10 upvotes · 17 November 2026

From Repositories to Training Environments: Scaling Long-Horizon Coding Agents with Environment Evolution Loop

Yang Tian, Tianyu Zheng, Xiyuan Yang, Yaxin Du, Chengran Yang, Bo Zhao, Jian Yang, Bryan Dai

This paper creates a system to train coding agents for long-term tasks by generating environments that provide timely and reliable feedback, and shows that this approach can significantly improve training results. Practitioners working on coding agent development might care about how to design more effective training environments.

Abstract

Training coding agents for long-horizon tasks requires environments with verifiable feedback. Existing pipelines often follow a linear process that ends with validation, establishing usability but providing little guidance on improving training value. Recursive self-improvement (RSI) often requires model rollouts or model training followed by downstream evaluation, with costs and feedback delays amplified on long-horizon tasks. How can we use timely, reliable proxy signals to improve training value and synthesize environments at scale? We present RepoFactory, an automated pipeline that converts real repositories into verifiable environments where agents build complete projects from scratch given a language specification. Its Environment Evolution Loop optimizes a multiplicative objective combining validity, structural coupling, and implementation scope through proxy feedback. RepoFactory produces 10,159 environments across 17 software domains. Training Qwen3.5-35B-A3B on these environments yields RepoHorizen-35B-A3B, which achieves gains of 22.6 percentage points on NL2Repo-Bench and 8.5 points on average across five benchmarks covering long-horizon coding and agentic tasks in real-world production settings, compared with 17.5 and 6.4 points without environment evolution. These results highlight the training value of RepoFactory environments and the additional gains from the Environment Evolution Loop.

View on Hugging Face →