task-decomposition

📁 d-o-hub/rust-self-learning-memory 📅 6 days ago
9
总安装量
9
周安装量
#31489
全站排名
安装命令
npx skills add https://github.com/d-o-hub/rust-self-learning-memory --skill task-decomposition

Agent 安装分布

opencode 9
claude-code 9
github-copilot 9
codex 9
kimi-cli 9
gemini-cli 9

Skill 文档

Task Decomposition

Break down complex tasks into atomic, actionable goals with clear dependencies.

When to Use

  • Complex user requests with multiple components
  • Multi-phase projects requiring coordination
  • Tasks that could benefit from parallel execution
  • Planning agent coordination strategies

Decomposition Framework

1. Requirements Analysis

  • Primary objective
  • Implicit requirements (quality, performance)
  • Constraints (time, resources)
  • Success criteria

2. Goal Hierarchy

Main Goal
├─ Sub-goal 1
│  ├─ Task 1.1 (atomic)
│  └─ Task 1.2 (atomic)
├─ Sub-goal 2
└─ Sub-goal 3

3. Dependency Types

Type Symbol Example
Sequential A → B → C B needs A’s output
Parallel A─┐ B─┐ C─┘ Independent, concurrent
Converging A─┐ B─┼─> D D needs A, B, C
Resource A, B Sequential or pooled

4. Success Criteria

For each task:

  • Input: What data/state is needed
  • Output: What artifacts will be produced
  • Quality: Performance, testing, docs requirements

Decomposition Patterns

Pattern Use Case
Layer-Based Architectural changes (data, logic, API, test, docs)
Feature-Based New features (MVP, error handling, optimization, integration)
Phase-Based Large projects (research, foundation, core, integration, polish)
Problem-Solution Debugging (reproduce, diagnose, design, fix, verify, prevent)

Quality Checklist

✓ Atomic and actionable ✓ Dependencies clearly identified ✓ Success criteria measurable ✓ No task too large (>4 hours) ✓ Parallelization opportunities identified

✗ Tasks too large or vague ✗ Missing dependencies ✗ Unclear success criteria ✗ Missing quality/testing tasks

Integration with GOAP

Task decomposition is Phase 1 of GOAP:

  1. Receive request
  2. Apply decomposition
  3. Create execution plan
  4. Execute with monitoring
  5. Report results