Ideas and scripts
How MORR writes for you
Every caption, hook, idea and script MORR writes is written from you: how you write, what has worked on your own channels, what you post about, and the hooks you study. It writes several takes in different kinds of opening, has an editor score them, rewrites the best, and throws out the phrases that give machine writing away. Then it tells you why the first one fits.
What it reads first
| It reads | From | So that |
|---|---|---|
| How you write | Your own posted captions, your scheduled ones, and scripts you wrote or read | Sentence length, lowercase or not, emoji, the words and openers you repeat. Measured, not guessed, and remeasured every few days. |
| What worked | Your connected channels' views, or reach where that is all a platform shares | Which kinds of opening, formats and lengths beat your own usual. A kind needs two posts before it counts, so one lucky post is not a pattern. |
| What you post about | Your pillars on the canvas and your series | Ideas land in your lanes, not generic ones. |
| Who you study | The creators behind your daily hooks, and hooks you used or saved | It borrows the device of a proven opening, never the words. |
| Your brand card | Settings: who you are, who you talk to, words to use and words never to use | Your never list wins over everything else. |
Kinds of opening
Each take opens a different way, so the options are real choices rather than three rewordings of one line. Every option is labeled with its kind.
| Kind | What it does |
|---|---|
| Contrarian | Names the advice everyone repeats and says why it is wrong for this viewer. |
| Specific number | Leads with a figure you can picture: a count, a price, a weight, a day. |
| Story in motion | Starts in the middle of the moment it went sideways. |
| I tried it | A specific thing, for a specific time, and the result is the promise. |
| Mistake | Names the costly mistake before the fix. |
| Before and after | Shows both ends and holds back how you got there. |
| Question with stakes | A question the viewer cannot answer yet, where the wrong answer costs them. |
Why this works for you
The line under an option names what it rests on, from your own numbers when there are some: Your "I tried it" openings average 2.1x your usual views (3 posts). It never quotes a number about your account that was not measured. With no channels connected yet it says what the kind of opening does instead.
Scripts
A short script is timed to its length at a natural reading pace: the hook in the first two seconds, a reason to keep watching, one concrete point per beat, the payoff, then one ask. Each beat carries the on-screen text for people watching muted and a note on the shot. A long video is planned first, with a re-hook at the start of every section, then written a section at a time against that plan.
When you give it little to go on
Two words and nothing else is a thin request. It still writes, but it says what it assumed, and when there is little on your account to go on it asks the one question that would change the piece most. Answer it in the topic and write again.
What it never does
It never invents a statistic, a study, a client result or a quote, and never makes a medical, legal or financial promise you could be held to. It holds every answer to one list of the phrases and shapes that give machine writing away, stacked emoji and walls of hashtags among them, and keeps the hashtag count sane for the platform. Ask writes captions to the same standard: one that still reads as generated goes back to be rewritten before it reaches your calendar.
Every rewrite costs one AI credit, the same as the first draft. See Batch for writing a month at a time, and Scripts for the tweaks on a script.