Loop/Prompt Library / Loops from builders on X
🔁 The builder can't touch the tests
One AI writes the code, a separate AI only runs the tests, and the builder is never allowed to change the tests to make them pass.
Use it when: You've caught the AI quietly editing a test so it would pass.
🛑 Stops when: All checks pass with the tests unchanged, or [5] cycles pass and it stops for you.
Copy this loop
Set up two helpers. Builder: may read and edit code, but may NOT edit anything in [test folder] and may NOT run the tests. Checker: may run tests, type checks, and lint, but may NOT edit any files. Before starting, record a fingerprint (checksum) of the test folder. Each cycle: the Builder makes a change, the Checker runs everything and reports failures, and the failures go back to the Builder. At the end, confirm the test folder fingerprint hasn't changed. If it has, throw away that work and tell me. Stop when all checks pass with unchanged tests, or after [5] cycles. Then show me the result.
Review every change the agent makes before you keep it. Set a budget or round limit — the prompt includes one.
Found on X · by @hitu_monke. 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.