How does the 'reverse review' process work?

It validates documentation in reverse order: verifying that tasks match the design, and then verifying the design matches the original requirements, ensuring no features are missed or unnecessarily added.

What is the ideal task size for this skill?

The skill targets tasks that take 20-40 minutes for an AI agent to complete. Tasks estimated at over 90 minutes are automatically flagged for further decomposition to ensure high success rates.

Does this skill support Test-Driven Development?

Yes, every task includes a dedicated TDD section that defines specific test file paths, test cases, and the 'Red-Green-Refactor' cycle for the AI agent to follow.

Where does the skill store the task data?

All tasks are stored as Markdown files within a local `docs/tasks/` directory, organized by phases (e.g., phase-1, phase-2) for easy version control and reference.

AI Task Planning & Decomposition

Name: AI Task Planning & Decomposition
Author: windschord

bywindschord

•

生产力与工作流

Transforms technical designs into granular, TDD-ready implementation tasks optimized for AI coding agents.

关于

The Task Planning skill bridges the gap between high-level technical design and execution by breaking down complex requirements into actionable, small-scale tasks. It manages a structured documentation system (docs/tasks/) that includes a central index and detailed task specifications with specific file paths, TDD (Test-Driven Development) steps, and clear acceptance criteria. By utilizing a unique 'reverse review' process, this skill ensures that every implementation task remains strictly aligned with technical designs and original project requirements, preventing scope creep and ensuring architectural integrity.

主要功能

1 GitHub stars
TDD-integrated implementation workflows with step-by-step instructions
Centralized progress tracking with phase-based task management
Automated task decomposition based on design specifications
Standardized task granularity (20-40 minute AI work windows)
Reverse review validation between tasks, design, and requirements

使用场景

Converting a fresh design document into a structured implementation roadmap
Maintaining document consistency when requirements or designs are updated mid-project
Generating precise, file-specific instructions for AI agents like Claude Code

关于

主要功能

1 GitHub stars
TDD-integrated implementation workflows with step-by-step instructions
Centralized progress tracking with phase-based task management
Automated task decomposition based on design specifications
Standardized task granularity (20-40 minute AI work windows)
Reverse review validation between tasks, design, and requirements

使用场景

Converting a fresh design document into a structured implementation roadmap
Maintaining document consistency when requirements or designs are updated mid-project
Generating precise, file-specific instructions for AI agents like Claude Code