Skip to main content

Introduction

wreck-it is a TUI agent harness that uses the GitHub Copilot SDK to perform Ralph Wiggum loops — continuous, external bash-style loops that run AI agent tasks to completion.

Ralph Wiggum. Cloud Scale.

wreck-it brings autonomous AI agent orchestration to your terminal. Define tasks, let the agents work, and watch your codebase evolve.

What is a Ralph Wiggum Loop?

The Ralph Wiggum Loop is a continuous execution pattern designed for AI agent workflows. Named after the Simpsons character famous for his persistence ("I'm helping!"), this pattern ensures tasks are completed through persistent iteration.

  • External Loop: Not an internal AI feature, but an external script running while true
  • Persistent Memory: Uses the filesystem (codebase) as memory rather than chat history
  • Workflow: Reads task file → Implements change → Runs tests → Commits code → Repeats
  • Safety: Includes max iterations limit to prevent infinite loops and excessive costs

Features

  • 🎨 TUI Interface — Beautiful terminal UI showing tasks, progress, and logs
  • 🔄 Continuous Execution — Runs until all tasks are complete or max iterations reached
  • 📝 Task Management — JSON-based task tracking with status persistence
  • 🧪 Automatic Testing — Runs tests after each task execution
  • 💾 Git Integration — Automatically commits successful changes
  • 🔒 Safety Limits — Configurable max iterations to prevent runaway costs
  • 🤖 Headless Mode — Run without TUI for CI/CD automation
  • ☁️ Cloud Agents — GitHub Models integration for cloud-scale agent execution
  • 🐕 Dogfooding — wreck-it develops itself via scheduled agent swarms