{"name":"Agentrooms: Multi-Agent Claude Code Orchestration Desktop App","description":"Agentrooms is a desktop application and API designed for multi-agent Claude Code orchestration. It enables users to coordinate both local and remote agents through an intuitive @mentions system, streamlining complex development workflows. This tool facilitates collaborative development by routing tasks to specialized agents, all while leveraging your existing Claude subscription without requiring additional API keys.","github":"https://github.com/baryhuang/agentrooms","url":"https://osrepos.com/repo/baryhuang-agentrooms","source":"osrepos.com","sourceDescription":"This repository profile is provided by osrepos.com, an open source repository discovery platform.","repositoryProfile":"https://osrepos.com/repo/baryhuang-agentrooms","generatedFor":"open source discovery and AI-assisted research","markdown":"https://osrepos.com/repo/baryhuang-agentrooms.md","json":"https://osrepos.com/repo/baryhuang-agentrooms.json","topics":["Claude API","Multi-Agent Systems","Desktop App","TypeScript","AI Development","MacOS App","Code Orchestration","Agent Planning"],"keywords":["Claude API","Multi-Agent Systems","Desktop App","TypeScript","AI Development","MacOS App","Code Orchestration","Agent Planning"],"stars":null,"summary":"Agentrooms is a desktop application and API designed for multi-agent Claude Code orchestration. It enables users to coordinate both local and remote agents through an intuitive @mentions system, streamlining complex development workflows. This tool facilitates collaborative development by routing tasks to specialized agents, all while leveraging your existing Claude subscription without requiring additional API keys.","content":"## Introduction\n\nAgentrooms is a desktop application and API designed for multi-agent Claude Code orchestration. It provides a collaborative development workspace where users can route tasks to specialized agents, coordinating complex workflows seamlessly. This tool supports both local and remote agents, enabling dynamic team environments without requiring additional API keys, as it leverages your existing Claude subscription via Claude CLI authentication.\n\nKey features include:\n*   **`@agent-name` mentions**: Direct execution without orchestration overhead.\n*   **Multi-agent workflows**: Automatic task decomposition and coordination.\n*   **Local + Remote agents**: Mix local agents with remote machines (e.g., Mac Mini browser agent, cloud instances).\n*   **No API key required**: Uses your Claude subscription through Claude CLI authentication.\n*   **Multi-agent orchestration**: Intelligent task coordination between specialized agents.\n*   **Custom API support**: Configure your own endpoint for private deployment.\n*   **Dynamic agents**: Add and remove agents via the web UI.\n\n## Installation\n\nTo get started with Agentrooms, ensure you meet the prerequisites and then choose your preferred installation method.\n\n### Prerequisites\n\n1.  **Claude Subscription**: Requires Claude Pro, Team, or Enterprise subscription.\n2.  **Install Claude CLI**: Download from the official [Claude Code documentation](https://docs.anthropic.com/en/docs/claude-code){:target=\"_blank\"}.\n3.  **Authenticate**: Run `claude auth login` and complete OAuth authentication. This connects Agentrooms to your Claude subscription, using OAuth tokens securely without needing API keys.\n\n### Option 1: Desktop App (Recommended)\n\n**Download Pre-built App:**\nVisit the [Releases page](https://github.com/baryhuang/claude-code-by-agents/releases){:target=\"_blank\"} to download the appropriate installer for your operating system.\n*   **Windows**: `claude-code-webui-windows-x64.exe` - Run the installer.\n*   **macOS Intel**: `claude-code-webui-macos-x64` - Drag to Applications folder.\n*   **macOS Apple Silicon**: `claude-code-webui-macos-arm64` - Drag to Applications folder.\n*   **Linux x64**: `claude-code-webui-linux-x64` - Make executable and run.\n*   **Linux ARM64**: `claude-code-webui-linux-arm64` - Make executable and run.\n\n**Important**: Remember to start the backend service separately as described in the \"Backend Setup\" section below.\n\n**Build from Source:**\n\nbash\ngit clone https://github.com/baryhuang/claude-code-by-agents.git\ncd claude-code-by-agents\nnpm install\nnpm run build:frontend\nnpm run dist:mac     # Creates macOS DMG in dist/ folder\nnpm run dist:win     # Creates Windows installer in dist/ folder\nnpm run dist:linux   # Creates Linux AppImage in dist/ folder\n\n\n### Option 2: Web Development\n\nbash\n# Start backend service\ncd backend && deno task dev        # Backend: http://localhost:8080\n\n# Start frontend (separate terminal)\ncd frontend && npm run dev         # Frontend: http://localhost:3000\n\n\n### Backend Setup (Required for Desktop Apps)\n\nThe desktop apps run frontend-only. You must start the backend service separately:\n\nbash\ngit clone https://github.com/baryhuang/claude-code-by-agents.git\ncd claude-code-by-agents\ncd backend && deno task dev\n# Backend will run on http://localhost:8080\n\n\nThe Agentrooms app will automatically connect to `localhost:8080`. If your backend is on a different port, you may need to update the frontend configuration.\n\n### Agent Setup (Optional)\n\nYou can start additional agent instances and configure them in the Settings UI:\n\nbash\ncd path/to/agent1 && deno task dev --port 8081   # Local agent\ncd path/to/agent2 && deno task dev --port 8082   # Local agent\n# Remote agents: Run on other machines, expose ports\n\n\n## Examples\n\nAgentrooms simplifies both single-agent and multi-agent workflows.\n\n### Single Agent Task\n\nTo direct a task to a specific agent, use `@agent-name` mentions:\n\n\n@api-agent add user authentication\n\n\nThis makes a direct HTTP call to the agent's endpoint, minimizing orchestration overhead.\n\n### Multi-Agent Task\n\nFor more complex projects, the orchestrator can manage multiple agents:\n\n\n\"Create full auth system with frontend and backend\"\n\n\nIn this scenario, the orchestrator analyzes the request, creates an execution plan, and coordinates file-based communication between agents. It automatically manages dependencies, ensuring a smooth workflow.\n\n## Why Use Agentrooms?\n\nAgentrooms stands out by offering a powerful, flexible, and user-friendly environment for AI-driven development with Claude Code.\n\n*   **Seamless Multi-Agent Coordination**: Automate task decomposition and intelligent coordination between specialized agents, significantly boosting productivity for complex projects.\n*   **Flexible Agent Deployment**: Integrate both local and remote agents into your workflow, allowing for diverse development setups and resource utilization.\n*   **Cost-Effective**: Leverage your existing Claude subscription without the need for additional API keys or separate billing, as all operations run through your authenticated Claude CLI session.\n*   **Intuitive Interaction**: The `@mentions` system provides a natural way to direct tasks and manage agent interactions, reducing manual orchestration efforts.\n*   **Customizable and Dynamic**: Easily add, remove, and configure agents via the web UI, and even set up custom API endpoints for private deployments.\n\n## Links\n\n*   **GitHub Repository**: [baryhuang/claude-code-by-agents](https://github.com/baryhuang/claude-code-by-agents){:target=\"_blank\"}\n*   **Releases Page**: [Download Agentrooms](https://github.com/baryhuang/claude-code-by-agents/releases){:target=\"_blank\"}\n*   **Claude Code Documentation**: [Learn more about Claude CLI](https://docs.anthropic.com/en/docs/claude-code){:target=\"_blank\"}","metrics":{"detailViews":11,"githubClicks":13},"dates":{"published":null,"modified":"2025-12-10T00:01:04.000Z"}}