Dudoxx MCP Task Manager

🚀 Overview

Dudoxx MCP Task Manager is an enhanced Model Context Protocol (MCP) server that revolutionizes task management for AI coding assistants. It ensures comprehensive context delivery, making every task execution precise and successful without ambiguity.

Key Features

• 🎯 Comprehensive Context Requirements: Every task includes files, dependencies, acceptance criteria
• 📊 Task Categorization: bugfix, feature, refactor, testing, documentation, infrastructure, research
• ⏱️ Complexity Estimation: From trivial to very-complex with time tracking
• ✅ Measurable Success: Clear acceptance criteria and expected outcomes
• 🔄 Approval Workflow: User validation at each step ensures quality
• 📁 File Tracking: Monitor planned vs actual file modifications

📋 Prerequisites

• Node.js 18+ (install via brew install node on macOS)
• Claude Desktop (install from https://claude.ai/desktop)
• npm or pnpm package manager

🛠️ Installation

1. Clone the Repository

git clone https://github.com/Dudoxx/dudoxx-mcp-task-manager.git
cd dudoxx-mcp-task-manager

2. Install Dependencies

npm install

3. Build the Project

npm run build

4. Configure Claude Desktop

Locate your Claude Desktop configuration file:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%/Claude/claude_desktop_config.json

Add the following configuration:

{
  "mcpServers": {
    "dudoxx-task-manager": {
      "command": "node",
      "args": ["/absolute/path/to/dudoxx-mcp-task-manager/dist/index.js"]
    }
  }
}

Note: Replace /absolute/path/to/dudoxx-mcp-task-manager with your actual path.

🎯 Enhanced Task Structure

Every task now includes comprehensive context:

{
  title: "Implement user authentication",
  description: "Add JWT-based authentication to Express API endpoints",
  context: {
    files: ["NEW: middleware/auth.js", "routes/users.js"],
    dependencies: ["jsonwebtoken", "bcryptjs"],
    currentState: "API has no authentication, all endpoints are public",
    acceptanceCriteria: [
      "JWT tokens generated on login",
      "Protected routes return 401 without valid token",
      "Passwords hashed with bcrypt",
      "Token refresh mechanism implemented"
    ],
    estimatedComplexity: "complex",
    category: "feature"
  },
  expectedOutcome: "Secure API with JWT authentication protecting sensitive endpoints"
}

🔧 Available Tools (12)

Core Workflow Tools

1. request_planning - Create comprehensive task plan with full context
2. get_next_task - Retrieve next task with complete execution context
3. mark_task_done - Complete task with evidence and time tracking
4. approve_task_completion - Validate task completion quality
5. approve_request_completion - Finalize entire request with summary

Information & Management Tools

6. list_requests - Overview of all requests with basic metrics
7. get_request_details - Comprehensive details of request(s) with all task info
8. open_task_details - Inspect specific task information
9. add_tasks_to_request - Extend request with new tasks
10. update_task - Modify task including context and outcomes
11. delete_task - Remove uncompleted task with validation
12. delete_request - Remove entire request (requires confirmation)

📖 Usage Example

Creating a Request with Full Context

// Request planning with comprehensive context
{
  "originalRequest": "Add real-time notifications to chat app",
  "projectContext": "React/Node.js chat using Socket.io, Redux state management, PostgreSQL database",
  "goals": [
    "Instant message notifications",
    "Cross-device synchronization",
    "Offline message queue"
  ],
  "constraints": [
    "Must not break existing WebSocket connections",
    "Support 1000+ messages/minute"
  ],
  "splitDetails": "Split into backend queue, frontend handler, and offline support for isolation",
  "tasks": [
    {
      "title": "Implement notification queue backend",
      "description": "Create Redis-backed queue for reliable notification delivery",
      "context": {
        "files": ["NEW: services/notificationQueue.js", "controllers/messageController.js"],
        "dependencies": ["bull", "redis"],
        "currentState": "Direct Socket.io delivery without queueing",
        "acceptanceCriteria": [
          "Queue created in Redis",
          "Exponential backoff for retries",
          "Failed deliveries logged",
          "Queue dashboard available"
        ],
        "estimatedComplexity": "complex",
        "category": "feature"
      },
      "expectedOutcome": "Robust notification system with delivery guarantees"
    }
  ]
}

🔄 Workflow Process

1. Planning Phase

   • AI creates request with request_planning
   • Provides comprehensive context for each task
   • Explains task decomposition strategy

2. Execution Phase

   • Retrieve task with get_next_task
   • Execute with full context (files, dependencies, criteria)
   • Mark complete with mark_task_done including evidence

3. Approval Phase

   • User validates with approve_task_completion
   • Continue to next task or revise if needed
   • Final approval with approve_request_completion

📊 Progress Visualization

The system provides detailed progress tables:

╔════════════════════════════════════════════════════════════╗
║ Project: Add user authentication (req-1)                    ║
║ Progress: ████████░░░░░░ 45% (3/7 tasks)                   ║
╠════════════════════════════════════════════════════════════╣
║ # │ Task                  │ Category │ Status │ Approval   ║
╟───┼───────────────────────┼──────────┼────────┼────────────╢
║ 1 │ Create user model     │ feature  │ ✅     │ ✅         ║
║ 2 │ Add password hashing  │ feature  │ ✅     │ ✅         ║
║ 3 │ Implement JWT tokens  │ feature  │ ✅     │ ⏳         ║
║ 4 │ Build login endpoint  │ feature  │ 🔄     │ -          ║
╚════════════════════════════════════════════════════════════╝

🐛 Debugging

Check MCP logs if you encounter issues:

# macOS
tail -n 50 -f ~/Library/Logs/Claude/mcp*.log

# View task persistence
cat ./tasks.json

🔨 Development

# Install dependencies
npm install

# Build TypeScript
npm run build

# Watch mode for development
npm run watch

# Type checking
npx tsc --noEmit

🏢 About Dudoxx

Dudoxx UG is a Hamburg-based technology company specializing in AI-enhanced development tools and medical technology solutions. This MCP Task Manager is part of our suite of tools designed to augment developer productivity through intelligent automation.

• Website: https://dudoxx.com
• Contact: [email protected]
• Location: Hamburg, Germany

📄 License

MIT License - see LICENSE file for details.

🙏 Acknowledgments

• Original MCP TaskManager concept by kazuph
• Built on Anthropic's Model Context Protocol
• Enhanced by Dudoxx UG for comprehensive context management

🚀 What's New in v2.0

• Enhanced Context System: Every task now requires comprehensive context
• Task Categories: Organize by type (bugfix, feature, etc.)
• Complexity Estimation: Better time and effort planning
• File Tracking: Monitor planned vs actual changes
• Delete Request Tool: Remove entire requests when needed
• Improved Meta Prompts: Clearer guidance for AI assistants
• Time Tracking: Monitor actual time spent on tasks

© 2025 Dudoxx UG - Enhancing AI-Assisted Development