cobusgreyling/loop-engineering — reverse-engineered prompt

I want a practical Loop Engineering repo for people using AI coding agents like Claude Code, Codex, Grok, and Cursor. Make it feel like a real handbook plus toolkit, not just a blog post. It should explain the idea of designing repeatable loops that prompt agents for you, show the core building blocks like scheduling, worktrees, skills, sub agents, connectors, and memory, and include a clean interactive showcase or pattern picker so someone can quickly choose a loop to start with.

Please add ready to use starters and examples for things like daily triage, PR babysitting, CI sweeping, dependency updates, changelog drafting, post merge cleanup, and issue triage. I also want simple command line tools to scaffold a starter, estimate token spend, and audit whether a repo is ready to run loops. Include safety guidance, budget and run log files, stories of wins and failures, and docs that compare how this works across different agent tools. If anything is unclear, look up current docs online and make sensible choices.