ralph-mode

Implements iterative cycles of coding, testing/typechecking, and committing verified changes.

15|3|Updated Jan 27, 2026
One-click install
npx skills add https://github.com/richardcb/oh-my-gemini --skill ralph-mode-richardcb
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ralph-mode
Source: https://github.com/richardcb/oh-my-gemini/tree/main/skills/ralph-mode
Command: npx skills add https://github.com/richardcb/oh-my-gemini --skill ralph-mode-richardcb

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a structured protocol for systematically completing tasks, especially when facing errors or needing to iterate on solutions. It ensures tasks are worked through methodically and reliably.

Core Features & Use Cases

  • Systematic Iteration: Guides the AI through implementing, verifying, and committing solutions.
  • Error Handling: Explicitly addresses how to handle failing tests or errors by analyzing, retrying, or documenting blockers.
  • Commitment Strategy: Ensures only passing code is committed, maintaining a working state.
  • Use Case: When building a new feature, the AI will follow this protocol to implement, test, and commit each part of the feature until it's fully complete and verified.

Quick Start

Activate ralph-mode to systematically complete the current task by implementing, verifying, and committing the solution.

Frequently Asked Questions about ralph-mode

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

FAQPage Schema
How do I ensure autonomous task completion in AI agents without breaking the codebase?

Autonomous task completion requires a persistent workflow that verifies code via tests and typechecking before committing. This ensures only passing code is committed, maintaining a working state while systematically iterating through feature implementation.

What is the best way to handle failing tests and errors during AI code iteration?

Handling failing tests during code iteration involves analyzing the failure, retrying the implementation, or documenting the blocker before proceeding. This structured error handling protocol ensures the agent systematically works through errors instead of abandoning the task.

How do I guide an AI agent to systematically implement, verify, and commit code?

To systematically implement, verify, and commit code, you need a structured protocol that guides the agent through iterative development. The workflow ensures each solution part passes verification tests before a code commit is executed.

Can I use a persistent workflow protocol for complex feature development?

Yes, a persistent workflow protocol is designed for complex feature development. When building a new feature, the protocol guides the agent to implement, test, and commit each part methodically until the feature is fully complete and verified.

What happens when an AI agent encounters an unresolved blocker during code commit?

When encountering an unresolved blocker during code commit, the agent addresses it by analyzing the failure, retrying the solution, or explicitly documenting the blocker. This ensures the task completion process proceeds reliably without committing broken code.