ralph — PRD変換 supreme-ralph, community, PRD変換, ide skills, prd.json形式, RALPH自主開発, コマンドラインインターフェース, プロジェクト状態表示, PRD検証, AIエージェント

v1.0.0

このスキルについて

開発エージェントがプロジェクトのワークフローと PRD 管理を効率化する必要がある場合に最適 RALPHスキルは、PRD Markdownファイルをprd.json形式に自動変換するツールです

機能

PRD Markdownファイル変換
prd.json形式対応
RALPH自主開発プロジェクト管理
コマンドラインインターフェース対応
プロジェクト状態表示
PRD検証

# Core Topics

doravidan doravidan
[0]
[0]
Updated: 1/26/2026

Killer-Skills Review

Decision support comes first. Repository text comes second.

Reference-Only Page Review Score: 10/11

This page remains useful for operators, but Killer-Skills treats it as reference material instead of a primary organic landing page.

Original recommendation layer Concrete use-case guidance Explicit limitations and caution Quality floor passed for review
Review Score
10/11
Quality Score
60
Canonical Locale
en
Detected Body Locale
en

開発エージェントがプロジェクトのワークフローと PRD 管理を効率化する必要がある場合に最適 RALPHスキルは、PRD Markdownファイルをprd.json形式に自動変換するツールです

このスキルを使用する理由

エージェントが PRD Markdown ファイルを prd.json 形式に変換し、prd.json を検証し、`/ralph --status` や `/ralph --analyze` などのトリガーを使用して自律的な開発を管理し、JSON 形式の互換性でプロジェクトのワークフローを合理化することを可能にします

おすすめ

開発エージェントがプロジェクトのワークフローと PRD 管理を効率化する必要がある場合に最適

実現可能なユースケース for ralph

PRD Markdown ファイルを prd.json に変換する
prd.json のエラーを検証する
`/ralph --analyze` を使用してプロジェクトを再分析して仕様を更新する

! セキュリティと制限

  • PRD Markdown ファイルが必要
  • prd.json 形式の変換と管理のみ

Why this page is reference-only

  • - Current locale does not satisfy the locale-governance contract.

Source Boundary

The section below is imported from the upstream repository and should be treated as secondary evidence. Use the Killer-Skills review above as the primary layer for fit, risk, and installation decisions.

After The Review

Decide The Next Action Before You Keep Reading Repository Material

Killer-Skills should not stop at opening repository instructions. It should help you decide whether to install this skill, when to cross-check against trusted collections, and when to move into workflow rollout.

Labs Demo

Browser Sandbox Environment

⚡️ Ready to unleash?

Experience this Agent in a zero-setup browser environment powered by WebContainers. No installation required.

Boot Container Sandbox

FAQ & Installation Steps

These questions and steps mirror the structured data on this page for better search understanding.

? Frequently Asked Questions

What is ralph?

開発エージェントがプロジェクトのワークフローと PRD 管理を効率化する必要がある場合に最適 RALPHスキルは、PRD Markdownファイルをprd.json形式に自動変換するツールです

How do I install ralph?

Run the command: npx killer-skills add doravidan/supreme-ralph/ralph. It works with Cursor, Windsurf, VS Code, Claude Code, and 19+ other IDEs.

What are the use cases for ralph?

Key use cases include: PRD Markdown ファイルを prd.json に変換する, prd.json のエラーを検証する, `/ralph --analyze` を使用してプロジェクトを再分析して仕様を更新する.

Which IDEs are compatible with ralph?

This skill is compatible with Cursor, Windsurf, VS Code, Trae, Claude Code, OpenClaw, Aider, Codex, OpenCode, Goose, Cline, Roo Code, Kiro, Augment Code, Continue, GitHub Copilot, Sourcegraph Cody, and Amazon Q Developer. Use the Killer-Skills CLI for universal one-command installation.

Are there any limitations for ralph?

PRD Markdown ファイルが必要. prd.json 形式の変換と管理のみ.

