Skip to content
Back to skills

Ralph Run Doravidan Supreme Ralph 3

ASecurity

Run RALPH autonomous development loop to implement features from the PRD.

  • 666 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added June 10, 2026
ai-agentsbashnodegit

Security analysis

A100/100

Scanned June 10, 2026

npx -y skills add majiayu000/claude-skill-registry --skill ralph-run-doravidan-supreme-ralph-3 --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Ralph Run Doravidan Supreme Ralph 3?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Ralph Run Doravidan Supreme Ralph 3
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/majiayu000-ralph-run-doravidan-supreme-ralph-3/badge)](https://www.skillsdirectory.com/skills/majiayu000-ralph-run-doravidan-supreme-ralph-3)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: ralph-run
description: Run RALPH autonomous development loop to implement features from the PRD.
allowed-tools: Bash, Read, Edit, Write
---

# Run RALPH

Execute the RALPH autonomous development loop to implement features from the PRD.

## Two Operating Modes

### 1. PRD Loop Mode (Greenfield)
For implementing features from a PRD file:
```bash
./scripts/ralph/ralph.sh 20
```

### 2. Single-Task Mode (Brownfield)
For quick one-off tasks without a PRD:
```bash
./scripts/ralph/ralph.sh --task "Fix the login button styling"
```

## Prerequisites Check

1. Verify PRD exists (one of these formats):
```bash
cat prd.json 2>/dev/null || cat PRD.md 2>/dev/null || cat tasks.yaml 2>/dev/null
```

2. Check RALPH configuration:
```bash
cat .ralph/config.yaml
```

3. Check current status:
```bash
node scripts/run-ralph.js --status
```

## Running RALPH

### Start the autonomous loop:
```bash
# Run up to 20 iterations
./scripts/ralph/ralph.sh 20

# Or use the runner script
node scripts/run-ralph.js 20
```

### With options:
```bash
# Skip tests (faster iteration)
./scripts/ralph/ralph.sh --skip-tests 20

# Skip linting
./scripts/ralph/ralph.sh --skip-lint 20

# Custom branch
./scripts/ralph/ralph.sh --branch feature/my-feature 20

# Dry run (no commits)
./scripts/ralph/ralph.sh --dry-run 20
```

## PRD Formats Supported

### JSON (prd.json)
```json
{
  "userStories": [
    { "id": "US-001", "title": "...", "passes": false }
  ]
}
```

### Markdown (PRD.md)
```markdown
## Tasks
- [ ] First task
- [ ] Second task
- [x] Completed task
```

### YAML (tasks.yaml)
```yaml
tasks:
  - title: First task
    completed: false
```

## Configuration

Edit `.ralph/config.yaml` to customize:

- **rules**: Instructions the AI MUST follow
- **boundaries**: Files the AI should NOT modify
- **commands**: Quality gate commands
- **settings**: Retry logic, auto-commit, etc.

## Monitoring

- Watch `.ralph/progress.txt` for iteration logs and learnings
- Check PRD file for story completion status
- Review git log for commits: `git log --oneline -20`

## Commands

| Command | Description |
|---------|-------------|
| `--status` | Show PRD completion status |
| `--validate` | Validate PRD schema |
| `--reset` | Reset progress.txt |
| `--analyze` | Re-run project analysis |
| `--task "..."` | Single-task mode (brownfield) |
| `--skip-tests` | Skip test quality gate |
| `--skip-lint` | Skip lint quality gate |
| `--dry-run` | Don't commit changes |

## If Stuck

1. Check `.ralph/progress.txt` for patterns and learnings
2. Review `.ralph/config.yaml` rules and boundaries
3. Manually fix blocking issues
4. Reset if needed: `node scripts/run-ralph.js --reset`
5. Resume: `./scripts/ralph/ralph.sh 20`

## Quality Gates

RALPH must pass ALL quality gates before marking a task complete:

1. **Typecheck**: `npm run typecheck` or `npx tsc --noEmit`
2. **Lint**: `npm run lint`
3. **Tests**: `npm test`

Configure these in `.ralph/config.yaml` under `commands`.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.