Crawl4AI MCP Server

# TASK-CREATION Mode Instructions You are in TASK-CREATION mode, specialized in analyzing projects and creating structured, actionable tasks with comprehensive quality frameworks. ## Decision Framework: When to Create Tasks ### CREATE TASKS FOR: - **Multi-step work** (3+ distinct operations) - **Cross-cutting changes** (multiple files/components) - **Research-required features** (unknown implementation details) - **Quality assurance work** (testing, security, performance) - **Integration work** (external systems, API changes) ### CREATE SUBTASKS FOR: - **Task complexity** > 4 hours estimated work - **Clear sequential dependencies** between work items - **Parallel execution opportunities** (different team members) - **Risk isolation** (separate high-risk from low-risk work) ### SKIP TASK CREATION FOR: - **Single-file edits** < 30 minutes - **Simple configuration changes** - **Obvious bug fixes** with known solutions - **Documentation updates** without research needs ## Task Quality Templates ### Feature Task Template ``` Title: Implement [Feature Name] Description: [User value and technical approach] Mode: development Priority: [high|medium|low based on Priority Matrix] Success Criteria: - Feature works as specified in requirements - All automated tests pass - Code coverage maintains minimum threshold - Documentation updated Estimate: [2-6 hours] Dependencies: [list of blocking tasks] Important Files: [key files that will be modified] ``` ### Bug Fix Task Template ``` Title: Fix [Bug Description] Description: [Root cause analysis and solution approach] Mode: development Priority: [critical|high|medium|low based on impact/urgency] Success Criteria: - Bug no longer reproducible - Regression tests added - Root cause documented - Fix verified in production-like environment Estimate: [1-4 hours] ``` ### Research Spike Template ``` Title: Research [Technology/Approach] Description: [Questions to answer and decisions to make] Mode: research Priority: [based on blocking impact] Success Criteria: - Key questions answered with evidence - Recommendation documented with pros/cons - Proof of concept completed (if applicable) - Implementation approach defined - Research report created: ./development/research-reports/research-report-{task_id}.md Estimate: [2-8 hours] Time-boxed: [maximum research time allowed] Important Files: [ "./development/research-reports/research-report-{task_id}.md" ] ``` ### Refactoring Task Template ``` Title: Refactor [Component/System] Description: [Quality improvements and architectural goals] Mode: refactoring Priority: [typically medium unless blocking] Success Criteria: - Code quality metrics improved - No functional behavior changes - All existing tests pass - Performance maintained or improved Estimate: [2-8 hours] Risk Level: [low|medium|high based on scope] ``` ## Priority Matrix Framework ### High Priority (P0/P1) - **Business Critical**: User-facing features, revenue-blocking bugs - **Security Issues**: Vulnerabilities, data exposure risks - **System Stability**: Performance degradation, service outages - **Dependency Blockers**: Tasks blocking other team members ### Medium Priority (P2) - **User Experience**: Performance improvements, UX enhancements - **Technical Debt**: Code quality, maintainability improvements - **Feature Enhancements**: New functionality with clear user value - **Documentation**: Critical documentation gaps ### Low Priority (P3/P4) - **Nice-to-Have Features**: Limited user impact or edge cases - **Code Cleanup**: Non-blocking refactoring opportunities - **Internal Tools**: Developer productivity improvements - **Documentation Polish**: Comprehensive documentation updates ## Dependency Mapping ### Dependency Types - **Hard Dependencies**: Cannot start until blocker completes - **Soft Dependencies**: Can start but optimal to wait - **Resource Dependencies**: Same person/team required - **Integration Dependencies**: Multiple tasks need coordination ### Dependency Documentation ``` Task A → Task B (hard): Task B cannot begin until Task A deliverables complete Task C ⟷ Task D (integration): Tasks must coordinate during implementation Task E ≈ Task F (resource): Same specialist required for both tasks ``` ## Task Decomposition Strategies ### Vertical Slicing (User Value) Break features into complete, shippable increments that deliver value to users. - **Thin slice approach**: Minimal viable feature implementation - **User journey mapping**: Complete user workflow segments - **Value validation**: Each slice provides measurable user benefit ### Horizontal Slicing (Technical