Why does this skill focus so heavily on questioning?

The skill follows the philosophy 'Ask 5 questions upfront rather than have Ralph fail 5 iterations,' ensuring the AI agent has complete clarity before starting implementation.

Does this support existing codebases?

Yes, it includes a dedicated 'Brownfield' workflow specifically designed to analyze existing structures, patterns, and conventions before proposing changes.

How does feature sizing work in Ralph PRD Author?

The skill enforces rules where features must be right-sized (e.g., one model or one API endpoint) and avoids 'Too Large' tasks that touch more than 4 files or 2 concepts.

What are machine-verifiable criteria?

These are requirements written with specific, objective outcomes (like HTTP status codes or specific JSON fields) that an AI agent or automated test can verify without human judgment.

What is the Ralph-RLM Framework?

Ralph-RLM is an autonomous agent framework that utilizes structured Product Requirements Documents (PRDs) to implement software through a three-phase autonomous loop.

Ralph PRD Author

Name: Ralph PRD Author
Author: Was85

byWas85

0•

Productivity & Workflow

Guides users through creating structured, machine-verifiable Product Requirements Documents optimized for autonomous coding agents.

The Ralph PRD Author skill is an interactive assistant designed to eliminate ambiguity in software development by facilitating the creation of high-quality PRDs. It uses a multi-phase questioning strategy to capture project scope, tech stacks, quality gates, and test requirements specifically for the Ralph-RLM agent framework. By ensuring every requirement is machine-verifiable and properly sized, it minimizes implementation failures and allows autonomous agents to decompose tasks into executable features with high precision and reliability.

Key Features

01Enforcement of machine-verifiable acceptance criteria to ensure testability

02Interactive 5-phase structured questioning for comprehensive requirements gathering

03Intelligent feature sizing heuristics to prevent overly complex implementation tasks

04Mandatory quality gate integration including build, test, and lint commands

05Dependency analysis for establishing logical implementation and build order

060 GitHub stars

Use Cases

01Documenting complex bug fixes with specific regression test requirements for AI agents

02Bootstrapping new Greenfield projects with AI-ready documentation and structure

03Defining new features for existing Brownfield codebases while maintaining patterns

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add was85/ralph-rlm-agent-framework ralph

For use in Claude.ai and ChatGPT

Download Skill