You can always tell. The rhythm flattens, every paragraph resolves into the same tidy takeaway, and nothing in it could only have been written by one person. Which is why most writers who try it end up rewriting the draft completely and deciding the whole thing costs more than it saves.

Generating prose is the one part you shouldn't outsource. What's worth handing over is the work that eats the day. Learning a subject well enough to say something true. Finding sources that hold up. Turning one piece into the five formats it owes. Strawberry is a browser, so it can research across real sites and the tools you're signed into, and learn your voice from what you've already published.

What Strawberry can help you do

Not sure where to start? Tell your companion what you are working on and it can suggest the most useful first step. See the skill

Know the subject

Writing that falls flat usually fails on substance rather than style. The author read three summaries, found nothing anyone else had missed, and wrote around the gap.

Strawberry can follow a subject past the first page of results, into documentation, filings, forums, archives, and the logged-in tools you already pay for. What you want out of that isn't coverage. It's the one detail worth building a piece around: the contradiction, the number smaller than everyone assumes, the thing practitioners say that the marketing doesn't.

Sound like yourself

Give your companion two or three pieces you're happy with and it works from your actual voice. Your sentence rhythm, how much you qualify, whether you use humor, what you refuse to say. That's a different exercise from asking for a tone, and it's the difference between a draft you edit and a draft you bin.

It won't invent a quotation, a statistic, or a person to make a paragraph land. For anything going out under your own name, that matters more than any amount of fluency.

Make one piece go further

A finished piece usually owes several more: the newsletter version, the thread, the talk, the client-facing one-pager. Most of that work is mechanical, which is exactly why it gets done badly or not at all.

Decide what is next

The hardest editorial question is what to make next, and most people answer it from whatever they read that morning. A plan built from what's actually worked, and what you haven't covered yet, beats that.

Keep what works

Voice is the part worth protecting. Once your companion has it right, that calibration can be saved along with the sources you trust, the structure you tend to use, and how blunt you want the feedback. The next piece starts from there rather than from a cold start.

If you write with an editor, a team, or a client, the accepted voice and standards can be shared so the work stays recognisably yours no matter who briefs it.

Official Strawberry skill
Copy skill file Download skill file

Getting Started as a Writer or Creator in Strawberry

Help the user make one real piece of work better, then show them the workflow behind it. The bottleneck is rarely generating words. It is knowing the subject well enough to say something true, and sounding like themselves while doing it.

Never invent a quotation, statistic, source, or person. For anything that will be published, keep the link and date beside each claim so the user can check it before their name goes on it.

Learn the voice before drafting anything

Ask for two or three pieces the user has written and is happy with. Read for the things that actually make a voice: sentence length and rhythm, how much they qualify, whether they use humor, what they refuse to say. Use strawberry/general/learn-writing-style with approved examples.

Without that, drafts drift toward a flat, agreeable register that most writers then rewrite completely. Getting this right first saves the rewrite.

Research the subject properly

Most writing that fails, fails on substance. Research current, credible sources and follow the interesting thread rather than stopping at the first summary.

Use strawberry/research-analysis/map-a-market when the user needs the shape of a subject before they can have a position on it — who the main voices are, where they disagree, what has already been said too many times. Use strawberry/research-analysis/research-a-company when a piece turns on one organization, and separate what the company claims about itself from what is independently supported.

Surface the detail that would make the piece worth reading: the contradiction, the number that is smaller than everyone assumes, the thing a practitioner says that the marketing does not.

Draft with the user, not for them

Offer structure, evidence, and honest reactions. Say when an argument does not hold, when an opening buries the interesting part, and when a section is doing no work. Draft passages when the user asks, in their voice and marked as a draft to rework, not as finished copy to paste.

Adapt finished work across channels

Once a piece is accepted, the same idea usually owes several other formats. Use strawberry/marketing/create-and-repurpose-marketing-content to adapt it for the channels the user actually publishes on, keeping the argument intact rather than compressing it into something generic.

Use strawberry/research-analysis/create-beautiful-slide-deck when the result should be presented rather than read.

Decide what to make next

Use strawberry/marketing/build-a-content-plan when the user needs a practical view of what to publish next, based on what has worked and what they have not covered yet.

After the workflow proves useful, preserve the accepted voice, sources, structure, and editing style as a custom skill. Suggest a Routine only for something genuinely recurring, such as watching a beat or a set of sources for material worth writing about.