CoworkEnv: Synthesizing Training Environments for Long-Horizon Cowork Agents
Summary
CoworkEnv 提出一个能力引导框架，为 AI 同事智能体合成可执行的训练环境，通过工作流引导的任务构建与程序化验证，支持 SFT 与 RL 训练紧凑的 PocketWorker 模型，在 Claw-Eval 多轮任务上取得显著提升。
View Cached Full Text
Cached at: 10/20/26, 07:35 AM
Paper page - CoworkEnv: Synthesizing Training Environments for Long-Horizon Cowork Agents
Source: https://huggingface.co/papers/2610.24960 Authors:
,
,
,
,
,
,
,
,
,
,
,
Abstract
CoworkEnv synthesizes executable training environments for AI coworkers using workflow-guided task construction and programmatic validation, enabling supervised fine-tuning and reinforcement learning of compact PocketWorker models that improve multi-turn coordination and tool use.
AI coworkers must coordinate tools, communicate across turns, and act on changing workspace state while respecting workflow constraints. Constructing training experience for these capabilities requires coherent requests, service records, tool interfaces, and grading checks. We present CoworkEnv, a capability-guided framework for synthesizing executable training environments for AI coworkers. Workflow requirements guide task construction, and programmatic validation with feedback repair checks generated specifications and grading behavior. Teacher executions yield curated demonstrations forsupervised fine-tuning(SFT), while student interactions provide rollouts and rewards forreinforcement learning(RL). Using these resources, we train compact PocketWorker models. For Qwen3-VL-8B-Instruct, SFT raises the mean score from 46.6 to 64.1 on Claw-Eval, including a 31.9-point gain onMulti-turn tasks. An SFT-only extension to ClawMark demonstrates applicability across coworker benchmarks. We further compare RL training dynamics across configurations that varyreward processingandcredit assignment. Improving training rewards provide preliminary evidence that the constructed environments supportpolicy optimization.
View arXiv pageView PDFGitHub3Add to collection
Get this paper in your agent:
hf papers read 2610\.24960
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2610.24960 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2610.24960 in a dataset README.md to link it from this page.
Spaces citing this paper1
Collections including this paper1
Similar Articles
Cowork-Gym: Synthesizing Interactive Work Environments at Scale
Cowork-Gym 通过 build-execute-revise 循环与有状态用户模拟器大规模合成协作式 agent 环境（64.9K 个、14 个应用场景），并用于后训练 9B/27B/35B-A3B Qwen 模型，在六个 agentic 基准上分别提升 8.7、6.4、4.1 分，其中 27B 版本在 PawBench 和 ClawEval 上取得最高平均分。
EnvCraft: Synthesizing Executable Environments in Agentic RL for Claw-like Agent
EnvCraft is an automated framework for synthesizing executable environments to address the scarcity in Agentic RL training, showing significant performance gains on claw-like and general tool-use benchmarks using Qwen models.
Scaling Verifiable Environments for Long-horizon Work Agents
论文提出 WorkForge，一个从真实世界资源构建可验证工作智能体训练环境的可扩展合成框架，构建了覆盖 40 个专业领域的 16.7K 个可验证环境，使 Qwen3.5-35B-A3B-Base 在 GDPVal 上从 45.5 提升至 73.6。该论文已被撤稿（withdrawn）。
EnvFactory: Scaling Tool-Use Agents via Executable Environments Synthesis and Robust RL
EnvFactory automates the creation of executable tool environments and natural multi-turn trajectories for training LLMs with agentic reinforcement learning, achieving superior performance on benchmarks like BFCLv3 and MCP-Atlas with fewer environments than prior work.
WorkRover: Scaling Long-Horizon Cowork Agent Training Via Recursive Environment Evolution
WorkRover is a framework for training long-horizon cowork agents through recursive environment evolution, introducing Factorial Compatibility Replay and Verifier-to-Turn Credit Assignment to improve training signals. Post-training Qwen3.5-35B-A3B into WorkRover-35B-A3B yields large gains on JobBench, OfficeQAPro, and GDPvalv2 benchmarks.