Trust your context, then check it
Even with great context, check drafts before they go out. A voice check is a short, reusable prompt that reviews any draft against your voice guide. You can review your own writing with it too.
The voice-check prompt
Copy this and fill in the brackets from your voice guide:
Review the draft below against my voice guide. Be strict.
My voice: [your three traits, "this, not that"]
Never use: [your words/phrases to avoid]
Audience: [your target audience]
For the draft:
1. Quote every sentence that breaks the voice, and say why.
2. List any words from my "never use" list.
3. Rate on-brand-ness from 1 to 10.
4. Rewrite only the sentences that failed, in my voice.
Draft:
[paste draft]Keep it handy
Two easy options:
- Prompt Builder. Paste the prompt (without a draft) into the Prompt Builder and refine it. Ace sharpens it into a Context / Clarity / Constraints version and keeps it under Library → Saved Prompts. Next time, click 💬 on that saved prompt: it opens Ask Ace with the prompt in the message box, and you paste your draft at the end.
- A plain note. Keep your filled-in version in your notes app and paste it when you need it.
Either way, keep your Brand Profile selected in Ask Ace when you run it.
Reading the result
- 8–10: publish, perhaps after one small edit.
- 5–7: apply the rewrites, then run the check once more.
- Below 5: something upstream is off. Is the right Brand Profile selected? Is the request missing context? Fix the cause, not just the draft.
When the same problem keeps coming back (say, the model keeps opening with a question), add it to Words/Phrases to Avoid or your compliance rules. Then it's handled on every message, not only in review.
Try it
Run the voice check on the bio you wrote in lesson 10 and note the score. You'll use this in the final lab.