[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-94561":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":9,"totalLinesOfCode":9,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":9,"subscribersCount":16,"size":16,"stars1d":16,"stars7d":16,"stars30d":17,"stars90d":16,"forks30d":16,"starsTrendScore":16,"compositeScore":18,"rankGlobal":9,"rankLanguage":9,"license":9,"archived":19,"fork":19,"defaultBranch":20,"hasWiki":19,"hasPages":19,"topics":9,"createdAt":9,"pushedAt":9,"updatedAt":21,"readmeContent":22,"aiSummary":23,"trendingCount":16,"starSnapshotCount":16,"syncStatus":24,"lastSyncTime":25,"discoverSource":26},94561,"kandev","kdlbs\u002Fkandev","kdlbs","AI Kanban & Development Environment. Orchestrate multiple agents, review changes, open PRs. Multi-provider, self-hostable, no telemetry.",null,"https:\u002F\u002Fgithub.com\u002Fkdlbs\u002Fkandev","Go",576,87,1,12,0,6,47.43,false,"main","2026-08-24 04:01:22","# Kandev\n\nManage and run tasks in parallel. Orchestrate agents. Review changes. Ship value.\n\n[Features](docs\u002Ffeatures.md) | [Workflows](docs\u002Fworkflow-tips.md) | [Run as a Service](docs\u002Frun-as-a-service.md) | [Debug Logs](docs\u002Fdebug-logs.md) | [Roadmap](docs\u002Froadmap.md) | [Contributing](CONTRIBUTING.md) | [Architecture](docs\u002FARCHITECTURE.md) | [Discord](https:\u002F\u002Fdiscord.gg\u002FgWdCPGcFCD)\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"docs\u002Fscreenshots\u002Freadme-intro.gif\" alt=\"Kandev Demo\">\n\u003C\u002Fp>\n\n[See screenshots](docs\u002Fscreenshots.md)\n\n## What\n\n\nKandev is a powerful tool for power users who want deeper control over how AI agents work: customize workflows, agent profiles, runtimes, prompts, and review gates to match your process.\n\nOrganize work across kanban and pipeline views with opinionated workflows and execute multiple tasks in parallel. Assign agents from any provider, and review their output in an integrated workspace - file editor, file tree, terminal, browser preview, and git changes in one place. Terminal agent TUIs are great for running agents, but reviewing and iterating on changes there doesn't scale.\n\nRun it locally or self-host it on your own infrastructure and access it from anywhere via [Tailscale](https:\u002F\u002Ftailscale.com\u002F) or any VPN.\n\nOpen source, multi-provider, no telemetry, not tied to any cloud.\n\n## Vision\n\n> **Humans stay in control.** Define tasks, build agentic workflows with gates, review every change, decide what ships.\n\n- **Review-first** - Humans support production systems. We need to understand (yet) and trust the code that gets deployed.\n- **Your workflow** - Every team is different, and not every developer uses AI the same way. Define workflows once, share them across the team, and give everyone a consistent process for working with agents - regardless of experience level.\n- **Remote agents** - Running multiple agents on a large codebase can quickly saturate a local machine. The goal is a single control plane: offload execution to servers, orchestrate from anywhere, including your phone.\n\n## Features\n\n- **Multi-agent support** - Claude Code, Codex, GitHub Copilot, Gemini CLI, Amp, Auggie, OpenCode, Cursor, Devin, Qwen, Factory Droid, iFlow, Kilocode, Pi, Kimi, AWS Kiro, Qoder, Trae, Oh My Pi, Grok, Hermes\n- **Parallel task execution** – start and manage multiple tasks from different sources simultaneously, boosting productivity with AI agents\n- **Integrated workspace** - Built-in terminal, code editor with LSP, git changes panel, embedded vscode and chat in one IDE-like view\n- **Kanban task management** - Drag-and-drop boards, columns, and workflow automation\n- **Agentic workflows** - Multi-step pipelines that mix-and-match agents per step - for example, Claude Code Opus to design a plan, GitHub Copilot Sonnet to implement it, and Codex GPT 5.4 to review the changes. See [docs\u002Fworkflow-tips.md](docs\u002Fworkflow-tips.md)\n- **Sub-tasks** - Agents can spawn sub-tasks that resume from the parent task's session. Useful for splitting a task that has grown too big, or producing several PRs from the same starting point.