How To Install

  1. 1. Open your terminal

    Open the terminal or command line in your project directory.

  2. 2. Run the install command

    Run: npx killer-skills add doravidan/supreme-ralph/ralph. The CLI will automatically detect your IDE or AI agent and configure the skill.

  3. 3. Start using the skill

    The skill is now active. Your AI agent can use ralph immediately in the current project.

! Reference-Only Mode

This page remains useful for installation and reference, but Killer-Skills no longer treats it as a primary indexable landing page. Read the review above before relying on the upstream repository instructions.

Upstream Repository Material

The section below is imported from the upstream repository and should be treated as secondary evidence. Use the Killer-Skills review above as the primary layer for fit, risk, and installation decisions.

Upstream Source

ralph

Install ralph, an AI agent skill for AI agent workflows and automation. Review the use cases, limitations, and setup path before rollout.

SKILL.md
Readonly
Upstream Repository Material
The section below is imported from the upstream repository and should be treated as secondary evidence. Use the Killer-Skills review above as the primary layer for fit, risk, and installation decisions.
Supporting Evidence

RALPH Skill - PRD Conversion & Management

Convert PRD Markdown files to prd.json format and manage RALPH autonomous development.

Triggers

This skill activates when:

  • /ralph - Show status and help
  • /ralph --status - Show detailed PRD status
  • /ralph --validate - Validate prd.json
  • /ralph --reset - Reset progress.txt for fresh start
  • /ralph --analyze - Re-analyze project and update specs
  • /ralph-convert <file> - Convert PRD markdown to prd.json

Commands

/ralph - Status & Help

Show current PRD status and available commands:

bash
1# Check if prd.json exists 2if [ -f prd.json ]; then 3 echo "=== RALPH Status ===" 4 cat prd.json | jq '{ 5 project: .project, 6 branch: .branchName, 7 total: (.userStories | length), 8 complete: ([.userStories[] | select(.passes == true)] | length), 9 remaining: ([.userStories[] | select(.passes == false)] | length) 10 }' 11 12 echo "" 13 echo "=== Stories ===" 14 cat prd.json | jq -r '.userStories[] | "\(.id): \(.title) [\(if .passes then "✓" else "○" end)]"' 15else 16 echo "No prd.json found." 17 echo "" 18 echo "Create one with:" 19 echo " /prd [feature description]" 20fi

/ralph --status - Detailed Status

bash
1echo "=== PRD Status ===" 2cat prd.json | jq '.' 3 4echo "" 5echo "=== Progress Log (last 50 lines) ===" 6tail -50 progress.txt 2>/dev/null || echo "No progress.txt found" 7 8echo "" 9echo "=== Git Status ===" 10git status --short 11git log --oneline -5

/ralph --validate - Validate PRD

Check prd.json for issues:

bash
1# Validation checks: 2# 1. JSON is valid 3# 2. Required fields exist 4# 3. All stories have acceptance criteria 5# 4. Stories have quality gate criteria 6# 5. Priorities are sequential 7# 6. Branch name follows convention

Validation Rules:

  • project - Required, non-empty string
  • branchName - Required, format: ralph/[slug]
  • userStories - Required, non-empty array
  • Each story must have:
    • id - Format: US-XXX
    • title - Non-empty string
    • acceptanceCriteria - Array with at least 2 items
    • priority - Number 1-10
    • passes - Boolean

Output:

Validating prd.json...

✓ JSON is valid
✓ Project name: [name]
✓ Branch: ralph/[slug]
✓ Stories: [N] total

Story Validation:
  US-001: [Title] ✓
  US-002: [Title] ✓
  ...

