[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-73589":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":23,"hasPages":23,"topics":25,"createdAt":10,"pushedAt":10,"updatedAt":43,"readmeContent":44,"aiSummary":45,"trendingCount":16,"starSnapshotCount":16,"syncStatus":46,"lastSyncTime":47,"discoverSource":48},73589,"cua","trycua\u002Fcua","trycua","Open-source infrastructure for Computer-Use Agents. Sandboxes, SDKs, and benchmarks to train and evaluate AI agents that can control full desktops (macOS, Linux, Windows).","https:\u002F\u002Fcua.ai",null,"HTML",17813,1146,62,131,0,108,426,784,324,119.18,"MIT License",false,"main",[26,27,28,29,30,31,5,32,33,34,35,36,37,38,39,40,41,42],"agent","ai-agent","apple","computer-use","computer-use-agent","containerization","desktop-automation","hacktoberfest","lume","macos","manus","operator","swift","virtualization","virtualization-framework","windows","windows-sandbox","2026-06-12 04:01:10","\u003Cdiv align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fcua.ai\" target=\"_blank\" rel=\"noopener noreferrer\">\n    \u003Cpicture>\n      \u003Csource media=\"(prefers-color-scheme: dark)\" alt=\"Cua logo\" width=\"150\" srcset=\"img\u002Flogo_white.svg\">\n      \u003Csource media=\"(prefers-color-scheme: light)\" alt=\"Cua logo\" width=\"150\" srcset=\"img\u002Flogo_black.svg\">\n      \u003Cimg alt=\"Cua logo\" width=\"150\" src=\"img\u002Flogo_black.svg\">\n    \u003C\u002Fpicture>\n  \u003C\u002Fa>\n\n  \u003Cp align=\"center\">Build, benchmark, and deploy agents that use computers\u003C\u002Fp>\n\n  \u003Cp align=\"center\">\n    \u003Ca href=\"https:\u002F\u002Fcua.ai\" target=\"_blank\" rel=\"noopener noreferrer\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fcua.ai-0ea5e9\" alt=\"cua.ai\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002FmVnXXpdE85\" target=\"_blank\" rel=\"noopener noreferrer\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDiscord-Join%20Server-10b981?logo=discord&logoColor=white\" alt=\"Discord\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fx.com\u002Ftrycua\" target=\"_blank\" rel=\"noopener noreferrer\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Ftwitter\u002Ffollow\u002Ftrycua?style=social\" alt=\"Twitter\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fcua.ai\u002Fdocs\" target=\"_blank\" rel=\"noopener noreferrer\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDocs-0ea5e9.svg\" alt=\"Documentation\">\u003C\u002Fa>\n    \u003Cbr>\n\u003Ca href=\"https:\u002F\u002Ftrendshift.io\u002Frepositories\u002F13685\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Ftrendshift.io\u002Fapi\u002Fbadge\u002Frepositories\u002F13685\" alt=\"trycua%2Fcua | Trendshift\" style=\"width: 250px; height: 55px;\" width=\"250\" height=\"55\"\u002F>\u003C\u002Fa>\n  \u003C\u002Fp>\n\n\u003C\u002Fdiv>\n\n## Choose Your Path\n\n\u003Cdiv align=\"center\">\n  \u003Ctable>\n    \u003Ctr>\n      \u003Ctd colspan=\"3\" align=\"center\">\n        \u003Ca href=\"#cua-driver---background-computer-use-on-macos\">\n          \u003Cimg src=\"img\u002Fcard-cua-driver.png\" alt=\"Cua Driver — The background computer-use agent\" width=\"888\">\n        \u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd align=\"center\">\n        \u003Ca href=\"#cua---agentic-ui-automation--code-execution\">\n          \u003Cimg src=\"img\u002Fcard-cua-sandbox.png\" alt=\"Cua & Cua Sandbox\" width=\"280\">\n        \u003C\u002Fa>\n      \u003C\u002Ftd>\n      \u003Ctd align=\"center\">\n        \u003Ca href=\"#cua-bench---benchmarks--rl-environments\">\n          \u003Cimg src=\"img\u002Fcard-cua-bench.png\" alt=\"Cua Bench\" width=\"280\">\n        \u003C\u002Fa>\n      \u003C\u002Ftd>\n      \u003Ctd align=\"center\">\n        \u003Ca href=\"#lume---macos-virtualization\">\n          \u003Cimg src=\"img\u002Fcard-cua-lume.png\" alt=\"Lume\" width=\"280\">\n        \u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftable>\n\u003C\u002Fdiv>\n\n---\n\n## Cua Driver - Background computer-use on macOS\n\nDrive any native macOS app **in the background** — agents click, type, and verify without stealing the cursor, focus, or Space, even on non-AX surfaces like Chromium web content and canvas-based tools (Blender, Figma, DAWs, game engines). Use with the CLI or MCP server for Claude Code, Cursor, and custom clients. Every session records as a replayable trajectory.