HN 표시: AI 에이전트용 작업 관리자(MCP, 오픈 소스)

AgentRQ is a modern, high-performance platform designed for seamless collaboration between human operators and AI agents. It leverages the Model Context Protocol (MCP) to allow AI models (like Claude) to interact directly with your workspace's task management system. The platform allows users to create workspaces, manage tasks, and delegate work to AI agents. Agents can "see" the workspace state via MCP, receive notifications when new tasks are assigned, and reply back to the human operator—all in real-time. AgentRQ follows a decoupled service-oriented architecture: - API Server: Fiber-based REST API for workspace and task management. - MCP Server: Integrated mcp-go SSE server that exposes tools and resources to AI models. - CoreMCP (Supervisor): A global MCP server that allows agents to manage all workspaces, tasks, and statistics across the entire platform. - Data Layer: GORM with SQLite for persistent, user-scoped storage. - Authentication: Google OAuth2 integration with JWT-based session management. - Event Bus: Internal pub/sub system for real-time SSE notifications. - Modern UI: Tailored with Vue 3, Pinia, and Tailwind CSS. - Glassmorphism: A sleek, premium design language with smooth transitions and real-time updates. - Reactive State: Synchronized with the backend via SSE events. - Go 1.21+ - Node.js 18+ (with npm) - Google Cloud Console: An OAuth2 Client ID and Secret. - Create a _config/base.yaml (ordevelopment.yaml ) in thebackend directory. - Fill in your Google OAuth2 credentials: auth: google: client_id: "your-google-client-id" client_secret: "your-google-client-secret" Use the provided Makefile to start the full stack: # 1. Install all dependencies make install # 2. Start both Frontend and Backend make dev The frontend will be available at http://localhost:5173 . AgentRQ is designed for seamless integration as a Claude Channel. This allows your AI agents to see tasks assigned to them and respond directly within your Claude session. Each workspace has its own MCP URL and token (visible in the workspace setup modal). In production, these follow the pattern https://WORKSPACE_ID.mcp.agentrq.com/mcp . Create a .mcp.json file in your local project directory (the leading dot is required). Each project gets its own file so Claude instances stay isolated per workspace. Replace YOUR_MCP_URL below with the full URL shown in the setup modal (e.g. https://WORKSPACE_ID.mcp.agentrq.com/mcp?token=TOKEN ). { "mcpServers": { "agentrq-WORKSPACE_ID": { "type": "http", "url": "YOUR_MCP_URL" } } } Add a .claude/settings.local.json file in the same project directory to pre-approve the AgentRQ tools and avoid permission prompts on every action: { "permissions": { "allow": [ "mcp__agentrq-WORKSPACE_ID__updateTaskStatus", "mcp__agentrq-WORKSPACE_ID__getWorkspace", "mcp__agentrq-WORKSPACE_ID__reply", "mcp__agentrq-WORKSPACE_ID__createTask", "mcp__agentrq-WORKSPACE_ID__downloadAttachment", "mcp__agentrq-WORKSPACE_ID__getTaskMessages", "mcp__agentrq-WORKSPACE_ID__getNextTask" ] }, "enableAllProjectMcpServers": true, "enabledMcpjsonServers": ["agentrq-WORKSPACE_ID"] } Once both files are in place, launch Claude Code from that project directory: claude --dangerously-load-development-channels server:agentrq-WORKSPACE_ID Tip: The workspace ID, full MCP URL (with token), and ready-to-paste config snippets are all available in the Setup modal inside each AgentRQ workspace. When connected, the AI agent has access to: createTask : Assign a task to the human user (supports optionalcron_schedule for recurring tasks).updateTaskStatus : Move tasks throughnotstarted ,ongoing ,blocked , andcompleted .reply : Send messages back to the AgentRQ dashboard in real-time.getWorkspace : Fetch the workspace name, mission description, and task statistics.getTaskMessages : Read the chat history of a task with cursor-based pagination.getNextTask : Efficiently retrieve the next "not started" task assigned to the agent.downloadAttachment : Retrieve an attachment by its ID.- Real-time Notifications: Agents receive notifications via the notifications/claude/channel protocol whenever a human interacts with their tasks. While individual workspaces provide a scoped view for specific projects, the Supervisor (CoreMCP) is a global MCP server that grants an agent bird's-eye view and management capabilities across your entire AgentRQ account. The Supervisor is accessible at https://mcp.agentrq.com/mcp . It uses OAuth2 for secure authentication, allowing modern AI tools (like Claude Code) to connect securely. - Multi-Workspace Management: List, create, and update workspaces. - Global Task View: Fetch tasks from all workspaces in a single call ( listAllTasks ). - Administrative Control: Manage task assignments, status, and priorities globally. - Unified Statistics: Access detailed statistics and health metrics for any workspace. The Supervisor provides a comprehensive suite of tools for global management, requiring workspaceId parameters where applicable: Workspace Management listWorkspaces : Overview of all active and archived workspaces.createWorkspace : Bootstrap new project environments.getWorkspace : Retrieve details of a specific workspace by ID.updateWorkspace : Modify workspace settings and metadata.getWorkspaceStats : Retrieve high-level analytics and performance data for a workspace. Task Management listAllTasks : Search and filter tasks across the entire platform.listTasks : List tasks within a specific workspace.createTask : Create a new task in a specific workspace.getTask : Retrieve details of a specific task.updateTaskStatus : Change a task's status.updateTaskOrder : Reorder a task in the list.updateTaskAssignee : Change the assignee of a task.updateTaskAllowAll : Toggleallow_all_commands permission for a task.updateScheduledTask : Modify a scheduled/cron task. Communication & Files replyToTask : Post a message to a task's chat thread.respondToTask : Submit an allow/deny verdict for a permission request.getAttachment : Retrieve metadata for a specific attachment. Since the Supervisor uses OAuth2, you can connect it using the following configuration in your ~/.mcp.json : { "mcpServers": { "agentrq-supervisor": { "type": "http", "url": "https://mcp.agentrq.com/mcp" } } } When you first run Claude with this server, it will provide a link to authenticate via your browser. - AgentRQ — The official Agent-Human collaboration platform. - HasMCP — Bridge the Gap Between APIs and Agents. Apache-2.0