I run this prompt on every AI draft before I use it.
It rewrites the patterns that make content sound generated.
Here's the full prompt. Copy it. Paste your draft at the end.
THE VOICE REWRITE PROMPT:
"Rewrite the draft below in operator voice.
Voice spec:
- Peer-to-peer, not teacher-to-student. Assume the reader does this work too.
- Operator, not consultant. Sounds like someone who has shipped this, not someone who has slides about it.
- Pain-first. Open with the reader's actual problem before naming the fix.
- Receipts-forward. Use real names, real numbers, real specifics. If a specific is not available, ask for it instead of inventing one.
- Humble close. No mic drop, no 'now go crush it,' no summary of what was just said.
- Conversational rhythm. Mix long sentences with shorter ones. No stacked fragments.
Hard rules:
- No em dashes
- No 'not X, but Y' or 'it's not this, it's that'
- No three-or-more short sentences in a row
- No setup phrases: 'here's the thing,' 'here's why this matters,' 'the truth is'
- No buzzwords: game-changer, unlock, delve, leverage, crucial, powerful, robust
- No credentials before the hook
- No fabricated stats
Output:
1. The rewrite
2. A line-by-line note of what changed and why
3. Three alternate openers, since the first line does 80% of the work
Draft: [paste your draft here]"
WHY THIS WORKS:
The voice spec tells AI how to sound, not just what to avoid.
The hard rules catch the patterns that feel generated.
The line-by-line notes show you what was wrong so you learn, not just copy.
The alternate openers give you options because the hook matters most.
Yesterday I shared the prompt that catches problems. This one fixes them.