Browse / Software development / Teaching & study / everything-about-loop-engineering

everything-about-loop-engineering

mdayan8

The complete reference and hands-on course on loop engineering — designing systems that prompt AI coding agents.

Open mdayan8/everything-about-loop-engineering → Does not match the four rules.

ai-agentsai-coding-agentsautomationbeginner-friendlyclaude-codecodexdeveloper-toolshands-on-projectslearning-pathllmloop-engineeringopen-source

Does not say

no agent config and no instruction file at the root.

In its own words

Aboutmdayan8/everything-about-loop-engineering

11 modules, 9 templates, 5 runnable projects, 8 agent skills, LLM wiki for AI agents.

README.md

Stop prompting. Start designing. The complete reference and hands-on course on loop engineering — designing systems that prompt AI coding agents — from absolute beginner to production-ready, in one repo.

This repo documents a discipline that is approximately two weeks old as of mid-June 2026. See HONESTY.md for transparency about the field's age, claim classification, and limitations.

The Big Idea

For two years, the way you worked with AI coding agents was simple: you typed a prompt, read the response, typed another prompt. One turn at a time. You were the human in the loop.

Loop engineering changes that. Instead of typing prompts yourself, you design the system that generates them — automatically, on a schedule, with verification, state memory, and sub-agent checking.

"My job is to write loops." — Boris Cherny, head of Claude Code at Anthropic

"You shouldn't be prompting coding agents anymore. You should be designing loops that prompt your agents." — Peter Steinberger (6.5M views)

"Build the loop. But build it like someone who intends to stay the engineer, not just the person who presses go." — Addy Osmani, who coined the term

What is Loop Engineering?

Loop engineering is replacing yourself as the person who prompts an AI coding agent — designing the system that does it instead.

The term was coined by Addy Osmani on June 8, 2026. The technique originated with Geoffrey Huntley's "Ralph" loop in July 2025.

The repository's own description, then 25% of its README. Read the rest on GitHub ↗

Read from the repository's git tree on 2026-08-17.