shape-to-ralph

Convert shaping artifacts into Ralph prd.json with quality gates and task files.

5|Updated Mar 5, 2026
One-click install
npx skills add https://github.com/adelrioj/claude-skills --skill shape-to-ralph
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: shape-to-ralph
Source: https://github.com/adelrioj/claude-skills/tree/main/skills/shape-to-ralph
Command: npx skills add https://github.com/adelrioj/claude-skills --skill shape-to-ralph

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq, git, codex, claude, and includes scripts (resource) components.

What problem does it solve?

Convert shaping outputs (breadboard and slices) directly into Ralph prd.json for autonomous execution. Bypasses plan generation — shaping artifacts ARE the spec.

Core Features & Use Cases

  • Convert shaping artifacts into Ralph's prd.json format, using the shaping doc as the source of truth.
  • Inject per-slice quality gates and wiring by mapping slices to user stories and generating progress and findings seeds.
  • Enable deterministic execution with two execution scripts (ralph.sh for Claude-based runs and ralph-codex.sh for Codex) and a standardized output layout for tasks/prd.json, tasks/progress.txt, and tasks/findings.md.

Quick Start

Run the shape-to-ralph workflow to convert your current shaping results into a Ralph PRD and start the autonomous loop.

Frequently Asked Questions about shape-to-ralph

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

FAQPage Schema
How do I convert shaping artifacts into Ralph prd.json for autonomous execution?

Converting shaping artifacts into Ralph prd.json requires mapping breadboard and slices to user stories, injecting quality gates, and generating standard task files like progress.txt and findings.md. This bypasses plan generation by using shaping artifacts directly as the spec.

What is the shape-to-ralph workflow for story mapping?

The shape-to-ralph workflow maps slices to Ralph user stories and generates standard task files including prd.json, progress.txt, and findings.md in the project root, enabling deterministic autonomous execution without regenerating requirements.

Do I need Claude or Codex to run Ralph autonomous execution scripts?

You need either Claude or Codex to run Ralph autonomous execution. The workflow provides ralph.sh for Claude-based runs and ralph-codex.sh for Codex, allowing deterministic execution using your preferred AI tool.

What dependencies are required to generate Ralph PRDs from shaping slices?

Generating Ralph PRDs from shaping slices requires jq, git, codex, and claude installed in your environment. These dependencies support artifact conversion, autonomous execution, and repository management throughout the workflow.

Why does autonomous execution fail when shaping artifacts are missing?

Autonomous execution fails without selected shape, breadboard, and slices because shaping artifacts ARE the spec. The workflow applies only when shaping assets exist with wiring that maps slices to Ralph user stories, bypassing plan generation entirely.

What's the best way to inject quality gates into Ralph user stories from slices?

The best way to inject quality gates is by mapping slices to Ralph user stories during artifact conversion. This generates prd.json with per-slice quality gates and wiring, alongside progress and findings seeds for end-to-end execution.