\n- **CLI passthrough** - Drop into raw agent CLI mode for direct terminal interaction with agents that support it, leveraging their full native TUI\n- **Workspace isolation** - Git worktrees prevent concurrent agents from conflicting\n- **Multi-repository tasks** - Span a single task across multiple repositories, with one worktree per repo, per-repo branches, per-repo PRs, and per-repo grouping in the Changes panel and review dialog\n- **Flexible runtimes** - Run agents as local processes, in isolated Docker containers, on remote servers via SSH, or in cloud executors like sprites.dev\n- **Runtime settings** - Executor profiles, secrets, custom prompts, utility agents, voice mode, and resource metrics are configurable from Settings\n- **Task-agent MCP** - Agents can create subtasks, target sibling repos, attach extra branches for multiple PRs, message other tasks, read conversations, and inspect related tasks\n- **External MCP** - Manage Kandev from outside coding agents over streamable HTTP or SSE, with copyable config snippets for popular agent CLIs\n- **Workflow portability** - Export and import workflows as portable YAML across workspaces or Kandev installs\n- **Session management** - Resume and review agent conversations\n- **Shareable task snapshots** - Publish redacted task conversation snapshots as secret GitHub Gists, with preview and revoke controls\n- **Stats** - Track your productivity with stats on the completed tasks, agent turns, etc\n\nSee [docs\u002Ffeatures.md](docs\u002Ffeatures.md) for the full feature inventory, including settings, secrets, custom prompts, and MCP task capabilities.\n\n### In progress: Office mode\n\nWe're working on **Office mode**, a feature-flagged autonomy layer for persistent agent teams. The direction is agent instances with roles and permissions, dashboards, inbox\u002Fapprovals, routines, task delegation, skills, memory, cost tracking, budgets, and workspace config sync. We'll document Office as a supported feature after it is live.\n\n## Integrations\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-181717?style=for-the-badge&logo=github&logoColor=white\" alt=\"GitHub\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fwww.atlassian.com\u002Fsoftware\u002Fjira\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FJira-0052CC?style=for-the-badge&logo=jira&logoColor=white\" alt=\"Jira\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Flinear.app\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLinear-5E6AD2?style=for-the-badge&logo=linear&logoColor=white\" alt=\"Linear\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fsentry.io\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSentry-362D59?style=for-the-badge&logo=sentry&logoColor=white\" alt=\"Sentry\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgitlab.com\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitLab-FC6D26?style=for-the-badge&logo=gitlab&logoColor=white\" alt=\"GitLab\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fslack.com\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSlack-4A154B?style=for-the-badge&logo=slack&logoColor=white\" alt=\"Slack\">\u003C\u002Fa>\n\u003C\u002Fp>\n\nConnect Kandev to the tools your team already uses — pull issues into the kanban, link tasks to PRs, and surface review activity inline.