⚠ Warnings:
  - US-003: Missing "Tests pass" in acceptance criteria
  - US-005: Large story (6+ criteria), consider splitting

✓ PRD is valid and ready for RALPH

/ralph --reset - Reset Progress

Reset progress.txt while preserving patterns:

bash
1# Archive current progress 2if [ -f progress.txt ]; then 3 mkdir -p archive/$(date +%Y-%m-%d) 4 cp progress.txt archive/$(date +%Y-%m-%d)/progress-backup.txt 5fi 6 7# Extract patterns section from current progress 8PATTERNS=$(sed -n '/## Codebase Patterns/,/^---$/p' progress.txt 2>/dev/null) 9 10# Get branch name from prd.json 11BRANCH=$(cat prd.json | jq -r '.branchName') 12 13# Create fresh progress.txt 14cat > progress.txt << EOF 15# Progress Log - $BRANCH 16 17Reset: $(date +%Y-%m-%d) 18 19$PATTERNS 20 21--- 22 23EOF 24 25echo "Progress reset. Previous progress archived."

/ralph --analyze - Re-analyze Project

Re-run project analysis and update specs:

bash
1# This triggers the project analyzer to refresh: 2# - PROJECT_SPEC.md 3# - scripts/ralph/CLAUDE.md context 4# - progress.txt patterns 5 6node scripts/run-ralph.js --analyze

/ralph-convert <file> - Convert PRD to JSON

Convert a PRD markdown file to prd.json:

Converting: tasks/prd-[feature].md

Reading PRD...
Extracting project info...
Parsing user stories...
Validating structure...

Generated prd.json:
- Project: [Feature Name]
- Branch: ralph/[feature-slug]
- Stories: [N] total

Initializing progress.txt...
Done!

Next: ./scripts/ralph/ralph.sh 20

PRD Conversion Process

Step 1: Read the PRD

bash
1cat tasks/prd-[feature-name].md

Step 2: Extract Information

Parse the PRD to extract:

  • Project name - From # PRD: [Name] title
  • Description - From ## Overview section
  • Project Context - From ## Project Context if present
  • User stories with:
    • ID (US-001, US-002, etc.)
    • Title
    • Description (As a... I want... So that...)
    • Acceptance criteria (bullet points)
    • Priority

Step 3: Generate prd.json

json
1{ 2 "project": "[Feature Name]", 3 "branchName": "ralph/[feature-slug]", 4 "description": "[Overview text]", 5 "createdAt": "[Today's date]", 6 "projectContext": { 7 "name": "[Project name from PROJECT_SPEC.md]", 8 "language": "[typescript/python/go]", 9 "framework": "[react/express/fastapi]", 10 "hasTypes": true, 11 "testFramework": "[vitest/pytest]" 12 }, 13 "existingPatterns": { 14 "moduleSystem": "[ES modules/CommonJS]", 15 "testFramework": "[vitest/jest/pytest]", 16 "linter": "[eslint/biome/ruff]" 17 }, 18 "userStories": [ 19 { 20 "id": "US-001", 21 "title": "[Story title]", 22 "description": "[Full story description]", 23 "acceptanceCriteria": [ 24 "[Criterion 1]", 25 "[Criterion 2]" 26 ], 27 "priority": 1, 28 "passes": false, 29 "notes": "" 30 } 31 ] 32}

Step 4: Archive Previous PRD

If prd.json already exists:

bash
1mkdir -p archive/$(date +%Y-%m-%d) 2cp prd.json archive/$(date +%Y-%m-%d)/prd-backup.json 3cp progress.txt archive/$(date +%Y-%m-%d)/progress-backup.txt 2>/dev/null

Step 5: Initialize Progress

Create fresh progress.txt:

markdown
1# Progress Log - ralph/[feature-slug] 2 3Started: [Date] 4Feature: [Feature description] 5 6## Project Context 7[From PROJECT_SPEC.md if available] 8 9## Codebase Patterns 10[Patterns from analysis or PROJECT_SPEC.md] 11 12## Quality Commands 13```bash 14[typecheck command] 15[lint command] 16[test command]


## Story Conversion Rules

### 1. Story Sizing

If a PRD story is too large, split it:
- Data model → separate story
- Backend logic → separate story
- API endpoint → separate story
- UI component → separate story
- Tests → integrated into each story

