REALITYWIPE

OLD SCRIPT DETECTED

I have to redo it until it’s right

Try this response:

I've read this enough times that I'm editing the memory of it, not the page.

Take the 60-second self-check

SEE THE PRACTICE

Build a response for this thought—and one move to prove it.

THE THOUGHT

I have to redo it until it’s right

YOUR RECORDED RESPONSE

I've read this enough times that I'm editing the memory of it, not the page.

ONE PRIVATE MOVE

Copy your current in-progress file or draft into a new note, timestamp it, and leave the original closed for two hours before rereading either.

The fourth revision of an email you already sent yourself twice

The version that's already fineYou wrote the email an hour ago. It said what you needed it to say. Then you reopened it, moved a sentence, reworded the middle paragraph, and put the sentence back where it started. The myth underneath this is that 'right' is a fixed destination you'll recognize the instant you land on it, so any version that still bothers you must not be it yet — even when the bothering has nothing left to point at.

Why the fourth pass feels more honest than the firstEach redo gives you something the first draft couldn't: a fresh read with fresh irritation to fix. That irritation feels like evidence the draft still needs work, so you keep going. But the irritation isn't measuring the draft anymore — it's measuring how many times you've now stared at the same sentences, which makes almost anything look slightly wrong on repeat viewing.

The lock-in testPick something you're currently on your third-or-later redo of. Copy the current version into a separate note, timestamp it, and don't touch the original for the next two hours. When you come back, read the locked copy once before touching anything. If you can't name a specific new problem within thirty seconds — not a feeling, an actual sentence — that's information about whether this redo was fixing something or just repeating the check.

TRACE · 01/04

How this script runs you

  • You reopen a finished file to 'just look at it again' and end up changing something back to how it was before.
  • You can't say what's wrong with the current version, only that you haven't looked at it enough times yet.
  • A version from two revisions ago sometimes reads better than the one you're on now, but you don't go back to it.

SOURCE_LOCATED · 02/04

The hidden rule underneath

The Perfection Standard

If I stop before I've reworked it enough times, I'll have let something wrong slip through on purpose — so the number of passes has to feel high enough, regardless of what any single pass actually finds.

FORGING_REPLACEMENT · 03/04

Replacement lines (record these)

  • I've read this enough times that I'm editing the memory of it, not the page.
  • One more pass without a named problem isn't a check, it's a habit.
  • The version from an hour ago said what I needed it to say.

Generated per person in the app — these are the flavor, not your script. Yours is built from your exact words.

INSTALL · 04/04

The protocol, on one card

When I think

I have to redo it until it’s right

I say

I've read this enough times that I'm editing the memory of it, not the page.

Then I do one thing

Copy your current in-progress file or draft into a new note, timestamp it, and leave the original closed for two hours before rereading either.

STATUS: READY_TO_INSTALL

Wipe this thought

4 minutes. Your voice. Free.

Start this exact wipe in the app →

The wipe protocol
  1. Write the thought exactly as it plays: "I have to redo it until it’s right". Word for word — the wipe targets the sentence, not the vibe.
  2. Trace the rule and the avoided action. What does this thought conveniently excuse you from doing?
  3. Record the replacement lines below in your own voice. Speak like you mean it — no recording, no install.
  4. Run the loop: play it every morning and night, log one proof action a day for 7 days.

Straight answers

What if the tenth redo actually does catch something the ninth one missed?

It sometimes will, which is exactly why the lock-in test asks you to name what changed rather than trust a general sense that another pass is warranted. A redo that finds a specific sentence problem is different from one that finds only a vague itch to keep going.

How is this different from normal editing, which is also repeated passes?

Normal editing stops when a pass turns up nothing new to fix. This pattern is the redo continuing past that point, driven by a feeling that not enough passes have happened yet rather than by anything the last pass actually surfaced.

What if I genuinely can't tell whether I'm still improving it or just repeating?

That's the exact question the two-hour lock-in answers without requiring you to trust your in-the-moment judgment. A specific, nameable problem found on a fresh read after a break is a different signal than the urge to reopen a file you already closed.

Related old scripts

Related self-checks

Related science

Researchers to explore

Related mechanisms

The research behind this script