Loop/Prompt Library / Loops from builders on X
🔁 A cheap judge decides when it's done
The strong AI does the work, and a smaller, cheaper AI answers the small 'is it done yet?' questions so you don't have to keep typing 'continue'.
Use it when: Long jobs keep stopping to ask small yes-or-no questions, and the big model is expensive to keep checking.
🛑 Stops when: The judge says the finish line is met and the final check agrees, or the round or budget limit is hit.
Copy this loop
Finish line: [one clear yes-or-no goal, e.g. 'all tests in folder X pass']. You (the main AI) do the work in rounds. After each round, a smaller, cheaper AI looks at the current state and answers only: 'Is the finish line met? yes/no, and if no, what's the next thing to fix?' Use the cheap judge for small yes/no questions like that; only wake up the main AI when there is real work. Never let the judge approve risky steps like deleting, deploying, or spending money; those come to me. Stop when the judge says yes and you confirm it with [the check command], or after [20] rounds or [budget], whichever comes first. Then show me the final result and the cost.
Review every change the agent makes before you keep it. Set a budget or round limit — the prompt includes one.
Found on X · by @L1vsun. Prompt rewritten in plain English by FixKit AI.
More loops like this
One useful thing every night
You give the AI a short mission, and each night it picks one task that moves you closer and leaves it for you to check in the morning.
Two watchers keep your agent healthy
Two separate AI helpers read your main agent's logs, find what broke, and must both agree on a fix before it is used.
Weekly skill cleanup
Once a week, the AI grades every saved skill or saved prompt it uses, removes the dead ones, and merges the repeats.
Overnight drafts that learn from results
Each night the AI drafts tomorrow's post from your content plan, and reads your notes about what worked before it writes the next one.
Split the work, keep only what passes
Splits a big job across several AI helpers working side by side, checks each one's work, and keeps only the work that passes.
Practice a website task until it's reliable
The AI tries a website chore, learns from each failed try, and keeps adjusting until it has steps that work every time.