ralph

Enforce persistent task loops with architect verification and fresh tests.

28|5|Updated Jan 26, 2026
One-click install
npx skills add https://github.com/MeroZemory/oh-my-droid --skill ralph-merozemory
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ralph
Source: https://github.com/MeroZemory/oh-my-droid/tree/main/skills/ralph
Command: npx skills add https://github.com/MeroZemory/oh-my-droid --skill ralph-merozemory

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ralph stops premature task termination by forcing persistent self-referential loops, architect oversight, and fresh verification so that every job truly finishes with evidence of completion.

Core Features & Use Cases

  • Persistent loop enforcement: Keep iterating on tasks with ultrawork parallelism until every requirement is satisfied and no pending todos remain.
  • Structured PRD support: Detect --prd flags to scaffold product requirement documents, user stories, and progress tracking before entering the Ralph loop.
  • Verification and cleanup: Mandate fresh builds/tests, architect verification, and /cancel state cleanup before claiming completion, ideal for complex product or infrastructure launches.

Quick Start

Ask Ralph to continue the task with ultrawork and architect verification while fresh tests run before promising completion.

Frequently Asked Questions about ralph

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

FAQPage Schema
How do I stop premature task termination in parallel execution workflows?

Mandate completion loop enforcement using architect verification to prevent premature task termination. Persistent self-referential loops keep iterating with ultrawork parallelism until every requirement passes fresh builds and tests.

How does architect verification work for complex software engineering tasks?

Architect verification requires fresh builds and tests alongside mandatory architect confirmation before claiming completion. This prevents incomplete work by enforcing oversight and routing specialist agents until no pending todos remain.

Can I scaffold product requirement documents before entering a task loop?

Yes, you can scaffold product requirement documents by detecting the --prd flag. This generates user stories and progress tracking structures before entering the persistent task loop for structured breakdown.

What's the best way to ensure background processing rules are followed until completion?

Enforce background processing rules by running persistent self-referential loops with ultrawork parallelism. The system mandates fresh verification runs and state cleanup before allowing any task cancellation or completion claim.

Does this task completion approach require specialist agent routing?

Yes, specialist agent routing is supported within the completion loop enforcement. The architect verification process routes tasks to specialist agents and requires fresh verification runs before allowing any cancellation.

Why does my software engineering task keep claiming completion before tests pass?

Tasks claim completion prematurely because they lack architect verification and fresh test enforcement. Forcing persistent loops with mandatory architect confirmation and fresh builds ensures no pending todos remain before finalizing.