ralph

Orchestrates Ralph loops for AI-driven development planning and execution via rp-cli and YOLO mode.

Updated Jan 2, 2026
One-click install
npx skills add https://github.com/carmandale/ralph-loop --skill ralph-carmandale
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ralph
Source: https://github.com/carmandale/ralph-loop/tree/main/pi-integration/skill
Command: npx skills add https://github.com/carmandale/ralph-loop --skill ralph-carmandale

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates the end-to-end process of planning, interviewing, reviewing, and executing structured Ralph loops for AI-driven development in YOLO mode, reducing manual coordination and ensuring repeatable workflows.

Core Features & Use Cases

  • Guided workflow: From initial research and interviews to final execution in YOLO mode.
  • Collaborative validation: Integrates with rp-cli for expert design and plan reviews.
  • End-to-end automation: Atomic tasks with integration points and codex reviews to accelerate delivery.

Quick Start

  • Start a Ralph loop plan for a feature: ralph plan for "<feature name>"
  • Draft tasks and conduct interviews, then trigger YOLO execution: ralph yolo
  • After execution, review Codex results and iterate as needed: codex review

Frequently Asked Questions about ralph

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

FAQPage Schema
How do I automate AI development planning and execution in YOLO mode?

AI development planning and execution in YOLO mode is automated by orchestrating structured Ralph loops, which handle research, interviews, and feature execution. You initiate a plan for a feature, draft atomic tasks, and trigger automated execution for repeatable outcomes.

What is a Ralph loop and how does it structure AI-driven development?

A Ralph loop structures AI-driven development by enforcing atomic task creation with integration points and codex-driven reviews. It guides the workflow from initial research and interviews through to final YOLO-mode execution, ensuring verifiable and repeatable delivery.

How do I start a Ralph loop plan for a new feature?

To start a Ralph loop plan for a new feature, use the command syntax 'ralph plan for' followed by your feature name. This initiates the guided workflow, allowing you to draft tasks and conduct interviews before triggering YOLO execution.

Does this workflow integrate expert design reviews before YOLO execution?

Yes, the workflow integrates collaborative validation through rp-cli for expert design and plan reviews before YOLO execution. This ensures that atomic tasks and integration steps are validated prior to automated, codex-driven delivery.

Can I review execution results and iterate after YOLO mode completes?

Yes, after YOLO mode execution completes, you can review codex results and iterate as needed. The command 'codex review' allows you to inspect the outcomes of the automated execution and make adjustments for future loops.

When should I use YOLO mode execution for AI development?

YOLO mode execution should be used when you need to accelerate delivery of AI-driven feature work through end-to-end automation. It is best suited after completing guided planning, interviews, and expert reviews to ensure repeatable, verifiable outcomes.