13 KiB

Raw Permalink Blame History

Task Master Project Initialization Guide

Overview

This guide provides step-by-step instructions for initializing a Task Master project to manage the development of the Directus Task Management Suite. This creates a tactical project management layer alongside the strategic BMad planning documents.

Quick Setup Commands

# Navigate to project directory
cd /Users/stevenhardge/Documents/projects/my-ai-projects/projects/directus-task-management

# Initialize Task Master in this directory
task-master init

# Create PRD for Task Master parsing
cp planning/01-bmad-prd.md .taskmaster/docs/prd.txt

# Parse PRD to generate initial tasks
task-master parse-prd .taskmaster/docs/prd.txt --num=15

# Analyze task complexity and expand key tasks
task-master analyze-complexity --research
task-master expand --all --research

Detailed Setup Process

Step 1: Initialize Task Master Project

Command:

cd /Users/stevenhardge/Documents/projects/my-ai-projects/projects/directus-task-management
task-master init

Expected Output:

✅ Task Master initialized successfully
✅ Created .taskmaster/ directory structure
✅ Created .taskmaster/config.json with default settings
✅ Created .taskmaster/tasks/tasks.json
✅ Created .taskmaster/docs/ directory

Verification:

ls -la .taskmaster/
# Should show: config.json, docs/, tasks/, reports/

Step 2: Prepare Project Requirements Document

Copy BMad PRD for Task Master:

# Copy the comprehensive PRD created in BMad planning
cp planning/01-bmad-prd.md .taskmaster/docs/prd.txt

# Verify content
head -20 .taskmaster/docs/prd.txt

Alternative: Create Focused PRD for Development If you prefer a more focused PRD specifically for the development work:

cat > .taskmaster/docs/prd.txt << 'EOF'
# Directus Task Management Suite - Development PRD

## Project Overview
Build a comprehensive task management system on Directus CMS that integrates with existing AI development ecosystem including Claude Code agents, BMad methodology, and Task Master CLI.

## Core Requirements

### Phase 1: Foundation (Weeks 1-2)
- Design and implement 10 Directus collections for task management
- Build 22 new MCP tools for programmatic task operations
- Create basic task CRUD operations via Directus admin interface
- Set up project hierarchy and customizable status workflows

### Phase 2: AI Integration (Weeks 3-4) 
- Implement AI-powered task creation using natural language prompts
- Build Claude Code agent context integration through MCP tools
- Add automated progress tracking from git events and agent activity
- Create task templates aligned with BMad methodology

### Phase 3: Workflow Integration (Weeks 5-6)
- Build BMad workflow templates and step tracking
- Implement Task Master bidirectional synchronization
- Add team collaboration features through web interface
- Create progress analytics dashboard

### Phase 4: Advanced Features (Weeks 7-8)
- Advanced dependency management with critical path analysis
- Time tracking and resource utilization analytics
- Performance optimizations for 10,000+ tasks
- Production deployment with monitoring

## Technical Architecture
- Platform: Directus CMS (existing instance at https://enias.zeabur.app)
- Integration: 22 new MCP tools extending existing 40+ tool ecosystem
- Database: 10 new collections with strategic indexing
- Sync: Bidirectional with Task Master using conflict resolution
- Performance: <100ms response times, cursor-based pagination

## Success Criteria
- 70% efficiency improvement in task management workflows
- 80% of tasks created through AI assistance
- 99.9% system uptime and data integrity
- Seamless integration with existing development tools

## Key Deliverables by Phase
1. **Phase 1**: Functional task management web interface
2. **Phase 2**: AI-powered task creation and agent integration  
3. **Phase 3**: BMad methodology support and Task Master sync
4. **Phase 4**: Analytics dashboard and production optimization
EOF

Step 3: Generate Initial Tasks

Parse PRD to Create Tasks:

# Generate tasks from PRD (15 high-level tasks recommended)
task-master parse-prd .taskmaster/docs/prd.txt --num=15 --research

# Alternative: Let Task Master determine optimal number
task-master parse-prd .taskmaster/docs/prd.txt --num=0 --research

Expected Output:

✅ Parsing PRD document: .taskmaster/docs/prd.txt
🔍 Using research mode for enhanced task generation
✅ Generated 15 high-level tasks from PRD content
✅ Tasks saved to .taskmaster/tasks/tasks.json
✅ Individual task files created in .taskmaster/tasks/

Verify Task Generation:

# List generated tasks
task-master list

# View task summary
task-master show 1

Step 4: Analyze and Expand Tasks

Run Complexity Analysis:

# Analyze all tasks for complexity and expansion recommendations
task-master analyze-complexity --research

# View complexity report
task-master complexity-report

Expand High-Complexity Tasks:

# Expand all tasks that meet complexity threshold
task-master expand --all --research

# Alternative: Expand specific high-complexity tasks
task-master expand --id=1 --research --num=5
task-master expand --id=5 --research --num=4
task-master expand --id=8 --research --num=6

Expected Results:

High-level tasks broken down into actionable subtasks
Implementation guidance for each subtask
Dependency relationships established
Time estimates provided

Step 5: Configure Development Workflow

Set Up Models (if not already configured):

# Configure AI models for task operations
task-master models --set-main claude-3-5-sonnet-20241022
task-master models --set-research perplexity-llama-3.1-sonar-large-128k-online
task-master models --set-fallback gpt-4o-mini

# Verify configuration
task-master models

Create Development Tags:

# Add relevant tags to tasks for better organization
task-master update-task --id=1 --prompt="Add tags: directus, mcp, collections, database"
task-master update-task --id=2 --prompt="Add tags: mcp, tools, api, integration"
task-master update-task --id=3 --prompt="Add tags: ai, nlp, claude, automation"

