planning-and-task-breakdown

Planning and Task Breakdown - Intelligent Task Decomposition and Planning

Skill Overview

Planning and Task Breakdown is an AI skill specifically designed to break complex work into actionable and verifiable tasks, helping you plan the complete process from requirements specifications to implementation tasks.

Applicable Scenarios

1. From Requirements Specifications to an Implementation Plan

When you have complete requirements documentation or clearly defined functional requirements, this skill can help transform abstract requirements into a concrete, actionable task list. It is particularly suitable for breaking down large functional modules into multiple small, independent, and testable implementation units.

2. Handling Ambiguous or Overly Complex Tasks

When a task feels too large, vague, or difficult to start directly, this skill transforms it into a clear implementation path through dependency analysis, task-size assessment, and the establishment of validation criteria. Each task is equipped with explicit acceptance criteria and verification steps.

3. Multi-Agent Collaboration and Parallelization

When work needs to be carried out in parallel across multiple AI agents or development sessions, this skill can identify independent task slices, plan safe parallelization strategies, and clarify which tasks must be executed sequentially. It helps teams collaborate efficiently while avoiding dependency conflicts.

Core Functions

1. Dependency Mapping and Task Ordering

By constructing a complete dependency graph, the skill identifies dependencies among database, API, frontend component, and other layers, ensuring that tasks are executed in the correct order. The system follows a bottom-up implementation principle: build the infrastructure first, then implement higher-level functionality.

2. Vertical Slicing and Task Size Control

The skill uses a vertical slicing approach, with each task containing a complete implementation path (such as database schema + API endpoint + UI component), rather than following traditional horizontal layering. Task-sizing guidelines keep the work within the S to M range (1–5 files), ensuring that each task can be completed and verified in a single focused session.

3. Checkpoints and Validation Criteria

Validation checkpoints are established at key stages, and each task is equipped with clear acceptance criteria and verification steps (tests passing, successful builds, and manual checks). This ensures that the system remains operational at all times and that issues can be identified and fixed early.

Frequently Asked Questions

What is a task dependency graph, and why is it needed?

A task dependency graph is a visual structure that shows the dependencies among tasks, such as “database schema → API model → API endpoint → frontend component.” It helps identify the implementation order and prevents teams from building features that depend on unfinished foundational components, thereby reducing rework and integration issues.

How do you determine whether a task is appropriately sized?

The skill provides clear task-sizing guidelines: XS (1 file), S (1–2 files), and M (3–5 files). If a task involves more than 5 files or requires multiple sessions to complete, it is too large and should be broken down further. An appropriately sized task should be completable within 2 hours, with no more than 3 acceptance criteria.

What is the difference between vertical slicing and horizontal slicing?

Horizontal slicing divides work by technical layer, completing all database work first, followed by all API work, and finally all UI work. This approach can lead to integration issues emerging late. Vertical slicing means that each task includes the complete implementation path for a feature, from the database to the UI. Each slice can function and be tested independently, making rapid validation and iteration easier.