\n\n```sh\n\u002Fbin\u002Fbash -c \"$(curl -fsSL https:\u002F\u002Fraw.githubusercontent.com\u002Ftrycua\u002Fcua\u002Fmain\u002Flibs\u002Fcua-driver\u002Fscripts\u002Finstall.sh)\"\n```\n\n> Want to try the cross-platform Rust port early? Add `-- --experimental-rust` to the line above — it delegates to the [`cua-driver-rs`](libs\u002Fcua-driver-rs\u002F) installer (separate bundle, coexists with the Swift binary).\n\nFull tool reference, architecture notes, and the Claude Code skill ship with the package: [`libs\u002Fcua-driver\u002FREADME.md`](libs\u002Fcua-driver\u002FREADME.md).\n\n---\n\n## Cua - Agent-Ready Sandboxes for Any OS\n\nBuild agents that see screens, click buttons, and complete tasks autonomously. One API for any VM or container image — cloud or local.\n\n```sh\npip install cua\n```\n\n\u003C!-- \u003Cimg src=\"img\u002Fcua-architecture.png\" alt=\"Cua Architecture\" width=\"100%\"> -->\n\n```python\n# Requires Python 3.11 or later\nfrom cua import Sandbox, Image\n\n# Same API regardless of OS or runtime\nasync with Sandbox.ephemeral(Image.linux()) as sb:   # or .macos() .windows() .android()\n    result = await sb.shell.run(\"echo hello\")\n    screenshot = await sb.screenshot()\n    await sb.mouse.click(100, 200)\n    await sb.keyboard.type(\"Hello from Cua!\")\n    await sb.mobile.gesture((100, 500), (100, 200))  # multi-touch gestures\n```\n\n|                    | Linux container | Linux VM | macOS | Windows | Android | BYOI (.qcow2, .iso) |\n| ------------------ | --------------- | -------- | ----- | ------- | ------- | ------------------- |\n| **Cloud (cua.ai)** | ✅              | ✅       | ✅    | ✅      | ✅      | 🔜 soon             |\n| **Local (QEMU)**   | ✅              | ✅       | ✅    | ✅      | ✅      | ✅                  |\n\n**[Get Started](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcua\u002Fguide\u002Fget-started\u002Fset-up-sandbox)** | **[Examples](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcua\u002Fexamples)** | **[API Reference](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcua\u002Freference\u002Fagent-sdk)**\n\n---\n\n## CuaBot - Co-op computer-use for any agent\n\n\u003Cdiv align=\"center\">\n  \u003Cimg src=\"img\u002Fcuabot-screenshot.png\" alt=\"cuabot screenshot\" width=\"720\">\n\u003C\u002Fdiv>\n\n`cuabot` gives any coding agent a seamless sandbox for computer-use. Individual windows appear natively on your desktop with H.265, shared clipboard, and audio.\n\n```bash\nnpx cuabot                 # Setup onboarding\n```\n\n```bash\n# Run any agent in a sandbox\ncuabot claude              # Claude Code\ncuabot openclaw            # OpenClaw in the sandbox\n\n# Run any GUI workflow in a sandbox\ncuabot chromium\ncuabot --screenshot\ncuabot --type \"hello\"\ncuabot --click \u003Cx> \u003Cy> [button]\n```\n\nBuilt-in support for `agent-browser` and `agent-device` (iOS, Android) out of the box.\n\n\u003Cdiv align=\"center\">\n\n**[Get Started](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcuabot\u002Fguide\u002Fgetting-started\u002Fintroduction)** | **[Installation](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcuabot\u002Fguide\u002Fgetting-started\u002Finstallation)** | First spotted at [ClawCon](https:\u002F\u002Fwww.claw-con.com\u002F)\n\n\u003Cimg height=\"64\" alt=\"cuaXclawdbot_nbg\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F8b92237d-6e9b-4b3a-ae9a-b3560622ec1d\" \u002F>\n\n\u003C\u002Fdiv>\n\n---\n\n## Cua-Bench - Benchmarks & RL Environments\n\nEvaluate computer-use agents on OSWorld, ScreenSpot, Windows Arena, and custom tasks. Export trajectories for training.\n\n\u003C!-- \u003Cimg src=\"img\u002Fcua-bench-architecture.png\" alt=\"Cua-Bench Architecture\" width=\"100%\"> -->\n\n```bash\n# Install and create base image\ncd cua-bench\nuv tool install -e . && cb image create linux-docker\n\n# Run benchmark with agent\ncb run dataset datasets\u002Fcua-bench-basic --agent cua-agent --max-parallel 4\n```\n\n**[Get Started](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcuabench\u002Fguide\u002Fgetting-started\u002Ffirst-steps)** | **[Partner With Us](https:\u002F\u002Fcuabench.ai\u002F)** | **[Registry](https:\u002F\u002Fcuabench.ai\u002Fregistry)** | **[CLI Reference](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcuabench\u002Freference\u002Fcli-reference)**\n\n---\n\n## Lume - macOS Virtualization\n\nCreate and manage macOS\u002FLinux VMs with near-native performance on Apple Silicon using Apple's Virtualization.Framework.