Expected Task Structure

After initialization, you should have a Task Master project with approximately:

High-Level Tasks (Examples)

Design Core Collection Schema - Database design for 10 collections
Implement MCP Tools Suite - 22 new tools for task operations
Build AI Task Creation - Natural language to task conversion
Create Task CRUD Interface - Basic web interface operations
Develop Status Workflow System - Customizable task statuses
Build Project Hierarchy - Multi-level project organization
Implement Agent Integration - Claude Code context provision
Create BMad Templates - Workflow methodology support
Build Task Master Sync - Bidirectional data synchronization
Develop Analytics Dashboard - Progress tracking and metrics
Implement Time Tracking - Resource utilization monitoring
Build Dependency Management - Task relationship handling
Create Team Collaboration - Multi-user workflow support
Performance Optimization - Scaling for large datasets
Production Deployment - Monitoring and reliability

Subtask Structure (Example for Task 1)

1. Design Core Collection Schema

1.1 Design projects collection with hierarchy support
1.2 Design tasks collection with rich metadata
1.3 Design task_statuses collection with workflow rules
1.4 Design task_dependencies for relationship management
1.5 Design integration collections for external systems
1.6 Create database indexes for performance optimization
1.7 Implement collection relationships and constraints

Daily Development Workflow

Starting Development Work

# Get next available task
task-master next

# View task details
task-master show <task-id>

# Mark task as in-progress
task-master set-status --id=<task-id> --status=in-progress

During Development

# Update task with progress notes
task-master update-subtask --id=<subtask-id> --prompt="Implementation notes: [describe progress]"

# Add time entries (optional)
task-master add-time-entry --id=<task-id> --hours=2.5 --description="Schema design work"

Completing Tasks

# Mark subtask as complete
task-master set-status --id=<subtask-id> --status=done

# When all subtasks complete, mark parent as done
task-master set-status --id=<task-id> --status=done

# Get next task
task-master next

Integration with BMad Planning

Linking Strategic and Tactical Levels

The Task Master project provides tactical execution tracking while the BMad planning documents provide strategic direction:

Strategic Level (BMad Documents):

Product vision and requirements (01-bmad-prd.md)
System architecture decisions (02-bmad-architecture.md)
Market research and analysis (03-bmad-research-analysis.md)
Technical validation (04-bmad-validation.md)

Tactical Level (Task Master):

Granular task breakdown and tracking
Daily development workflow management
Progress monitoring and velocity tracking
Implementation-level dependencies

Cross-Reference Pattern

Reference Strategic Context in Tasks:

# Link tactical tasks to strategic planning
task-master update-task --id=1 --prompt="Reference: See architecture decisions in planning/02-bmad-architecture.md for collection design patterns"

task-master update-task --id=3 --prompt="Reference: AI integration patterns documented in planning/03-bmad-research-analysis.md section on AI Integration Research"

Update Strategic Documents with Tactical Progress:

# Periodically update BMad documents with implementation findings
# (Manual process - updates to planning/*.md files as insights emerge)

Maintenance and Updates

Regular Maintenance

# Weekly: Update task progress and velocity
task-master generate  # Regenerate task files from JSON

# Monthly: Analyze project health
task-master complexity-report
task-master validate-dependencies

Adding New Requirements

# For new requirements discovered during development
task-master add-task --prompt="[New requirement description]" --research

# For urgent bug fixes or issues
task-master add-task --prompt="Fix: [Issue description]" --priority=high

Integration Testing

# Validate Task Master integration with project
task-master validate-dependencies
task-master fix-dependencies  # If issues found

Troubleshooting

Common Issues

1. Task Master Command Not Found:

# Install Task Master globally
npm install -g task-master-ai

# Verify installation
task-master --version

2. API Keys Not Configured:

# Check environment variables
echo $ANTHROPIC_API_KEY
echo $PERPLEXITY_API_KEY

# Configure in .env file or Task Master config
task-master models --setup

3. PRD Parsing Fails:

# Verify PRD file exists and is readable
cat .taskmaster/docs/prd.txt | head -10

# Try parsing with different parameters
task-master parse-prd .taskmaster/docs/prd.txt --num=10 --force

4. Tasks Not Expanding:

# Check complexity analysis first
task-master analyze-complexity

# Try expanding specific tasks
task-master expand --id=1 --force --num=5

Success Validation

After setup, verify the Task Master project is working correctly:

✅ Validation Checklist

Task Master initialized in project directory
PRD successfully parsed into tasks
10-20 high-level tasks generated
Complexity analysis completed
High-complexity tasks expanded into subtasks
Can view and navigate task hierarchy
Can update task status and progress
Dependencies are properly linked
Tasks reference strategic planning documents

📊 Expected Metrics

Total Tasks: 15-25 high-level tasks
Total Subtasks: 50-100 subtasks (after expansion)
Task Complexity: Mix of trivial (20%), minor (40%), major (30%), critical (10%)
Dependencies: 20-30% of tasks have dependencies
Estimated Hours: 200-400 hours total project effort

🔗 Strategic-Tactical Alignment

Each major BMad planning phase has corresponding Task Master tasks
Implementation tasks reference appropriate strategic planning documents
Tactical progress can inform strategic planning updates
Both levels support the overall project delivery timeline

This setup creates a comprehensive tactical project management layer that complements the strategic BMad planning, enabling efficient development workflow while maintaining visibility into both strategic direction and tactical execution progress.

13 KiB Raw Permalink Blame History