# Loop Engineering: Orchestrating AI Agents with Practical Patterns and Tools

This repository profile is provided by osrepos.com, an open source repository discovery platform.

Source: osrepos.com
Repository profile: https://osrepos.com/repo/cobusgreyling-loop-engineering
Generated for open source discovery and AI-assisted research.

Loop Engineering is a GitHub repository offering practical patterns, starters, and CLI tools for building robust AI coding agent systems. It shifts the focus from individual prompt crafting to designing control systems that orchestrate agents over time. This project empowers developers to create autonomous, iterative AI workflows for various development tasks.

GitHub: https://github.com/cobusgreyling/loop-engineering
OSRepos URL: https://osrepos.com/repo/cobusgreyling-loop-engineering

## Summary

Loop Engineering is a GitHub repository offering practical patterns, starters, and CLI tools for building robust AI coding agent systems. It shifts the focus from individual prompt crafting to designing control systems that orchestrate agents over time. This project empowers developers to create autonomous, iterative AI workflows for various development tasks.

## Topics

- agentic-ai
- ai-agents
- loop-engineering
- automation
- prompt-engineering
- javascript
- devtools
- llm

## Repository Information

Last analyzed by OSRepos: Thu Jun 25 2026 01:48:49 GMT+0100 (Western European Summer Time)
Detail views: 2
GitHub clicks: 1

## Safety Notice

OSRepos shares public repositories for knowledge and discovery only. Review source code, dependencies, licenses, and security implications before running or installing anything.

## Content

## Introduction
Loop Engineering is an innovative project that redefines how developers interact with AI coding agents. Instead of manually prompting agents, this repository provides the tools and patterns to design entire systems that automatically prompt and orchestrate AI agents. Inspired by the insights of Addy Osmani and Boris Cherny, it offers a comprehensive approach to building autonomous, iterative AI workflows.

## Installation
Getting started with Loop Engineering is straightforward, leveraging `npx` for its command-line tools. You can quickly scaffold projects, estimate costs, and audit your loops:

bash
# 1. Scaffold a starter pattern (e.g., daily-triage with Grok)
npx @cobusgreyling/loop-init . --pattern daily-triage --tool grok

# 2. Estimate token spend for your chosen cadence
npx @cobusgreyling/loop-cost --pattern daily-triage --level L1

# 3. Audit your loop's readiness, including budget and run-log scores
npx @cobusgreyling/loop-audit . --suggest


These CLIs are published to npm and can be used without cloning the repository. For development from source, refer to the project's `CONTRIBUTING.md`.

## Examples
Loop Engineering provides a rich set of production-ready patterns and starters for various AI coding agents. You can explore interactive examples and specific implementations for tools like Grok, Claude Code, Codex, and GitHub Actions.

The project includes patterns such as:
*   Daily Triage
*   PR Babysitter
*   CI Sweeper
*   Dependency Sweeper
*   Changelog Drafter
*   Post-Merge Cleanup
*   Issue Triage

An [interactive pattern picker](https://cobusgreyling.github.io/loop-engineering/#interactive) is available to help you choose the right loop for your needs.

## Why Use It
The core philosophy behind Loop Engineering is to move beyond individual prompts and embrace the design of control systems for AI agents. As Peter Steinberger and Boris Cherny highlight, the leverage point has shifted to orchestrating agents over time, allowing AI to iterate on goals recursively.

The project introduces 'The Five Building Blocks + Memory' for designing robust loops:
*   **Automations / Scheduling**: For discovery and triage on a cadence.
*   **Worktrees**: Enabling safe parallel execution.
*   **Skills**: Providing persistent project knowledge.
*   **Plugins & Connectors**: For integrating with real-world tools (MCP).
*   **Sub-agents**: Facilitating maker/checker splits.
*   **+ Memory / State**: A durable spine for managing state outside conversations.

By adopting Loop Engineering, developers can achieve greater automation, consistency, and efficiency in their AI-driven development workflows, while also understanding the caveats and safety considerations involved.

## Links
Explore Loop Engineering further through these resources:

*   [GitHub Repository](https://github.com/cobusgreyling/loop-engineering)
*   [Interactive Showcase](https://cobusgreyling.github.io/loop-engineering/)
*   [Loop Engineering Essay (Substack)](https://cobusgreyling.substack.com/p/loop-engineering)
*   [Canonical Essay by Addy Osmani](https://addyosmani.com/blog/loop-engineering/)