一句话：提出 Comfrey，在运行时缓解 LLM 使能软件中由 LLM/RAG 输出与组件要求冲突引发的集成失效。
论文关注集成 LLM agent 与检索增强生成（RAG）的软件中广泛存在的集成失效问题。其根源在于 LLM 输出不受限制，而软件组件对其输入有严格要求；二者不匹配时，即便 LLM/RAG 的响应看似正确，也可能触发软件的错误行为。针对这一问题，作者提出 Comfrey，目标是在运行时缓解此类集成失效。摘要仅给出问题定位与系统名称，未披露具体技术机制、实现细节以及评估方式与实验结果。
Due to the unrestricted outputs of LLMs and strict requirements of software components, integration failures are widespread in software that incorporates LLM agents and retrieval-augmented generation (RAG). Even seemingly correct LLM/RAG responses can trigger software misbehaviors if they violate these requirements.
@inproceedings{paperbot11856,
title = {Comfrey: Mitigating Integration Failures in LLM-enabled Software at Run-Time},
author = {Yuchen Shao and Yuheng Huang and Jiazhen Zou and Yuling Shi and Long Yang and Lei Ma and Ting Su and Chengcheng Wan},
booktitle = {ICSE},
year = {2026},
doi = {10.1145/3744916.3787847}
}