Reliable Docker-based environment construction is a dominant bottleneck for scaling execution-grounded training and evaluation of software engineering agents. We introduce DockSmith, a specialized agentic Docker builder designed to address this challenge. DockSmith treats environment construction not only as a preprocessing step, but as a core agentic capability that exercises long-horizon tool use, dependency reasoning, and failure recovery, yielding supervision that transfers beyond Docker building itself. DockSmith is trained on large-scale, execution-grounded Docker-building trajectories produced by a SWE-Factory-style pipeline augmented with a loop-detection controller and a cross-task success memory. Training a 30B-A3B model on these trajectories achieves open-source state-of-the-art performance on Multi-Docker-Eval, with 39.72% Fail-to-Pass and 58.28% Commit Rate. Moreover, DockSmith improves out-of-distribution performance on SWE-bench Verified, SWE-bench Multilingual, and Terminal-Bench 2.0, demonstrating broader agentic benefits of environment construction.
@article{arxiv.2602.00592,
title = {DockSmith: Scaling Reliable Coding Environments via an Agentic Docker Builder},
author = {Jiaran Zhang and Luck Ma and Fanqi Wan and Di Qi and Xu Zhao and Jieyi Hou and Zhe Xie and Mengqiang Ren and Xin Wu and Zhewei Huang and Liangyu Chen and Qi Han and Xiangyu Zhang},
journal= {arXiv preprint arXiv:2602.00592},
year = {2026}
}