wtf is Loop Engineer & how to setup for real

The video defines a 'loop engineer' as someone who designs systems where AI agents operate in autonomous loops, rather than just manually prompting them. The approach centers on four core components: triggers (cron, webhooks), file structure (memory/logs/artifacts), tools (LLM-accessible skills), and verification (automated testing and codebase hardening). A significant focus is placed on the 'loop contract', a structured way to define a loop's goal, workflow, task backlog, and timeline, which allows for persistent, cross-session work across multiple AI agents. The author also shares a 'codebase harness' template to automate the setup process for new projects. Throughout the video, the author uses their own company's workflows for support, SEO, and product growth as real-world examples, emphasizing how these loops can compound and integrate by sharing the same file system.