skills.rest
SanHsien
SanHsienCOMMUNITY
Explore all skills by this author ā†’
š•@Hsien_3GitHub

planning-and-task-breakdown

Decomposes specs into ordered, verifiable implementation tasks with acceptance criteria.

1|Updated Sep 4, 2026
Search Tags:#planning#task breakdown#acceptance criteria#implementation plan#spec#dependency graph#vertical slicing
One-click install
npx skills add https://github.com/SanHsien/agent-skills --skill planning-and-task-breakdown-sanhsien
Or copy as Structured Prompt for Agentā–¼
Please help me install this Agent Skill.
Skill: planning-and-task-breakdown
Source: https://github.com/SanHsien/agent-skills/tree/main/skills/planning-and-task-breakdown
Command: npx skills add https://github.com/SanHsien/agent-skills --skill planning-and-task-breakdown-sanhsien

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Large or vague development tasks cause AI agents to produce tangled, unverifiable work. This Skill turns a spec or clear requirements into an ordered list of small, testable tasks with explicit acceptance criteria, dependencies, and checkpoints. ## Core Features & Use Cases - Dependency-aware task ordering: Maps the dependency graph (schema ā†’ API ā†’ UI) and orders tasks bottom-up so foundations are built first. - Vertical slicing: Breaks work into complete feature slices (e.g., "user can log in") instead of horizontal layers, so every task delivers working functionality. - Structured task format: Each task includes acceptance criteria, verification steps, dependencies, files likely touched, and size estimates (XSā€“XL), with checkpoints every 2-3 tasks. - Use Case: You have a spec for a task-management app. The Skill produces tasks/plan.md with architecture decisions and risks, plus tasks/todo.md listing tasks like "Task 1: User can create an account" with testable acceptance criteria, ready for the /build command to execute. ## Quick Start Ask the agent to read the spec and break the feature into an ordered task list saved to tasks/plan.md and tasks/todo.md.

Frequently Asked Questions about planning-and-task-breakdown

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I break a large feature into tasks for an AI coding agent?ā–¼

Map the dependency graph first, then slice vertically so each task delivers one complete working feature path. Give every task acceptance criteria, verification steps, dependencies, and a size estimate of five files or fewer.

What is vertical slicing in task breakdown?ā–¼

Vertical slicing builds one complete feature path (schema, API, and UI together) per task instead of building all layers separately. Each slice leaves the system in a working, testable state.

How big should a task be for an AI agent to implement reliably?ā–¼

Tasks touching one to five files (S or M size) work best. Anything larger than about eight files, spanning two subsystems, or needing more than three acceptance criteria bullets should be broken down further.

Can tasks be tracked in GitHub Issues or Jira instead of a markdown file?ā–¼

Yes. If the project's agent rules or the user designate an external tracker, create one tracker item per task with acceptance criteria in the body and dependencies via the tracker's linking mechanism, and note the tracker in tasks/plan.md.

What happens if tasks/plan.md already exists with unchecked tasks?ā–¼

The Skill never overwrites an incomplete plan. If the existing unchecked tasks belong to different work, it stops and asks the user whether to finish the old plan, discard it, or write the new plan elsewhere.

When should I skip formal task breakdown?ā–¼

Skip it for single-file changes with obvious scope, or when the spec already contains well-defined tasks. Planning adds value only when work is large, vague, parallelized, or has unclear implementation order.

Related Skills

Explore More Software Engineering Skillsā†’
Codagent-AICodagent-AI
30

spec

Convert unstructured project proposals into user-approved requirement specification files.

Official
Advanced
authrain-cloud-abdullahformuliauthrain-cloud-abdullahformuli

incremental-implementation

Implements code changes in thin, tested vertical slices with per-increment commits.

Community
Intermediate
Ibrahim-3dIbrahim-3d
371

verification-before-completion

Enforce evidence-based verification before marking tasks complete.

Community
Advanced
rmansrmans

scaffold-new-prototype

Create prototype documents with unique PROTO-### IDs and index registration.

Community
Intermediate