FormAct: Agentic Source Editing for Rich-Format Document Generation
Eugene Yu, Xingxing Zhang, Yuan Xia, Tao Ge, XWang, FNU Kartik, Vishwas Suryanarayanan, Cheng Yang
Abstract
Rich-format documents are essential for everyday operations yet costly to author, motivating the need for automated generation to enhance productivity. To this end, we present FormAct, an agentic system that generates professional rich-format documents from scratch. FormAct operates on an HTML source representation and performs iterative source refinement with an *editing agent* that invokes a suite of tools, including a syntax-aware source editor and a template retriever, and a *review agent* that critiques rendered pages to guide refinement. Additionally, we incorporate edit-triggered context compression to maintain a bounded working context and keep multi-round editing efficient. To support development and evaluation, we introduce RichDocBench for end-to-end generation, and RichDocFuzz to evaluate formatting-error recognition for reviewer agents. Through extensive automated evaluation and blind human-preference studies, we show that FormAct consistently outperforms strong baselines, including Codex-CLI, with particularly strong improvements in generating error-free, professional rich-format documents.
BibTeX
@inproceedings{
yu2026formact,
title={FormAct: Agentic Source Editing for Rich-Format Document Generation},
author={Eugene J. Yu and Xingxing Zhang and Yuan Xia and Tao Ge and Xun Wang and FNU Kartik and Vishwas Suryanarayanan and Cheng Yang and Amanda Jiang and Jiayu Ding and Xiangyu Wong and Tengchao Lv and Lei Cui and Si-Qing Chen and Furu Wei and Sujian Li},
booktitle={Forty-third International Conference on Machine Learning},
year={2026},
url={https://openreview.net/forum?id=n7Ta0YEcgw}
}