Generate specs, implementation plans, prompts, and loop scripts for iterative development.
npx skills add https://github.com/hmemcpy/ralph-wiggum --skill ralph-hmemcpy
Please help me install this Agent Skill. Skill: ralph Source: https://github.com/hmemcpy/ralph-wiggum/tree/main/agents/ralph Command: npx skills add https://github.com/hmemcpy/ralph-wiggum --skill ralph-hmemcpy
The Ralph Planning Skill reduces friction in AI-driven development by coordinating clarifying questions, plan generation, and artifact creation (specs, implementation plans, prompts, and loop scripts) in a single, repeatable flow.
Use the skill to kick off an interactive planning session. Run /skill ralph [optional/path/to/plan.md] to base the plan on a provided plan file or the current conversation context. The skill will generate the necessary artifacts and a reusable loop for iterative development.
High-intent search queries and answers about installing and using this skill.
Ralph automates interactive planning by collecting requirements through clarifying questions, then generates specs, implementation plans, and build prompts in a single workflow. Run `/skill ralph` to start an interactive session that outputs structured artifacts for iterative AI development.
Yes. Ralph creates repeatable loop scripts and infrastructure artifacts—including planning and build prompts—designed to coordinate multiple agents like Amp and Claude Code across iterative development cycles. The generated loop.sh enables autonomous iteration without manual intervention.
Ralph condenses requirements gathering, spec generation, and loop assembly into one flow. Provide optional context via a plan.md file or conversation, and Ralph delivers all artifacts—specs, implementation plans, prompts, and executable loop scripts—immediately.
Ralph is agent-agnostic and designed to work with multiple agents including Amp and Claude Code. It generates metadata-driven artifacts and loop infrastructure compatible with any agent, making it flexible across different iterative development setups.
Ralph presents A/B/C/D option sets to interactively narrow scope, define validation criteria, and lock in requirements before artifact generation. This structured interview ensures the generated specs and plans align with your actual development constraints.
Yes. Ralph includes an optional Oracle review step that analyzes your architecture before proceeding to plan and artifact generation, letting you validate the approach before committing to the full workflow.
fredericpoitevinAsk clarifying questions to define research goals and scope before starting.
javierhbrAutomate iterative openspec task implementation with Ralph Wiggum loops.
liangdabiaoCoordinate long-running AI tasks with state persistence and checkpoints.
Bliz243Initialize Ralph workflow scaffolds with loop.sh, prompts, and specs directory.
