[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-74682":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":14,"stars7d":16,"stars30d":17,"stars90d":15,"forks30d":15,"starsTrendScore":18,"compositeScore":19,"rankGlobal":9,"rankLanguage":9,"license":20,"archived":21,"fork":21,"defaultBranch":22,"hasWiki":23,"hasPages":21,"topics":24,"createdAt":9,"pushedAt":9,"updatedAt":25,"readmeContent":26,"aiSummary":27,"trendingCount":15,"starSnapshotCount":15,"syncStatus":28,"lastSyncTime":29,"discoverSource":30},74682,"OpenSquirrel","Infatoshi\u002FOpenSquirrel","Infatoshi","For people who get distracted by agents. A native Rust\u002FGPUI control plane for running Claude Code, Codex, Cursor, and OpenCode side by side — because if you're going to be squirrely, you might as well optimize for it.",null,"Rust",1362,83,8,1,0,5,9,3,18.77,"MIT License",false,"main",true,[],"2026-06-12 02:03:26","\u003Cp align=\"center\">\n  \u003Cimg src=\"assets\u002Flogo.png\" width=\"128\" alt=\"OpenSquirrel\">\n\u003C\u002Fp>\n\n\u003Ch1 align=\"center\">OpenSquirrel\u003C\u002Fh1>\n\n\u003Cp align=\"center\">\n  \u003Cstrong>This project is no longer actively maintained.\u003C\u002Fstrong>\u003Cbr>\n  Fork it, customize it, make it yours.\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"assets\u002Fscreenshot-main.png\" width=\"800\" alt=\"OpenSquirrel screenshot\">\n\u003C\u002Fp>\n\n## What it was\n\nA native, GPU-rendered tiling manager for AI coding agents. Rust + GPUI. Run Claude Code, Codex, Cursor, Gemini, and OpenCode side by side with automatic sub-agent delegation, remote machine targeting via SSH, and persistent sessions.\n\n## Why it's archived\n\nAfter a week of building this, I arrived at a simpler conclusion: **you don't need a custom GUI to orchestrate AI agents.**\n\nHere's what I learned:\n\n**The terminal already won.** Every AI coding CLI (Claude Code, Codex, Cursor Agent, Gemini CLI, OpenCode) ships with a polished terminal TUI. Building a Rust GUI that parses their JSON output and re-renders it will always be worse than just... using the native TUI. Users are comfortable in their terminal. They don't want a new window.\n\n**Delegation is a prompt, not a product.** The entire coordinator\u002Fworker delegation system -- spawning sub-agents across runtimes, collecting results, feeding them back -- can be done with 4 lines in a `CLAUDE.md` file telling the model to use `cursor agent --print` or `codex exec` via Bash. No orchestration daemon needed. No hooks. No middleware. Claude Code's Agent tool already handles internal delegation. For external CLIs, just run them headless.\n\n**Token tracking already exists.** [CodexBar](https:\u002F\u002Fgithub.com\u002Fsteipete\u002FCodexBar) sits in your macOS menu bar and tracks usage across Claude, Codex, Cursor, Gemini, and more by reading their local data files. No need to build this into a GUI.\n\n**The architectural mismatch.** Using Claude Code (a Node\u002FBun process) to build and iterate on a Rust GPU application through JSON stream parsing is a bizarre feedback loop. The model is trained on terminal interactions, not on debugging GPUI render pipelines. Every feature took 10x longer than it should have because the tooling fought the workflow.\n\n**Models aren't good enough yet for opinionated UX.** Nobody knows the right workflow for multi-agent coding. Building a rigid UI around one workflow locks you in. The terminal is infinitely flexible. Wait for patterns to emerge before building products around them.\n\n## What works instead\n\nThe setup I actually use now:\n\n- **Terminal**: Ghostty (or whatever you prefer)\n- **Agents**: Run them directly -- `claude`, `codex`, `cursor agent`, `gemini`\n- **Delegation**: Instructions in `~\u002F.claude\u002FCLAUDE.md` telling Claude to run external CLIs via Bash when asked\n- **Token tracking**: CodexBar (menu bar app, reads local files)\n- **Multi-agent**: Just open multiple terminal tabs\u002Fpanes\n\nThat's it. No custom software. The orchestration layer is a config file.\n\n## If you want to use or fork this\n\nThe code works. The `legacy` branch at commit `78f1bf2` has the full feature set:\n\n- Multi-agent grid with auto-layout\n- Coordinator\u002Fworker delegation across runtimes\n- Remote SSH targeting with tmux session persistence\n- Reusable UI components (FuzzyList, modal builders, selectable rows)\n- Model picker with fuzzy search (Cmd+M)\n- Token\u002Fcost tracking per agent\n- 7 themes, persistent state, MCP integration\n- 93 passing tests\n\n### Build & run\n\n```bash\ncargo build --release\n# Run as .app bundle:\ncp target\u002Frelease\u002Fopensquirrel dist\u002FOpenSquirrel.app\u002FContents\u002FMacOS\u002FOpenSquirrel-bin\nopen dist\u002FOpenSquirrel.app\n```\n\nRequires Rust 1.85+ and macOS (Metal GPU). Linux (Vulkan) compiles and tests pass.\n\n### Config\n\n`~\u002F.osq\u002Fconfig.toml` -- runtimes, machines, MCPs, themes, settings.\n\n### Supported runtimes\n\n| Runtime | CLI | Mode |\n|---------|-----|------|\n| Claude Code | `claude` | Persistent multi-turn |\n| Codex | `codex` | One-shot |\n| Cursor Agent | `cursor agent` | One-shot |\n| Gemini CLI | `gemini` | One-shot |\n| OpenCode | `opencode` | One-shot |\n\n## License\n\nMIT\n","OpenSquirrel 是一个原生的 GPU 渲染平铺管理器，旨在为 AI 编码代理提供并行运行环境。项目采用 Rust 语言开发，并利用 GPU 加速渲染，支持 Claude Code、Codex、Cursor 和 OpenCode 等多种 AI 编码工具的同时运行与自动子代理委派功能。此外，它还具备通过 SSH 连接远程机器的能力以及持久会话支持。尽管该项目已不再积极维护，但其设计理念适用于需要在单一界面上高效管理和协调多个 AI 编码助手的开发者或团队。然而，根据项目作者的经验分享，对于大多数用户而言，直接使用终端及其内置的 TUI 可能是更简单且有效的选择。",2,"2026-06-11 03:50:25","high_star"]