Tell the AI how your store thinks. Once.
Set your brand voice, target customers, SEO preferences, forbidden claims and writing examples on one page. Every AI suggestion across your catalog follows them. You still review each one.
One page that steers every suggestion.
Brand voice & audience
Describe how your store sounds and who it speaks to, whether that's clear and warm or clinical and premium. Every draft follows that tone.
Target customers
Tell the AI who buys from you and what they actually care about. Suggestions sell to the right person, not a generic shopper.
SEO preferences
Set the keywords that matter, your naming conventions and how titles should be built. It's your search strategy, written down once.
Words & claims to avoid
Hard guardrails, not hints. Your no-go list applies to every prompt, and you still review every suggestion before it saves.
Writing examples
Paste copy you're proud of. The AI imitates your best work instead of a generic template.
A live prompt preview
See exactly what the AI is told before it writes a word. There's no black box between your settings and the output.
Set it up in minutes. It pays off on every field.
Describe your store
Share your brand voice, audience and SEO preferences in your own words.
Set guardrails
List the words and claims that must never appear in your copy.
Add examples
Show the AI the copy you already love, so it imitates you.
Every suggestion follows
From then on, each draft is shaped by your context. You still review every single one.
You stay in control.
Yours alone
Your context stays in your workspace. It's never shared with other workspaces and never used for anyone else's store.
Change it any time
Update your voice or guardrails and new suggestions follow right away. There's no retraining and no waiting.
Nothing rewrites itself
Setting a context never rewrites existing copy. Drafts happen only when you ask for them.
Enforced on every prompt
Guardrails aren't a polite request. They're applied to every prompt, on every field, in every language.