\n\n## Supported ACP Agents\n\n| Agent | Package \u002F command |\n|:-------:|:----------:|\n| **Claude Code** | `@agentclientprotocol\u002Fclaude-agent-acp` |\n| **Codex** | `@agentclientprotocol\u002Fcodex-acp` |\n| **GitHub Copilot** | `@github\u002Fcopilot` |\n| **Gemini CLI** | `@google\u002Fgemini-cli` |\n| **Amp** | `amp-acp` |\n| **Auggie** | `@augmentcode\u002Fauggie` |\n| **OpenCode** | `opencode-ai` |\n| **Cursor** | `cursor-agent` *(requires Cursor Pro)* |\n| **Devin** | `devin` *(install Devin CLI from Devin Desktop or standalone installer)* |\n| **Qwen** | `@qwen-code\u002Fqwen-code` |\n| **Factory Droid** | `droid` |\n| **iFlow (beta)** | `@iflow-ai\u002Fiflow-cli` |\n| **Kilocode** | `@kilocode\u002Fcli` |\n| **Pi** | `pi-acp` |\n| **Kimi** | `kimi` *(install Kimi CLI from Moonshot AI)* |\n| **Kiro** | `kiro-cli-chat` *(install Kiro CLI from AWS)* |\n| **Qoder** | `qodercli` *(install Qoder CLI)* |\n| **Trae** | `traecli` *(install Trae IDE CLI)* |\n| **Oh My Pi** | `omp` *(install `@oh-my-pi\u002Fpi-coding-agent` with Bun)* |\n| **Grok** | `grok` *(install `@xai-official\u002Fgrok` with npm)* |\n| **Hermes** | `hermes` *(install with the official Hermes installer)* |\n\n> All agents communicate via [ACP](https:\u002F\u002Fagentclientprotocol.com) (Agent Client Protocol). Some agents support ACP natively, while others use ACP adapter packages that bridge their native protocols. **CLI Passthrough mode** is available when an integration provides a passthrough command. If your agent isn't supported yet, open an issue or submit a PR with the integration. See [Adding a New Agent CLI](docs\u002Fadd-agent-cli.md) for a step-by-step guide.\n\nKandev does not pin the managed npm runtimes for Claude, Codex, OpenCode,\nCopilot, or Gemini. Normal launches can reuse npm's best-effort execution\ncache. To deliberately fetch an upstream release, use **Update agent** in\n**Settings > Agents**; Kandev updates the host runtime and refreshes its\nadvertised models and modes for future sessions without restarting active\nsessions. See [Agents and Profiles](docs\u002Fpublic\u002Fagents-and-profiles.md#update-a-managed-agent-runtime).\n\n### Bring your own TUI agents\n\nKandev can run any agent CLI as a TUI inside a terminal, even when it does not speak ACP. Add the command in agent profile settings and the task starts that agent inside a PTY terminal. CLI passthrough keeps the agent's native terminal UX available while still giving Kandev task tracking, worktrees, review, and session management.\n\n## Supported Executors\n\n| Executor | Description |\n|:--------:|-------------|\n| **Local Process** | Runs the agent as a local process on the host machine |\n| **Docker** | Runs the agent in an isolated Docker container |\n| **SSH** | Runs the agent on a remote server over SSH |\n| **Sprites** | Runs the agent in a remote cloud environment via [sprites.dev](https:\u002F\u002Fsprites.dev) |\n\nExecutors support profiles for reusable runtime configuration: prepare scripts, environment variables, credentials, and settings. Worktree-based tasks can also attach multiple repositories or multiple branches from the same repository, which lets one task produce several PRs when the work needs it.\n\n## Quick Start\n\n### Homebrew (macOS, Linux)\n\n```bash\nbrew install kdlbs\u002Fkandev\u002Fkandev\nkandev\n```\n\n### Scoop (Windows)\n\n```bash\nscoop bucket add kandev https:\u002F\u002Fgithub.com\u002Fkdlbs\u002Fscoop-kandev\nscoop install kandev\nkandev\n```\n\nInstalls the native bundle — no Node.js required.