ralph

Enforce a persistence loop until tasks are architect-verified.

Updated Jan 4, 2026
One-click install
npx skills add https://github.com/YeaChan05/Remittance --skill ralph-yeachan05
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ralph
Source: https://github.com/YeaChan05/Remittance/tree/main/.codex/skills/ralph
Command: npx skills add https://github.com/YeaChan05/Remittance --skill ralph-yeachan05

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ralph provides a persistence loop that keeps working on a task until it is genuinely complete and architect-verified, preventing partial implementations and silent failures.

Core Features & Use Cases

  • Multi-iteration task handling with session persistence across retries to ensure eventual completeness.
  • Automatic retry on failure with mandatory verification before completion.
  • Parallel delegation and architect sign-off at the end to guarantee quality and auditable outcomes.
  • Use Case: When you have a long-running, complex task that requires repeated progress checks and formal verification, Ralph ensures the task is fully done before reporting success.

Quick Start

Instruct Ralph to loop on a task until it is completed and architect-verified.

Frequently Asked Questions about ralph

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

FAQPage Schema
How do I ensure complex task completion with automatic retry on failure?

To ensure complex task completion with automatic retry on failure, you need a persistence loop that continues work until the task is fully completed and verified by an architect. This prevents partial implementations and silent failures during multi-iteration tasks.

What is a persistence loop for multi-iteration task management?

A persistence loop for multi-iteration task management is a mechanism that handles session persistence across retries, enforcing repeated progress checks until a task is genuinely complete. It guarantees eventual completeness through mandatory verification before final completion.

Does architect sign-off verification work for parallel delegation across agents?

Yes, architect sign-off verification works for parallel delegation by applying concurrent execution of independent steps across agents. This guarantees quality and produces auditable outcomes through formal verification at the end of the task.

How do I stop silent failures in long-running automation tasks?

To stop silent failures in long-running automation tasks, apply a persistence loop that enforces automatic retry on failure and mandatory verification. This prevents partial implementations by requiring explicit iteration control until the task is genuinely complete.

When do I need session persistence for task verification?

You need session persistence for task verification when handling long-running, complex tasks that require repeated progress checks and formal verification. It ensures eventual completeness by maintaining state across retries until architect sign-off is achieved.