ralph

Coordinate PRD-driven task execution until all user stories pass verification.

1|1|Updated Apr 13, 2026
One-click install
npx skills add https://github.com/Cheggin/request-for-startups --skill ralph-cheggin
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ralph
Source: https://github.com/Cheggin/request-for-startups/tree/main/skills/ralph
Command: npx skills add https://github.com/Cheggin/request-for-startups --skill ralph-cheggin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ralph addresses the challenge of ensuring long-running tasks are completed with explicit, reviewer-verified acceptance criteria. It provides a PRD-driven loop that persists across retries and iterations so nothing falls through the cracks.

Core Features & Use Cases

  • PRD-driven iteration: splits work into discrete user stories with verifiable criteria and tracks progress across retries.
  • Session persistence and automatic retry: keeps state and re-attempts work until all conditions are met.
  • Fresh reviewer verification: requires up-to-date sign-off against acceptance criteria before completion.
  • Progress visibility: maintains prd.json and progress.txt to reveal learnings and patterns over time.
  • Multi-task orchestration: supports coordinating complex tasks that span multiple iterations.

Quick Start

Describe your task to begin an iterative, verifier-backed execution flow with Ralph.

Frequently Asked Questions about ralph

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

FAQPage Schema
How do I automate long-running task execution with automatic retries until acceptance criteria pass?

Automating long-running task execution with retries requires a PRD-driven loop that persists across iterations, splitting work into discrete user stories with verifiable criteria and re-attempting until all conditions are met.

What is PRD-driven task orchestration and when do I need it for multi-iteration workflows?

PRD-driven task orchestration coordinates complex multi-iteration workflows by enforcing structured acceptance criteria, progress tracking, and mandatory verification before completion, ensuring nothing falls through the cracks.

How do I ensure reviewer sign-off is required before completing a complex workflow?

Ensuring reviewer sign-off requires an execution flow that demands fresh, up-to-date verification against acceptance criteria before a task can be marked complete, preventing unreviewed deliveries.

Can I track progress and learnings across multiple task iterations automatically?

Tracking progress across iterations is supported by maintaining dedicated state files like prd.json and progress.txt, which automatically log execution state, learnings, and patterns over time.

Does task orchestration work for complex tasks spanning multiple retries without losing state?

Task orchestration for multi-retry tasks uses session persistence to keep state intact, automatically re-attempting work until all user stories pass verification and explicit reviewer sign-off is achieved.

Why does long-running task automation fail without structured acceptance criteria?

Long-running task automation fails without structured acceptance criteria because there is no verifiable baseline for success, making it impossible to enforce mandatory verification or trigger automatic retries effectively.