Loop/Prompt Library / Loops from builders on X
🔁 Review and fix until nothing's left
The AI reviews its own changes, fixes what it finds, and reviews again, assuming there's still a bug until a pass comes back clean.
Use it when: The AI says 'looks good' too fast, and bugs still get through.
🛑 Stops when: A full review pass finds no real problems, or the round limit is hit.
Copy this loop
Review the changes on [branch]. Assume there IS at least one real bug, and look hard: wrong logic, missing cases, errors that are silently ignored, security holes. List each problem with file, line, and why it matters. Fix every real one, run the tests, then review the whole change again from the top. Don't invent problems just to have something to report; if you're not sure, say so. Big design questions go on a list for me instead of being changed. Stop when a full pass finds no real problems, or after [5] rounds. Show me what you fixed each round.
Review every change the agent makes before you keep it. Set a budget or round limit — the prompt includes one.
Found on X · by @steipete. 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.