Loop/Prompt Library / Loops from builders on X
🔁 Cheap checks first, wake the big AI only when needed
A simple, cheap check runs often (is my site up? anything urgent in email?) and only wakes the expensive AI when there's a real decision to make.
Use it when: You want things watched all day, but don't want to pay a big AI to look at nothing.
🛑 Stops when: Each check ends in 'all fine' or one wake-up. Anything risky waits for you.
Copy this loop
Set up two levels. Level 1 (cheap, frequent): every [15] minutes, check [list, e.g. 'site loads in under 2 seconds', 'API answers']. Use a simple script or a small model. Write one line to [log] each time. Stay quiet if everything looks normal. Level 2 (smart, rare): only if a level-1 check fails [2] times in a row, wake the main AI with the details. The main AI explains what's wrong and suggests a fix. It may only look, not change anything. Never send, post, buy, pay, delete, or sign anything; I do those myself. Stop each check after its one-line result. Cap level-2 wake-ups at [5] a day; after that, just message me.
Review every change the agent makes before you keep it. Set a budget or round limit — the prompt includes one.
Found on X · by @clawpowered. 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.