### 2. Priority Assignment

Assign priorities based on dependencies:

| Priority | Category | Examples |
|----------|----------|----------|
| 1 | Foundation | Schema, types, interfaces |
| 2 | Core Logic | Services, business logic |
| 3 | API/Backend | Routes, controllers, middleware |
| 4 | UI Components | Forms, displays, interactions |
| 5 | Polish | Optimization, edge cases, docs |

### 3. Required Acceptance Criteria

Always ensure these criteria exist based on tech stack:

**TypeScript/JavaScript:**
- "TypeScript compiles without errors" or "No type errors"
- "ESLint/Biome passes"
- "Tests pass"

**Python:**
- "Type hints complete"
- "Ruff/Pylint passes"
- "Pytest passes"

**Go:**
- "`go build` succeeds"
- "`golangci-lint` passes"
- "`go test ./...` passes"

**For specific story types:**
- UI stories: "Verify in browser", "Accessible"
- API stories: "Response format correct", "Error handling complete"
- Auth stories: "Security best practices followed"

### 4. Branch Naming

Convert feature name to slug:
- "User Authentication" → `ralph/user-authentication`
- "Dark Mode Toggle" → `ralph/dark-mode-toggle`
- Use lowercase, replace spaces with hyphens
- Max 30 characters

## Output Format

After conversion:

=== PRD Converted ===

Project: [Feature Name] Branch: ralph/[feature-slug] Stories: [N] total

Story Summary:

  1. US-001: [Title] (Priority 1) - Foundation
  2. US-002: [Title] (Priority 2) - Core logic
  3. US-003: [Title] (Priority 3) - API ...

Files Created/Updated:

  • prd.json
  • progress.txt

Next Steps:

  1. Review prd.json for accuracy
  2. Create branch: git checkout -b ralph/[feature-slug]
  3. Start RALPH: ./scripts/ralph/ralph.sh 20

## Example Conversion

**Input** (`tasks/prd-user-auth.md`):
```markdown
# PRD: User Authentication

## Overview
Add user authentication with email/password login.

## User Stories

### US-001: Create user model
**As a** developer
**I want** a User model with proper types
**So that** I can store user data securely

**Acceptance Criteria:**
- [ ] User interface with id, email, passwordHash
- [ ] Validation for email format
- [ ] TypeScript compiles

**Priority:** 1

Output (prd.json):

json
1{ 2 "project": "User Authentication", 3 "branchName": "ralph/user-auth", 4 "description": "Add user authentication with email/password login.", 5 "createdAt": "2026-01-22", 6 "userStories": [ 7 { 8 "id": "US-001", 9 "title": "Create user model", 10 "description": "As a developer, I want a User model with proper types so that I can store user data securely", 11 "acceptanceCriteria": [ 12 "User interface with id, email, passwordHash", 13 "Validation for email format", 14 "TypeScript compiles" 15 ], 16 "priority": 1, 17 "passes": false, 18 "notes": "" 19 } 20 ] 21}

Integration with RALPH

After prd.json is created:

  1. Create feature branch:

    bash
    1git checkout -b ralph/[feature-slug]
  2. Start RALPH:

    bash
    1./scripts/ralph/ralph.sh 20
  3. Monitor progress:

    bash
    1tail -f progress.txt 2cat prd.json | jq '.userStories[] | {id, title, passes}'
  4. When complete:

    bash
    1git log --oneline 2# Review changes, create PR

関連スキル

Looking for an alternative to ralph or another community skill for your workflow? Explore these related open-source skills.

すべて表示

openclaw-release-maintainer

Logo of openclaw
openclaw

Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

333.8k
0
AI

widget-generator

Logo of f
f

カスタマイズ可能なウィジェットプラグインをprompts.chatのフィードシステム用に生成する

149.6k
0
AI

flags

Logo of vercel
vercel

React フレームワーク

138.4k
0
ブラウザ

pr-review

Logo of pytorch
pytorch

Pythonにおけるテンソルと動的ニューラルネットワーク(強力なGPUアクセラレーション)

98.6k
0
開発者