ralph

Run PRD-driven persistence loops until all acceptance criteria pass reviewer verification.

Updated May 5, 2026
One-click install
npx skills add https://github.com/HyperionBurn/searchv1beta --skill ralph-hyperionburn
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ralph
Source: https://github.com/HyperionBurn/searchv1beta/tree/main/.github/skills/ralph
Command: npx skills add https://github.com/HyperionBurn/searchv1beta --skill ralph-hyperionburn

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ralph solves the problem of incomplete work by running a PRD-driven persistence loop that continues until every user story’s acceptance criteria are verified as passing, including reviewer confirmation and regression re-checks.

Core Features & Use Cases

  • PRD-scaffolded execution: Generates a PRD scaffold when none exists and converts requirements into task-specific, testable acceptance criteria.
  • Story-by-story persistence: Repeatedly selects the next highest-priority story that has not passed, delegates implementation, then re-verifies until all stories pass.
  • Verified completion with cleanup: Performs optional post-review cleanup and re-runs regression tests to prevent “done but broken” outcomes; uses reviewer verification for final sign-off.
  • Use case: When you need guaranteed completion for a multi-step feature or refactor that may require several iterations, Ralph keeps working until the PRD evidence proves everything is correct.

Quick Start

Activate Ralph by telling the assistant: ralph, don't stop, must complete, finish this, keep going until done.

Frequently Asked Questions about ralph

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

FAQPage Schema
How do I verify task completion against PRD acceptance criteria?

PRD-driven task completion verifies each user story by running iterative execution loops that check acceptance criteria until all stories pass reviewer-verified requirements. It continuously selects and implements the highest-priority incomplete story, then re-verifies.

What's the best way to automate regression testing after code cleanup?

Automated regression testing after cleanup uses a persistence loop that re-runs verification checks to prevent broken outcomes. It performs post-review cleanup passes and re-verifies all acceptance criteria to ensure previous functionality remains intact.

Can I generate a PRD scaffold automatically for iterative engineering tasks?

PRD scaffold generation automatically creates a product requirements document when none exists, converting requirements into task-specific, testable acceptance criteria for multi-iteration engineering workflows requiring evidence-based validation.

How do you handle incomplete user stories in a multi-iteration development workflow?

Incomplete user stories are handled through story-by-story persistence that repeatedly delegates implementation and re-verifies results. The system tracks progress and continues selecting the next highest-priority failing story until all pass reviewer confirmation.

Does PRD-driven verification require manual reviewer confirmation for final sign-off?

PRD-driven verification requires manual reviewer confirmation for final sign-off to ensure evidence-based validation. The system performs delegated implementation, criterion checks, optional cleanup, and regression re-verification before presenting results for reviewer approval.