npx claudepluginhub stavarengo/ralph-wiggum-loopAutonomous software development plugin implementing Geoffrey Huntley's Ralph Wiggum technique using Claude Code native features
Official prompts.chat marketplace - AI prompts, skills, and tools for Claude Code
Behavioral guidelines to reduce common LLM coding mistakes, derived from Andrej Karpathy's observations
Claude Code plugins for the Slidev presentation framework
Ralph is a Claude Code plugin that implements Geoffrey Huntley's "Ralph Wiggum as a software engineer" technique for autonomous software development using fresh context per iteration.
Ralph is inspired by Geoffrey Huntley's insight that you can achieve autonomous software development by giving an AI agent:
The name "Ralph Wiggum" (from The Simpsons) represents this philosophy: simple, focused, consistent execution without overthinking. Ralph doesn't need to remember the big picture - he just needs clear instructions for the current task.
Phase 1 (This Implementation): Interactive Mode
This is a Claude Code plugin using native agent features:
Phase 2 (Future): Full Autonomy
Future standalone CLI tool with:
Phase 1 provides immediate value while validating the core technique. Phase 2 will scale it to true autonomy.
# Add the marketplace
/plugin marketplace add stavarengo/ralph-wiggum-loop
# Install Ralph
/plugin install ralph@ralph-marketplace
# Initialize in your project
/ralph:init
# 1. Initialize Ralph in your project
/ralph:init
# 2. Customize the templates (optional)
# Edit docs/ai/ralph/PROMPT.md - Ralph's iteration instructions
# Edit docs/ai/ralph/AGENTS.md - Your build/test commands
# Edit docs/ai/ralph/fix_plan.md - Your task list
# 3. Start autonomous development
/ralph:start
# Ralph will now:
# - Pick tasks from fix_plan.md (highest priority first)
# - Implement, test, document, commit each task
# - Continue until complete or you stop it
# 4. Stop when needed
/ralph:stop
Ralph provides 5 skills for different stages of autonomous development:
/ralph:init - Initialize RalphSets up Ralph in your project by creating the directory structure and copying templates.
What it does:
docs/ai/ralph/ directoryPROMPT.md, fix_plan.md, AGENTS.mdCLAUDE.md symlink for auto-loading contextstatus.json for iteration tracking.gitignore to exclude Ralph's working directoryAfter initialization, your project will have:
your-project/
├── docs/ai/ralph/
│ ├── PROMPT.md # Ralph's iteration instructions
│ ├── fix_plan.md # Task list and priorities
│ ├── AGENTS.md # Build/test instructions + learnings
│ └── status.json # Iteration tracking
├── CLAUDE.md # Symlink → docs/ai/ralph/AGENTS.md
└── .gitignore # Updated to exclude docs/ai/ralph/
Next steps:
fix_plan.md to add your tasks (in priority order)AGENTS.md to add your build/test commands/ralph:start to begin autonomous development/ralph:start - Start Autonomous LoopBegins the autonomous iteration loop that runs continuously until all tasks are complete.
What it does:
Example output:
Ralph started. Beginning autonomous iteration loop...
=========================================
Starting iteration 1
=========================================
Iteration 1: Implemented user authentication
- Files changed: src/auth.ts, tests/auth.test.ts
- Tests: PASS (all 12 tests passing)
- Next: Add password reset flow
=========================================
Starting iteration 2
=========================================
Iteration 2: Implemented password reset flow
- Files changed: src/reset.ts, tests/reset.test.ts
- Tests: PASS (all 15 tests passing)
- Next: RALPH_COMPLETE
RALPH_COMPLETE signal detected. All tasks complete!
Ralph completed successfully!
- Total iterations: 2
- Status: complete
- All tasks in fix_plan.md have been completed.