💻 SETA: Scaling Environments for Terminal Agents - Environments

Toonflow 是开源一站式 AI 短剧创作工具，将小说、剧本快速转化为动画短剧。集成 AI 编剧、智能分镜、角色与视频生成，跨平台桌面端轻量部署，助力创作者低成本批量产出视觉内容。Toonflow is an open-source AI tool that turns stories and scripts into animated short dramas. Features AI…

The lightweight framework for building agents

Benchmarking Agents on Real-World Terminal Tasks

OmniDance

Code for the paper "Self-Compacting Language Model Agents"

The roadmap of long-horizon agents

Infinite Interactive World Rollout on a Single Desktop GPU

🔥 Quo Vadis, World Modeling? Towards Interactive World Proxies for Continually Improving Agents

Qwen-CUA: Native Computer Use for (Almost) Everything — a screenshot-driven agent that operates computers with keyboard and mouse, jointly developed by the Qwen Team and XLang Lab.

🐧 Harness for RSI. AI Build Better AI

The long-horizon computer-use harness. Run AI agents across desktop apps and the CLI for extended periods while preserving task state and making reliable progress on complex workflows. Features fre…

Long Horizon Terminal Benchmark with Dense Reward Grading

Agents' Last Exam

A benchmark for general-purpose terminal-use agents.

WeaveBench: A Long-Horizon, Real-World Benchmark for Computer-Use Agents with Hybrid Interfaces

Data recipes and robust infrastructure for training AI agents

A Unified Memory Benchmark Suite for Memory-Augmented Agents

Automate the build, execution and test of GitHub repositories across programming languages and operating systems.

Training terminal-agents

Official Implementation of "CLI-Gym: Scalable CLI Task Generation via Agentic Environment Inversion"

MiMo Code: Where Models and Agents Co-Evolve

Resources and paper list for 'Scaling Environments for Agents'. This repository accompanies our survey on how environments contribute to agent intelligence.