ralph-plan

Decompose PRD user stories into structured implementation subtasks with dependencies.

Updated Jan 22, 2026
One-click install
npx skills add https://github.com/doravidan/supreme-ralph --skill ralph-plan
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ralph-plan
Source: https://github.com/doravidan/supreme-ralph/tree/main/templates/global/skills/ralph-plan
Command: npx skills add https://github.com/doravidan/supreme-ralph --skill ralph-plan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps teams translate PRD user stories into concrete, ordered implementation tasks using the Planner agent, reducing ambiguity and accelerating delivery.

Core Features & Use Cases

  • Subtask decomposition: Break PRD narratives into atomic, testable subtasks with clear files to create/modify.
  • Context-aware planning: Leverages PRD data (prd.json), PROJECT_SPEC.md, CLAUDE.md, and memory insights to prioritize work and surface risks.
  • Execution-ready plans: Outputs a structured plan with IDs, scopes, dependencies, acceptance criteria, and complexity estimates.

Quick Start

Run /ralph-plan to generate a full implementation plan from prd.json. You can also plan a single story with /ralph-plan US-001 or refresh incomplete plans with /ralph-plan --refresh.

Frequently Asked Questions about ralph-plan

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

FAQPage Schema
How do I break down PRD user stories into actionable implementation subtasks?

You can decompose PRD user stories into actionable implementation subtasks using the ralph-plan Skill, which generates structured tasks with IDs, file scopes, dependencies, and acceptance criteria.

How does context-aware task generation from product requirements work?

Context-aware task generation from product requirements leverages prd.json, PROJECT_SPEC.md, CLAUDE.md, and memory insights to prioritize work, surface risks, and output structured subtask schemas with dependency graphs.

Can I generate an implementation plan for a single PRD user story instead of the whole project?

Yes, you can generate an implementation plan for a single PRD user story by running the command with a specific story ID like US-001, or refresh incomplete plans with a refresh flag.

What's the best way to estimate complexity and identify parallelizable tasks from a PRD?

To estimate complexity and identify parallelizable tasks from a PRD, use a Planner agent that outputs structured schemas including estimated_complexity fields and parallelizable flags for dependency graph mapping.

Do I need a prd.json file to generate structured subtasks with dependency graphs?

Yes, a prd.json file is required to generate structured subtasks with dependency graphs, as the Planner agent uses it alongside PROJECT_SPEC.md and CLAUDE.md for context-aware task decomposition.

What information is included in a structured subtask schema for implementation planning?

A structured subtask schema for implementation planning includes id, storyId, title, description, files_to_create, files_to_modify, dependencies, acceptance_criteria, estimated_complexity, and parallelizable fields.

Related Skills