Spawn Task Skill
Create detailed task prompts that another agent (or future self) can execute without additional context.
When to Use
-
Breaking down complex features
-
Creating implementation specs
-
When user says "spawn a task for..."
-
Planning multi-phase work
Task Discovery
Gather from user:
-
Task Description: What needs to be done?
-
Priority: High / Medium / Low
-
Scope: What's included/excluded?
-
Timeline: Expected completion (optional)
Current State Analysis
Check related files
find src -name "relevant" -type f
Check documentation
ls docs/features/
Check GitHub issues
gh issue list --search "[related term]"
Task Prompt Template
Create docs/tasks/[TaskName].md :
🎯 Task: [Task Name]
Objective: [Clear, measurable goal] Priority: [High/Medium/Low] Scope: [What's included/excluded]
📋 Requirements
Functional Requirements
- [REQ-001] [Requirement with acceptance criteria]
Technical Requirements
- [TECH-001] [Technical specification]
🏗️ Implementation Plan
Phase 1: Setup
- [Setup task]
Phase 2: Core Implementation
- [Core task]
Phase 3: Enhancement
- [Enhancement task]
Phase 4: Testing & Documentation
- [Testing task]
📁 Files to Create/Modify
| File | Action | Purpose |
|---|---|---|
src/features/[Feature]/... | Create | [Purpose] |
✅ Success Criteria
Code Quality
- TypeScript compliant (no
any) - Passes ESLint
- Unit tests with >80% coverage
Performance
- Bundle size increase < 5KB
- Load time < 1.5s
Functionality
- All requirements implemented
- Mobile responsive
🔗 Dependencies
Depends on: [Other components] Used by: [Components that use this]
🚀 Getting Started
- Read this task prompt completely
- Review related files
- Begin with Phase 1
- Provide progress updates after each phase
Options After Spawning
-
A) Start implementing now
-
B) Create GitHub Issue to track
-
C) Save task for later
Task Categories
-
🆕 Feature: New functionality
-
🐛 Bug Fix: Fixing broken behavior
-
♻️ Refactor: Improving existing code
-
📚 Documentation: Writing docs
Priority Guidelines
-
High: Blocking other work, user-facing bugs
-
Medium: Important features, non-blocking issues
-
Low: Nice-to-haves, minor improvements