ralph-initialize
Initialize a new project with Ralph methodology by creating prd.json backlog, progress.md log, and tech-stack configuration. Use when starting a new project or setting up Ralph for the first time.
Best use case
ralph-initialize is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Initialize a new project with Ralph methodology by creating prd.json backlog, progress.md log, and tech-stack configuration. Use when starting a new project or setting up Ralph for the first time.
Teams using ralph-initialize should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/ralph-initialize/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How ralph-initialize Compares
| Feature / Agent | ralph-initialize | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Initialize a new project with Ralph methodology by creating prd.json backlog, progress.md log, and tech-stack configuration. Use when starting a new project or setting up Ralph for the first time.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Ralph Project Initialization This skill guides you through initializing a project with the Ralph Wiggum Methodology for incremental feature development. ## Prerequisites Check Before starting, verify: - Git repository is initialized - No existing `prd.json` or `progress.md` (or user approves overwrite) ## Initialization Steps ### 1. Environment Sanity Check Check if `prd.json` and `progress.md` exist. If they do, ask the user if they want to overwrite them. If not, exit. ### 2. Goal Acquisition Ask the user: "What do you want to build? Please describe the project, its goals, and key features in detail." Wait for the user's response before proceeding. ### 3. Technical Architecture Check `.windsurf/rules/tech-stack.md`. If it doesn't exist, create it. Ask the user: "What is your preferred tech stack (Language, Framework, Testing Library)?" Write the response into `.windsurf/rules/tech-stack.md` using the template provided in this skill folder. ### 4. Backlog Generation Analyze the user's project description. Break the project down into small, atomic implementation tasks. Populate `prd.json` with these tasks using the template provided. **Constraints:** - Ensure every task has a `passes: false` status - Include clear acceptance criteria for each task - The first task should always be "Setup basic project structure/repository scaffolding" ## Output Files Use these template files from this skill folder: - `.windsurf/skills/ralph-initialize/prd-template.json` - Product Requirements Document structure - `.windsurf/skills/ralph-initialize/progress-template.md` - Progress log format - `.windsurf/skills/ralph-initialize/tech-stack-template.md` - Tech stack documentation format ## Success Criteria Initialization is complete when: - ✅ `prd.json` exists with valid backlog - ✅ `progress.md` exists with initial entry - ✅ `.windsurf/rules/tech-stack.md` exists with tech stack details - ✅ All tasks in backlog have `passes: false`