\n\n### NPX\n\n```bash\nnpx kandev@latest\n```\n\nRequires npm 7+. Installs the CLI plus a platform-matched runtime bundle in one step.\n\n### NPM (global)\n\n```bash\nnpm install -g kandev@latest\nkandev\n```\n\n### Updates\n\n- `brew upgrade kandev`\n- `scoop update kandev`\n- `npx kandev@latest` (always uses the latest published version)\n- `npm install -g kandev@latest`\n\nTo try the current prerelease from `main`, use npm's `nightly` tag:\n\n```bash\nnpx -y kandev@nightly\n# or\nnpm install -g kandev@nightly\n```\n\n`npx` runs a transient Nightly without changing a global package. The global install replaces your\ninstalled Kandev until you reinstall `kandev@latest`.\n\nNightlies are best-effort builds scheduled for 12:00 UTC when `main` has changed since the latest\nstable release. They do not move `latest`, and no Homebrew, Scoop, or Desktop nightly channel is\npublished. A verified Kandev-managed npm\u002Fnpx user service can switch between Stable and Nightly in\n**Settings > System > Updates**. Other install types expose only Stable in that page; npm\u002Fnpx users\ncan still invoke `@nightly` directly with the commands above.\n\nThe worktrees and sqlite db live in `~\u002F.kandev` by default. Works on macOS, Linux, and Windows (WSL or native).\n\n### From Source\n\n```bash\n# Clone the repository\ngit clone git@github.com:kdlbs\u002Fkandev.git\ncd kandev\n\n# Start in production mode\nmake start\n```\n\n**Prerequisites:** Go 1.26.0, Node.js 24, and pnpm 9.15.9 (pinned in `mise.toml`). Docker is optional and needed only for container runtimes.\n\n## High level architecture\n\n```mermaid\ngraph LR\n    UI[Web UI] --> Backend\n\n\n        Backend[\"Backend (Orchestrator)\"]\n\n\n    Backend --> W1 & W2 & W3 & W4\n\n    subgraph W1[Local Process]\n        Agent1[Agent CLI] --- WT1[Worktree]\n    end\n\n    subgraph W2[Docker Container]\n        Agent2[Agent CLI] --- WT2[Worktree]\n    end\n\n    subgraph W3[SSH Remote]\n        Agent3[Agent CLI] --- WT3[Worktree]\n    end\n\n    subgraph W4[Sprites Cloud]\n        Agent4[Agent CLI] --- WT4[Worktree]\n    end\n```\n\nWe also want to add support for this remote runtime:\n- **K8s operator** - run agents in a Kubernetes cluster, with auto-scaling and resource management.\n\n\u003Cdetails>\n\u003Csummary>\u003Cstrong>Development\u003C\u002Fstrong>\u003C\u002Fsummary>\n\n### Project Structure\n\n```\napps\u002F\n├── backend\u002F    # Go backend (orchestrator, lifecycle, agentctl, WS gateway)\n├── web\u002F        # Vite\u002FReact frontend (SPA, Zustand, real-time subscriptions)\n├── cli\u002F        # CLI tool (npx kandev launcher)\n├── desktop\u002F    # Tauri desktop shell\n└── packages\u002F   # Shared UI components & types\n```\n\n### Prerequisites\n\n- [mise](https:\u002F\u002Fmise.jdx.dev\u002F) for the pinned toolchain in `mise.toml` (`make bootstrap` installs it when missing)\n- Go 1.26.0, Node.js 24, pnpm 9.15.9 (installed by `make bootstrap`)\n- Docker (optional)\n\n### Environment Setup\n\n```bash\n# Install mise tools, workspace dependencies, and git hooks\nmake bootstrap\n\n# Also install Playwright Chromium browser\u002Fdeps for E2E\nmake bootstrap-e2e\n```\n\n### Running Dev Servers\n\n```bash\n# Start everything (backend + frontend with auto ports)\nmake dev\n\n# Or run separately\nmake dev-backend    # Backend on :38429\nmake dev-web        # Frontend on :37429\n```\n\n### Testing & Linting\n\n```bash\nmake test           # Run all tests (backend + web)\nmake lint           # Run all linters\nmake typecheck      # TypeScript type checking\nmake fmt            # Format all code\n```\n\n### Pre-commit Hooks\n\n```bash\n# Installed and wired by make bootstrap\nmake doctor\n```\n\n\u003C\u002Fdetails>\n\n## Comparison to Other Tools\n\nThere are a few similar tools in this space, and new ones appearing everyday. Here's what sets this one apart:\n\n- **Server-first architecture** - The core app runs as a server you can access from any device, including your phone.