Earlier this month I wrote WTF Is a Loop? Peter Steinberger vs. Boris Cherny, which did 3.6M views on what a loop even is. This is the sequel, and it answers the next question: which loops do people actually run? I pointed /last30days (@slashlast30days) at exactly that, eight searches across X, TikTok, Reddit, YouTube, and GitHub. Below are the fifteen best that came back, every one attributed, with real engagement where it went viral and honest provenance where it came from a curated catalog, rewritten as a command you can paste tonight if there wasn't one. First, the one distinction almost everyone gets wrong. First, the three commands (this is where everyone trips) The cleanest way to say it surfaced from a creator named inyourhandmedia on TikTok: Goal is keep working until the outcome is achieved. Loop is keep repeating a task while I am here. Routine is keep working while I am gone. In real commands: `/goal <condition>` runs until a verifiable condition is true, then stops. A separate fast model checks after every turn whether you are actually done. This is the one for "fix it until the tests pass," and it is the one command both tools share: Claude Code shipped it in v2.1.139 (docs), and Codex shipped its own in CLI v0.128.0 with set, pause, resume, and clear controls (docs). `/loop <interval> <prompt>` repeats on a timer while your session is open, like `/loop 5m check the deploy`. It is for watching something, hands on, right now (docs). Codex has no `/loop` command yet; its equivalent is `codex exec` wrapped in a shell loop, or a minute-interval Thread Automation in the Codex app. `/schedule <description>` creates a cloud Routine that runs while your laptop is closed, like `/schedule daily PR review at 9am` (docs). This is the "while I sleep" one. Codex's equivalent is Automations in the Codex app: standalone, project, or thread automations on a daily, weekly, or custom cron schedule, with results landing in a Triage inbox (local now, cloud execution rolling out). One trap that comes up constantly in the threads: there is no `/routine` command in either tool. In Claude Code the scheduler is `/schedule`; in Codex it is Automations in the app. Get the verb right and every loop below just works. Eleven of these came straight off X, TikTok, Reddit, and GitHub with the engagement attached; the last four I lifted from the one catalog worth raiding, and I have flagged exactly which. 1. The build-test-fix pair (loop) The single most-demoed loop in the whole pull, from a creator named raycfu whose walkthrough did 43,587 views and 1,040 comments on Instagram. Two agents: a builder that writes code and a checker that runs the tests, types, and lint and reports exactly what broke. They pass work back and forth until it is clean. The whole pitch is the pain it kills: a one-shot agent ships its bugs. 2. Boris's verifier loop (loop) This is the loop in Boris Cherny's own words, and the highest-engagement description of it (@bcherny, 781 likes): run Claude Code plus an advanced model plus a verifier in a loop, feed it tasks, and remove bottlenecks as you go. The verifier is the part everyone skips. Without it you are just trusting the agent. 3. The loop-engineer starter (harness) The most-watched actually-set-one-up video, from AI Jason (15,436 views, 537 likes), ships a free loop-engineer template: a codebase harness plus a knowledge template you clone, point at your repo, and run, so you are not wiring the build, observe, verify, and stop scaffolding from scratch. The fastest on-ramp if you want a loop running tonight without designing the plumbing yourself. 4. The five-minute repository maintainer (loop) Peter Steinberger, who merged 859 PRs across his repos in the last 30 days at a 95 percent acceptance rate, runs this one on a tight timer: every five minutes while he is working, the agent does one small, verified piece of upkeep. What to clean is the agent's call, not a hardcoded script. That decision is the whole point. 5. The plan-generate-verify-fix loop (goal) A creator named qbuilder (4,560 views, 125 likes on TikTok) runs a bounded version that fixes the runaway problem cold: plan, generate, verify, fix, repeat, with everything saved to files and a hard cap of five iterations. You only read the final version. The cap is what makes it safe to leave. 6. roborev, the post-commit reviewer (shipped tool) roborev is a free, open-source code-review tool (a Go binary; install from roborev.io) that surfaced in the research through Dan Kornas. It installs a git hook so every commit triggers a background review, then feeds the findings into an agentic fix loop while the context is still warm. The launch tweet was quiet at 20 likes, but the repo is the real signal: 1,410 stars and commits shipping the same day I write this. It is the installable version of the one thing this whole piece argues is the hard part, a verifier living inside the loop, and it plugs into Claude Code, Codex, and Gemini CLI. 7. T