiv preprint arXiv:2603.27065},
year = {2026}
}
Comments
10 pages, 4 figures,