\n- **Remote runtimes** - Run agents on remote servers via SSH, Docker hosts, and cloud environments, not just your local machine.\n- **Multi-provider** - Use Claude Code, Codex, Copilot, Gemini, Amp, Auggie, OpenCode, Cursor, Devin, Qwen, Droid, iFlow, Kilocode, Pi, Kimi, Kiro, Qoder, Trae, Oh My Pi, Grok, and Hermes side by side. Not locked to one vendor.\n- **CLI passthrough and chat** - Interact with agents through structured chat messages or, where supported, drop into raw CLI mode for full agent TUI capabilities.\n- **Open source and self-hostable** - No vendor lock-in, no telemetry, runs on your infrastructure.\n\n## Contributing\n\nContributions are welcome! Please read [CONTRIBUTING.md](CONTRIBUTING.md) before opening a PR.\n\nSee the [issue tracker](https:\u002F\u002Fgithub.com\u002Fkdlbs\u002Fkandev\u002Fissues) for open tasks, or join our [Discord](https:\u002F\u002Fdiscord.gg\u002FgWdCPGcFCD) to chat with maintainers and other contributors.\n\n## Acknowledgments\n\nBuilt with these excellent open-source projects:\n\n[Monaco Editor](https:\u002F\u002Fmicrosoft.github.io\u002Fmonaco-editor\u002F) · [Tiptap](https:\u002F\u002Ftiptap.dev\u002F) · [xterm.js](https:\u002F\u002Fxtermjs.org\u002F) · [dockview](https:\u002F\u002Fdockview.dev\u002F) · [CodeMirror](https:\u002F\u002Fcodemirror.net\u002F) · [dnd-kit](https:\u002F\u002Fdndkit.com\u002F) · [Mermaid](https:\u002F\u002Fmermaid.js.org\u002F) · [Recharts](https:\u002F\u002Frecharts.org\u002F) · [TanStack Table](https:\u002F\u002Ftanstack.com\u002Ftable) · [Zustand](https:\u002F\u002Fzustand.docs.pmnd.rs\u002F) · [Shadcn\u002FUI](https:\u002F\u002Fui.shadcn.com\u002F) · [Radix UI](https:\u002F\u002Fwww.radix-ui.com\u002F) · [Tailwind CSS](https:\u002F\u002Ftailwindcss.com\u002F)\n\n## License\n\n[AGPL-3.0](LICENSE)\n\n## Star History\n\n\u003Ca href=\"https:\u002F\u002Fwww.star-history.com\u002F?repos=kdlbs%2Fkandev&type=date&legend=top-left\">\n \u003Cpicture>\n   \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fchart?repos=kdlbs\u002Fkandev&type=date&theme=dark&legend=top-left&sealed_token=VcdQXTkas2odX-D8DZINZOcBVuA6ZfkugIzgWgnAi0lqW3gwdqlh_Ei77fi2bdXwc3HnCFz3NQAYtRnoKxspsck-pIQxTdEwEIveR27Y3k8cKBWJxYGOCFqYhHiJGcECnrrMU_NhWcWWTWaz0t5cX-3Fm-1BOouTejaUEy9w78CSuoakpeUJby66XlCB\" \u002F>\n   \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fchart?repos=kdlbs\u002Fkandev&type=date&legend=top-left&sealed_token=VcdQXTkas2odX-D8DZINZOcBVuA6ZfkugIzgWgnAi0lqW3gwdqlh_Ei77fi2bdXwc3HnCFz3NQAYtRnoKxspsck-pIQxTdEwEIveR27Y3k8cKBWJxYGOCFqYhHiJGcECnrrMU_NhWcWWTWaz0t5cX-3Fm-1BOouTejaUEy9w78CSuoakpeUJby66XlCB\" \u002F>\n   \u003Cimg alt=\"Star History Chart\" src=\"https:\u002F\u002Fapi.star-history.com\u002Fchart?repos=kdlbs\u002Fkandev&type=date&legend=top-left&sealed_token=VcdQXTkas2odX-D8DZINZOcBVuA6ZfkugIzgWgnAi0lqW3gwdqlh_Ei77fi2bdXwc3HnCFz3NQAYtRnoKxspsck-pIQxTdEwEIveR27Y3k8cKBWJxYGOCFqYhHiJGcECnrrMU_NhWcWWTWaz0t5cX-3Fm-1BOouTejaUEy9w78CSuoakpeUJby66XlCB\" \u002F>\n \u003C\u002Fpicture>\n\u003C\u002Fa>\n","Kandev 是一个开源的 AI 工作流编排与开发环境，用于并行管理多智能体任务、可视化审查代码变更并自动化 Pull Request 提交。核心功能包括：支持 Claude、Gemini、Qwen 等 20+ LLM 提供商的多代理协同；集成式工作区（含终端、LSP 编辑器、Git 变更面板和浏览器预览）；看板式任务管理与可定制的多步 agentic 流程（如设计-实现-审查分阶段执行）；本地运行或自托管，无遥测，兼容 Tailscale 远程访问。适用于需要人工深度介入 AI 开发闭环的团队，尤其适合中大型代码库的渐进式 AI 协作、工程化 AI 应用落地及 DevOps 流程增强场景。",2,"2026-08-12 02:30:05","trending"]