Audit it before I tweet

Why use it

The audit on a long post asks whether anyone gets past the first line. On a tweet nobody has that problem, because the whole thing is visible at once. The risk moves to context: a tweet arrives with no title, no thread and no idea who you are, and then gets screenshotted and quoted.

What it does

It reads the draft with everything around it hidden, checks the claims before the phrasing because a tweet carries a mistake further than a post does, then runs the screenshot test: how this reads quoted by somebody who dislikes you. Then it asks what the length cut, because the first thing trimmed to fit is usually the qualifier that made the claim true.

What you get

• The tweet read as a stranger meets it, with no bio and no thread to explain it
• A claim check on every number, every story and anyone identifiable
• The screenshot test, which is the normal life of a tweet that travels
• At most three fixes in order, and a straight answer on whether to post it

What it will not do

It will not predict performance, and it will not give you a score. Reading a draft against a handful of pasted posts is a read-through, not a measurement.

What changes when VoiceMoat is connected

Everything above is craft. The step that needs to know you is whether it still sounds like you, and short form makes borrowing somebody else's register both easier and more obvious.

• score_voice_match Scores the draft from 0 to 100 against your trained profile and names the lines that miss, which is where the audit stops being an opinion.
• get_voice_profile Is per platform, so a tweet is judged against how you write tweets rather than against your long-form voice.
• improve_post Takes the draft through a deeper pass when the audit finds three real problems rather than one.
• publish_post Sends it once you are happy, behind a preview and a one-time confirmation.