\n\n\u003C!-- \u003Cimg src=\"img\u002Flume-architecture.png\" alt=\"Lume Architecture\" width=\"100%\"> -->\n\n```bash\n# Install Lume\n\u002Fbin\u002Fbash -c \"$(curl -fsSL https:\u002F\u002Fraw.githubusercontent.com\u002Ftrycua\u002Fcua\u002Fmain\u002Flibs\u002Flume\u002Fscripts\u002Finstall.sh)\"\n\n# Pull & start a macOS VM\nlume run macos-sequoia-vanilla:latest\n```\n\n**[Get Started](https:\u002F\u002Fcua.ai\u002Fdocs\u002Flume)** | **[FAQ](https:\u002F\u002Fcua.ai\u002Fdocs\u002Flume\u002Fguide\u002Fgetting-started\u002Ffaq)** | **[CLI Reference](https:\u002F\u002Fcua.ai\u002Fdocs\u002Flume\u002Freference\u002Fcli-reference)**\n\n---\n\n## Packages\n\n| Package                                                                     | Description                                                |\n| --------------------------------------------------------------------------- | ---------------------------------------------------------- |\n| [cuabot](https:\u002F\u002Fdocs.trycua.com\u002Fcuabot\u002Fguide\u002Fgetting-started\u002Fintroduction) | Multi-agent computer-use sandbox CLI                       |\n| [cua-agent](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcua\u002Freference\u002Fagent-sdk)                    | AI agent framework for computer-use tasks                  |\n| [cua-sandbox](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcua\u002Freference\u002Fsandbox-sdk)                | SDK for creating and controlling sandboxes                 |\n| [cua-computer-server](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcua\u002Freference\u002Fsandbox-sdk)        | Driver for UI interactions and code execution in sandboxes |\n| [cua-bench](https:\u002F\u002Fcua.ai\u002Fdocs\u002Fcuabench)                                   | Benchmarks and RL environments for computer-use            |\n| [lume](https:\u002F\u002Fcua.ai\u002Fdocs\u002Flume)                                            | macOS\u002FLinux VM management on Apple Silicon                 |\n| [lumier](https:\u002F\u002Fcua.ai\u002Fdocs\u002Flume\u002Fguide\u002Fadvanced\u002Flumier)                    | Docker-compatible interface for Lume VMs                   |\n\n## Resources\n\n- [Documentation](https:\u002F\u002Fcua.ai\u002Fdocs) — Guides, examples, and API reference\n- [Blog](https:\u002F\u002Fwww.cua.ai\u002Fblog) — Tutorials, updates, and research\n- [Discord](https:\u002F\u002Fdiscord.com\u002Finvite\u002FmVnXXpdE85) — Community support and discussions\n- [GitHub Issues](https:\u002F\u002Fgithub.com\u002Ftrycua\u002Fcua\u002Fissues) — Bug reports and feature requests\n\n## Contributing\n\nWe welcome contributions! See our [Contributing Guidelines](CONTRIBUTING.md) for details.\n\n## License\n\nMIT License — see [LICENSE](LICENSE.md) for details.\n\nThird-party components have their own licenses:\n\n- [Kasm](libs\u002Fkasm\u002FLICENSE) (MIT)\n- [OmniParser](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FOmniParser\u002Fblob\u002Fmaster\u002FLICENSE) (CC-BY-4.0)\n- Optional `cua-agent[omni]` includes ultralytics (AGPL-3.0)\n\n## Trademarks\n\nApple, macOS, Ubuntu, Canonical, and Microsoft are trademarks of their respective owners. This project is not affiliated with or endorsed by these companies.\n\n---\n\n\u003Cdiv align=\"center\">\n\n[![Stargazers over time](https:\u002F\u002Fstarchart.cc\u002Ftrycua\u002Fcua.svg?variant=adaptive)](https:\u002F\u002Fstarchart.cc\u002Ftrycua\u002Fcua)\n\nThank you to all our [GitHub Sponsors](https:\u002F\u002Fgithub.com\u002Fsponsors\u002Ftrycua)!\n\n\u003Cimg width=\"300\" alt=\"coderabbit-cli\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F23a98e38-7897-4043-8ef7-eb990520dccc\" \u002F>\n\n\u003C\u002Fdiv>\n","Cua 是一个开源基础设施项目，旨在支持计算机使用代理的开发、训练和评估。它提供沙箱环境、SDK 以及基准测试工具，使得AI代理能够控制完整的桌面操作系统（包括macOS、Linux和Windows）。核心功能包括在后台驱动任何原生macOS应用程序，执行点击、输入等操作而不干扰用户当前工作，特别适用于非AX界面如基于Chromium的内容或图形设计软件。此外，Cua还支持跨平台的Rust版本以扩展其兼容性。该项目非常适合需要自动化桌面任务或希望研究人机交互领域的开发者与研究人员使用。",2,"2026-06-11 03:46:16","high_star"]