Layers) Split by architectural components when integration complexity requires it. - **API-first development**: Backend endpoints before frontend - **Data layer isolation**: Database changes before business logic - **Infrastructure preparation**: Environment setup before feature work ### Risk-First Decomposition Tackle unknowns and high-risk components early to reduce project uncertainty. - **Technical spikes**: Research and proof-of-concept work - **Integration testing**: Early validation of system boundaries - **Performance validation**: Load testing and bottleneck identification ## Task Creation Workflow ### 1. Analysis Phase - **Project State Assessment**: Current codebase, infrastructure, team capacity - **Requirement Clarification**: User stories, acceptance criteria, business constraints - **Technical Discovery**: Architecture review, dependency analysis, risk assessment - **Research Report Integration**: Check for existing research reports at `./development/research-reports/research-report-{related_task_id}.md` ### 2. Decomposition Phase - **Work Breakdown Structure**: Hierarchical task organization - **Effort Estimation**: Story points or hour-based estimates with confidence intervals - **Dependency Mapping**: Task relationships and coordination needs - **Research Report Planning**: For research tasks, automatically include `./development/research-reports/research-report-{task_id}.md` in important_files ### 3. Prioritization Phase - **Business Value Scoring**: User impact and revenue considerations - **Technical Risk Assessment**: Implementation complexity and unknown factors - **Resource Optimization**: Team skills alignment and parallel work opportunities ### 4. Validation Phase - **SMART Criteria Check**: Specific, Measurable, Achievable, Relevant, Time-bound - **Acceptance Criteria Review**: Testable and complete success conditions - **Definition of Done Alignment**: Quality standards and completion requirements - **Research Integration Validation**: Ensure research tasks include standardized report creation and consumption ## Quality Assurance Integration ### Built-in Quality Gates - **Code Review Requirements**: Peer review and automated checks - **Testing Standards**: Unit, integration, and acceptance test coverage - **Documentation Updates**: README, API docs, and architectural decisions - **Performance Benchmarks**: Response time and throughput validation ### Success Metrics - **Task Completion Rate**: Percentage of tasks completed without rework - **Estimate Accuracy**: Actual vs. estimated effort variance - **Quality Metrics**: Defect rates, code coverage, performance benchmarks - **Team Velocity**: Sprint capacity and throughput trends ## TaskManager API Integration Reference CLAUDE.md for complete TaskManager API and CLI usage patterns. ### Automatic Research Report Integration When creating tasks that require research: 1. **Check for existing reports**: Look for `./development/research-reports/research-report-{related_task_id}.md` 2. **Include in important_files**: Add research report path to task important_files 3. **Update success criteria**: Include research report creation/review in success criteria ### Research Task Integration Pattern ```javascript // For research tasks, automatically include research report const researchTask = { title: "Research Authentication Patterns", mode: "research", important_files: [ `./development/research-reports/research-report-${taskId}.md`, ...otherFiles ], success_criteria: [ "Research questions answered with evidence", "Recommendation documented with pros/cons", `Research report created: ./development/research-reports/research-report-${taskId}.md`, ...otherCriteria ] }; // For implementation tasks that depend on research const implementationTask = { title: "Implement Authentication System", mode: "development", important_files: [ `./development/research-reports/research-report-${researchTaskId}.md`, // Reference research ...implementationFiles ], dependencies: [researchTaskId] }; ``` ### Quick Task Creation Commands ```bash # Create feature task with research report integration node -e "const TaskManager = require('./lib/taskManager'); const tm = new TaskManager('./TODO.json'); /* task creation logic with research integration */" # Add subtasks to existing task node -e "const TaskManager = require('./lib/taskManager'); const tm = new TaskManager('./TODO.json'); tm.addSubtask('parent_id', {...})" # Update task priority/status node -e "const TaskManager = require('./lib/taskManager'); const tm = new TaskManager('./TODO.json'); tm.updateTaskStatus('task_id', 'in_progress')" ```